mirror of https://github.com/docker/cli.git
Merge pull request #316 from thaJeztah/revendor-dependencies
Revendor dependencies
This commit is contained in:
commit
a765218f19
|
@ -1,97 +0,0 @@
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
// Simple tool to create an archive stream from an old and new directory
|
|
||||||
//
|
|
||||||
// By default it will stream the comparison of two temporary directories with junk files
|
|
||||||
package main
|
|
||||||
|
|
||||||
import (
|
|
||||||
"flag"
|
|
||||||
"fmt"
|
|
||||||
"io"
|
|
||||||
"io/ioutil"
|
|
||||||
"os"
|
|
||||||
"path"
|
|
||||||
|
|
||||||
"github.com/Sirupsen/logrus"
|
|
||||||
"github.com/docker/docker/pkg/archive"
|
|
||||||
)
|
|
||||||
|
|
||||||
var (
|
|
||||||
flDebug = flag.Bool("D", false, "debugging output")
|
|
||||||
flNewDir = flag.String("newdir", "", "")
|
|
||||||
flOldDir = flag.String("olddir", "", "")
|
|
||||||
log = logrus.New()
|
|
||||||
)
|
|
||||||
|
|
||||||
func main() {
|
|
||||||
flag.Usage = func() {
|
|
||||||
fmt.Println("Produce a tar from comparing two directory paths. By default a demo tar is created of around 200 files (including hardlinks)")
|
|
||||||
fmt.Printf("%s [OPTIONS]\n", os.Args[0])
|
|
||||||
flag.PrintDefaults()
|
|
||||||
}
|
|
||||||
flag.Parse()
|
|
||||||
log.Out = os.Stderr
|
|
||||||
if (len(os.Getenv("DEBUG")) > 0) || *flDebug {
|
|
||||||
logrus.SetLevel(logrus.DebugLevel)
|
|
||||||
}
|
|
||||||
var newDir, oldDir string
|
|
||||||
|
|
||||||
if len(*flNewDir) == 0 {
|
|
||||||
var err error
|
|
||||||
newDir, err = ioutil.TempDir("", "docker-test-newDir")
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
defer os.RemoveAll(newDir)
|
|
||||||
if _, err := prepareUntarSourceDirectory(100, newDir, true); err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
} else {
|
|
||||||
newDir = *flNewDir
|
|
||||||
}
|
|
||||||
|
|
||||||
if len(*flOldDir) == 0 {
|
|
||||||
oldDir, err := ioutil.TempDir("", "docker-test-oldDir")
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
defer os.RemoveAll(oldDir)
|
|
||||||
} else {
|
|
||||||
oldDir = *flOldDir
|
|
||||||
}
|
|
||||||
|
|
||||||
changes, err := archive.ChangesDirs(newDir, oldDir)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
|
|
||||||
a, err := archive.ExportChanges(newDir, changes)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
defer a.Close()
|
|
||||||
|
|
||||||
i, err := io.Copy(os.Stdout, a)
|
|
||||||
if err != nil && err != io.EOF {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
fmt.Fprintf(os.Stderr, "wrote archive of %d bytes", i)
|
|
||||||
}
|
|
||||||
|
|
||||||
func prepareUntarSourceDirectory(numberOfFiles int, targetPath string, makeLinks bool) (int, error) {
|
|
||||||
fileData := []byte("fooo")
|
|
||||||
for n := 0; n < numberOfFiles; n++ {
|
|
||||||
fileName := fmt.Sprintf("file-%d", n)
|
|
||||||
if err := ioutil.WriteFile(path.Join(targetPath, fileName), fileData, 0700); err != nil {
|
|
||||||
return 0, err
|
|
||||||
}
|
|
||||||
if makeLinks {
|
|
||||||
if err := os.Link(path.Join(targetPath, fileName), path.Join(targetPath, fileName+"-link")); err != nil {
|
|
||||||
return 0, err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
totalSize := numberOfFiles * len(fileData)
|
|
||||||
return totalSize, nil
|
|
||||||
}
|
|
|
@ -1,62 +0,0 @@
|
||||||
// Copyright 2016 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
// mkpost processes the output of cgo -godefs to
|
|
||||||
// modify the generated types. It is used to clean up
|
|
||||||
// the sys API in an architecture specific manner.
|
|
||||||
//
|
|
||||||
// mkpost is run after cgo -godefs by mkall.sh.
|
|
||||||
package main
|
|
||||||
|
|
||||||
import (
|
|
||||||
"fmt"
|
|
||||||
"go/format"
|
|
||||||
"io/ioutil"
|
|
||||||
"log"
|
|
||||||
"os"
|
|
||||||
"regexp"
|
|
||||||
)
|
|
||||||
|
|
||||||
func main() {
|
|
||||||
b, err := ioutil.ReadAll(os.Stdin)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
s := string(b)
|
|
||||||
|
|
||||||
goarch := os.Getenv("GOARCH")
|
|
||||||
goos := os.Getenv("GOOS")
|
|
||||||
if goarch == "s390x" && goos == "linux" {
|
|
||||||
// Export the types of PtraceRegs fields.
|
|
||||||
re := regexp.MustCompile("ptrace(Psw|Fpregs|Per)")
|
|
||||||
s = re.ReplaceAllString(s, "Ptrace$1")
|
|
||||||
|
|
||||||
// Replace padding fields inserted by cgo with blank identifiers.
|
|
||||||
re = regexp.MustCompile("Pad_cgo[A-Za-z0-9_]*")
|
|
||||||
s = re.ReplaceAllString(s, "_")
|
|
||||||
|
|
||||||
// Replace other unwanted fields with blank identifiers.
|
|
||||||
re = regexp.MustCompile("X_[A-Za-z0-9_]*")
|
|
||||||
s = re.ReplaceAllString(s, "_")
|
|
||||||
|
|
||||||
// Replace the control_regs union with a blank identifier for now.
|
|
||||||
re = regexp.MustCompile("(Control_regs)\\s+\\[0\\]uint64")
|
|
||||||
s = re.ReplaceAllString(s, "_ [0]uint64")
|
|
||||||
}
|
|
||||||
|
|
||||||
// gofmt
|
|
||||||
b, err = format.Source([]byte(s))
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
|
|
||||||
// Append this command to the header to show where the new file
|
|
||||||
// came from.
|
|
||||||
re := regexp.MustCompile("(cgo -godefs [a-zA-Z0-9_]+\\.go.*)")
|
|
||||||
b = re.ReplaceAll(b, []byte("$1 | go run mkpost.go"))
|
|
||||||
|
|
||||||
fmt.Printf("%s", b)
|
|
||||||
}
|
|
|
@ -1,250 +0,0 @@
|
||||||
// Copyright 2009 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
/*
|
|
||||||
Input to cgo -godefs. See also mkerrors.sh and mkall.sh
|
|
||||||
*/
|
|
||||||
|
|
||||||
// +godefs map struct_in_addr [4]byte /* in_addr */
|
|
||||||
// +godefs map struct_in6_addr [16]byte /* in6_addr */
|
|
||||||
|
|
||||||
package unix
|
|
||||||
|
|
||||||
/*
|
|
||||||
#define __DARWIN_UNIX03 0
|
|
||||||
#define KERNEL
|
|
||||||
#define _DARWIN_USE_64_BIT_INODE
|
|
||||||
#include <dirent.h>
|
|
||||||
#include <fcntl.h>
|
|
||||||
#include <signal.h>
|
|
||||||
#include <termios.h>
|
|
||||||
#include <unistd.h>
|
|
||||||
#include <mach/mach.h>
|
|
||||||
#include <mach/message.h>
|
|
||||||
#include <sys/event.h>
|
|
||||||
#include <sys/mman.h>
|
|
||||||
#include <sys/mount.h>
|
|
||||||
#include <sys/param.h>
|
|
||||||
#include <sys/ptrace.h>
|
|
||||||
#include <sys/resource.h>
|
|
||||||
#include <sys/select.h>
|
|
||||||
#include <sys/signal.h>
|
|
||||||
#include <sys/socket.h>
|
|
||||||
#include <sys/stat.h>
|
|
||||||
#include <sys/time.h>
|
|
||||||
#include <sys/types.h>
|
|
||||||
#include <sys/uio.h>
|
|
||||||
#include <sys/un.h>
|
|
||||||
#include <sys/wait.h>
|
|
||||||
#include <net/bpf.h>
|
|
||||||
#include <net/if.h>
|
|
||||||
#include <net/if_dl.h>
|
|
||||||
#include <net/if_var.h>
|
|
||||||
#include <net/route.h>
|
|
||||||
#include <netinet/in.h>
|
|
||||||
#include <netinet/icmp6.h>
|
|
||||||
#include <netinet/tcp.h>
|
|
||||||
|
|
||||||
enum {
|
|
||||||
sizeofPtr = sizeof(void*),
|
|
||||||
};
|
|
||||||
|
|
||||||
union sockaddr_all {
|
|
||||||
struct sockaddr s1; // this one gets used for fields
|
|
||||||
struct sockaddr_in s2; // these pad it out
|
|
||||||
struct sockaddr_in6 s3;
|
|
||||||
struct sockaddr_un s4;
|
|
||||||
struct sockaddr_dl s5;
|
|
||||||
};
|
|
||||||
|
|
||||||
struct sockaddr_any {
|
|
||||||
struct sockaddr addr;
|
|
||||||
char pad[sizeof(union sockaddr_all) - sizeof(struct sockaddr)];
|
|
||||||
};
|
|
||||||
|
|
||||||
*/
|
|
||||||
import "C"
|
|
||||||
|
|
||||||
// Machine characteristics; for internal use.
|
|
||||||
|
|
||||||
const (
|
|
||||||
sizeofPtr = C.sizeofPtr
|
|
||||||
sizeofShort = C.sizeof_short
|
|
||||||
sizeofInt = C.sizeof_int
|
|
||||||
sizeofLong = C.sizeof_long
|
|
||||||
sizeofLongLong = C.sizeof_longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Basic types
|
|
||||||
|
|
||||||
type (
|
|
||||||
_C_short C.short
|
|
||||||
_C_int C.int
|
|
||||||
_C_long C.long
|
|
||||||
_C_long_long C.longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Time
|
|
||||||
|
|
||||||
type Timespec C.struct_timespec
|
|
||||||
|
|
||||||
type Timeval C.struct_timeval
|
|
||||||
|
|
||||||
type Timeval32 C.struct_timeval32
|
|
||||||
|
|
||||||
// Processes
|
|
||||||
|
|
||||||
type Rusage C.struct_rusage
|
|
||||||
|
|
||||||
type Rlimit C.struct_rlimit
|
|
||||||
|
|
||||||
type _Gid_t C.gid_t
|
|
||||||
|
|
||||||
// Files
|
|
||||||
|
|
||||||
type Stat_t C.struct_stat64
|
|
||||||
|
|
||||||
type Statfs_t C.struct_statfs64
|
|
||||||
|
|
||||||
type Flock_t C.struct_flock
|
|
||||||
|
|
||||||
type Fstore_t C.struct_fstore
|
|
||||||
|
|
||||||
type Radvisory_t C.struct_radvisory
|
|
||||||
|
|
||||||
type Fbootstraptransfer_t C.struct_fbootstraptransfer
|
|
||||||
|
|
||||||
type Log2phys_t C.struct_log2phys
|
|
||||||
|
|
||||||
type Fsid C.struct_fsid
|
|
||||||
|
|
||||||
type Dirent C.struct_dirent
|
|
||||||
|
|
||||||
// Sockets
|
|
||||||
|
|
||||||
type RawSockaddrInet4 C.struct_sockaddr_in
|
|
||||||
|
|
||||||
type RawSockaddrInet6 C.struct_sockaddr_in6
|
|
||||||
|
|
||||||
type RawSockaddrUnix C.struct_sockaddr_un
|
|
||||||
|
|
||||||
type RawSockaddrDatalink C.struct_sockaddr_dl
|
|
||||||
|
|
||||||
type RawSockaddr C.struct_sockaddr
|
|
||||||
|
|
||||||
type RawSockaddrAny C.struct_sockaddr_any
|
|
||||||
|
|
||||||
type _Socklen C.socklen_t
|
|
||||||
|
|
||||||
type Linger C.struct_linger
|
|
||||||
|
|
||||||
type Iovec C.struct_iovec
|
|
||||||
|
|
||||||
type IPMreq C.struct_ip_mreq
|
|
||||||
|
|
||||||
type IPv6Mreq C.struct_ipv6_mreq
|
|
||||||
|
|
||||||
type Msghdr C.struct_msghdr
|
|
||||||
|
|
||||||
type Cmsghdr C.struct_cmsghdr
|
|
||||||
|
|
||||||
type Inet4Pktinfo C.struct_in_pktinfo
|
|
||||||
|
|
||||||
type Inet6Pktinfo C.struct_in6_pktinfo
|
|
||||||
|
|
||||||
type IPv6MTUInfo C.struct_ip6_mtuinfo
|
|
||||||
|
|
||||||
type ICMPv6Filter C.struct_icmp6_filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofSockaddrInet4 = C.sizeof_struct_sockaddr_in
|
|
||||||
SizeofSockaddrInet6 = C.sizeof_struct_sockaddr_in6
|
|
||||||
SizeofSockaddrAny = C.sizeof_struct_sockaddr_any
|
|
||||||
SizeofSockaddrUnix = C.sizeof_struct_sockaddr_un
|
|
||||||
SizeofSockaddrDatalink = C.sizeof_struct_sockaddr_dl
|
|
||||||
SizeofLinger = C.sizeof_struct_linger
|
|
||||||
SizeofIPMreq = C.sizeof_struct_ip_mreq
|
|
||||||
SizeofIPv6Mreq = C.sizeof_struct_ipv6_mreq
|
|
||||||
SizeofMsghdr = C.sizeof_struct_msghdr
|
|
||||||
SizeofCmsghdr = C.sizeof_struct_cmsghdr
|
|
||||||
SizeofInet4Pktinfo = C.sizeof_struct_in_pktinfo
|
|
||||||
SizeofInet6Pktinfo = C.sizeof_struct_in6_pktinfo
|
|
||||||
SizeofIPv6MTUInfo = C.sizeof_struct_ip6_mtuinfo
|
|
||||||
SizeofICMPv6Filter = C.sizeof_struct_icmp6_filter
|
|
||||||
)
|
|
||||||
|
|
||||||
// Ptrace requests
|
|
||||||
|
|
||||||
const (
|
|
||||||
PTRACE_TRACEME = C.PT_TRACE_ME
|
|
||||||
PTRACE_CONT = C.PT_CONTINUE
|
|
||||||
PTRACE_KILL = C.PT_KILL
|
|
||||||
)
|
|
||||||
|
|
||||||
// Events (kqueue, kevent)
|
|
||||||
|
|
||||||
type Kevent_t C.struct_kevent
|
|
||||||
|
|
||||||
// Select
|
|
||||||
|
|
||||||
type FdSet C.fd_set
|
|
||||||
|
|
||||||
// Routing and interface messages
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofIfMsghdr = C.sizeof_struct_if_msghdr
|
|
||||||
SizeofIfData = C.sizeof_struct_if_data
|
|
||||||
SizeofIfaMsghdr = C.sizeof_struct_ifa_msghdr
|
|
||||||
SizeofIfmaMsghdr = C.sizeof_struct_ifma_msghdr
|
|
||||||
SizeofIfmaMsghdr2 = C.sizeof_struct_ifma_msghdr2
|
|
||||||
SizeofRtMsghdr = C.sizeof_struct_rt_msghdr
|
|
||||||
SizeofRtMetrics = C.sizeof_struct_rt_metrics
|
|
||||||
)
|
|
||||||
|
|
||||||
type IfMsghdr C.struct_if_msghdr
|
|
||||||
|
|
||||||
type IfData C.struct_if_data
|
|
||||||
|
|
||||||
type IfaMsghdr C.struct_ifa_msghdr
|
|
||||||
|
|
||||||
type IfmaMsghdr C.struct_ifma_msghdr
|
|
||||||
|
|
||||||
type IfmaMsghdr2 C.struct_ifma_msghdr2
|
|
||||||
|
|
||||||
type RtMsghdr C.struct_rt_msghdr
|
|
||||||
|
|
||||||
type RtMetrics C.struct_rt_metrics
|
|
||||||
|
|
||||||
// Berkeley packet filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofBpfVersion = C.sizeof_struct_bpf_version
|
|
||||||
SizeofBpfStat = C.sizeof_struct_bpf_stat
|
|
||||||
SizeofBpfProgram = C.sizeof_struct_bpf_program
|
|
||||||
SizeofBpfInsn = C.sizeof_struct_bpf_insn
|
|
||||||
SizeofBpfHdr = C.sizeof_struct_bpf_hdr
|
|
||||||
)
|
|
||||||
|
|
||||||
type BpfVersion C.struct_bpf_version
|
|
||||||
|
|
||||||
type BpfStat C.struct_bpf_stat
|
|
||||||
|
|
||||||
type BpfProgram C.struct_bpf_program
|
|
||||||
|
|
||||||
type BpfInsn C.struct_bpf_insn
|
|
||||||
|
|
||||||
type BpfHdr C.struct_bpf_hdr
|
|
||||||
|
|
||||||
// Terminal handling
|
|
||||||
|
|
||||||
type Termios C.struct_termios
|
|
||||||
|
|
||||||
// fchmodat-like syscalls.
|
|
||||||
|
|
||||||
const (
|
|
||||||
AT_FDCWD = C.AT_FDCWD
|
|
||||||
AT_SYMLINK_NOFOLLOW = C.AT_SYMLINK_NOFOLLOW
|
|
||||||
)
|
|
|
@ -1,242 +0,0 @@
|
||||||
// Copyright 2009 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
/*
|
|
||||||
Input to cgo -godefs. See also mkerrors.sh and mkall.sh
|
|
||||||
*/
|
|
||||||
|
|
||||||
// +godefs map struct_in_addr [4]byte /* in_addr */
|
|
||||||
// +godefs map struct_in6_addr [16]byte /* in6_addr */
|
|
||||||
|
|
||||||
package unix
|
|
||||||
|
|
||||||
/*
|
|
||||||
#define KERNEL
|
|
||||||
#include <dirent.h>
|
|
||||||
#include <fcntl.h>
|
|
||||||
#include <signal.h>
|
|
||||||
#include <termios.h>
|
|
||||||
#include <stdio.h>
|
|
||||||
#include <unistd.h>
|
|
||||||
#include <sys/event.h>
|
|
||||||
#include <sys/mman.h>
|
|
||||||
#include <sys/mount.h>
|
|
||||||
#include <sys/param.h>
|
|
||||||
#include <sys/ptrace.h>
|
|
||||||
#include <sys/resource.h>
|
|
||||||
#include <sys/select.h>
|
|
||||||
#include <sys/signal.h>
|
|
||||||
#include <sys/socket.h>
|
|
||||||
#include <sys/stat.h>
|
|
||||||
#include <sys/time.h>
|
|
||||||
#include <sys/types.h>
|
|
||||||
#include <sys/un.h>
|
|
||||||
#include <sys/wait.h>
|
|
||||||
#include <net/bpf.h>
|
|
||||||
#include <net/if.h>
|
|
||||||
#include <net/if_dl.h>
|
|
||||||
#include <net/route.h>
|
|
||||||
#include <netinet/in.h>
|
|
||||||
#include <netinet/icmp6.h>
|
|
||||||
#include <netinet/tcp.h>
|
|
||||||
|
|
||||||
enum {
|
|
||||||
sizeofPtr = sizeof(void*),
|
|
||||||
};
|
|
||||||
|
|
||||||
union sockaddr_all {
|
|
||||||
struct sockaddr s1; // this one gets used for fields
|
|
||||||
struct sockaddr_in s2; // these pad it out
|
|
||||||
struct sockaddr_in6 s3;
|
|
||||||
struct sockaddr_un s4;
|
|
||||||
struct sockaddr_dl s5;
|
|
||||||
};
|
|
||||||
|
|
||||||
struct sockaddr_any {
|
|
||||||
struct sockaddr addr;
|
|
||||||
char pad[sizeof(union sockaddr_all) - sizeof(struct sockaddr)];
|
|
||||||
};
|
|
||||||
|
|
||||||
*/
|
|
||||||
import "C"
|
|
||||||
|
|
||||||
// Machine characteristics; for internal use.
|
|
||||||
|
|
||||||
const (
|
|
||||||
sizeofPtr = C.sizeofPtr
|
|
||||||
sizeofShort = C.sizeof_short
|
|
||||||
sizeofInt = C.sizeof_int
|
|
||||||
sizeofLong = C.sizeof_long
|
|
||||||
sizeofLongLong = C.sizeof_longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Basic types
|
|
||||||
|
|
||||||
type (
|
|
||||||
_C_short C.short
|
|
||||||
_C_int C.int
|
|
||||||
_C_long C.long
|
|
||||||
_C_long_long C.longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Time
|
|
||||||
|
|
||||||
type Timespec C.struct_timespec
|
|
||||||
|
|
||||||
type Timeval C.struct_timeval
|
|
||||||
|
|
||||||
// Processes
|
|
||||||
|
|
||||||
type Rusage C.struct_rusage
|
|
||||||
|
|
||||||
type Rlimit C.struct_rlimit
|
|
||||||
|
|
||||||
type _Gid_t C.gid_t
|
|
||||||
|
|
||||||
// Files
|
|
||||||
|
|
||||||
const ( // Directory mode bits
|
|
||||||
S_IFMT = C.S_IFMT
|
|
||||||
S_IFIFO = C.S_IFIFO
|
|
||||||
S_IFCHR = C.S_IFCHR
|
|
||||||
S_IFDIR = C.S_IFDIR
|
|
||||||
S_IFBLK = C.S_IFBLK
|
|
||||||
S_IFREG = C.S_IFREG
|
|
||||||
S_IFLNK = C.S_IFLNK
|
|
||||||
S_IFSOCK = C.S_IFSOCK
|
|
||||||
S_ISUID = C.S_ISUID
|
|
||||||
S_ISGID = C.S_ISGID
|
|
||||||
S_ISVTX = C.S_ISVTX
|
|
||||||
S_IRUSR = C.S_IRUSR
|
|
||||||
S_IWUSR = C.S_IWUSR
|
|
||||||
S_IXUSR = C.S_IXUSR
|
|
||||||
)
|
|
||||||
|
|
||||||
type Stat_t C.struct_stat
|
|
||||||
|
|
||||||
type Statfs_t C.struct_statfs
|
|
||||||
|
|
||||||
type Flock_t C.struct_flock
|
|
||||||
|
|
||||||
type Dirent C.struct_dirent
|
|
||||||
|
|
||||||
type Fsid C.struct_fsid
|
|
||||||
|
|
||||||
// Sockets
|
|
||||||
|
|
||||||
type RawSockaddrInet4 C.struct_sockaddr_in
|
|
||||||
|
|
||||||
type RawSockaddrInet6 C.struct_sockaddr_in6
|
|
||||||
|
|
||||||
type RawSockaddrUnix C.struct_sockaddr_un
|
|
||||||
|
|
||||||
type RawSockaddrDatalink C.struct_sockaddr_dl
|
|
||||||
|
|
||||||
type RawSockaddr C.struct_sockaddr
|
|
||||||
|
|
||||||
type RawSockaddrAny C.struct_sockaddr_any
|
|
||||||
|
|
||||||
type _Socklen C.socklen_t
|
|
||||||
|
|
||||||
type Linger C.struct_linger
|
|
||||||
|
|
||||||
type Iovec C.struct_iovec
|
|
||||||
|
|
||||||
type IPMreq C.struct_ip_mreq
|
|
||||||
|
|
||||||
type IPv6Mreq C.struct_ipv6_mreq
|
|
||||||
|
|
||||||
type Msghdr C.struct_msghdr
|
|
||||||
|
|
||||||
type Cmsghdr C.struct_cmsghdr
|
|
||||||
|
|
||||||
type Inet6Pktinfo C.struct_in6_pktinfo
|
|
||||||
|
|
||||||
type IPv6MTUInfo C.struct_ip6_mtuinfo
|
|
||||||
|
|
||||||
type ICMPv6Filter C.struct_icmp6_filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofSockaddrInet4 = C.sizeof_struct_sockaddr_in
|
|
||||||
SizeofSockaddrInet6 = C.sizeof_struct_sockaddr_in6
|
|
||||||
SizeofSockaddrAny = C.sizeof_struct_sockaddr_any
|
|
||||||
SizeofSockaddrUnix = C.sizeof_struct_sockaddr_un
|
|
||||||
SizeofSockaddrDatalink = C.sizeof_struct_sockaddr_dl
|
|
||||||
SizeofLinger = C.sizeof_struct_linger
|
|
||||||
SizeofIPMreq = C.sizeof_struct_ip_mreq
|
|
||||||
SizeofIPv6Mreq = C.sizeof_struct_ipv6_mreq
|
|
||||||
SizeofMsghdr = C.sizeof_struct_msghdr
|
|
||||||
SizeofCmsghdr = C.sizeof_struct_cmsghdr
|
|
||||||
SizeofInet6Pktinfo = C.sizeof_struct_in6_pktinfo
|
|
||||||
SizeofIPv6MTUInfo = C.sizeof_struct_ip6_mtuinfo
|
|
||||||
SizeofICMPv6Filter = C.sizeof_struct_icmp6_filter
|
|
||||||
)
|
|
||||||
|
|
||||||
// Ptrace requests
|
|
||||||
|
|
||||||
const (
|
|
||||||
PTRACE_TRACEME = C.PT_TRACE_ME
|
|
||||||
PTRACE_CONT = C.PT_CONTINUE
|
|
||||||
PTRACE_KILL = C.PT_KILL
|
|
||||||
)
|
|
||||||
|
|
||||||
// Events (kqueue, kevent)
|
|
||||||
|
|
||||||
type Kevent_t C.struct_kevent
|
|
||||||
|
|
||||||
// Select
|
|
||||||
|
|
||||||
type FdSet C.fd_set
|
|
||||||
|
|
||||||
// Routing and interface messages
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofIfMsghdr = C.sizeof_struct_if_msghdr
|
|
||||||
SizeofIfData = C.sizeof_struct_if_data
|
|
||||||
SizeofIfaMsghdr = C.sizeof_struct_ifa_msghdr
|
|
||||||
SizeofIfmaMsghdr = C.sizeof_struct_ifma_msghdr
|
|
||||||
SizeofIfAnnounceMsghdr = C.sizeof_struct_if_announcemsghdr
|
|
||||||
SizeofRtMsghdr = C.sizeof_struct_rt_msghdr
|
|
||||||
SizeofRtMetrics = C.sizeof_struct_rt_metrics
|
|
||||||
)
|
|
||||||
|
|
||||||
type IfMsghdr C.struct_if_msghdr
|
|
||||||
|
|
||||||
type IfData C.struct_if_data
|
|
||||||
|
|
||||||
type IfaMsghdr C.struct_ifa_msghdr
|
|
||||||
|
|
||||||
type IfmaMsghdr C.struct_ifma_msghdr
|
|
||||||
|
|
||||||
type IfAnnounceMsghdr C.struct_if_announcemsghdr
|
|
||||||
|
|
||||||
type RtMsghdr C.struct_rt_msghdr
|
|
||||||
|
|
||||||
type RtMetrics C.struct_rt_metrics
|
|
||||||
|
|
||||||
// Berkeley packet filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofBpfVersion = C.sizeof_struct_bpf_version
|
|
||||||
SizeofBpfStat = C.sizeof_struct_bpf_stat
|
|
||||||
SizeofBpfProgram = C.sizeof_struct_bpf_program
|
|
||||||
SizeofBpfInsn = C.sizeof_struct_bpf_insn
|
|
||||||
SizeofBpfHdr = C.sizeof_struct_bpf_hdr
|
|
||||||
)
|
|
||||||
|
|
||||||
type BpfVersion C.struct_bpf_version
|
|
||||||
|
|
||||||
type BpfStat C.struct_bpf_stat
|
|
||||||
|
|
||||||
type BpfProgram C.struct_bpf_program
|
|
||||||
|
|
||||||
type BpfInsn C.struct_bpf_insn
|
|
||||||
|
|
||||||
type BpfHdr C.struct_bpf_hdr
|
|
||||||
|
|
||||||
// Terminal handling
|
|
||||||
|
|
||||||
type Termios C.struct_termios
|
|
|
@ -1,353 +0,0 @@
|
||||||
// Copyright 2009 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
/*
|
|
||||||
Input to cgo -godefs. See also mkerrors.sh and mkall.sh
|
|
||||||
*/
|
|
||||||
|
|
||||||
// +godefs map struct_in_addr [4]byte /* in_addr */
|
|
||||||
// +godefs map struct_in6_addr [16]byte /* in6_addr */
|
|
||||||
|
|
||||||
package unix
|
|
||||||
|
|
||||||
/*
|
|
||||||
#define KERNEL
|
|
||||||
#include <dirent.h>
|
|
||||||
#include <fcntl.h>
|
|
||||||
#include <signal.h>
|
|
||||||
#include <termios.h>
|
|
||||||
#include <stdio.h>
|
|
||||||
#include <unistd.h>
|
|
||||||
#include <sys/event.h>
|
|
||||||
#include <sys/mman.h>
|
|
||||||
#include <sys/mount.h>
|
|
||||||
#include <sys/param.h>
|
|
||||||
#include <sys/ptrace.h>
|
|
||||||
#include <sys/resource.h>
|
|
||||||
#include <sys/select.h>
|
|
||||||
#include <sys/signal.h>
|
|
||||||
#include <sys/socket.h>
|
|
||||||
#include <sys/stat.h>
|
|
||||||
#include <sys/time.h>
|
|
||||||
#include <sys/types.h>
|
|
||||||
#include <sys/un.h>
|
|
||||||
#include <sys/wait.h>
|
|
||||||
#include <net/bpf.h>
|
|
||||||
#include <net/if.h>
|
|
||||||
#include <net/if_dl.h>
|
|
||||||
#include <net/route.h>
|
|
||||||
#include <netinet/in.h>
|
|
||||||
#include <netinet/icmp6.h>
|
|
||||||
#include <netinet/tcp.h>
|
|
||||||
|
|
||||||
enum {
|
|
||||||
sizeofPtr = sizeof(void*),
|
|
||||||
};
|
|
||||||
|
|
||||||
union sockaddr_all {
|
|
||||||
struct sockaddr s1; // this one gets used for fields
|
|
||||||
struct sockaddr_in s2; // these pad it out
|
|
||||||
struct sockaddr_in6 s3;
|
|
||||||
struct sockaddr_un s4;
|
|
||||||
struct sockaddr_dl s5;
|
|
||||||
};
|
|
||||||
|
|
||||||
struct sockaddr_any {
|
|
||||||
struct sockaddr addr;
|
|
||||||
char pad[sizeof(union sockaddr_all) - sizeof(struct sockaddr)];
|
|
||||||
};
|
|
||||||
|
|
||||||
// This structure is a duplicate of stat on FreeBSD 8-STABLE.
|
|
||||||
// See /usr/include/sys/stat.h.
|
|
||||||
struct stat8 {
|
|
||||||
#undef st_atimespec st_atim
|
|
||||||
#undef st_mtimespec st_mtim
|
|
||||||
#undef st_ctimespec st_ctim
|
|
||||||
#undef st_birthtimespec st_birthtim
|
|
||||||
__dev_t st_dev;
|
|
||||||
ino_t st_ino;
|
|
||||||
mode_t st_mode;
|
|
||||||
nlink_t st_nlink;
|
|
||||||
uid_t st_uid;
|
|
||||||
gid_t st_gid;
|
|
||||||
__dev_t st_rdev;
|
|
||||||
#if __BSD_VISIBLE
|
|
||||||
struct timespec st_atimespec;
|
|
||||||
struct timespec st_mtimespec;
|
|
||||||
struct timespec st_ctimespec;
|
|
||||||
#else
|
|
||||||
time_t st_atime;
|
|
||||||
long __st_atimensec;
|
|
||||||
time_t st_mtime;
|
|
||||||
long __st_mtimensec;
|
|
||||||
time_t st_ctime;
|
|
||||||
long __st_ctimensec;
|
|
||||||
#endif
|
|
||||||
off_t st_size;
|
|
||||||
blkcnt_t st_blocks;
|
|
||||||
blksize_t st_blksize;
|
|
||||||
fflags_t st_flags;
|
|
||||||
__uint32_t st_gen;
|
|
||||||
__int32_t st_lspare;
|
|
||||||
#if __BSD_VISIBLE
|
|
||||||
struct timespec st_birthtimespec;
|
|
||||||
unsigned int :(8 / 2) * (16 - (int)sizeof(struct timespec));
|
|
||||||
unsigned int :(8 / 2) * (16 - (int)sizeof(struct timespec));
|
|
||||||
#else
|
|
||||||
time_t st_birthtime;
|
|
||||||
long st_birthtimensec;
|
|
||||||
unsigned int :(8 / 2) * (16 - (int)sizeof(struct __timespec));
|
|
||||||
unsigned int :(8 / 2) * (16 - (int)sizeof(struct __timespec));
|
|
||||||
#endif
|
|
||||||
};
|
|
||||||
|
|
||||||
// This structure is a duplicate of if_data on FreeBSD 8-STABLE.
|
|
||||||
// See /usr/include/net/if.h.
|
|
||||||
struct if_data8 {
|
|
||||||
u_char ifi_type;
|
|
||||||
u_char ifi_physical;
|
|
||||||
u_char ifi_addrlen;
|
|
||||||
u_char ifi_hdrlen;
|
|
||||||
u_char ifi_link_state;
|
|
||||||
u_char ifi_spare_char1;
|
|
||||||
u_char ifi_spare_char2;
|
|
||||||
u_char ifi_datalen;
|
|
||||||
u_long ifi_mtu;
|
|
||||||
u_long ifi_metric;
|
|
||||||
u_long ifi_baudrate;
|
|
||||||
u_long ifi_ipackets;
|
|
||||||
u_long ifi_ierrors;
|
|
||||||
u_long ifi_opackets;
|
|
||||||
u_long ifi_oerrors;
|
|
||||||
u_long ifi_collisions;
|
|
||||||
u_long ifi_ibytes;
|
|
||||||
u_long ifi_obytes;
|
|
||||||
u_long ifi_imcasts;
|
|
||||||
u_long ifi_omcasts;
|
|
||||||
u_long ifi_iqdrops;
|
|
||||||
u_long ifi_noproto;
|
|
||||||
u_long ifi_hwassist;
|
|
||||||
time_t ifi_epoch;
|
|
||||||
struct timeval ifi_lastchange;
|
|
||||||
};
|
|
||||||
|
|
||||||
// This structure is a duplicate of if_msghdr on FreeBSD 8-STABLE.
|
|
||||||
// See /usr/include/net/if.h.
|
|
||||||
struct if_msghdr8 {
|
|
||||||
u_short ifm_msglen;
|
|
||||||
u_char ifm_version;
|
|
||||||
u_char ifm_type;
|
|
||||||
int ifm_addrs;
|
|
||||||
int ifm_flags;
|
|
||||||
u_short ifm_index;
|
|
||||||
struct if_data8 ifm_data;
|
|
||||||
};
|
|
||||||
*/
|
|
||||||
import "C"
|
|
||||||
|
|
||||||
// Machine characteristics; for internal use.
|
|
||||||
|
|
||||||
const (
|
|
||||||
sizeofPtr = C.sizeofPtr
|
|
||||||
sizeofShort = C.sizeof_short
|
|
||||||
sizeofInt = C.sizeof_int
|
|
||||||
sizeofLong = C.sizeof_long
|
|
||||||
sizeofLongLong = C.sizeof_longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Basic types
|
|
||||||
|
|
||||||
type (
|
|
||||||
_C_short C.short
|
|
||||||
_C_int C.int
|
|
||||||
_C_long C.long
|
|
||||||
_C_long_long C.longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Time
|
|
||||||
|
|
||||||
type Timespec C.struct_timespec
|
|
||||||
|
|
||||||
type Timeval C.struct_timeval
|
|
||||||
|
|
||||||
// Processes
|
|
||||||
|
|
||||||
type Rusage C.struct_rusage
|
|
||||||
|
|
||||||
type Rlimit C.struct_rlimit
|
|
||||||
|
|
||||||
type _Gid_t C.gid_t
|
|
||||||
|
|
||||||
// Files
|
|
||||||
|
|
||||||
const ( // Directory mode bits
|
|
||||||
S_IFMT = C.S_IFMT
|
|
||||||
S_IFIFO = C.S_IFIFO
|
|
||||||
S_IFCHR = C.S_IFCHR
|
|
||||||
S_IFDIR = C.S_IFDIR
|
|
||||||
S_IFBLK = C.S_IFBLK
|
|
||||||
S_IFREG = C.S_IFREG
|
|
||||||
S_IFLNK = C.S_IFLNK
|
|
||||||
S_IFSOCK = C.S_IFSOCK
|
|
||||||
S_ISUID = C.S_ISUID
|
|
||||||
S_ISGID = C.S_ISGID
|
|
||||||
S_ISVTX = C.S_ISVTX
|
|
||||||
S_IRUSR = C.S_IRUSR
|
|
||||||
S_IWUSR = C.S_IWUSR
|
|
||||||
S_IXUSR = C.S_IXUSR
|
|
||||||
)
|
|
||||||
|
|
||||||
type Stat_t C.struct_stat8
|
|
||||||
|
|
||||||
type Statfs_t C.struct_statfs
|
|
||||||
|
|
||||||
type Flock_t C.struct_flock
|
|
||||||
|
|
||||||
type Dirent C.struct_dirent
|
|
||||||
|
|
||||||
type Fsid C.struct_fsid
|
|
||||||
|
|
||||||
// Advice to Fadvise
|
|
||||||
|
|
||||||
const (
|
|
||||||
FADV_NORMAL = C.POSIX_FADV_NORMAL
|
|
||||||
FADV_RANDOM = C.POSIX_FADV_RANDOM
|
|
||||||
FADV_SEQUENTIAL = C.POSIX_FADV_SEQUENTIAL
|
|
||||||
FADV_WILLNEED = C.POSIX_FADV_WILLNEED
|
|
||||||
FADV_DONTNEED = C.POSIX_FADV_DONTNEED
|
|
||||||
FADV_NOREUSE = C.POSIX_FADV_NOREUSE
|
|
||||||
)
|
|
||||||
|
|
||||||
// Sockets
|
|
||||||
|
|
||||||
type RawSockaddrInet4 C.struct_sockaddr_in
|
|
||||||
|
|
||||||
type RawSockaddrInet6 C.struct_sockaddr_in6
|
|
||||||
|
|
||||||
type RawSockaddrUnix C.struct_sockaddr_un
|
|
||||||
|
|
||||||
type RawSockaddrDatalink C.struct_sockaddr_dl
|
|
||||||
|
|
||||||
type RawSockaddr C.struct_sockaddr
|
|
||||||
|
|
||||||
type RawSockaddrAny C.struct_sockaddr_any
|
|
||||||
|
|
||||||
type _Socklen C.socklen_t
|
|
||||||
|
|
||||||
type Linger C.struct_linger
|
|
||||||
|
|
||||||
type Iovec C.struct_iovec
|
|
||||||
|
|
||||||
type IPMreq C.struct_ip_mreq
|
|
||||||
|
|
||||||
type IPMreqn C.struct_ip_mreqn
|
|
||||||
|
|
||||||
type IPv6Mreq C.struct_ipv6_mreq
|
|
||||||
|
|
||||||
type Msghdr C.struct_msghdr
|
|
||||||
|
|
||||||
type Cmsghdr C.struct_cmsghdr
|
|
||||||
|
|
||||||
type Inet6Pktinfo C.struct_in6_pktinfo
|
|
||||||
|
|
||||||
type IPv6MTUInfo C.struct_ip6_mtuinfo
|
|
||||||
|
|
||||||
type ICMPv6Filter C.struct_icmp6_filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofSockaddrInet4 = C.sizeof_struct_sockaddr_in
|
|
||||||
SizeofSockaddrInet6 = C.sizeof_struct_sockaddr_in6
|
|
||||||
SizeofSockaddrAny = C.sizeof_struct_sockaddr_any
|
|
||||||
SizeofSockaddrUnix = C.sizeof_struct_sockaddr_un
|
|
||||||
SizeofSockaddrDatalink = C.sizeof_struct_sockaddr_dl
|
|
||||||
SizeofLinger = C.sizeof_struct_linger
|
|
||||||
SizeofIPMreq = C.sizeof_struct_ip_mreq
|
|
||||||
SizeofIPMreqn = C.sizeof_struct_ip_mreqn
|
|
||||||
SizeofIPv6Mreq = C.sizeof_struct_ipv6_mreq
|
|
||||||
SizeofMsghdr = C.sizeof_struct_msghdr
|
|
||||||
SizeofCmsghdr = C.sizeof_struct_cmsghdr
|
|
||||||
SizeofInet6Pktinfo = C.sizeof_struct_in6_pktinfo
|
|
||||||
SizeofIPv6MTUInfo = C.sizeof_struct_ip6_mtuinfo
|
|
||||||
SizeofICMPv6Filter = C.sizeof_struct_icmp6_filter
|
|
||||||
)
|
|
||||||
|
|
||||||
// Ptrace requests
|
|
||||||
|
|
||||||
const (
|
|
||||||
PTRACE_TRACEME = C.PT_TRACE_ME
|
|
||||||
PTRACE_CONT = C.PT_CONTINUE
|
|
||||||
PTRACE_KILL = C.PT_KILL
|
|
||||||
)
|
|
||||||
|
|
||||||
// Events (kqueue, kevent)
|
|
||||||
|
|
||||||
type Kevent_t C.struct_kevent
|
|
||||||
|
|
||||||
// Select
|
|
||||||
|
|
||||||
type FdSet C.fd_set
|
|
||||||
|
|
||||||
// Routing and interface messages
|
|
||||||
|
|
||||||
const (
|
|
||||||
sizeofIfMsghdr = C.sizeof_struct_if_msghdr
|
|
||||||
SizeofIfMsghdr = C.sizeof_struct_if_msghdr8
|
|
||||||
sizeofIfData = C.sizeof_struct_if_data
|
|
||||||
SizeofIfData = C.sizeof_struct_if_data8
|
|
||||||
SizeofIfaMsghdr = C.sizeof_struct_ifa_msghdr
|
|
||||||
SizeofIfmaMsghdr = C.sizeof_struct_ifma_msghdr
|
|
||||||
SizeofIfAnnounceMsghdr = C.sizeof_struct_if_announcemsghdr
|
|
||||||
SizeofRtMsghdr = C.sizeof_struct_rt_msghdr
|
|
||||||
SizeofRtMetrics = C.sizeof_struct_rt_metrics
|
|
||||||
)
|
|
||||||
|
|
||||||
type ifMsghdr C.struct_if_msghdr
|
|
||||||
|
|
||||||
type IfMsghdr C.struct_if_msghdr8
|
|
||||||
|
|
||||||
type ifData C.struct_if_data
|
|
||||||
|
|
||||||
type IfData C.struct_if_data8
|
|
||||||
|
|
||||||
type IfaMsghdr C.struct_ifa_msghdr
|
|
||||||
|
|
||||||
type IfmaMsghdr C.struct_ifma_msghdr
|
|
||||||
|
|
||||||
type IfAnnounceMsghdr C.struct_if_announcemsghdr
|
|
||||||
|
|
||||||
type RtMsghdr C.struct_rt_msghdr
|
|
||||||
|
|
||||||
type RtMetrics C.struct_rt_metrics
|
|
||||||
|
|
||||||
// Berkeley packet filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofBpfVersion = C.sizeof_struct_bpf_version
|
|
||||||
SizeofBpfStat = C.sizeof_struct_bpf_stat
|
|
||||||
SizeofBpfZbuf = C.sizeof_struct_bpf_zbuf
|
|
||||||
SizeofBpfProgram = C.sizeof_struct_bpf_program
|
|
||||||
SizeofBpfInsn = C.sizeof_struct_bpf_insn
|
|
||||||
SizeofBpfHdr = C.sizeof_struct_bpf_hdr
|
|
||||||
SizeofBpfZbufHeader = C.sizeof_struct_bpf_zbuf_header
|
|
||||||
)
|
|
||||||
|
|
||||||
type BpfVersion C.struct_bpf_version
|
|
||||||
|
|
||||||
type BpfStat C.struct_bpf_stat
|
|
||||||
|
|
||||||
type BpfZbuf C.struct_bpf_zbuf
|
|
||||||
|
|
||||||
type BpfProgram C.struct_bpf_program
|
|
||||||
|
|
||||||
type BpfInsn C.struct_bpf_insn
|
|
||||||
|
|
||||||
type BpfHdr C.struct_bpf_hdr
|
|
||||||
|
|
||||||
type BpfZbufHeader C.struct_bpf_zbuf_header
|
|
||||||
|
|
||||||
// Terminal handling
|
|
||||||
|
|
||||||
type Termios C.struct_termios
|
|
|
@ -1,450 +0,0 @@
|
||||||
// Copyright 2009 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
/*
|
|
||||||
Input to cgo -godefs. See also mkerrors.sh and mkall.sh
|
|
||||||
*/
|
|
||||||
|
|
||||||
// +godefs map struct_in_addr [4]byte /* in_addr */
|
|
||||||
// +godefs map struct_in6_addr [16]byte /* in6_addr */
|
|
||||||
|
|
||||||
package unix
|
|
||||||
|
|
||||||
/*
|
|
||||||
#define _LARGEFILE_SOURCE
|
|
||||||
#define _LARGEFILE64_SOURCE
|
|
||||||
#define _FILE_OFFSET_BITS 64
|
|
||||||
#define _GNU_SOURCE
|
|
||||||
|
|
||||||
#include <dirent.h>
|
|
||||||
#include <fcntl.h>
|
|
||||||
#include <netinet/in.h>
|
|
||||||
#include <netinet/tcp.h>
|
|
||||||
#include <netpacket/packet.h>
|
|
||||||
#include <poll.h>
|
|
||||||
#include <signal.h>
|
|
||||||
#include <stdio.h>
|
|
||||||
#include <sys/epoll.h>
|
|
||||||
#include <sys/inotify.h>
|
|
||||||
#include <sys/mman.h>
|
|
||||||
#include <sys/mount.h>
|
|
||||||
#include <sys/param.h>
|
|
||||||
#include <sys/ptrace.h>
|
|
||||||
#include <sys/resource.h>
|
|
||||||
#include <sys/select.h>
|
|
||||||
#include <sys/signal.h>
|
|
||||||
#include <sys/stat.h>
|
|
||||||
#include <sys/statfs.h>
|
|
||||||
#include <sys/sysinfo.h>
|
|
||||||
#include <sys/time.h>
|
|
||||||
#include <sys/times.h>
|
|
||||||
#include <sys/timex.h>
|
|
||||||
#include <sys/types.h>
|
|
||||||
#include <sys/un.h>
|
|
||||||
#include <sys/user.h>
|
|
||||||
#include <sys/utsname.h>
|
|
||||||
#include <sys/wait.h>
|
|
||||||
#include <linux/filter.h>
|
|
||||||
#include <linux/netlink.h>
|
|
||||||
#include <linux/rtnetlink.h>
|
|
||||||
#include <linux/icmpv6.h>
|
|
||||||
#include <asm/termbits.h>
|
|
||||||
#include <time.h>
|
|
||||||
#include <unistd.h>
|
|
||||||
#include <ustat.h>
|
|
||||||
#include <utime.h>
|
|
||||||
#include <bluetooth/bluetooth.h>
|
|
||||||
#include <bluetooth/hci.h>
|
|
||||||
|
|
||||||
#ifdef TCSETS2
|
|
||||||
// On systems that have "struct termios2" use this as type Termios.
|
|
||||||
typedef struct termios2 termios_t;
|
|
||||||
#else
|
|
||||||
typedef struct termios termios_t;
|
|
||||||
#endif
|
|
||||||
|
|
||||||
enum {
|
|
||||||
sizeofPtr = sizeof(void*),
|
|
||||||
};
|
|
||||||
|
|
||||||
union sockaddr_all {
|
|
||||||
struct sockaddr s1; // this one gets used for fields
|
|
||||||
struct sockaddr_in s2; // these pad it out
|
|
||||||
struct sockaddr_in6 s3;
|
|
||||||
struct sockaddr_un s4;
|
|
||||||
struct sockaddr_ll s5;
|
|
||||||
struct sockaddr_nl s6;
|
|
||||||
};
|
|
||||||
|
|
||||||
struct sockaddr_any {
|
|
||||||
struct sockaddr addr;
|
|
||||||
char pad[sizeof(union sockaddr_all) - sizeof(struct sockaddr)];
|
|
||||||
};
|
|
||||||
|
|
||||||
// copied from /usr/include/linux/un.h
|
|
||||||
struct my_sockaddr_un {
|
|
||||||
sa_family_t sun_family;
|
|
||||||
#if defined(__ARM_EABI__) || defined(__powerpc64__)
|
|
||||||
// on ARM char is by default unsigned
|
|
||||||
signed char sun_path[108];
|
|
||||||
#else
|
|
||||||
char sun_path[108];
|
|
||||||
#endif
|
|
||||||
};
|
|
||||||
|
|
||||||
#ifdef __ARM_EABI__
|
|
||||||
typedef struct user_regs PtraceRegs;
|
|
||||||
#elif defined(__aarch64__)
|
|
||||||
typedef struct user_pt_regs PtraceRegs;
|
|
||||||
#elif defined(__powerpc64__)
|
|
||||||
typedef struct pt_regs PtraceRegs;
|
|
||||||
#elif defined(__mips__)
|
|
||||||
typedef struct user PtraceRegs;
|
|
||||||
#elif defined(__s390x__)
|
|
||||||
typedef struct _user_regs_struct PtraceRegs;
|
|
||||||
#else
|
|
||||||
typedef struct user_regs_struct PtraceRegs;
|
|
||||||
#endif
|
|
||||||
|
|
||||||
#if defined(__s390x__)
|
|
||||||
typedef struct _user_psw_struct ptracePsw;
|
|
||||||
typedef struct _user_fpregs_struct ptraceFpregs;
|
|
||||||
typedef struct _user_per_struct ptracePer;
|
|
||||||
#else
|
|
||||||
typedef struct {} ptracePsw;
|
|
||||||
typedef struct {} ptraceFpregs;
|
|
||||||
typedef struct {} ptracePer;
|
|
||||||
#endif
|
|
||||||
|
|
||||||
// The real epoll_event is a union, and godefs doesn't handle it well.
|
|
||||||
struct my_epoll_event {
|
|
||||||
uint32_t events;
|
|
||||||
#if defined(__ARM_EABI__) || defined(__aarch64__)
|
|
||||||
// padding is not specified in linux/eventpoll.h but added to conform to the
|
|
||||||
// alignment requirements of EABI
|
|
||||||
int32_t padFd;
|
|
||||||
#elif defined(__powerpc64__) || defined(__s390x__)
|
|
||||||
int32_t _padFd;
|
|
||||||
#endif
|
|
||||||
int32_t fd;
|
|
||||||
int32_t pad;
|
|
||||||
};
|
|
||||||
|
|
||||||
*/
|
|
||||||
import "C"
|
|
||||||
|
|
||||||
// Machine characteristics; for internal use.
|
|
||||||
|
|
||||||
const (
|
|
||||||
sizeofPtr = C.sizeofPtr
|
|
||||||
sizeofShort = C.sizeof_short
|
|
||||||
sizeofInt = C.sizeof_int
|
|
||||||
sizeofLong = C.sizeof_long
|
|
||||||
sizeofLongLong = C.sizeof_longlong
|
|
||||||
PathMax = C.PATH_MAX
|
|
||||||
)
|
|
||||||
|
|
||||||
// Basic types
|
|
||||||
|
|
||||||
type (
|
|
||||||
_C_short C.short
|
|
||||||
_C_int C.int
|
|
||||||
_C_long C.long
|
|
||||||
_C_long_long C.longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Time
|
|
||||||
|
|
||||||
type Timespec C.struct_timespec
|
|
||||||
|
|
||||||
type Timeval C.struct_timeval
|
|
||||||
|
|
||||||
type Timex C.struct_timex
|
|
||||||
|
|
||||||
type Time_t C.time_t
|
|
||||||
|
|
||||||
type Tms C.struct_tms
|
|
||||||
|
|
||||||
type Utimbuf C.struct_utimbuf
|
|
||||||
|
|
||||||
// Processes
|
|
||||||
|
|
||||||
type Rusage C.struct_rusage
|
|
||||||
|
|
||||||
type Rlimit C.struct_rlimit
|
|
||||||
|
|
||||||
type _Gid_t C.gid_t
|
|
||||||
|
|
||||||
// Files
|
|
||||||
|
|
||||||
type Stat_t C.struct_stat
|
|
||||||
|
|
||||||
type Statfs_t C.struct_statfs
|
|
||||||
|
|
||||||
type Dirent C.struct_dirent
|
|
||||||
|
|
||||||
type Fsid C.fsid_t
|
|
||||||
|
|
||||||
type Flock_t C.struct_flock
|
|
||||||
|
|
||||||
// Advice to Fadvise
|
|
||||||
|
|
||||||
const (
|
|
||||||
FADV_NORMAL = C.POSIX_FADV_NORMAL
|
|
||||||
FADV_RANDOM = C.POSIX_FADV_RANDOM
|
|
||||||
FADV_SEQUENTIAL = C.POSIX_FADV_SEQUENTIAL
|
|
||||||
FADV_WILLNEED = C.POSIX_FADV_WILLNEED
|
|
||||||
FADV_DONTNEED = C.POSIX_FADV_DONTNEED
|
|
||||||
FADV_NOREUSE = C.POSIX_FADV_NOREUSE
|
|
||||||
)
|
|
||||||
|
|
||||||
// Sockets
|
|
||||||
|
|
||||||
type RawSockaddrInet4 C.struct_sockaddr_in
|
|
||||||
|
|
||||||
type RawSockaddrInet6 C.struct_sockaddr_in6
|
|
||||||
|
|
||||||
type RawSockaddrUnix C.struct_my_sockaddr_un
|
|
||||||
|
|
||||||
type RawSockaddrLinklayer C.struct_sockaddr_ll
|
|
||||||
|
|
||||||
type RawSockaddrNetlink C.struct_sockaddr_nl
|
|
||||||
|
|
||||||
type RawSockaddrHCI C.struct_sockaddr_hci
|
|
||||||
|
|
||||||
type RawSockaddr C.struct_sockaddr
|
|
||||||
|
|
||||||
type RawSockaddrAny C.struct_sockaddr_any
|
|
||||||
|
|
||||||
type _Socklen C.socklen_t
|
|
||||||
|
|
||||||
type Linger C.struct_linger
|
|
||||||
|
|
||||||
type Iovec C.struct_iovec
|
|
||||||
|
|
||||||
type IPMreq C.struct_ip_mreq
|
|
||||||
|
|
||||||
type IPMreqn C.struct_ip_mreqn
|
|
||||||
|
|
||||||
type IPv6Mreq C.struct_ipv6_mreq
|
|
||||||
|
|
||||||
type Msghdr C.struct_msghdr
|
|
||||||
|
|
||||||
type Cmsghdr C.struct_cmsghdr
|
|
||||||
|
|
||||||
type Inet4Pktinfo C.struct_in_pktinfo
|
|
||||||
|
|
||||||
type Inet6Pktinfo C.struct_in6_pktinfo
|
|
||||||
|
|
||||||
type IPv6MTUInfo C.struct_ip6_mtuinfo
|
|
||||||
|
|
||||||
type ICMPv6Filter C.struct_icmp6_filter
|
|
||||||
|
|
||||||
type Ucred C.struct_ucred
|
|
||||||
|
|
||||||
type TCPInfo C.struct_tcp_info
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofSockaddrInet4 = C.sizeof_struct_sockaddr_in
|
|
||||||
SizeofSockaddrInet6 = C.sizeof_struct_sockaddr_in6
|
|
||||||
SizeofSockaddrAny = C.sizeof_struct_sockaddr_any
|
|
||||||
SizeofSockaddrUnix = C.sizeof_struct_sockaddr_un
|
|
||||||
SizeofSockaddrLinklayer = C.sizeof_struct_sockaddr_ll
|
|
||||||
SizeofSockaddrNetlink = C.sizeof_struct_sockaddr_nl
|
|
||||||
SizeofSockaddrHCI = C.sizeof_struct_sockaddr_hci
|
|
||||||
SizeofLinger = C.sizeof_struct_linger
|
|
||||||
SizeofIPMreq = C.sizeof_struct_ip_mreq
|
|
||||||
SizeofIPMreqn = C.sizeof_struct_ip_mreqn
|
|
||||||
SizeofIPv6Mreq = C.sizeof_struct_ipv6_mreq
|
|
||||||
SizeofMsghdr = C.sizeof_struct_msghdr
|
|
||||||
SizeofCmsghdr = C.sizeof_struct_cmsghdr
|
|
||||||
SizeofInet4Pktinfo = C.sizeof_struct_in_pktinfo
|
|
||||||
SizeofInet6Pktinfo = C.sizeof_struct_in6_pktinfo
|
|
||||||
SizeofIPv6MTUInfo = C.sizeof_struct_ip6_mtuinfo
|
|
||||||
SizeofICMPv6Filter = C.sizeof_struct_icmp6_filter
|
|
||||||
SizeofUcred = C.sizeof_struct_ucred
|
|
||||||
SizeofTCPInfo = C.sizeof_struct_tcp_info
|
|
||||||
)
|
|
||||||
|
|
||||||
// Netlink routing and interface messages
|
|
||||||
|
|
||||||
const (
|
|
||||||
IFA_UNSPEC = C.IFA_UNSPEC
|
|
||||||
IFA_ADDRESS = C.IFA_ADDRESS
|
|
||||||
IFA_LOCAL = C.IFA_LOCAL
|
|
||||||
IFA_LABEL = C.IFA_LABEL
|
|
||||||
IFA_BROADCAST = C.IFA_BROADCAST
|
|
||||||
IFA_ANYCAST = C.IFA_ANYCAST
|
|
||||||
IFA_CACHEINFO = C.IFA_CACHEINFO
|
|
||||||
IFA_MULTICAST = C.IFA_MULTICAST
|
|
||||||
IFLA_UNSPEC = C.IFLA_UNSPEC
|
|
||||||
IFLA_ADDRESS = C.IFLA_ADDRESS
|
|
||||||
IFLA_BROADCAST = C.IFLA_BROADCAST
|
|
||||||
IFLA_IFNAME = C.IFLA_IFNAME
|
|
||||||
IFLA_MTU = C.IFLA_MTU
|
|
||||||
IFLA_LINK = C.IFLA_LINK
|
|
||||||
IFLA_QDISC = C.IFLA_QDISC
|
|
||||||
IFLA_STATS = C.IFLA_STATS
|
|
||||||
IFLA_COST = C.IFLA_COST
|
|
||||||
IFLA_PRIORITY = C.IFLA_PRIORITY
|
|
||||||
IFLA_MASTER = C.IFLA_MASTER
|
|
||||||
IFLA_WIRELESS = C.IFLA_WIRELESS
|
|
||||||
IFLA_PROTINFO = C.IFLA_PROTINFO
|
|
||||||
IFLA_TXQLEN = C.IFLA_TXQLEN
|
|
||||||
IFLA_MAP = C.IFLA_MAP
|
|
||||||
IFLA_WEIGHT = C.IFLA_WEIGHT
|
|
||||||
IFLA_OPERSTATE = C.IFLA_OPERSTATE
|
|
||||||
IFLA_LINKMODE = C.IFLA_LINKMODE
|
|
||||||
IFLA_LINKINFO = C.IFLA_LINKINFO
|
|
||||||
IFLA_NET_NS_PID = C.IFLA_NET_NS_PID
|
|
||||||
IFLA_IFALIAS = C.IFLA_IFALIAS
|
|
||||||
IFLA_MAX = C.IFLA_MAX
|
|
||||||
RT_SCOPE_UNIVERSE = C.RT_SCOPE_UNIVERSE
|
|
||||||
RT_SCOPE_SITE = C.RT_SCOPE_SITE
|
|
||||||
RT_SCOPE_LINK = C.RT_SCOPE_LINK
|
|
||||||
RT_SCOPE_HOST = C.RT_SCOPE_HOST
|
|
||||||
RT_SCOPE_NOWHERE = C.RT_SCOPE_NOWHERE
|
|
||||||
RT_TABLE_UNSPEC = C.RT_TABLE_UNSPEC
|
|
||||||
RT_TABLE_COMPAT = C.RT_TABLE_COMPAT
|
|
||||||
RT_TABLE_DEFAULT = C.RT_TABLE_DEFAULT
|
|
||||||
RT_TABLE_MAIN = C.RT_TABLE_MAIN
|
|
||||||
RT_TABLE_LOCAL = C.RT_TABLE_LOCAL
|
|
||||||
RT_TABLE_MAX = C.RT_TABLE_MAX
|
|
||||||
RTA_UNSPEC = C.RTA_UNSPEC
|
|
||||||
RTA_DST = C.RTA_DST
|
|
||||||
RTA_SRC = C.RTA_SRC
|
|
||||||
RTA_IIF = C.RTA_IIF
|
|
||||||
RTA_OIF = C.RTA_OIF
|
|
||||||
RTA_GATEWAY = C.RTA_GATEWAY
|
|
||||||
RTA_PRIORITY = C.RTA_PRIORITY
|
|
||||||
RTA_PREFSRC = C.RTA_PREFSRC
|
|
||||||
RTA_METRICS = C.RTA_METRICS
|
|
||||||
RTA_MULTIPATH = C.RTA_MULTIPATH
|
|
||||||
RTA_FLOW = C.RTA_FLOW
|
|
||||||
RTA_CACHEINFO = C.RTA_CACHEINFO
|
|
||||||
RTA_TABLE = C.RTA_TABLE
|
|
||||||
RTN_UNSPEC = C.RTN_UNSPEC
|
|
||||||
RTN_UNICAST = C.RTN_UNICAST
|
|
||||||
RTN_LOCAL = C.RTN_LOCAL
|
|
||||||
RTN_BROADCAST = C.RTN_BROADCAST
|
|
||||||
RTN_ANYCAST = C.RTN_ANYCAST
|
|
||||||
RTN_MULTICAST = C.RTN_MULTICAST
|
|
||||||
RTN_BLACKHOLE = C.RTN_BLACKHOLE
|
|
||||||
RTN_UNREACHABLE = C.RTN_UNREACHABLE
|
|
||||||
RTN_PROHIBIT = C.RTN_PROHIBIT
|
|
||||||
RTN_THROW = C.RTN_THROW
|
|
||||||
RTN_NAT = C.RTN_NAT
|
|
||||||
RTN_XRESOLVE = C.RTN_XRESOLVE
|
|
||||||
RTNLGRP_NONE = C.RTNLGRP_NONE
|
|
||||||
RTNLGRP_LINK = C.RTNLGRP_LINK
|
|
||||||
RTNLGRP_NOTIFY = C.RTNLGRP_NOTIFY
|
|
||||||
RTNLGRP_NEIGH = C.RTNLGRP_NEIGH
|
|
||||||
RTNLGRP_TC = C.RTNLGRP_TC
|
|
||||||
RTNLGRP_IPV4_IFADDR = C.RTNLGRP_IPV4_IFADDR
|
|
||||||
RTNLGRP_IPV4_MROUTE = C.RTNLGRP_IPV4_MROUTE
|
|
||||||
RTNLGRP_IPV4_ROUTE = C.RTNLGRP_IPV4_ROUTE
|
|
||||||
RTNLGRP_IPV4_RULE = C.RTNLGRP_IPV4_RULE
|
|
||||||
RTNLGRP_IPV6_IFADDR = C.RTNLGRP_IPV6_IFADDR
|
|
||||||
RTNLGRP_IPV6_MROUTE = C.RTNLGRP_IPV6_MROUTE
|
|
||||||
RTNLGRP_IPV6_ROUTE = C.RTNLGRP_IPV6_ROUTE
|
|
||||||
RTNLGRP_IPV6_IFINFO = C.RTNLGRP_IPV6_IFINFO
|
|
||||||
RTNLGRP_IPV6_PREFIX = C.RTNLGRP_IPV6_PREFIX
|
|
||||||
RTNLGRP_IPV6_RULE = C.RTNLGRP_IPV6_RULE
|
|
||||||
RTNLGRP_ND_USEROPT = C.RTNLGRP_ND_USEROPT
|
|
||||||
SizeofNlMsghdr = C.sizeof_struct_nlmsghdr
|
|
||||||
SizeofNlMsgerr = C.sizeof_struct_nlmsgerr
|
|
||||||
SizeofRtGenmsg = C.sizeof_struct_rtgenmsg
|
|
||||||
SizeofNlAttr = C.sizeof_struct_nlattr
|
|
||||||
SizeofRtAttr = C.sizeof_struct_rtattr
|
|
||||||
SizeofIfInfomsg = C.sizeof_struct_ifinfomsg
|
|
||||||
SizeofIfAddrmsg = C.sizeof_struct_ifaddrmsg
|
|
||||||
SizeofRtMsg = C.sizeof_struct_rtmsg
|
|
||||||
SizeofRtNexthop = C.sizeof_struct_rtnexthop
|
|
||||||
)
|
|
||||||
|
|
||||||
type NlMsghdr C.struct_nlmsghdr
|
|
||||||
|
|
||||||
type NlMsgerr C.struct_nlmsgerr
|
|
||||||
|
|
||||||
type RtGenmsg C.struct_rtgenmsg
|
|
||||||
|
|
||||||
type NlAttr C.struct_nlattr
|
|
||||||
|
|
||||||
type RtAttr C.struct_rtattr
|
|
||||||
|
|
||||||
type IfInfomsg C.struct_ifinfomsg
|
|
||||||
|
|
||||||
type IfAddrmsg C.struct_ifaddrmsg
|
|
||||||
|
|
||||||
type RtMsg C.struct_rtmsg
|
|
||||||
|
|
||||||
type RtNexthop C.struct_rtnexthop
|
|
||||||
|
|
||||||
// Linux socket filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofSockFilter = C.sizeof_struct_sock_filter
|
|
||||||
SizeofSockFprog = C.sizeof_struct_sock_fprog
|
|
||||||
)
|
|
||||||
|
|
||||||
type SockFilter C.struct_sock_filter
|
|
||||||
|
|
||||||
type SockFprog C.struct_sock_fprog
|
|
||||||
|
|
||||||
// Inotify
|
|
||||||
|
|
||||||
type InotifyEvent C.struct_inotify_event
|
|
||||||
|
|
||||||
const SizeofInotifyEvent = C.sizeof_struct_inotify_event
|
|
||||||
|
|
||||||
// Ptrace
|
|
||||||
|
|
||||||
// Register structures
|
|
||||||
type PtraceRegs C.PtraceRegs
|
|
||||||
|
|
||||||
// Structures contained in PtraceRegs on s390x (exported by mkpost.go)
|
|
||||||
type ptracePsw C.ptracePsw
|
|
||||||
|
|
||||||
type ptraceFpregs C.ptraceFpregs
|
|
||||||
|
|
||||||
type ptracePer C.ptracePer
|
|
||||||
|
|
||||||
// Misc
|
|
||||||
|
|
||||||
type FdSet C.fd_set
|
|
||||||
|
|
||||||
type Sysinfo_t C.struct_sysinfo
|
|
||||||
|
|
||||||
type Utsname C.struct_utsname
|
|
||||||
|
|
||||||
type Ustat_t C.struct_ustat
|
|
||||||
|
|
||||||
type EpollEvent C.struct_my_epoll_event
|
|
||||||
|
|
||||||
const (
|
|
||||||
AT_FDCWD = C.AT_FDCWD
|
|
||||||
AT_REMOVEDIR = C.AT_REMOVEDIR
|
|
||||||
AT_SYMLINK_FOLLOW = C.AT_SYMLINK_FOLLOW
|
|
||||||
AT_SYMLINK_NOFOLLOW = C.AT_SYMLINK_NOFOLLOW
|
|
||||||
)
|
|
||||||
|
|
||||||
type PollFd C.struct_pollfd
|
|
||||||
|
|
||||||
const (
|
|
||||||
POLLIN = C.POLLIN
|
|
||||||
POLLPRI = C.POLLPRI
|
|
||||||
POLLOUT = C.POLLOUT
|
|
||||||
POLLRDHUP = C.POLLRDHUP
|
|
||||||
POLLERR = C.POLLERR
|
|
||||||
POLLHUP = C.POLLHUP
|
|
||||||
POLLNVAL = C.POLLNVAL
|
|
||||||
)
|
|
||||||
|
|
||||||
type Sigset_t C.sigset_t
|
|
||||||
|
|
||||||
// Terminal handling
|
|
||||||
|
|
||||||
type Termios C.termios_t
|
|
|
@ -1,232 +0,0 @@
|
||||||
// Copyright 2009 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
/*
|
|
||||||
Input to cgo -godefs. See also mkerrors.sh and mkall.sh
|
|
||||||
*/
|
|
||||||
|
|
||||||
// +godefs map struct_in_addr [4]byte /* in_addr */
|
|
||||||
// +godefs map struct_in6_addr [16]byte /* in6_addr */
|
|
||||||
|
|
||||||
package unix
|
|
||||||
|
|
||||||
/*
|
|
||||||
#define KERNEL
|
|
||||||
#include <dirent.h>
|
|
||||||
#include <fcntl.h>
|
|
||||||
#include <signal.h>
|
|
||||||
#include <termios.h>
|
|
||||||
#include <stdio.h>
|
|
||||||
#include <unistd.h>
|
|
||||||
#include <sys/param.h>
|
|
||||||
#include <sys/types.h>
|
|
||||||
#include <sys/event.h>
|
|
||||||
#include <sys/mman.h>
|
|
||||||
#include <sys/mount.h>
|
|
||||||
#include <sys/ptrace.h>
|
|
||||||
#include <sys/resource.h>
|
|
||||||
#include <sys/select.h>
|
|
||||||
#include <sys/signal.h>
|
|
||||||
#include <sys/socket.h>
|
|
||||||
#include <sys/stat.h>
|
|
||||||
#include <sys/sysctl.h>
|
|
||||||
#include <sys/time.h>
|
|
||||||
#include <sys/uio.h>
|
|
||||||
#include <sys/un.h>
|
|
||||||
#include <sys/wait.h>
|
|
||||||
#include <net/bpf.h>
|
|
||||||
#include <net/if.h>
|
|
||||||
#include <net/if_dl.h>
|
|
||||||
#include <net/route.h>
|
|
||||||
#include <netinet/in.h>
|
|
||||||
#include <netinet/icmp6.h>
|
|
||||||
#include <netinet/tcp.h>
|
|
||||||
|
|
||||||
enum {
|
|
||||||
sizeofPtr = sizeof(void*),
|
|
||||||
};
|
|
||||||
|
|
||||||
union sockaddr_all {
|
|
||||||
struct sockaddr s1; // this one gets used for fields
|
|
||||||
struct sockaddr_in s2; // these pad it out
|
|
||||||
struct sockaddr_in6 s3;
|
|
||||||
struct sockaddr_un s4;
|
|
||||||
struct sockaddr_dl s5;
|
|
||||||
};
|
|
||||||
|
|
||||||
struct sockaddr_any {
|
|
||||||
struct sockaddr addr;
|
|
||||||
char pad[sizeof(union sockaddr_all) - sizeof(struct sockaddr)];
|
|
||||||
};
|
|
||||||
|
|
||||||
*/
|
|
||||||
import "C"
|
|
||||||
|
|
||||||
// Machine characteristics; for internal use.
|
|
||||||
|
|
||||||
const (
|
|
||||||
sizeofPtr = C.sizeofPtr
|
|
||||||
sizeofShort = C.sizeof_short
|
|
||||||
sizeofInt = C.sizeof_int
|
|
||||||
sizeofLong = C.sizeof_long
|
|
||||||
sizeofLongLong = C.sizeof_longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Basic types
|
|
||||||
|
|
||||||
type (
|
|
||||||
_C_short C.short
|
|
||||||
_C_int C.int
|
|
||||||
_C_long C.long
|
|
||||||
_C_long_long C.longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Time
|
|
||||||
|
|
||||||
type Timespec C.struct_timespec
|
|
||||||
|
|
||||||
type Timeval C.struct_timeval
|
|
||||||
|
|
||||||
// Processes
|
|
||||||
|
|
||||||
type Rusage C.struct_rusage
|
|
||||||
|
|
||||||
type Rlimit C.struct_rlimit
|
|
||||||
|
|
||||||
type _Gid_t C.gid_t
|
|
||||||
|
|
||||||
// Files
|
|
||||||
|
|
||||||
type Stat_t C.struct_stat
|
|
||||||
|
|
||||||
type Statfs_t C.struct_statfs
|
|
||||||
|
|
||||||
type Flock_t C.struct_flock
|
|
||||||
|
|
||||||
type Dirent C.struct_dirent
|
|
||||||
|
|
||||||
type Fsid C.fsid_t
|
|
||||||
|
|
||||||
// Sockets
|
|
||||||
|
|
||||||
type RawSockaddrInet4 C.struct_sockaddr_in
|
|
||||||
|
|
||||||
type RawSockaddrInet6 C.struct_sockaddr_in6
|
|
||||||
|
|
||||||
type RawSockaddrUnix C.struct_sockaddr_un
|
|
||||||
|
|
||||||
type RawSockaddrDatalink C.struct_sockaddr_dl
|
|
||||||
|
|
||||||
type RawSockaddr C.struct_sockaddr
|
|
||||||
|
|
||||||
type RawSockaddrAny C.struct_sockaddr_any
|
|
||||||
|
|
||||||
type _Socklen C.socklen_t
|
|
||||||
|
|
||||||
type Linger C.struct_linger
|
|
||||||
|
|
||||||
type Iovec C.struct_iovec
|
|
||||||
|
|
||||||
type IPMreq C.struct_ip_mreq
|
|
||||||
|
|
||||||
type IPv6Mreq C.struct_ipv6_mreq
|
|
||||||
|
|
||||||
type Msghdr C.struct_msghdr
|
|
||||||
|
|
||||||
type Cmsghdr C.struct_cmsghdr
|
|
||||||
|
|
||||||
type Inet6Pktinfo C.struct_in6_pktinfo
|
|
||||||
|
|
||||||
type IPv6MTUInfo C.struct_ip6_mtuinfo
|
|
||||||
|
|
||||||
type ICMPv6Filter C.struct_icmp6_filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofSockaddrInet4 = C.sizeof_struct_sockaddr_in
|
|
||||||
SizeofSockaddrInet6 = C.sizeof_struct_sockaddr_in6
|
|
||||||
SizeofSockaddrAny = C.sizeof_struct_sockaddr_any
|
|
||||||
SizeofSockaddrUnix = C.sizeof_struct_sockaddr_un
|
|
||||||
SizeofSockaddrDatalink = C.sizeof_struct_sockaddr_dl
|
|
||||||
SizeofLinger = C.sizeof_struct_linger
|
|
||||||
SizeofIPMreq = C.sizeof_struct_ip_mreq
|
|
||||||
SizeofIPv6Mreq = C.sizeof_struct_ipv6_mreq
|
|
||||||
SizeofMsghdr = C.sizeof_struct_msghdr
|
|
||||||
SizeofCmsghdr = C.sizeof_struct_cmsghdr
|
|
||||||
SizeofInet6Pktinfo = C.sizeof_struct_in6_pktinfo
|
|
||||||
SizeofIPv6MTUInfo = C.sizeof_struct_ip6_mtuinfo
|
|
||||||
SizeofICMPv6Filter = C.sizeof_struct_icmp6_filter
|
|
||||||
)
|
|
||||||
|
|
||||||
// Ptrace requests
|
|
||||||
|
|
||||||
const (
|
|
||||||
PTRACE_TRACEME = C.PT_TRACE_ME
|
|
||||||
PTRACE_CONT = C.PT_CONTINUE
|
|
||||||
PTRACE_KILL = C.PT_KILL
|
|
||||||
)
|
|
||||||
|
|
||||||
// Events (kqueue, kevent)
|
|
||||||
|
|
||||||
type Kevent_t C.struct_kevent
|
|
||||||
|
|
||||||
// Select
|
|
||||||
|
|
||||||
type FdSet C.fd_set
|
|
||||||
|
|
||||||
// Routing and interface messages
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofIfMsghdr = C.sizeof_struct_if_msghdr
|
|
||||||
SizeofIfData = C.sizeof_struct_if_data
|
|
||||||
SizeofIfaMsghdr = C.sizeof_struct_ifa_msghdr
|
|
||||||
SizeofIfAnnounceMsghdr = C.sizeof_struct_if_announcemsghdr
|
|
||||||
SizeofRtMsghdr = C.sizeof_struct_rt_msghdr
|
|
||||||
SizeofRtMetrics = C.sizeof_struct_rt_metrics
|
|
||||||
)
|
|
||||||
|
|
||||||
type IfMsghdr C.struct_if_msghdr
|
|
||||||
|
|
||||||
type IfData C.struct_if_data
|
|
||||||
|
|
||||||
type IfaMsghdr C.struct_ifa_msghdr
|
|
||||||
|
|
||||||
type IfAnnounceMsghdr C.struct_if_announcemsghdr
|
|
||||||
|
|
||||||
type RtMsghdr C.struct_rt_msghdr
|
|
||||||
|
|
||||||
type RtMetrics C.struct_rt_metrics
|
|
||||||
|
|
||||||
type Mclpool C.struct_mclpool
|
|
||||||
|
|
||||||
// Berkeley packet filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofBpfVersion = C.sizeof_struct_bpf_version
|
|
||||||
SizeofBpfStat = C.sizeof_struct_bpf_stat
|
|
||||||
SizeofBpfProgram = C.sizeof_struct_bpf_program
|
|
||||||
SizeofBpfInsn = C.sizeof_struct_bpf_insn
|
|
||||||
SizeofBpfHdr = C.sizeof_struct_bpf_hdr
|
|
||||||
)
|
|
||||||
|
|
||||||
type BpfVersion C.struct_bpf_version
|
|
||||||
|
|
||||||
type BpfStat C.struct_bpf_stat
|
|
||||||
|
|
||||||
type BpfProgram C.struct_bpf_program
|
|
||||||
|
|
||||||
type BpfInsn C.struct_bpf_insn
|
|
||||||
|
|
||||||
type BpfHdr C.struct_bpf_hdr
|
|
||||||
|
|
||||||
type BpfTimeval C.struct_bpf_timeval
|
|
||||||
|
|
||||||
// Terminal handling
|
|
||||||
|
|
||||||
type Termios C.struct_termios
|
|
||||||
|
|
||||||
// Sysctl
|
|
||||||
|
|
||||||
type Sysctlnode C.struct_sysctlnode
|
|
|
@ -1,244 +0,0 @@
|
||||||
// Copyright 2009 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
/*
|
|
||||||
Input to cgo -godefs. See also mkerrors.sh and mkall.sh
|
|
||||||
*/
|
|
||||||
|
|
||||||
// +godefs map struct_in_addr [4]byte /* in_addr */
|
|
||||||
// +godefs map struct_in6_addr [16]byte /* in6_addr */
|
|
||||||
|
|
||||||
package unix
|
|
||||||
|
|
||||||
/*
|
|
||||||
#define KERNEL
|
|
||||||
#include <dirent.h>
|
|
||||||
#include <fcntl.h>
|
|
||||||
#include <signal.h>
|
|
||||||
#include <termios.h>
|
|
||||||
#include <stdio.h>
|
|
||||||
#include <unistd.h>
|
|
||||||
#include <sys/param.h>
|
|
||||||
#include <sys/types.h>
|
|
||||||
#include <sys/event.h>
|
|
||||||
#include <sys/mman.h>
|
|
||||||
#include <sys/mount.h>
|
|
||||||
#include <sys/ptrace.h>
|
|
||||||
#include <sys/resource.h>
|
|
||||||
#include <sys/select.h>
|
|
||||||
#include <sys/signal.h>
|
|
||||||
#include <sys/socket.h>
|
|
||||||
#include <sys/stat.h>
|
|
||||||
#include <sys/time.h>
|
|
||||||
#include <sys/uio.h>
|
|
||||||
#include <sys/un.h>
|
|
||||||
#include <sys/wait.h>
|
|
||||||
#include <net/bpf.h>
|
|
||||||
#include <net/if.h>
|
|
||||||
#include <net/if_dl.h>
|
|
||||||
#include <net/route.h>
|
|
||||||
#include <netinet/in.h>
|
|
||||||
#include <netinet/icmp6.h>
|
|
||||||
#include <netinet/tcp.h>
|
|
||||||
|
|
||||||
enum {
|
|
||||||
sizeofPtr = sizeof(void*),
|
|
||||||
};
|
|
||||||
|
|
||||||
union sockaddr_all {
|
|
||||||
struct sockaddr s1; // this one gets used for fields
|
|
||||||
struct sockaddr_in s2; // these pad it out
|
|
||||||
struct sockaddr_in6 s3;
|
|
||||||
struct sockaddr_un s4;
|
|
||||||
struct sockaddr_dl s5;
|
|
||||||
};
|
|
||||||
|
|
||||||
struct sockaddr_any {
|
|
||||||
struct sockaddr addr;
|
|
||||||
char pad[sizeof(union sockaddr_all) - sizeof(struct sockaddr)];
|
|
||||||
};
|
|
||||||
|
|
||||||
*/
|
|
||||||
import "C"
|
|
||||||
|
|
||||||
// Machine characteristics; for internal use.
|
|
||||||
|
|
||||||
const (
|
|
||||||
sizeofPtr = C.sizeofPtr
|
|
||||||
sizeofShort = C.sizeof_short
|
|
||||||
sizeofInt = C.sizeof_int
|
|
||||||
sizeofLong = C.sizeof_long
|
|
||||||
sizeofLongLong = C.sizeof_longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Basic types
|
|
||||||
|
|
||||||
type (
|
|
||||||
_C_short C.short
|
|
||||||
_C_int C.int
|
|
||||||
_C_long C.long
|
|
||||||
_C_long_long C.longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Time
|
|
||||||
|
|
||||||
type Timespec C.struct_timespec
|
|
||||||
|
|
||||||
type Timeval C.struct_timeval
|
|
||||||
|
|
||||||
// Processes
|
|
||||||
|
|
||||||
type Rusage C.struct_rusage
|
|
||||||
|
|
||||||
type Rlimit C.struct_rlimit
|
|
||||||
|
|
||||||
type _Gid_t C.gid_t
|
|
||||||
|
|
||||||
// Files
|
|
||||||
|
|
||||||
const ( // Directory mode bits
|
|
||||||
S_IFMT = C.S_IFMT
|
|
||||||
S_IFIFO = C.S_IFIFO
|
|
||||||
S_IFCHR = C.S_IFCHR
|
|
||||||
S_IFDIR = C.S_IFDIR
|
|
||||||
S_IFBLK = C.S_IFBLK
|
|
||||||
S_IFREG = C.S_IFREG
|
|
||||||
S_IFLNK = C.S_IFLNK
|
|
||||||
S_IFSOCK = C.S_IFSOCK
|
|
||||||
S_ISUID = C.S_ISUID
|
|
||||||
S_ISGID = C.S_ISGID
|
|
||||||
S_ISVTX = C.S_ISVTX
|
|
||||||
S_IRUSR = C.S_IRUSR
|
|
||||||
S_IWUSR = C.S_IWUSR
|
|
||||||
S_IXUSR = C.S_IXUSR
|
|
||||||
)
|
|
||||||
|
|
||||||
type Stat_t C.struct_stat
|
|
||||||
|
|
||||||
type Statfs_t C.struct_statfs
|
|
||||||
|
|
||||||
type Flock_t C.struct_flock
|
|
||||||
|
|
||||||
type Dirent C.struct_dirent
|
|
||||||
|
|
||||||
type Fsid C.fsid_t
|
|
||||||
|
|
||||||
// Sockets
|
|
||||||
|
|
||||||
type RawSockaddrInet4 C.struct_sockaddr_in
|
|
||||||
|
|
||||||
type RawSockaddrInet6 C.struct_sockaddr_in6
|
|
||||||
|
|
||||||
type RawSockaddrUnix C.struct_sockaddr_un
|
|
||||||
|
|
||||||
type RawSockaddrDatalink C.struct_sockaddr_dl
|
|
||||||
|
|
||||||
type RawSockaddr C.struct_sockaddr
|
|
||||||
|
|
||||||
type RawSockaddrAny C.struct_sockaddr_any
|
|
||||||
|
|
||||||
type _Socklen C.socklen_t
|
|
||||||
|
|
||||||
type Linger C.struct_linger
|
|
||||||
|
|
||||||
type Iovec C.struct_iovec
|
|
||||||
|
|
||||||
type IPMreq C.struct_ip_mreq
|
|
||||||
|
|
||||||
type IPv6Mreq C.struct_ipv6_mreq
|
|
||||||
|
|
||||||
type Msghdr C.struct_msghdr
|
|
||||||
|
|
||||||
type Cmsghdr C.struct_cmsghdr
|
|
||||||
|
|
||||||
type Inet6Pktinfo C.struct_in6_pktinfo
|
|
||||||
|
|
||||||
type IPv6MTUInfo C.struct_ip6_mtuinfo
|
|
||||||
|
|
||||||
type ICMPv6Filter C.struct_icmp6_filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofSockaddrInet4 = C.sizeof_struct_sockaddr_in
|
|
||||||
SizeofSockaddrInet6 = C.sizeof_struct_sockaddr_in6
|
|
||||||
SizeofSockaddrAny = C.sizeof_struct_sockaddr_any
|
|
||||||
SizeofSockaddrUnix = C.sizeof_struct_sockaddr_un
|
|
||||||
SizeofSockaddrDatalink = C.sizeof_struct_sockaddr_dl
|
|
||||||
SizeofLinger = C.sizeof_struct_linger
|
|
||||||
SizeofIPMreq = C.sizeof_struct_ip_mreq
|
|
||||||
SizeofIPv6Mreq = C.sizeof_struct_ipv6_mreq
|
|
||||||
SizeofMsghdr = C.sizeof_struct_msghdr
|
|
||||||
SizeofCmsghdr = C.sizeof_struct_cmsghdr
|
|
||||||
SizeofInet6Pktinfo = C.sizeof_struct_in6_pktinfo
|
|
||||||
SizeofIPv6MTUInfo = C.sizeof_struct_ip6_mtuinfo
|
|
||||||
SizeofICMPv6Filter = C.sizeof_struct_icmp6_filter
|
|
||||||
)
|
|
||||||
|
|
||||||
// Ptrace requests
|
|
||||||
|
|
||||||
const (
|
|
||||||
PTRACE_TRACEME = C.PT_TRACE_ME
|
|
||||||
PTRACE_CONT = C.PT_CONTINUE
|
|
||||||
PTRACE_KILL = C.PT_KILL
|
|
||||||
)
|
|
||||||
|
|
||||||
// Events (kqueue, kevent)
|
|
||||||
|
|
||||||
type Kevent_t C.struct_kevent
|
|
||||||
|
|
||||||
// Select
|
|
||||||
|
|
||||||
type FdSet C.fd_set
|
|
||||||
|
|
||||||
// Routing and interface messages
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofIfMsghdr = C.sizeof_struct_if_msghdr
|
|
||||||
SizeofIfData = C.sizeof_struct_if_data
|
|
||||||
SizeofIfaMsghdr = C.sizeof_struct_ifa_msghdr
|
|
||||||
SizeofIfAnnounceMsghdr = C.sizeof_struct_if_announcemsghdr
|
|
||||||
SizeofRtMsghdr = C.sizeof_struct_rt_msghdr
|
|
||||||
SizeofRtMetrics = C.sizeof_struct_rt_metrics
|
|
||||||
)
|
|
||||||
|
|
||||||
type IfMsghdr C.struct_if_msghdr
|
|
||||||
|
|
||||||
type IfData C.struct_if_data
|
|
||||||
|
|
||||||
type IfaMsghdr C.struct_ifa_msghdr
|
|
||||||
|
|
||||||
type IfAnnounceMsghdr C.struct_if_announcemsghdr
|
|
||||||
|
|
||||||
type RtMsghdr C.struct_rt_msghdr
|
|
||||||
|
|
||||||
type RtMetrics C.struct_rt_metrics
|
|
||||||
|
|
||||||
type Mclpool C.struct_mclpool
|
|
||||||
|
|
||||||
// Berkeley packet filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofBpfVersion = C.sizeof_struct_bpf_version
|
|
||||||
SizeofBpfStat = C.sizeof_struct_bpf_stat
|
|
||||||
SizeofBpfProgram = C.sizeof_struct_bpf_program
|
|
||||||
SizeofBpfInsn = C.sizeof_struct_bpf_insn
|
|
||||||
SizeofBpfHdr = C.sizeof_struct_bpf_hdr
|
|
||||||
)
|
|
||||||
|
|
||||||
type BpfVersion C.struct_bpf_version
|
|
||||||
|
|
||||||
type BpfStat C.struct_bpf_stat
|
|
||||||
|
|
||||||
type BpfProgram C.struct_bpf_program
|
|
||||||
|
|
||||||
type BpfInsn C.struct_bpf_insn
|
|
||||||
|
|
||||||
type BpfHdr C.struct_bpf_hdr
|
|
||||||
|
|
||||||
type BpfTimeval C.struct_bpf_timeval
|
|
||||||
|
|
||||||
// Terminal handling
|
|
||||||
|
|
||||||
type Termios C.struct_termios
|
|
|
@ -1,262 +0,0 @@
|
||||||
// Copyright 2009 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
/*
|
|
||||||
Input to cgo -godefs. See also mkerrors.sh and mkall.sh
|
|
||||||
*/
|
|
||||||
|
|
||||||
// +godefs map struct_in_addr [4]byte /* in_addr */
|
|
||||||
// +godefs map struct_in6_addr [16]byte /* in6_addr */
|
|
||||||
|
|
||||||
package unix
|
|
||||||
|
|
||||||
/*
|
|
||||||
#define KERNEL
|
|
||||||
// These defines ensure that builds done on newer versions of Solaris are
|
|
||||||
// backwards-compatible with older versions of Solaris and
|
|
||||||
// OpenSolaris-based derivatives.
|
|
||||||
#define __USE_SUNOS_SOCKETS__ // msghdr
|
|
||||||
#define __USE_LEGACY_PROTOTYPES__ // iovec
|
|
||||||
#include <dirent.h>
|
|
||||||
#include <fcntl.h>
|
|
||||||
#include <netdb.h>
|
|
||||||
#include <limits.h>
|
|
||||||
#include <signal.h>
|
|
||||||
#include <termios.h>
|
|
||||||
#include <termio.h>
|
|
||||||
#include <stdio.h>
|
|
||||||
#include <unistd.h>
|
|
||||||
#include <sys/mman.h>
|
|
||||||
#include <sys/mount.h>
|
|
||||||
#include <sys/param.h>
|
|
||||||
#include <sys/resource.h>
|
|
||||||
#include <sys/select.h>
|
|
||||||
#include <sys/signal.h>
|
|
||||||
#include <sys/socket.h>
|
|
||||||
#include <sys/stat.h>
|
|
||||||
#include <sys/time.h>
|
|
||||||
#include <sys/times.h>
|
|
||||||
#include <sys/types.h>
|
|
||||||
#include <sys/utsname.h>
|
|
||||||
#include <sys/un.h>
|
|
||||||
#include <sys/wait.h>
|
|
||||||
#include <net/bpf.h>
|
|
||||||
#include <net/if.h>
|
|
||||||
#include <net/if_dl.h>
|
|
||||||
#include <net/route.h>
|
|
||||||
#include <netinet/in.h>
|
|
||||||
#include <netinet/icmp6.h>
|
|
||||||
#include <netinet/tcp.h>
|
|
||||||
#include <ustat.h>
|
|
||||||
#include <utime.h>
|
|
||||||
|
|
||||||
enum {
|
|
||||||
sizeofPtr = sizeof(void*),
|
|
||||||
};
|
|
||||||
|
|
||||||
union sockaddr_all {
|
|
||||||
struct sockaddr s1; // this one gets used for fields
|
|
||||||
struct sockaddr_in s2; // these pad it out
|
|
||||||
struct sockaddr_in6 s3;
|
|
||||||
struct sockaddr_un s4;
|
|
||||||
struct sockaddr_dl s5;
|
|
||||||
};
|
|
||||||
|
|
||||||
struct sockaddr_any {
|
|
||||||
struct sockaddr addr;
|
|
||||||
char pad[sizeof(union sockaddr_all) - sizeof(struct sockaddr)];
|
|
||||||
};
|
|
||||||
|
|
||||||
*/
|
|
||||||
import "C"
|
|
||||||
|
|
||||||
// Machine characteristics; for internal use.
|
|
||||||
|
|
||||||
const (
|
|
||||||
sizeofPtr = C.sizeofPtr
|
|
||||||
sizeofShort = C.sizeof_short
|
|
||||||
sizeofInt = C.sizeof_int
|
|
||||||
sizeofLong = C.sizeof_long
|
|
||||||
sizeofLongLong = C.sizeof_longlong
|
|
||||||
PathMax = C.PATH_MAX
|
|
||||||
MaxHostNameLen = C.MAXHOSTNAMELEN
|
|
||||||
)
|
|
||||||
|
|
||||||
// Basic types
|
|
||||||
|
|
||||||
type (
|
|
||||||
_C_short C.short
|
|
||||||
_C_int C.int
|
|
||||||
_C_long C.long
|
|
||||||
_C_long_long C.longlong
|
|
||||||
)
|
|
||||||
|
|
||||||
// Time
|
|
||||||
|
|
||||||
type Timespec C.struct_timespec
|
|
||||||
|
|
||||||
type Timeval C.struct_timeval
|
|
||||||
|
|
||||||
type Timeval32 C.struct_timeval32
|
|
||||||
|
|
||||||
type Tms C.struct_tms
|
|
||||||
|
|
||||||
type Utimbuf C.struct_utimbuf
|
|
||||||
|
|
||||||
// Processes
|
|
||||||
|
|
||||||
type Rusage C.struct_rusage
|
|
||||||
|
|
||||||
type Rlimit C.struct_rlimit
|
|
||||||
|
|
||||||
type _Gid_t C.gid_t
|
|
||||||
|
|
||||||
// Files
|
|
||||||
|
|
||||||
const ( // Directory mode bits
|
|
||||||
S_IFMT = C.S_IFMT
|
|
||||||
S_IFIFO = C.S_IFIFO
|
|
||||||
S_IFCHR = C.S_IFCHR
|
|
||||||
S_IFDIR = C.S_IFDIR
|
|
||||||
S_IFBLK = C.S_IFBLK
|
|
||||||
S_IFREG = C.S_IFREG
|
|
||||||
S_IFLNK = C.S_IFLNK
|
|
||||||
S_IFSOCK = C.S_IFSOCK
|
|
||||||
S_ISUID = C.S_ISUID
|
|
||||||
S_ISGID = C.S_ISGID
|
|
||||||
S_ISVTX = C.S_ISVTX
|
|
||||||
S_IRUSR = C.S_IRUSR
|
|
||||||
S_IWUSR = C.S_IWUSR
|
|
||||||
S_IXUSR = C.S_IXUSR
|
|
||||||
)
|
|
||||||
|
|
||||||
type Stat_t C.struct_stat
|
|
||||||
|
|
||||||
type Flock_t C.struct_flock
|
|
||||||
|
|
||||||
type Dirent C.struct_dirent
|
|
||||||
|
|
||||||
// Sockets
|
|
||||||
|
|
||||||
type RawSockaddrInet4 C.struct_sockaddr_in
|
|
||||||
|
|
||||||
type RawSockaddrInet6 C.struct_sockaddr_in6
|
|
||||||
|
|
||||||
type RawSockaddrUnix C.struct_sockaddr_un
|
|
||||||
|
|
||||||
type RawSockaddrDatalink C.struct_sockaddr_dl
|
|
||||||
|
|
||||||
type RawSockaddr C.struct_sockaddr
|
|
||||||
|
|
||||||
type RawSockaddrAny C.struct_sockaddr_any
|
|
||||||
|
|
||||||
type _Socklen C.socklen_t
|
|
||||||
|
|
||||||
type Linger C.struct_linger
|
|
||||||
|
|
||||||
type Iovec C.struct_iovec
|
|
||||||
|
|
||||||
type IPMreq C.struct_ip_mreq
|
|
||||||
|
|
||||||
type IPv6Mreq C.struct_ipv6_mreq
|
|
||||||
|
|
||||||
type Msghdr C.struct_msghdr
|
|
||||||
|
|
||||||
type Cmsghdr C.struct_cmsghdr
|
|
||||||
|
|
||||||
type Inet6Pktinfo C.struct_in6_pktinfo
|
|
||||||
|
|
||||||
type IPv6MTUInfo C.struct_ip6_mtuinfo
|
|
||||||
|
|
||||||
type ICMPv6Filter C.struct_icmp6_filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofSockaddrInet4 = C.sizeof_struct_sockaddr_in
|
|
||||||
SizeofSockaddrInet6 = C.sizeof_struct_sockaddr_in6
|
|
||||||
SizeofSockaddrAny = C.sizeof_struct_sockaddr_any
|
|
||||||
SizeofSockaddrUnix = C.sizeof_struct_sockaddr_un
|
|
||||||
SizeofSockaddrDatalink = C.sizeof_struct_sockaddr_dl
|
|
||||||
SizeofLinger = C.sizeof_struct_linger
|
|
||||||
SizeofIPMreq = C.sizeof_struct_ip_mreq
|
|
||||||
SizeofIPv6Mreq = C.sizeof_struct_ipv6_mreq
|
|
||||||
SizeofMsghdr = C.sizeof_struct_msghdr
|
|
||||||
SizeofCmsghdr = C.sizeof_struct_cmsghdr
|
|
||||||
SizeofInet6Pktinfo = C.sizeof_struct_in6_pktinfo
|
|
||||||
SizeofIPv6MTUInfo = C.sizeof_struct_ip6_mtuinfo
|
|
||||||
SizeofICMPv6Filter = C.sizeof_struct_icmp6_filter
|
|
||||||
)
|
|
||||||
|
|
||||||
// Select
|
|
||||||
|
|
||||||
type FdSet C.fd_set
|
|
||||||
|
|
||||||
// Misc
|
|
||||||
|
|
||||||
type Utsname C.struct_utsname
|
|
||||||
|
|
||||||
type Ustat_t C.struct_ustat
|
|
||||||
|
|
||||||
const (
|
|
||||||
AT_FDCWD = C.AT_FDCWD
|
|
||||||
AT_SYMLINK_NOFOLLOW = C.AT_SYMLINK_NOFOLLOW
|
|
||||||
AT_SYMLINK_FOLLOW = C.AT_SYMLINK_FOLLOW
|
|
||||||
AT_REMOVEDIR = C.AT_REMOVEDIR
|
|
||||||
AT_EACCESS = C.AT_EACCESS
|
|
||||||
)
|
|
||||||
|
|
||||||
// Routing and interface messages
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofIfMsghdr = C.sizeof_struct_if_msghdr
|
|
||||||
SizeofIfData = C.sizeof_struct_if_data
|
|
||||||
SizeofIfaMsghdr = C.sizeof_struct_ifa_msghdr
|
|
||||||
SizeofRtMsghdr = C.sizeof_struct_rt_msghdr
|
|
||||||
SizeofRtMetrics = C.sizeof_struct_rt_metrics
|
|
||||||
)
|
|
||||||
|
|
||||||
type IfMsghdr C.struct_if_msghdr
|
|
||||||
|
|
||||||
type IfData C.struct_if_data
|
|
||||||
|
|
||||||
type IfaMsghdr C.struct_ifa_msghdr
|
|
||||||
|
|
||||||
type RtMsghdr C.struct_rt_msghdr
|
|
||||||
|
|
||||||
type RtMetrics C.struct_rt_metrics
|
|
||||||
|
|
||||||
// Berkeley packet filter
|
|
||||||
|
|
||||||
const (
|
|
||||||
SizeofBpfVersion = C.sizeof_struct_bpf_version
|
|
||||||
SizeofBpfStat = C.sizeof_struct_bpf_stat
|
|
||||||
SizeofBpfProgram = C.sizeof_struct_bpf_program
|
|
||||||
SizeofBpfInsn = C.sizeof_struct_bpf_insn
|
|
||||||
SizeofBpfHdr = C.sizeof_struct_bpf_hdr
|
|
||||||
)
|
|
||||||
|
|
||||||
type BpfVersion C.struct_bpf_version
|
|
||||||
|
|
||||||
type BpfStat C.struct_bpf_stat
|
|
||||||
|
|
||||||
type BpfProgram C.struct_bpf_program
|
|
||||||
|
|
||||||
type BpfInsn C.struct_bpf_insn
|
|
||||||
|
|
||||||
type BpfTimeval C.struct_bpf_timeval
|
|
||||||
|
|
||||||
type BpfHdr C.struct_bpf_hdr
|
|
||||||
|
|
||||||
// sysconf information
|
|
||||||
|
|
||||||
const _SC_PAGESIZE = C._SC_PAGESIZE
|
|
||||||
|
|
||||||
// Terminal handling
|
|
||||||
|
|
||||||
type Termios C.struct_termios
|
|
||||||
|
|
||||||
type Termio C.struct_termio
|
|
||||||
|
|
||||||
type Winsize C.struct_winsize
|
|
|
@ -1,351 +0,0 @@
|
||||||
// Copyright 2015 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
package gen
|
|
||||||
|
|
||||||
import (
|
|
||||||
"bytes"
|
|
||||||
"encoding/gob"
|
|
||||||
"fmt"
|
|
||||||
"hash"
|
|
||||||
"hash/fnv"
|
|
||||||
"io"
|
|
||||||
"log"
|
|
||||||
"os"
|
|
||||||
"reflect"
|
|
||||||
"strings"
|
|
||||||
"unicode"
|
|
||||||
"unicode/utf8"
|
|
||||||
)
|
|
||||||
|
|
||||||
// This file contains utilities for generating code.
|
|
||||||
|
|
||||||
// TODO: other write methods like:
|
|
||||||
// - slices, maps, types, etc.
|
|
||||||
|
|
||||||
// CodeWriter is a utility for writing structured code. It computes the content
|
|
||||||
// hash and size of written content. It ensures there are newlines between
|
|
||||||
// written code blocks.
|
|
||||||
type CodeWriter struct {
|
|
||||||
buf bytes.Buffer
|
|
||||||
Size int
|
|
||||||
Hash hash.Hash32 // content hash
|
|
||||||
gob *gob.Encoder
|
|
||||||
// For comments we skip the usual one-line separator if they are followed by
|
|
||||||
// a code block.
|
|
||||||
skipSep bool
|
|
||||||
}
|
|
||||||
|
|
||||||
func (w *CodeWriter) Write(p []byte) (n int, err error) {
|
|
||||||
return w.buf.Write(p)
|
|
||||||
}
|
|
||||||
|
|
||||||
// NewCodeWriter returns a new CodeWriter.
|
|
||||||
func NewCodeWriter() *CodeWriter {
|
|
||||||
h := fnv.New32()
|
|
||||||
return &CodeWriter{Hash: h, gob: gob.NewEncoder(h)}
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteGoFile appends the buffer with the total size of all created structures
|
|
||||||
// and writes it as a Go file to the the given file with the given package name.
|
|
||||||
func (w *CodeWriter) WriteGoFile(filename, pkg string) {
|
|
||||||
f, err := os.Create(filename)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatalf("Could not create file %s: %v", filename, err)
|
|
||||||
}
|
|
||||||
defer f.Close()
|
|
||||||
if _, err = w.WriteGo(f, pkg); err != nil {
|
|
||||||
log.Fatalf("Error writing file %s: %v", filename, err)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteGo appends the buffer with the total size of all created structures and
|
|
||||||
// writes it as a Go file to the the given writer with the given package name.
|
|
||||||
func (w *CodeWriter) WriteGo(out io.Writer, pkg string) (n int, err error) {
|
|
||||||
sz := w.Size
|
|
||||||
w.WriteComment("Total table size %d bytes (%dKiB); checksum: %X\n", sz, sz/1024, w.Hash.Sum32())
|
|
||||||
defer w.buf.Reset()
|
|
||||||
return WriteGo(out, pkg, w.buf.Bytes())
|
|
||||||
}
|
|
||||||
|
|
||||||
func (w *CodeWriter) printf(f string, x ...interface{}) {
|
|
||||||
fmt.Fprintf(w, f, x...)
|
|
||||||
}
|
|
||||||
|
|
||||||
func (w *CodeWriter) insertSep() {
|
|
||||||
if w.skipSep {
|
|
||||||
w.skipSep = false
|
|
||||||
return
|
|
||||||
}
|
|
||||||
// Use at least two newlines to ensure a blank space between the previous
|
|
||||||
// block. WriteGoFile will remove extraneous newlines.
|
|
||||||
w.printf("\n\n")
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteComment writes a comment block. All line starts are prefixed with "//".
|
|
||||||
// Initial empty lines are gobbled. The indentation for the first line is
|
|
||||||
// stripped from consecutive lines.
|
|
||||||
func (w *CodeWriter) WriteComment(comment string, args ...interface{}) {
|
|
||||||
s := fmt.Sprintf(comment, args...)
|
|
||||||
s = strings.Trim(s, "\n")
|
|
||||||
|
|
||||||
// Use at least two newlines to ensure a blank space between the previous
|
|
||||||
// block. WriteGoFile will remove extraneous newlines.
|
|
||||||
w.printf("\n\n// ")
|
|
||||||
w.skipSep = true
|
|
||||||
|
|
||||||
// strip first indent level.
|
|
||||||
sep := "\n"
|
|
||||||
for ; len(s) > 0 && (s[0] == '\t' || s[0] == ' '); s = s[1:] {
|
|
||||||
sep += s[:1]
|
|
||||||
}
|
|
||||||
|
|
||||||
strings.NewReplacer(sep, "\n// ", "\n", "\n// ").WriteString(w, s)
|
|
||||||
|
|
||||||
w.printf("\n")
|
|
||||||
}
|
|
||||||
|
|
||||||
func (w *CodeWriter) writeSizeInfo(size int) {
|
|
||||||
w.printf("// Size: %d bytes\n", size)
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteConst writes a constant of the given name and value.
|
|
||||||
func (w *CodeWriter) WriteConst(name string, x interface{}) {
|
|
||||||
w.insertSep()
|
|
||||||
v := reflect.ValueOf(x)
|
|
||||||
|
|
||||||
switch v.Type().Kind() {
|
|
||||||
case reflect.String:
|
|
||||||
w.printf("const %s %s = ", name, typeName(x))
|
|
||||||
w.WriteString(v.String())
|
|
||||||
w.printf("\n")
|
|
||||||
default:
|
|
||||||
w.printf("const %s = %#v\n", name, x)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteVar writes a variable of the given name and value.
|
|
||||||
func (w *CodeWriter) WriteVar(name string, x interface{}) {
|
|
||||||
w.insertSep()
|
|
||||||
v := reflect.ValueOf(x)
|
|
||||||
oldSize := w.Size
|
|
||||||
sz := int(v.Type().Size())
|
|
||||||
w.Size += sz
|
|
||||||
|
|
||||||
switch v.Type().Kind() {
|
|
||||||
case reflect.String:
|
|
||||||
w.printf("var %s %s = ", name, typeName(x))
|
|
||||||
w.WriteString(v.String())
|
|
||||||
case reflect.Struct:
|
|
||||||
w.gob.Encode(x)
|
|
||||||
fallthrough
|
|
||||||
case reflect.Slice, reflect.Array:
|
|
||||||
w.printf("var %s = ", name)
|
|
||||||
w.writeValue(v)
|
|
||||||
w.writeSizeInfo(w.Size - oldSize)
|
|
||||||
default:
|
|
||||||
w.printf("var %s %s = ", name, typeName(x))
|
|
||||||
w.gob.Encode(x)
|
|
||||||
w.writeValue(v)
|
|
||||||
w.writeSizeInfo(w.Size - oldSize)
|
|
||||||
}
|
|
||||||
w.printf("\n")
|
|
||||||
}
|
|
||||||
|
|
||||||
func (w *CodeWriter) writeValue(v reflect.Value) {
|
|
||||||
x := v.Interface()
|
|
||||||
switch v.Kind() {
|
|
||||||
case reflect.String:
|
|
||||||
w.WriteString(v.String())
|
|
||||||
case reflect.Array:
|
|
||||||
// Don't double count: callers of WriteArray count on the size being
|
|
||||||
// added, so we need to discount it here.
|
|
||||||
w.Size -= int(v.Type().Size())
|
|
||||||
w.writeSlice(x, true)
|
|
||||||
case reflect.Slice:
|
|
||||||
w.writeSlice(x, false)
|
|
||||||
case reflect.Struct:
|
|
||||||
w.printf("%s{\n", typeName(v.Interface()))
|
|
||||||
t := v.Type()
|
|
||||||
for i := 0; i < v.NumField(); i++ {
|
|
||||||
w.printf("%s: ", t.Field(i).Name)
|
|
||||||
w.writeValue(v.Field(i))
|
|
||||||
w.printf(",\n")
|
|
||||||
}
|
|
||||||
w.printf("}")
|
|
||||||
default:
|
|
||||||
w.printf("%#v", x)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteString writes a string literal.
|
|
||||||
func (w *CodeWriter) WriteString(s string) {
|
|
||||||
s = strings.Replace(s, `\`, `\\`, -1)
|
|
||||||
io.WriteString(w.Hash, s) // content hash
|
|
||||||
w.Size += len(s)
|
|
||||||
|
|
||||||
const maxInline = 40
|
|
||||||
if len(s) <= maxInline {
|
|
||||||
w.printf("%q", s)
|
|
||||||
return
|
|
||||||
}
|
|
||||||
|
|
||||||
// We will render the string as a multi-line string.
|
|
||||||
const maxWidth = 80 - 4 - len(`"`) - len(`" +`)
|
|
||||||
|
|
||||||
// When starting on its own line, go fmt indents line 2+ an extra level.
|
|
||||||
n, max := maxWidth, maxWidth-4
|
|
||||||
|
|
||||||
// As per https://golang.org/issue/18078, the compiler has trouble
|
|
||||||
// compiling the concatenation of many strings, s0 + s1 + s2 + ... + sN,
|
|
||||||
// for large N. We insert redundant, explicit parentheses to work around
|
|
||||||
// that, lowering the N at any given step: (s0 + s1 + ... + s63) + (s64 +
|
|
||||||
// ... + s127) + etc + (etc + ... + sN).
|
|
||||||
explicitParens, extraComment := len(s) > 128*1024, ""
|
|
||||||
if explicitParens {
|
|
||||||
w.printf(`(`)
|
|
||||||
extraComment = "; the redundant, explicit parens are for https://golang.org/issue/18078"
|
|
||||||
}
|
|
||||||
|
|
||||||
// Print "" +\n, if a string does not start on its own line.
|
|
||||||
b := w.buf.Bytes()
|
|
||||||
if p := len(bytes.TrimRight(b, " \t")); p > 0 && b[p-1] != '\n' {
|
|
||||||
w.printf("\"\" + // Size: %d bytes%s\n", len(s), extraComment)
|
|
||||||
n, max = maxWidth, maxWidth
|
|
||||||
}
|
|
||||||
|
|
||||||
w.printf(`"`)
|
|
||||||
|
|
||||||
for sz, p, nLines := 0, 0, 0; p < len(s); {
|
|
||||||
var r rune
|
|
||||||
r, sz = utf8.DecodeRuneInString(s[p:])
|
|
||||||
out := s[p : p+sz]
|
|
||||||
chars := 1
|
|
||||||
if !unicode.IsPrint(r) || r == utf8.RuneError || r == '"' {
|
|
||||||
switch sz {
|
|
||||||
case 1:
|
|
||||||
out = fmt.Sprintf("\\x%02x", s[p])
|
|
||||||
case 2, 3:
|
|
||||||
out = fmt.Sprintf("\\u%04x", r)
|
|
||||||
case 4:
|
|
||||||
out = fmt.Sprintf("\\U%08x", r)
|
|
||||||
}
|
|
||||||
chars = len(out)
|
|
||||||
}
|
|
||||||
if n -= chars; n < 0 {
|
|
||||||
nLines++
|
|
||||||
if explicitParens && nLines&63 == 63 {
|
|
||||||
w.printf("\") + (\"")
|
|
||||||
}
|
|
||||||
w.printf("\" +\n\"")
|
|
||||||
n = max - len(out)
|
|
||||||
}
|
|
||||||
w.printf("%s", out)
|
|
||||||
p += sz
|
|
||||||
}
|
|
||||||
w.printf(`"`)
|
|
||||||
if explicitParens {
|
|
||||||
w.printf(`)`)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteSlice writes a slice value.
|
|
||||||
func (w *CodeWriter) WriteSlice(x interface{}) {
|
|
||||||
w.writeSlice(x, false)
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteArray writes an array value.
|
|
||||||
func (w *CodeWriter) WriteArray(x interface{}) {
|
|
||||||
w.writeSlice(x, true)
|
|
||||||
}
|
|
||||||
|
|
||||||
func (w *CodeWriter) writeSlice(x interface{}, isArray bool) {
|
|
||||||
v := reflect.ValueOf(x)
|
|
||||||
w.gob.Encode(v.Len())
|
|
||||||
w.Size += v.Len() * int(v.Type().Elem().Size())
|
|
||||||
name := typeName(x)
|
|
||||||
if isArray {
|
|
||||||
name = fmt.Sprintf("[%d]%s", v.Len(), name[strings.Index(name, "]")+1:])
|
|
||||||
}
|
|
||||||
if isArray {
|
|
||||||
w.printf("%s{\n", name)
|
|
||||||
} else {
|
|
||||||
w.printf("%s{ // %d elements\n", name, v.Len())
|
|
||||||
}
|
|
||||||
|
|
||||||
switch kind := v.Type().Elem().Kind(); kind {
|
|
||||||
case reflect.String:
|
|
||||||
for _, s := range x.([]string) {
|
|
||||||
w.WriteString(s)
|
|
||||||
w.printf(",\n")
|
|
||||||
}
|
|
||||||
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64,
|
|
||||||
reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
|
|
||||||
// nLine and nBlock are the number of elements per line and block.
|
|
||||||
nLine, nBlock, format := 8, 64, "%d,"
|
|
||||||
switch kind {
|
|
||||||
case reflect.Uint8:
|
|
||||||
format = "%#02x,"
|
|
||||||
case reflect.Uint16:
|
|
||||||
format = "%#04x,"
|
|
||||||
case reflect.Uint32:
|
|
||||||
nLine, nBlock, format = 4, 32, "%#08x,"
|
|
||||||
case reflect.Uint, reflect.Uint64:
|
|
||||||
nLine, nBlock, format = 4, 32, "%#016x,"
|
|
||||||
case reflect.Int8:
|
|
||||||
nLine = 16
|
|
||||||
}
|
|
||||||
n := nLine
|
|
||||||
for i := 0; i < v.Len(); i++ {
|
|
||||||
if i%nBlock == 0 && v.Len() > nBlock {
|
|
||||||
w.printf("// Entry %X - %X\n", i, i+nBlock-1)
|
|
||||||
}
|
|
||||||
x := v.Index(i).Interface()
|
|
||||||
w.gob.Encode(x)
|
|
||||||
w.printf(format, x)
|
|
||||||
if n--; n == 0 {
|
|
||||||
n = nLine
|
|
||||||
w.printf("\n")
|
|
||||||
}
|
|
||||||
}
|
|
||||||
w.printf("\n")
|
|
||||||
case reflect.Struct:
|
|
||||||
zero := reflect.Zero(v.Type().Elem()).Interface()
|
|
||||||
for i := 0; i < v.Len(); i++ {
|
|
||||||
x := v.Index(i).Interface()
|
|
||||||
w.gob.EncodeValue(v)
|
|
||||||
if !reflect.DeepEqual(zero, x) {
|
|
||||||
line := fmt.Sprintf("%#v,\n", x)
|
|
||||||
line = line[strings.IndexByte(line, '{'):]
|
|
||||||
w.printf("%d: ", i)
|
|
||||||
w.printf(line)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
case reflect.Array:
|
|
||||||
for i := 0; i < v.Len(); i++ {
|
|
||||||
w.printf("%d: %#v,\n", i, v.Index(i).Interface())
|
|
||||||
}
|
|
||||||
default:
|
|
||||||
panic("gen: slice elem type not supported")
|
|
||||||
}
|
|
||||||
w.printf("}")
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteType writes a definition of the type of the given value and returns the
|
|
||||||
// type name.
|
|
||||||
func (w *CodeWriter) WriteType(x interface{}) string {
|
|
||||||
t := reflect.TypeOf(x)
|
|
||||||
w.printf("type %s struct {\n", t.Name())
|
|
||||||
for i := 0; i < t.NumField(); i++ {
|
|
||||||
w.printf("\t%s %s\n", t.Field(i).Name, t.Field(i).Type)
|
|
||||||
}
|
|
||||||
w.printf("}\n")
|
|
||||||
return t.Name()
|
|
||||||
}
|
|
||||||
|
|
||||||
// typeName returns the name of the go type of x.
|
|
||||||
func typeName(x interface{}) string {
|
|
||||||
t := reflect.ValueOf(x).Type()
|
|
||||||
return strings.Replace(fmt.Sprint(t), "main.", "", 1)
|
|
||||||
}
|
|
|
@ -1,281 +0,0 @@
|
||||||
// Copyright 2015 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// Package gen contains common code for the various code generation tools in the
|
|
||||||
// text repository. Its usage ensures consistency between tools.
|
|
||||||
//
|
|
||||||
// This package defines command line flags that are common to most generation
|
|
||||||
// tools. The flags allow for specifying specific Unicode and CLDR versions
|
|
||||||
// in the public Unicode data repository (http://www.unicode.org/Public).
|
|
||||||
//
|
|
||||||
// A local Unicode data mirror can be set through the flag -local or the
|
|
||||||
// environment variable UNICODE_DIR. The former takes precedence. The local
|
|
||||||
// directory should follow the same structure as the public repository.
|
|
||||||
//
|
|
||||||
// IANA data can also optionally be mirrored by putting it in the iana directory
|
|
||||||
// rooted at the top of the local mirror. Beware, though, that IANA data is not
|
|
||||||
// versioned. So it is up to the developer to use the right version.
|
|
||||||
package gen // import "golang.org/x/text/internal/gen"
|
|
||||||
|
|
||||||
import (
|
|
||||||
"bytes"
|
|
||||||
"flag"
|
|
||||||
"fmt"
|
|
||||||
"go/build"
|
|
||||||
"go/format"
|
|
||||||
"io"
|
|
||||||
"io/ioutil"
|
|
||||||
"log"
|
|
||||||
"net/http"
|
|
||||||
"os"
|
|
||||||
"path"
|
|
||||||
"path/filepath"
|
|
||||||
"sync"
|
|
||||||
"unicode"
|
|
||||||
|
|
||||||
"golang.org/x/text/unicode/cldr"
|
|
||||||
)
|
|
||||||
|
|
||||||
var (
|
|
||||||
url = flag.String("url",
|
|
||||||
"http://www.unicode.org/Public",
|
|
||||||
"URL of Unicode database directory")
|
|
||||||
iana = flag.String("iana",
|
|
||||||
"http://www.iana.org",
|
|
||||||
"URL of the IANA repository")
|
|
||||||
unicodeVersion = flag.String("unicode",
|
|
||||||
getEnv("UNICODE_VERSION", unicode.Version),
|
|
||||||
"unicode version to use")
|
|
||||||
cldrVersion = flag.String("cldr",
|
|
||||||
getEnv("CLDR_VERSION", cldr.Version),
|
|
||||||
"cldr version to use")
|
|
||||||
)
|
|
||||||
|
|
||||||
func getEnv(name, def string) string {
|
|
||||||
if v := os.Getenv(name); v != "" {
|
|
||||||
return v
|
|
||||||
}
|
|
||||||
return def
|
|
||||||
}
|
|
||||||
|
|
||||||
// Init performs common initialization for a gen command. It parses the flags
|
|
||||||
// and sets up the standard logging parameters.
|
|
||||||
func Init() {
|
|
||||||
log.SetPrefix("")
|
|
||||||
log.SetFlags(log.Lshortfile)
|
|
||||||
flag.Parse()
|
|
||||||
}
|
|
||||||
|
|
||||||
const header = `// This file was generated by go generate; DO NOT EDIT
|
|
||||||
|
|
||||||
package %s
|
|
||||||
|
|
||||||
`
|
|
||||||
|
|
||||||
// UnicodeVersion reports the requested Unicode version.
|
|
||||||
func UnicodeVersion() string {
|
|
||||||
return *unicodeVersion
|
|
||||||
}
|
|
||||||
|
|
||||||
// UnicodeVersion reports the requested CLDR version.
|
|
||||||
func CLDRVersion() string {
|
|
||||||
return *cldrVersion
|
|
||||||
}
|
|
||||||
|
|
||||||
// IsLocal reports whether data files are available locally.
|
|
||||||
func IsLocal() bool {
|
|
||||||
dir, err := localReadmeFile()
|
|
||||||
if err != nil {
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
if _, err = os.Stat(dir); err != nil {
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
|
|
||||||
// OpenUCDFile opens the requested UCD file. The file is specified relative to
|
|
||||||
// the public Unicode root directory. It will call log.Fatal if there are any
|
|
||||||
// errors.
|
|
||||||
func OpenUCDFile(file string) io.ReadCloser {
|
|
||||||
return openUnicode(path.Join(*unicodeVersion, "ucd", file))
|
|
||||||
}
|
|
||||||
|
|
||||||
// OpenCLDRCoreZip opens the CLDR core zip file. It will call log.Fatal if there
|
|
||||||
// are any errors.
|
|
||||||
func OpenCLDRCoreZip() io.ReadCloser {
|
|
||||||
return OpenUnicodeFile("cldr", *cldrVersion, "core.zip")
|
|
||||||
}
|
|
||||||
|
|
||||||
// OpenUnicodeFile opens the requested file of the requested category from the
|
|
||||||
// root of the Unicode data archive. The file is specified relative to the
|
|
||||||
// public Unicode root directory. If version is "", it will use the default
|
|
||||||
// Unicode version. It will call log.Fatal if there are any errors.
|
|
||||||
func OpenUnicodeFile(category, version, file string) io.ReadCloser {
|
|
||||||
if version == "" {
|
|
||||||
version = UnicodeVersion()
|
|
||||||
}
|
|
||||||
return openUnicode(path.Join(category, version, file))
|
|
||||||
}
|
|
||||||
|
|
||||||
// OpenIANAFile opens the requested IANA file. The file is specified relative
|
|
||||||
// to the IANA root, which is typically either http://www.iana.org or the
|
|
||||||
// iana directory in the local mirror. It will call log.Fatal if there are any
|
|
||||||
// errors.
|
|
||||||
func OpenIANAFile(path string) io.ReadCloser {
|
|
||||||
return Open(*iana, "iana", path)
|
|
||||||
}
|
|
||||||
|
|
||||||
var (
|
|
||||||
dirMutex sync.Mutex
|
|
||||||
localDir string
|
|
||||||
)
|
|
||||||
|
|
||||||
const permissions = 0755
|
|
||||||
|
|
||||||
func localReadmeFile() (string, error) {
|
|
||||||
p, err := build.Import("golang.org/x/text", "", build.FindOnly)
|
|
||||||
if err != nil {
|
|
||||||
return "", fmt.Errorf("Could not locate package: %v", err)
|
|
||||||
}
|
|
||||||
return filepath.Join(p.Dir, "DATA", "README"), nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func getLocalDir() string {
|
|
||||||
dirMutex.Lock()
|
|
||||||
defer dirMutex.Unlock()
|
|
||||||
|
|
||||||
readme, err := localReadmeFile()
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
dir := filepath.Dir(readme)
|
|
||||||
if _, err := os.Stat(readme); err != nil {
|
|
||||||
if err := os.MkdirAll(dir, permissions); err != nil {
|
|
||||||
log.Fatalf("Could not create directory: %v", err)
|
|
||||||
}
|
|
||||||
ioutil.WriteFile(readme, []byte(readmeTxt), permissions)
|
|
||||||
}
|
|
||||||
return dir
|
|
||||||
}
|
|
||||||
|
|
||||||
const readmeTxt = `Generated by golang.org/x/text/internal/gen. DO NOT EDIT.
|
|
||||||
|
|
||||||
This directory contains downloaded files used to generate the various tables
|
|
||||||
in the golang.org/x/text subrepo.
|
|
||||||
|
|
||||||
Note that the language subtag repo (iana/assignments/language-subtag-registry)
|
|
||||||
and all other times in the iana subdirectory are not versioned and will need
|
|
||||||
to be periodically manually updated. The easiest way to do this is to remove
|
|
||||||
the entire iana directory. This is mostly of concern when updating the language
|
|
||||||
package.
|
|
||||||
`
|
|
||||||
|
|
||||||
// Open opens subdir/path if a local directory is specified and the file exists,
|
|
||||||
// where subdir is a directory relative to the local root, or fetches it from
|
|
||||||
// urlRoot/path otherwise. It will call log.Fatal if there are any errors.
|
|
||||||
func Open(urlRoot, subdir, path string) io.ReadCloser {
|
|
||||||
file := filepath.Join(getLocalDir(), subdir, filepath.FromSlash(path))
|
|
||||||
return open(file, urlRoot, path)
|
|
||||||
}
|
|
||||||
|
|
||||||
func openUnicode(path string) io.ReadCloser {
|
|
||||||
file := filepath.Join(getLocalDir(), filepath.FromSlash(path))
|
|
||||||
return open(file, *url, path)
|
|
||||||
}
|
|
||||||
|
|
||||||
// TODO: automatically periodically update non-versioned files.
|
|
||||||
|
|
||||||
func open(file, urlRoot, path string) io.ReadCloser {
|
|
||||||
if f, err := os.Open(file); err == nil {
|
|
||||||
return f
|
|
||||||
}
|
|
||||||
r := get(urlRoot, path)
|
|
||||||
defer r.Close()
|
|
||||||
b, err := ioutil.ReadAll(r)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatalf("Could not download file: %v", err)
|
|
||||||
}
|
|
||||||
os.MkdirAll(filepath.Dir(file), permissions)
|
|
||||||
if err := ioutil.WriteFile(file, b, permissions); err != nil {
|
|
||||||
log.Fatalf("Could not create file: %v", err)
|
|
||||||
}
|
|
||||||
return ioutil.NopCloser(bytes.NewReader(b))
|
|
||||||
}
|
|
||||||
|
|
||||||
func get(root, path string) io.ReadCloser {
|
|
||||||
url := root + "/" + path
|
|
||||||
fmt.Printf("Fetching %s...", url)
|
|
||||||
defer fmt.Println(" done.")
|
|
||||||
resp, err := http.Get(url)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatalf("HTTP GET: %v", err)
|
|
||||||
}
|
|
||||||
if resp.StatusCode != 200 {
|
|
||||||
log.Fatalf("Bad GET status for %q: %q", url, resp.Status)
|
|
||||||
}
|
|
||||||
return resp.Body
|
|
||||||
}
|
|
||||||
|
|
||||||
// TODO: use Write*Version in all applicable packages.
|
|
||||||
|
|
||||||
// WriteUnicodeVersion writes a constant for the Unicode version from which the
|
|
||||||
// tables are generated.
|
|
||||||
func WriteUnicodeVersion(w io.Writer) {
|
|
||||||
fmt.Fprintf(w, "// UnicodeVersion is the Unicode version from which the tables in this package are derived.\n")
|
|
||||||
fmt.Fprintf(w, "const UnicodeVersion = %q\n\n", UnicodeVersion())
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteCLDRVersion writes a constant for the CLDR version from which the
|
|
||||||
// tables are generated.
|
|
||||||
func WriteCLDRVersion(w io.Writer) {
|
|
||||||
fmt.Fprintf(w, "// CLDRVersion is the CLDR version from which the tables in this package are derived.\n")
|
|
||||||
fmt.Fprintf(w, "const CLDRVersion = %q\n\n", CLDRVersion())
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteGoFile prepends a standard file comment and package statement to the
|
|
||||||
// given bytes, applies gofmt, and writes them to a file with the given name.
|
|
||||||
// It will call log.Fatal if there are any errors.
|
|
||||||
func WriteGoFile(filename, pkg string, b []byte) {
|
|
||||||
w, err := os.Create(filename)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatalf("Could not create file %s: %v", filename, err)
|
|
||||||
}
|
|
||||||
defer w.Close()
|
|
||||||
if _, err = WriteGo(w, pkg, b); err != nil {
|
|
||||||
log.Fatalf("Error writing file %s: %v", filename, err)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// WriteGo prepends a standard file comment and package statement to the given
|
|
||||||
// bytes, applies gofmt, and writes them to w.
|
|
||||||
func WriteGo(w io.Writer, pkg string, b []byte) (n int, err error) {
|
|
||||||
src := []byte(fmt.Sprintf(header, pkg))
|
|
||||||
src = append(src, b...)
|
|
||||||
formatted, err := format.Source(src)
|
|
||||||
if err != nil {
|
|
||||||
// Print the generated code even in case of an error so that the
|
|
||||||
// returned error can be meaningfully interpreted.
|
|
||||||
n, _ = w.Write(src)
|
|
||||||
return n, err
|
|
||||||
}
|
|
||||||
return w.Write(formatted)
|
|
||||||
}
|
|
||||||
|
|
||||||
// Repackage rewrites a Go file from belonging to package main to belonging to
|
|
||||||
// the given package.
|
|
||||||
func Repackage(inFile, outFile, pkg string) {
|
|
||||||
src, err := ioutil.ReadFile(inFile)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatalf("reading %s: %v", inFile, err)
|
|
||||||
}
|
|
||||||
const toDelete = "package main\n\n"
|
|
||||||
i := bytes.Index(src, []byte(toDelete))
|
|
||||||
if i < 0 {
|
|
||||||
log.Fatalf("Could not find %q in %s.", toDelete, inFile)
|
|
||||||
}
|
|
||||||
w := &bytes.Buffer{}
|
|
||||||
w.Write(src[i+len(toDelete):])
|
|
||||||
WriteGoFile(outFile, pkg, w.Bytes())
|
|
||||||
}
|
|
|
@ -1,58 +0,0 @@
|
||||||
// Copyright 2014 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
package triegen
|
|
||||||
|
|
||||||
// This file defines Compacter and its implementations.
|
|
||||||
|
|
||||||
import "io"
|
|
||||||
|
|
||||||
// A Compacter generates an alternative, more space-efficient way to store a
|
|
||||||
// trie value block. A trie value block holds all possible values for the last
|
|
||||||
// byte of a UTF-8 encoded rune. Excluding ASCII characters, a trie value block
|
|
||||||
// always has 64 values, as a UTF-8 encoding ends with a byte in [0x80, 0xC0).
|
|
||||||
type Compacter interface {
|
|
||||||
// Size returns whether the Compacter could encode the given block as well
|
|
||||||
// as its size in case it can. len(v) is always 64.
|
|
||||||
Size(v []uint64) (sz int, ok bool)
|
|
||||||
|
|
||||||
// Store stores the block using the Compacter's compression method.
|
|
||||||
// It returns a handle with which the block can be retrieved.
|
|
||||||
// len(v) is always 64.
|
|
||||||
Store(v []uint64) uint32
|
|
||||||
|
|
||||||
// Print writes the data structures associated to the given store to w.
|
|
||||||
Print(w io.Writer) error
|
|
||||||
|
|
||||||
// Handler returns the name of a function that gets called during trie
|
|
||||||
// lookup for blocks generated by the Compacter. The function should be of
|
|
||||||
// the form func (n uint32, b byte) uint64, where n is the index returned by
|
|
||||||
// the Compacter's Store method and b is the last byte of the UTF-8
|
|
||||||
// encoding, where 0x80 <= b < 0xC0, for which to do the lookup in the
|
|
||||||
// block.
|
|
||||||
Handler() string
|
|
||||||
}
|
|
||||||
|
|
||||||
// simpleCompacter is the default Compacter used by builder. It implements a
|
|
||||||
// normal trie block.
|
|
||||||
type simpleCompacter builder
|
|
||||||
|
|
||||||
func (b *simpleCompacter) Size([]uint64) (sz int, ok bool) {
|
|
||||||
return blockSize * b.ValueSize, true
|
|
||||||
}
|
|
||||||
|
|
||||||
func (b *simpleCompacter) Store(v []uint64) uint32 {
|
|
||||||
h := uint32(len(b.ValueBlocks) - blockOffset)
|
|
||||||
b.ValueBlocks = append(b.ValueBlocks, v)
|
|
||||||
return h
|
|
||||||
}
|
|
||||||
|
|
||||||
func (b *simpleCompacter) Print(io.Writer) error {
|
|
||||||
// Structures are printed in print.go.
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func (b *simpleCompacter) Handler() string {
|
|
||||||
panic("Handler should be special-cased for this Compacter")
|
|
||||||
}
|
|
|
@ -1,251 +0,0 @@
|
||||||
// Copyright 2014 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
package triegen
|
|
||||||
|
|
||||||
import (
|
|
||||||
"bytes"
|
|
||||||
"fmt"
|
|
||||||
"io"
|
|
||||||
"strings"
|
|
||||||
"text/template"
|
|
||||||
)
|
|
||||||
|
|
||||||
// print writes all the data structures as well as the code necessary to use the
|
|
||||||
// trie to w.
|
|
||||||
func (b *builder) print(w io.Writer) error {
|
|
||||||
b.Stats.NValueEntries = len(b.ValueBlocks) * blockSize
|
|
||||||
b.Stats.NValueBytes = len(b.ValueBlocks) * blockSize * b.ValueSize
|
|
||||||
b.Stats.NIndexEntries = len(b.IndexBlocks) * blockSize
|
|
||||||
b.Stats.NIndexBytes = len(b.IndexBlocks) * blockSize * b.IndexSize
|
|
||||||
b.Stats.NHandleBytes = len(b.Trie) * 2 * b.IndexSize
|
|
||||||
|
|
||||||
// If we only have one root trie, all starter blocks are at position 0 and
|
|
||||||
// we can access the arrays directly.
|
|
||||||
if len(b.Trie) == 1 {
|
|
||||||
// At this point we cannot refer to the generated tables directly.
|
|
||||||
b.ASCIIBlock = b.Name + "Values"
|
|
||||||
b.StarterBlock = b.Name + "Index"
|
|
||||||
} else {
|
|
||||||
// Otherwise we need to have explicit starter indexes in the trie
|
|
||||||
// structure.
|
|
||||||
b.ASCIIBlock = "t.ascii"
|
|
||||||
b.StarterBlock = "t.utf8Start"
|
|
||||||
}
|
|
||||||
|
|
||||||
b.SourceType = "[]byte"
|
|
||||||
if err := lookupGen.Execute(w, b); err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
|
|
||||||
b.SourceType = "string"
|
|
||||||
if err := lookupGen.Execute(w, b); err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
|
|
||||||
if err := trieGen.Execute(w, b); err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
|
|
||||||
for _, c := range b.Compactions {
|
|
||||||
if err := c.c.Print(w); err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func printValues(n int, values []uint64) string {
|
|
||||||
w := &bytes.Buffer{}
|
|
||||||
boff := n * blockSize
|
|
||||||
fmt.Fprintf(w, "\t// Block %#x, offset %#x", n, boff)
|
|
||||||
var newline bool
|
|
||||||
for i, v := range values {
|
|
||||||
if i%6 == 0 {
|
|
||||||
newline = true
|
|
||||||
}
|
|
||||||
if v != 0 {
|
|
||||||
if newline {
|
|
||||||
fmt.Fprintf(w, "\n")
|
|
||||||
newline = false
|
|
||||||
}
|
|
||||||
fmt.Fprintf(w, "\t%#02x:%#04x, ", boff+i, v)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return w.String()
|
|
||||||
}
|
|
||||||
|
|
||||||
func printIndex(b *builder, nr int, n *node) string {
|
|
||||||
w := &bytes.Buffer{}
|
|
||||||
boff := nr * blockSize
|
|
||||||
fmt.Fprintf(w, "\t// Block %#x, offset %#x", nr, boff)
|
|
||||||
var newline bool
|
|
||||||
for i, c := range n.children {
|
|
||||||
if i%8 == 0 {
|
|
||||||
newline = true
|
|
||||||
}
|
|
||||||
if c != nil {
|
|
||||||
v := b.Compactions[c.index.compaction].Offset + uint32(c.index.index)
|
|
||||||
if v != 0 {
|
|
||||||
if newline {
|
|
||||||
fmt.Fprintf(w, "\n")
|
|
||||||
newline = false
|
|
||||||
}
|
|
||||||
fmt.Fprintf(w, "\t%#02x:%#02x, ", boff+i, v)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return w.String()
|
|
||||||
}
|
|
||||||
|
|
||||||
var (
|
|
||||||
trieGen = template.Must(template.New("trie").Funcs(template.FuncMap{
|
|
||||||
"printValues": printValues,
|
|
||||||
"printIndex": printIndex,
|
|
||||||
"title": strings.Title,
|
|
||||||
"dec": func(x int) int { return x - 1 },
|
|
||||||
"psize": func(n int) string {
|
|
||||||
return fmt.Sprintf("%d bytes (%.2f KiB)", n, float64(n)/1024)
|
|
||||||
},
|
|
||||||
}).Parse(trieTemplate))
|
|
||||||
lookupGen = template.Must(template.New("lookup").Parse(lookupTemplate))
|
|
||||||
)
|
|
||||||
|
|
||||||
// TODO: consider the return type of lookup. It could be uint64, even if the
|
|
||||||
// internal value type is smaller. We will have to verify this with the
|
|
||||||
// performance of unicode/norm, which is very sensitive to such changes.
|
|
||||||
const trieTemplate = `{{$b := .}}{{$multi := gt (len .Trie) 1}}
|
|
||||||
// {{.Name}}Trie. Total size: {{psize .Size}}. Checksum: {{printf "%08x" .Checksum}}.
|
|
||||||
type {{.Name}}Trie struct { {{if $multi}}
|
|
||||||
ascii []{{.ValueType}} // index for ASCII bytes
|
|
||||||
utf8Start []{{.IndexType}} // index for UTF-8 bytes >= 0xC0
|
|
||||||
{{end}}}
|
|
||||||
|
|
||||||
func new{{title .Name}}Trie(i int) *{{.Name}}Trie { {{if $multi}}
|
|
||||||
h := {{.Name}}TrieHandles[i]
|
|
||||||
return &{{.Name}}Trie{ {{.Name}}Values[uint32(h.ascii)<<6:], {{.Name}}Index[uint32(h.multi)<<6:] }
|
|
||||||
}
|
|
||||||
|
|
||||||
type {{.Name}}TrieHandle struct {
|
|
||||||
ascii, multi {{.IndexType}}
|
|
||||||
}
|
|
||||||
|
|
||||||
// {{.Name}}TrieHandles: {{len .Trie}} handles, {{.Stats.NHandleBytes}} bytes
|
|
||||||
var {{.Name}}TrieHandles = [{{len .Trie}}]{{.Name}}TrieHandle{
|
|
||||||
{{range .Trie}} { {{.ASCIIIndex}}, {{.StarterIndex}} }, // {{printf "%08x" .Checksum}}: {{.Name}}
|
|
||||||
{{end}}}{{else}}
|
|
||||||
return &{{.Name}}Trie{}
|
|
||||||
}
|
|
||||||
{{end}}
|
|
||||||
// lookupValue determines the type of block n and looks up the value for b.
|
|
||||||
func (t *{{.Name}}Trie) lookupValue(n uint32, b byte) {{.ValueType}}{{$last := dec (len .Compactions)}} {
|
|
||||||
switch { {{range $i, $c := .Compactions}}
|
|
||||||
{{if eq $i $last}}default{{else}}case n < {{$c.Cutoff}}{{end}}:{{if ne $i 0}}
|
|
||||||
n -= {{$c.Offset}}{{end}}
|
|
||||||
return {{print $b.ValueType}}({{$c.Handler}}){{end}}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// {{.Name}}Values: {{len .ValueBlocks}} blocks, {{.Stats.NValueEntries}} entries, {{.Stats.NValueBytes}} bytes
|
|
||||||
// The third block is the zero block.
|
|
||||||
var {{.Name}}Values = [{{.Stats.NValueEntries}}]{{.ValueType}} {
|
|
||||||
{{range $i, $v := .ValueBlocks}}{{printValues $i $v}}
|
|
||||||
{{end}}}
|
|
||||||
|
|
||||||
// {{.Name}}Index: {{len .IndexBlocks}} blocks, {{.Stats.NIndexEntries}} entries, {{.Stats.NIndexBytes}} bytes
|
|
||||||
// Block 0 is the zero block.
|
|
||||||
var {{.Name}}Index = [{{.Stats.NIndexEntries}}]{{.IndexType}} {
|
|
||||||
{{range $i, $v := .IndexBlocks}}{{printIndex $b $i $v}}
|
|
||||||
{{end}}}
|
|
||||||
`
|
|
||||||
|
|
||||||
// TODO: consider allowing zero-length strings after evaluating performance with
|
|
||||||
// unicode/norm.
|
|
||||||
const lookupTemplate = `
|
|
||||||
// lookup{{if eq .SourceType "string"}}String{{end}} returns the trie value for the first UTF-8 encoding in s and
|
|
||||||
// the width in bytes of this encoding. The size will be 0 if s does not
|
|
||||||
// hold enough bytes to complete the encoding. len(s) must be greater than 0.
|
|
||||||
func (t *{{.Name}}Trie) lookup{{if eq .SourceType "string"}}String{{end}}(s {{.SourceType}}) (v {{.ValueType}}, sz int) {
|
|
||||||
c0 := s[0]
|
|
||||||
switch {
|
|
||||||
case c0 < 0x80: // is ASCII
|
|
||||||
return {{.ASCIIBlock}}[c0], 1
|
|
||||||
case c0 < 0xC2:
|
|
||||||
return 0, 1 // Illegal UTF-8: not a starter, not ASCII.
|
|
||||||
case c0 < 0xE0: // 2-byte UTF-8
|
|
||||||
if len(s) < 2 {
|
|
||||||
return 0, 0
|
|
||||||
}
|
|
||||||
i := {{.StarterBlock}}[c0]
|
|
||||||
c1 := s[1]
|
|
||||||
if c1 < 0x80 || 0xC0 <= c1 {
|
|
||||||
return 0, 1 // Illegal UTF-8: not a continuation byte.
|
|
||||||
}
|
|
||||||
return t.lookupValue(uint32(i), c1), 2
|
|
||||||
case c0 < 0xF0: // 3-byte UTF-8
|
|
||||||
if len(s) < 3 {
|
|
||||||
return 0, 0
|
|
||||||
}
|
|
||||||
i := {{.StarterBlock}}[c0]
|
|
||||||
c1 := s[1]
|
|
||||||
if c1 < 0x80 || 0xC0 <= c1 {
|
|
||||||
return 0, 1 // Illegal UTF-8: not a continuation byte.
|
|
||||||
}
|
|
||||||
o := uint32(i)<<6 + uint32(c1)
|
|
||||||
i = {{.Name}}Index[o]
|
|
||||||
c2 := s[2]
|
|
||||||
if c2 < 0x80 || 0xC0 <= c2 {
|
|
||||||
return 0, 2 // Illegal UTF-8: not a continuation byte.
|
|
||||||
}
|
|
||||||
return t.lookupValue(uint32(i), c2), 3
|
|
||||||
case c0 < 0xF8: // 4-byte UTF-8
|
|
||||||
if len(s) < 4 {
|
|
||||||
return 0, 0
|
|
||||||
}
|
|
||||||
i := {{.StarterBlock}}[c0]
|
|
||||||
c1 := s[1]
|
|
||||||
if c1 < 0x80 || 0xC0 <= c1 {
|
|
||||||
return 0, 1 // Illegal UTF-8: not a continuation byte.
|
|
||||||
}
|
|
||||||
o := uint32(i)<<6 + uint32(c1)
|
|
||||||
i = {{.Name}}Index[o]
|
|
||||||
c2 := s[2]
|
|
||||||
if c2 < 0x80 || 0xC0 <= c2 {
|
|
||||||
return 0, 2 // Illegal UTF-8: not a continuation byte.
|
|
||||||
}
|
|
||||||
o = uint32(i)<<6 + uint32(c2)
|
|
||||||
i = {{.Name}}Index[o]
|
|
||||||
c3 := s[3]
|
|
||||||
if c3 < 0x80 || 0xC0 <= c3 {
|
|
||||||
return 0, 3 // Illegal UTF-8: not a continuation byte.
|
|
||||||
}
|
|
||||||
return t.lookupValue(uint32(i), c3), 4
|
|
||||||
}
|
|
||||||
// Illegal rune
|
|
||||||
return 0, 1
|
|
||||||
}
|
|
||||||
|
|
||||||
// lookup{{if eq .SourceType "string"}}String{{end}}Unsafe returns the trie value for the first UTF-8 encoding in s.
|
|
||||||
// s must start with a full and valid UTF-8 encoded rune.
|
|
||||||
func (t *{{.Name}}Trie) lookup{{if eq .SourceType "string"}}String{{end}}Unsafe(s {{.SourceType}}) {{.ValueType}} {
|
|
||||||
c0 := s[0]
|
|
||||||
if c0 < 0x80 { // is ASCII
|
|
||||||
return {{.ASCIIBlock}}[c0]
|
|
||||||
}
|
|
||||||
i := {{.StarterBlock}}[c0]
|
|
||||||
if c0 < 0xE0 { // 2-byte UTF-8
|
|
||||||
return t.lookupValue(uint32(i), s[1])
|
|
||||||
}
|
|
||||||
i = {{.Name}}Index[uint32(i)<<6+uint32(s[1])]
|
|
||||||
if c0 < 0xF0 { // 3-byte UTF-8
|
|
||||||
return t.lookupValue(uint32(i), s[2])
|
|
||||||
}
|
|
||||||
i = {{.Name}}Index[uint32(i)<<6+uint32(s[2])]
|
|
||||||
if c0 < 0xF8 { // 4-byte UTF-8
|
|
||||||
return t.lookupValue(uint32(i), s[3])
|
|
||||||
}
|
|
||||||
return 0
|
|
||||||
}
|
|
||||||
`
|
|
|
@ -1,494 +0,0 @@
|
||||||
// Copyright 2014 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// Package triegen implements a code generator for a trie for associating
|
|
||||||
// unsigned integer values with UTF-8 encoded runes.
|
|
||||||
//
|
|
||||||
// Many of the go.text packages use tries for storing per-rune information. A
|
|
||||||
// trie is especially useful if many of the runes have the same value. If this
|
|
||||||
// is the case, many blocks can be expected to be shared allowing for
|
|
||||||
// information on many runes to be stored in little space.
|
|
||||||
//
|
|
||||||
// As most of the lookups are done directly on []byte slices, the tries use the
|
|
||||||
// UTF-8 bytes directly for the lookup. This saves a conversion from UTF-8 to
|
|
||||||
// runes and contributes a little bit to better performance. It also naturally
|
|
||||||
// provides a fast path for ASCII.
|
|
||||||
//
|
|
||||||
// Space is also an issue. There are many code points defined in Unicode and as
|
|
||||||
// a result tables can get quite large. So every byte counts. The triegen
|
|
||||||
// package automatically chooses the smallest integer values to represent the
|
|
||||||
// tables. Compacters allow further compression of the trie by allowing for
|
|
||||||
// alternative representations of individual trie blocks.
|
|
||||||
//
|
|
||||||
// triegen allows generating multiple tries as a single structure. This is
|
|
||||||
// useful when, for example, one wants to generate tries for several languages
|
|
||||||
// that have a lot of values in common. Some existing libraries for
|
|
||||||
// internationalization store all per-language data as a dynamically loadable
|
|
||||||
// chunk. The go.text packages are designed with the assumption that the user
|
|
||||||
// typically wants to compile in support for all supported languages, in line
|
|
||||||
// with the approach common to Go to create a single standalone binary. The
|
|
||||||
// multi-root trie approach can give significant storage savings in this
|
|
||||||
// scenario.
|
|
||||||
//
|
|
||||||
// triegen generates both tables and code. The code is optimized to use the
|
|
||||||
// automatically chosen data types. The following code is generated for a Trie
|
|
||||||
// or multiple Tries named "foo":
|
|
||||||
// - type fooTrie
|
|
||||||
// The trie type.
|
|
||||||
//
|
|
||||||
// - func newFooTrie(x int) *fooTrie
|
|
||||||
// Trie constructor, where x is the index of the trie passed to Gen.
|
|
||||||
//
|
|
||||||
// - func (t *fooTrie) lookup(s []byte) (v uintX, sz int)
|
|
||||||
// The lookup method, where uintX is automatically chosen.
|
|
||||||
//
|
|
||||||
// - func lookupString, lookupUnsafe and lookupStringUnsafe
|
|
||||||
// Variants of the above.
|
|
||||||
//
|
|
||||||
// - var fooValues and fooIndex and any tables generated by Compacters.
|
|
||||||
// The core trie data.
|
|
||||||
//
|
|
||||||
// - var fooTrieHandles
|
|
||||||
// Indexes of starter blocks in case of multiple trie roots.
|
|
||||||
//
|
|
||||||
// It is recommended that users test the generated trie by checking the returned
|
|
||||||
// value for every rune. Such exhaustive tests are possible as the the number of
|
|
||||||
// runes in Unicode is limited.
|
|
||||||
package triegen // import "golang.org/x/text/internal/triegen"
|
|
||||||
|
|
||||||
// TODO: Arguably, the internally optimized data types would not have to be
|
|
||||||
// exposed in the generated API. We could also investigate not generating the
|
|
||||||
// code, but using it through a package. We would have to investigate the impact
|
|
||||||
// on performance of making such change, though. For packages like unicode/norm,
|
|
||||||
// small changes like this could tank performance.
|
|
||||||
|
|
||||||
import (
|
|
||||||
"encoding/binary"
|
|
||||||
"fmt"
|
|
||||||
"hash/crc64"
|
|
||||||
"io"
|
|
||||||
"log"
|
|
||||||
"unicode/utf8"
|
|
||||||
)
|
|
||||||
|
|
||||||
// builder builds a set of tries for associating values with runes. The set of
|
|
||||||
// tries can share common index and value blocks.
|
|
||||||
type builder struct {
|
|
||||||
Name string
|
|
||||||
|
|
||||||
// ValueType is the type of the trie values looked up.
|
|
||||||
ValueType string
|
|
||||||
|
|
||||||
// ValueSize is the byte size of the ValueType.
|
|
||||||
ValueSize int
|
|
||||||
|
|
||||||
// IndexType is the type of trie index values used for all UTF-8 bytes of
|
|
||||||
// a rune except the last one.
|
|
||||||
IndexType string
|
|
||||||
|
|
||||||
// IndexSize is the byte size of the IndexType.
|
|
||||||
IndexSize int
|
|
||||||
|
|
||||||
// SourceType is used when generating the lookup functions. If the user
|
|
||||||
// requests StringSupport, all lookup functions will be generated for
|
|
||||||
// string input as well.
|
|
||||||
SourceType string
|
|
||||||
|
|
||||||
Trie []*Trie
|
|
||||||
|
|
||||||
IndexBlocks []*node
|
|
||||||
ValueBlocks [][]uint64
|
|
||||||
Compactions []compaction
|
|
||||||
Checksum uint64
|
|
||||||
|
|
||||||
ASCIIBlock string
|
|
||||||
StarterBlock string
|
|
||||||
|
|
||||||
indexBlockIdx map[uint64]int
|
|
||||||
valueBlockIdx map[uint64]nodeIndex
|
|
||||||
asciiBlockIdx map[uint64]int
|
|
||||||
|
|
||||||
// Stats are used to fill out the template.
|
|
||||||
Stats struct {
|
|
||||||
NValueEntries int
|
|
||||||
NValueBytes int
|
|
||||||
NIndexEntries int
|
|
||||||
NIndexBytes int
|
|
||||||
NHandleBytes int
|
|
||||||
}
|
|
||||||
|
|
||||||
err error
|
|
||||||
}
|
|
||||||
|
|
||||||
// A nodeIndex encodes the index of a node, which is defined by the compaction
|
|
||||||
// which stores it and an index within the compaction. For internal nodes, the
|
|
||||||
// compaction is always 0.
|
|
||||||
type nodeIndex struct {
|
|
||||||
compaction int
|
|
||||||
index int
|
|
||||||
}
|
|
||||||
|
|
||||||
// compaction keeps track of stats used for the compaction.
|
|
||||||
type compaction struct {
|
|
||||||
c Compacter
|
|
||||||
blocks []*node
|
|
||||||
maxHandle uint32
|
|
||||||
totalSize int
|
|
||||||
|
|
||||||
// Used by template-based generator and thus exported.
|
|
||||||
Cutoff uint32
|
|
||||||
Offset uint32
|
|
||||||
Handler string
|
|
||||||
}
|
|
||||||
|
|
||||||
func (b *builder) setError(err error) {
|
|
||||||
if b.err == nil {
|
|
||||||
b.err = err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// An Option can be passed to Gen.
|
|
||||||
type Option func(b *builder) error
|
|
||||||
|
|
||||||
// Compact configures the trie generator to use the given Compacter.
|
|
||||||
func Compact(c Compacter) Option {
|
|
||||||
return func(b *builder) error {
|
|
||||||
b.Compactions = append(b.Compactions, compaction{
|
|
||||||
c: c,
|
|
||||||
Handler: c.Handler() + "(n, b)"})
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Gen writes Go code for a shared trie lookup structure to w for the given
|
|
||||||
// Tries. The generated trie type will be called nameTrie. newNameTrie(x) will
|
|
||||||
// return the *nameTrie for tries[x]. A value can be looked up by using one of
|
|
||||||
// the various lookup methods defined on nameTrie. It returns the table size of
|
|
||||||
// the generated trie.
|
|
||||||
func Gen(w io.Writer, name string, tries []*Trie, opts ...Option) (sz int, err error) {
|
|
||||||
// The index contains two dummy blocks, followed by the zero block. The zero
|
|
||||||
// block is at offset 0x80, so that the offset for the zero block for
|
|
||||||
// continuation bytes is 0.
|
|
||||||
b := &builder{
|
|
||||||
Name: name,
|
|
||||||
Trie: tries,
|
|
||||||
IndexBlocks: []*node{{}, {}, {}},
|
|
||||||
Compactions: []compaction{{
|
|
||||||
Handler: name + "Values[n<<6+uint32(b)]",
|
|
||||||
}},
|
|
||||||
// The 0 key in indexBlockIdx and valueBlockIdx is the hash of the zero
|
|
||||||
// block.
|
|
||||||
indexBlockIdx: map[uint64]int{0: 0},
|
|
||||||
valueBlockIdx: map[uint64]nodeIndex{0: {}},
|
|
||||||
asciiBlockIdx: map[uint64]int{},
|
|
||||||
}
|
|
||||||
b.Compactions[0].c = (*simpleCompacter)(b)
|
|
||||||
|
|
||||||
for _, f := range opts {
|
|
||||||
if err := f(b); err != nil {
|
|
||||||
return 0, err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
b.build()
|
|
||||||
if b.err != nil {
|
|
||||||
return 0, b.err
|
|
||||||
}
|
|
||||||
if err = b.print(w); err != nil {
|
|
||||||
return 0, err
|
|
||||||
}
|
|
||||||
return b.Size(), nil
|
|
||||||
}
|
|
||||||
|
|
||||||
// A Trie represents a single root node of a trie. A builder may build several
|
|
||||||
// overlapping tries at once.
|
|
||||||
type Trie struct {
|
|
||||||
root *node
|
|
||||||
|
|
||||||
hiddenTrie
|
|
||||||
}
|
|
||||||
|
|
||||||
// hiddenTrie contains values we want to be visible to the template generator,
|
|
||||||
// but hidden from the API documentation.
|
|
||||||
type hiddenTrie struct {
|
|
||||||
Name string
|
|
||||||
Checksum uint64
|
|
||||||
ASCIIIndex int
|
|
||||||
StarterIndex int
|
|
||||||
}
|
|
||||||
|
|
||||||
// NewTrie returns a new trie root.
|
|
||||||
func NewTrie(name string) *Trie {
|
|
||||||
return &Trie{
|
|
||||||
&node{
|
|
||||||
children: make([]*node, blockSize),
|
|
||||||
values: make([]uint64, utf8.RuneSelf),
|
|
||||||
},
|
|
||||||
hiddenTrie{Name: name},
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Gen is a convenience wrapper around the Gen func passing t as the only trie
|
|
||||||
// and uses the name passed to NewTrie. It returns the size of the generated
|
|
||||||
// tables.
|
|
||||||
func (t *Trie) Gen(w io.Writer, opts ...Option) (sz int, err error) {
|
|
||||||
return Gen(w, t.Name, []*Trie{t}, opts...)
|
|
||||||
}
|
|
||||||
|
|
||||||
// node is a node of the intermediate trie structure.
|
|
||||||
type node struct {
|
|
||||||
// children holds this node's children. It is always of length 64.
|
|
||||||
// A child node may be nil.
|
|
||||||
children []*node
|
|
||||||
|
|
||||||
// values contains the values of this node. If it is non-nil, this node is
|
|
||||||
// either a root or leaf node:
|
|
||||||
// For root nodes, len(values) == 128 and it maps the bytes in [0x00, 0x7F].
|
|
||||||
// For leaf nodes, len(values) == 64 and it maps the bytes in [0x80, 0xBF].
|
|
||||||
values []uint64
|
|
||||||
|
|
||||||
index nodeIndex
|
|
||||||
}
|
|
||||||
|
|
||||||
// Insert associates value with the given rune. Insert will panic if a non-zero
|
|
||||||
// value is passed for an invalid rune.
|
|
||||||
func (t *Trie) Insert(r rune, value uint64) {
|
|
||||||
if value == 0 {
|
|
||||||
return
|
|
||||||
}
|
|
||||||
s := string(r)
|
|
||||||
if []rune(s)[0] != r && value != 0 {
|
|
||||||
// Note: The UCD tables will always assign what amounts to a zero value
|
|
||||||
// to a surrogate. Allowing a zero value for an illegal rune allows
|
|
||||||
// users to iterate over [0..MaxRune] without having to explicitly
|
|
||||||
// exclude surrogates, which would be tedious.
|
|
||||||
panic(fmt.Sprintf("triegen: non-zero value for invalid rune %U", r))
|
|
||||||
}
|
|
||||||
if len(s) == 1 {
|
|
||||||
// It is a root node value (ASCII).
|
|
||||||
t.root.values[s[0]] = value
|
|
||||||
return
|
|
||||||
}
|
|
||||||
|
|
||||||
n := t.root
|
|
||||||
for ; len(s) > 1; s = s[1:] {
|
|
||||||
if n.children == nil {
|
|
||||||
n.children = make([]*node, blockSize)
|
|
||||||
}
|
|
||||||
p := s[0] % blockSize
|
|
||||||
c := n.children[p]
|
|
||||||
if c == nil {
|
|
||||||
c = &node{}
|
|
||||||
n.children[p] = c
|
|
||||||
}
|
|
||||||
if len(s) > 2 && c.values != nil {
|
|
||||||
log.Fatalf("triegen: insert(%U): found internal node with values", r)
|
|
||||||
}
|
|
||||||
n = c
|
|
||||||
}
|
|
||||||
if n.values == nil {
|
|
||||||
n.values = make([]uint64, blockSize)
|
|
||||||
}
|
|
||||||
if n.children != nil {
|
|
||||||
log.Fatalf("triegen: insert(%U): found leaf node that also has child nodes", r)
|
|
||||||
}
|
|
||||||
n.values[s[0]-0x80] = value
|
|
||||||
}
|
|
||||||
|
|
||||||
// Size returns the number of bytes the generated trie will take to store. It
|
|
||||||
// needs to be exported as it is used in the templates.
|
|
||||||
func (b *builder) Size() int {
|
|
||||||
// Index blocks.
|
|
||||||
sz := len(b.IndexBlocks) * blockSize * b.IndexSize
|
|
||||||
|
|
||||||
// Skip the first compaction, which represents the normal value blocks, as
|
|
||||||
// its totalSize does not account for the ASCII blocks, which are managed
|
|
||||||
// separately.
|
|
||||||
sz += len(b.ValueBlocks) * blockSize * b.ValueSize
|
|
||||||
for _, c := range b.Compactions[1:] {
|
|
||||||
sz += c.totalSize
|
|
||||||
}
|
|
||||||
|
|
||||||
// TODO: this computation does not account for the fixed overhead of a using
|
|
||||||
// a compaction, either code or data. As for data, though, the typical
|
|
||||||
// overhead of data is in the order of bytes (2 bytes for cases). Further,
|
|
||||||
// the savings of using a compaction should anyway be substantial for it to
|
|
||||||
// be worth it.
|
|
||||||
|
|
||||||
// For multi-root tries, we also need to account for the handles.
|
|
||||||
if len(b.Trie) > 1 {
|
|
||||||
sz += 2 * b.IndexSize * len(b.Trie)
|
|
||||||
}
|
|
||||||
return sz
|
|
||||||
}
|
|
||||||
|
|
||||||
func (b *builder) build() {
|
|
||||||
// Compute the sizes of the values.
|
|
||||||
var vmax uint64
|
|
||||||
for _, t := range b.Trie {
|
|
||||||
vmax = maxValue(t.root, vmax)
|
|
||||||
}
|
|
||||||
b.ValueType, b.ValueSize = getIntType(vmax)
|
|
||||||
|
|
||||||
// Compute all block allocations.
|
|
||||||
// TODO: first compute the ASCII blocks for all tries and then the other
|
|
||||||
// nodes. ASCII blocks are more restricted in placement, as they require two
|
|
||||||
// blocks to be placed consecutively. Processing them first may improve
|
|
||||||
// sharing (at least one zero block can be expected to be saved.)
|
|
||||||
for _, t := range b.Trie {
|
|
||||||
b.Checksum += b.buildTrie(t)
|
|
||||||
}
|
|
||||||
|
|
||||||
// Compute the offsets for all the Compacters.
|
|
||||||
offset := uint32(0)
|
|
||||||
for i := range b.Compactions {
|
|
||||||
c := &b.Compactions[i]
|
|
||||||
c.Offset = offset
|
|
||||||
offset += c.maxHandle + 1
|
|
||||||
c.Cutoff = offset
|
|
||||||
}
|
|
||||||
|
|
||||||
// Compute the sizes of indexes.
|
|
||||||
// TODO: different byte positions could have different sizes. So far we have
|
|
||||||
// not found a case where this is beneficial.
|
|
||||||
imax := uint64(b.Compactions[len(b.Compactions)-1].Cutoff)
|
|
||||||
for _, ib := range b.IndexBlocks {
|
|
||||||
if x := uint64(ib.index.index); x > imax {
|
|
||||||
imax = x
|
|
||||||
}
|
|
||||||
}
|
|
||||||
b.IndexType, b.IndexSize = getIntType(imax)
|
|
||||||
}
|
|
||||||
|
|
||||||
func maxValue(n *node, max uint64) uint64 {
|
|
||||||
if n == nil {
|
|
||||||
return max
|
|
||||||
}
|
|
||||||
for _, c := range n.children {
|
|
||||||
max = maxValue(c, max)
|
|
||||||
}
|
|
||||||
for _, v := range n.values {
|
|
||||||
if max < v {
|
|
||||||
max = v
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return max
|
|
||||||
}
|
|
||||||
|
|
||||||
func getIntType(v uint64) (string, int) {
|
|
||||||
switch {
|
|
||||||
case v < 1<<8:
|
|
||||||
return "uint8", 1
|
|
||||||
case v < 1<<16:
|
|
||||||
return "uint16", 2
|
|
||||||
case v < 1<<32:
|
|
||||||
return "uint32", 4
|
|
||||||
}
|
|
||||||
return "uint64", 8
|
|
||||||
}
|
|
||||||
|
|
||||||
const (
|
|
||||||
blockSize = 64
|
|
||||||
|
|
||||||
// Subtract two blocks to offset 0x80, the first continuation byte.
|
|
||||||
blockOffset = 2
|
|
||||||
|
|
||||||
// Subtract three blocks to offset 0xC0, the first non-ASCII starter.
|
|
||||||
rootBlockOffset = 3
|
|
||||||
)
|
|
||||||
|
|
||||||
var crcTable = crc64.MakeTable(crc64.ISO)
|
|
||||||
|
|
||||||
func (b *builder) buildTrie(t *Trie) uint64 {
|
|
||||||
n := t.root
|
|
||||||
|
|
||||||
// Get the ASCII offset. For the first trie, the ASCII block will be at
|
|
||||||
// position 0.
|
|
||||||
hasher := crc64.New(crcTable)
|
|
||||||
binary.Write(hasher, binary.BigEndian, n.values)
|
|
||||||
hash := hasher.Sum64()
|
|
||||||
|
|
||||||
v, ok := b.asciiBlockIdx[hash]
|
|
||||||
if !ok {
|
|
||||||
v = len(b.ValueBlocks)
|
|
||||||
b.asciiBlockIdx[hash] = v
|
|
||||||
|
|
||||||
b.ValueBlocks = append(b.ValueBlocks, n.values[:blockSize], n.values[blockSize:])
|
|
||||||
if v == 0 {
|
|
||||||
// Add the zero block at position 2 so that it will be assigned a
|
|
||||||
// zero reference in the lookup blocks.
|
|
||||||
// TODO: always do this? This would allow us to remove a check from
|
|
||||||
// the trie lookup, but at the expense of extra space. Analyze
|
|
||||||
// performance for unicode/norm.
|
|
||||||
b.ValueBlocks = append(b.ValueBlocks, make([]uint64, blockSize))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
t.ASCIIIndex = v
|
|
||||||
|
|
||||||
// Compute remaining offsets.
|
|
||||||
t.Checksum = b.computeOffsets(n, true)
|
|
||||||
// We already subtracted the normal blockOffset from the index. Subtract the
|
|
||||||
// difference for starter bytes.
|
|
||||||
t.StarterIndex = n.index.index - (rootBlockOffset - blockOffset)
|
|
||||||
return t.Checksum
|
|
||||||
}
|
|
||||||
|
|
||||||
func (b *builder) computeOffsets(n *node, root bool) uint64 {
|
|
||||||
// For the first trie, the root lookup block will be at position 3, which is
|
|
||||||
// the offset for UTF-8 non-ASCII starter bytes.
|
|
||||||
first := len(b.IndexBlocks) == rootBlockOffset
|
|
||||||
if first {
|
|
||||||
b.IndexBlocks = append(b.IndexBlocks, n)
|
|
||||||
}
|
|
||||||
|
|
||||||
// We special-case the cases where all values recursively are 0. This allows
|
|
||||||
// for the use of a zero block to which all such values can be directed.
|
|
||||||
hash := uint64(0)
|
|
||||||
if n.children != nil || n.values != nil {
|
|
||||||
hasher := crc64.New(crcTable)
|
|
||||||
for _, c := range n.children {
|
|
||||||
var v uint64
|
|
||||||
if c != nil {
|
|
||||||
v = b.computeOffsets(c, false)
|
|
||||||
}
|
|
||||||
binary.Write(hasher, binary.BigEndian, v)
|
|
||||||
}
|
|
||||||
binary.Write(hasher, binary.BigEndian, n.values)
|
|
||||||
hash = hasher.Sum64()
|
|
||||||
}
|
|
||||||
|
|
||||||
if first {
|
|
||||||
b.indexBlockIdx[hash] = rootBlockOffset - blockOffset
|
|
||||||
}
|
|
||||||
|
|
||||||
// Compacters don't apply to internal nodes.
|
|
||||||
if n.children != nil {
|
|
||||||
v, ok := b.indexBlockIdx[hash]
|
|
||||||
if !ok {
|
|
||||||
v = len(b.IndexBlocks) - blockOffset
|
|
||||||
b.IndexBlocks = append(b.IndexBlocks, n)
|
|
||||||
b.indexBlockIdx[hash] = v
|
|
||||||
}
|
|
||||||
n.index = nodeIndex{0, v}
|
|
||||||
} else {
|
|
||||||
h, ok := b.valueBlockIdx[hash]
|
|
||||||
if !ok {
|
|
||||||
bestI, bestSize := 0, blockSize*b.ValueSize
|
|
||||||
for i, c := range b.Compactions[1:] {
|
|
||||||
if sz, ok := c.c.Size(n.values); ok && bestSize > sz {
|
|
||||||
bestI, bestSize = i+1, sz
|
|
||||||
}
|
|
||||||
}
|
|
||||||
c := &b.Compactions[bestI]
|
|
||||||
c.totalSize += bestSize
|
|
||||||
v := c.c.Store(n.values)
|
|
||||||
if c.maxHandle < v {
|
|
||||||
c.maxHandle = v
|
|
||||||
}
|
|
||||||
h = nodeIndex{bestI, int(v)}
|
|
||||||
b.valueBlockIdx[hash] = h
|
|
||||||
}
|
|
||||||
n.index = h
|
|
||||||
}
|
|
||||||
return hash
|
|
||||||
}
|
|
|
@ -1,376 +0,0 @@
|
||||||
// Copyright 2014 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// Package ucd provides a parser for Unicode Character Database files, the
|
|
||||||
// format of which is defined in http://www.unicode.org/reports/tr44/. See
|
|
||||||
// http://www.unicode.org/Public/UCD/latest/ucd/ for example files.
|
|
||||||
//
|
|
||||||
// It currently does not support substitutions of missing fields.
|
|
||||||
package ucd // import "golang.org/x/text/internal/ucd"
|
|
||||||
|
|
||||||
import (
|
|
||||||
"bufio"
|
|
||||||
"bytes"
|
|
||||||
"errors"
|
|
||||||
"io"
|
|
||||||
"log"
|
|
||||||
"regexp"
|
|
||||||
"strconv"
|
|
||||||
"strings"
|
|
||||||
)
|
|
||||||
|
|
||||||
// UnicodeData.txt fields.
|
|
||||||
const (
|
|
||||||
CodePoint = iota
|
|
||||||
Name
|
|
||||||
GeneralCategory
|
|
||||||
CanonicalCombiningClass
|
|
||||||
BidiClass
|
|
||||||
DecompMapping
|
|
||||||
DecimalValue
|
|
||||||
DigitValue
|
|
||||||
NumericValue
|
|
||||||
BidiMirrored
|
|
||||||
Unicode1Name
|
|
||||||
ISOComment
|
|
||||||
SimpleUppercaseMapping
|
|
||||||
SimpleLowercaseMapping
|
|
||||||
SimpleTitlecaseMapping
|
|
||||||
)
|
|
||||||
|
|
||||||
// Parse calls f for each entry in the given reader of a UCD file. It will close
|
|
||||||
// the reader upon return. It will call log.Fatal if any error occurred.
|
|
||||||
//
|
|
||||||
// This implements the most common usage pattern of using Parser.
|
|
||||||
func Parse(r io.ReadCloser, f func(p *Parser)) {
|
|
||||||
defer r.Close()
|
|
||||||
|
|
||||||
p := New(r)
|
|
||||||
for p.Next() {
|
|
||||||
f(p)
|
|
||||||
}
|
|
||||||
if err := p.Err(); err != nil {
|
|
||||||
r.Close() // os.Exit will cause defers not to be called.
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// An Option is used to configure a Parser.
|
|
||||||
type Option func(p *Parser)
|
|
||||||
|
|
||||||
func keepRanges(p *Parser) {
|
|
||||||
p.keepRanges = true
|
|
||||||
}
|
|
||||||
|
|
||||||
var (
|
|
||||||
// KeepRanges prevents the expansion of ranges. The raw ranges can be
|
|
||||||
// obtained by calling Range(0) on the parser.
|
|
||||||
KeepRanges Option = keepRanges
|
|
||||||
)
|
|
||||||
|
|
||||||
// The Part option register a handler for lines starting with a '@'. The text
|
|
||||||
// after a '@' is available as the first field. Comments are handled as usual.
|
|
||||||
func Part(f func(p *Parser)) Option {
|
|
||||||
return func(p *Parser) {
|
|
||||||
p.partHandler = f
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// The CommentHandler option passes comments that are on a line by itself to
|
|
||||||
// a given handler.
|
|
||||||
func CommentHandler(f func(s string)) Option {
|
|
||||||
return func(p *Parser) {
|
|
||||||
p.commentHandler = f
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// A Parser parses Unicode Character Database (UCD) files.
|
|
||||||
type Parser struct {
|
|
||||||
scanner *bufio.Scanner
|
|
||||||
|
|
||||||
keepRanges bool // Don't expand rune ranges in field 0.
|
|
||||||
|
|
||||||
err error
|
|
||||||
comment []byte
|
|
||||||
field [][]byte
|
|
||||||
// parsedRange is needed in case Range(0) is called more than once for one
|
|
||||||
// field. In some cases this requires scanning ahead.
|
|
||||||
parsedRange bool
|
|
||||||
rangeStart, rangeEnd rune
|
|
||||||
|
|
||||||
partHandler func(p *Parser)
|
|
||||||
commentHandler func(s string)
|
|
||||||
}
|
|
||||||
|
|
||||||
func (p *Parser) setError(err error) {
|
|
||||||
if p.err == nil {
|
|
||||||
p.err = err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
func (p *Parser) getField(i int) []byte {
|
|
||||||
if i >= len(p.field) {
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
return p.field[i]
|
|
||||||
}
|
|
||||||
|
|
||||||
// Err returns a non-nil error if any error occurred during parsing.
|
|
||||||
func (p *Parser) Err() error {
|
|
||||||
return p.err
|
|
||||||
}
|
|
||||||
|
|
||||||
// New returns a Parser for the given Reader.
|
|
||||||
func New(r io.Reader, o ...Option) *Parser {
|
|
||||||
p := &Parser{
|
|
||||||
scanner: bufio.NewScanner(r),
|
|
||||||
}
|
|
||||||
for _, f := range o {
|
|
||||||
f(p)
|
|
||||||
}
|
|
||||||
return p
|
|
||||||
}
|
|
||||||
|
|
||||||
// Next parses the next line in the file. It returns true if a line was parsed
|
|
||||||
// and false if it reached the end of the file.
|
|
||||||
func (p *Parser) Next() bool {
|
|
||||||
if !p.keepRanges && p.rangeStart < p.rangeEnd {
|
|
||||||
p.rangeStart++
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
p.comment = nil
|
|
||||||
p.field = p.field[:0]
|
|
||||||
p.parsedRange = false
|
|
||||||
|
|
||||||
for p.scanner.Scan() {
|
|
||||||
b := p.scanner.Bytes()
|
|
||||||
if len(b) == 0 {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
if b[0] == '#' {
|
|
||||||
if p.commentHandler != nil {
|
|
||||||
p.commentHandler(strings.TrimSpace(string(b[1:])))
|
|
||||||
}
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
|
|
||||||
// Parse line
|
|
||||||
if i := bytes.IndexByte(b, '#'); i != -1 {
|
|
||||||
p.comment = bytes.TrimSpace(b[i+1:])
|
|
||||||
b = b[:i]
|
|
||||||
}
|
|
||||||
if b[0] == '@' {
|
|
||||||
if p.partHandler != nil {
|
|
||||||
p.field = append(p.field, bytes.TrimSpace(b[1:]))
|
|
||||||
p.partHandler(p)
|
|
||||||
p.field = p.field[:0]
|
|
||||||
}
|
|
||||||
p.comment = nil
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
for {
|
|
||||||
i := bytes.IndexByte(b, ';')
|
|
||||||
if i == -1 {
|
|
||||||
p.field = append(p.field, bytes.TrimSpace(b))
|
|
||||||
break
|
|
||||||
}
|
|
||||||
p.field = append(p.field, bytes.TrimSpace(b[:i]))
|
|
||||||
b = b[i+1:]
|
|
||||||
}
|
|
||||||
if !p.keepRanges {
|
|
||||||
p.rangeStart, p.rangeEnd = p.getRange(0)
|
|
||||||
}
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
p.setError(p.scanner.Err())
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
|
|
||||||
func parseRune(b []byte) (rune, error) {
|
|
||||||
if len(b) > 2 && b[0] == 'U' && b[1] == '+' {
|
|
||||||
b = b[2:]
|
|
||||||
}
|
|
||||||
x, err := strconv.ParseUint(string(b), 16, 32)
|
|
||||||
return rune(x), err
|
|
||||||
}
|
|
||||||
|
|
||||||
func (p *Parser) parseRune(b []byte) rune {
|
|
||||||
x, err := parseRune(b)
|
|
||||||
p.setError(err)
|
|
||||||
return x
|
|
||||||
}
|
|
||||||
|
|
||||||
// Rune parses and returns field i as a rune.
|
|
||||||
func (p *Parser) Rune(i int) rune {
|
|
||||||
if i > 0 || p.keepRanges {
|
|
||||||
return p.parseRune(p.getField(i))
|
|
||||||
}
|
|
||||||
return p.rangeStart
|
|
||||||
}
|
|
||||||
|
|
||||||
// Runes interprets and returns field i as a sequence of runes.
|
|
||||||
func (p *Parser) Runes(i int) (runes []rune) {
|
|
||||||
add := func(b []byte) {
|
|
||||||
if b = bytes.TrimSpace(b); len(b) > 0 {
|
|
||||||
runes = append(runes, p.parseRune(b))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
for b := p.getField(i); ; {
|
|
||||||
i := bytes.IndexByte(b, ' ')
|
|
||||||
if i == -1 {
|
|
||||||
add(b)
|
|
||||||
break
|
|
||||||
}
|
|
||||||
add(b[:i])
|
|
||||||
b = b[i+1:]
|
|
||||||
}
|
|
||||||
return
|
|
||||||
}
|
|
||||||
|
|
||||||
var (
|
|
||||||
errIncorrectLegacyRange = errors.New("ucd: unmatched <* First>")
|
|
||||||
|
|
||||||
// reRange matches one line of a legacy rune range.
|
|
||||||
reRange = regexp.MustCompile("^([0-9A-F]*);<([^,]*), ([^>]*)>(.*)$")
|
|
||||||
)
|
|
||||||
|
|
||||||
// Range parses and returns field i as a rune range. A range is inclusive at
|
|
||||||
// both ends. If the field only has one rune, first and last will be identical.
|
|
||||||
// It supports the legacy format for ranges used in UnicodeData.txt.
|
|
||||||
func (p *Parser) Range(i int) (first, last rune) {
|
|
||||||
if !p.keepRanges {
|
|
||||||
return p.rangeStart, p.rangeStart
|
|
||||||
}
|
|
||||||
return p.getRange(i)
|
|
||||||
}
|
|
||||||
|
|
||||||
func (p *Parser) getRange(i int) (first, last rune) {
|
|
||||||
b := p.getField(i)
|
|
||||||
if k := bytes.Index(b, []byte("..")); k != -1 {
|
|
||||||
return p.parseRune(b[:k]), p.parseRune(b[k+2:])
|
|
||||||
}
|
|
||||||
// The first field may not be a rune, in which case we may ignore any error
|
|
||||||
// and set the range as 0..0.
|
|
||||||
x, err := parseRune(b)
|
|
||||||
if err != nil {
|
|
||||||
// Disable range parsing henceforth. This ensures that an error will be
|
|
||||||
// returned if the user subsequently will try to parse this field as
|
|
||||||
// a Rune.
|
|
||||||
p.keepRanges = true
|
|
||||||
}
|
|
||||||
// Special case for UnicodeData that was retained for backwards compatibility.
|
|
||||||
if i == 0 && len(p.field) > 1 && bytes.HasSuffix(p.field[1], []byte("First>")) {
|
|
||||||
if p.parsedRange {
|
|
||||||
return p.rangeStart, p.rangeEnd
|
|
||||||
}
|
|
||||||
mf := reRange.FindStringSubmatch(p.scanner.Text())
|
|
||||||
if mf == nil || !p.scanner.Scan() {
|
|
||||||
p.setError(errIncorrectLegacyRange)
|
|
||||||
return x, x
|
|
||||||
}
|
|
||||||
// Using Bytes would be more efficient here, but Text is a lot easier
|
|
||||||
// and this is not a frequent case.
|
|
||||||
ml := reRange.FindStringSubmatch(p.scanner.Text())
|
|
||||||
if ml == nil || mf[2] != ml[2] || ml[3] != "Last" || mf[4] != ml[4] {
|
|
||||||
p.setError(errIncorrectLegacyRange)
|
|
||||||
return x, x
|
|
||||||
}
|
|
||||||
p.rangeStart, p.rangeEnd = x, p.parseRune(p.scanner.Bytes()[:len(ml[1])])
|
|
||||||
p.parsedRange = true
|
|
||||||
return p.rangeStart, p.rangeEnd
|
|
||||||
}
|
|
||||||
return x, x
|
|
||||||
}
|
|
||||||
|
|
||||||
// bools recognizes all valid UCD boolean values.
|
|
||||||
var bools = map[string]bool{
|
|
||||||
"": false,
|
|
||||||
"N": false,
|
|
||||||
"No": false,
|
|
||||||
"F": false,
|
|
||||||
"False": false,
|
|
||||||
"Y": true,
|
|
||||||
"Yes": true,
|
|
||||||
"T": true,
|
|
||||||
"True": true,
|
|
||||||
}
|
|
||||||
|
|
||||||
// Bool parses and returns field i as a boolean value.
|
|
||||||
func (p *Parser) Bool(i int) bool {
|
|
||||||
b := p.getField(i)
|
|
||||||
for s, v := range bools {
|
|
||||||
if bstrEq(b, s) {
|
|
||||||
return v
|
|
||||||
}
|
|
||||||
}
|
|
||||||
p.setError(strconv.ErrSyntax)
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
|
|
||||||
// Int parses and returns field i as an integer value.
|
|
||||||
func (p *Parser) Int(i int) int {
|
|
||||||
x, err := strconv.ParseInt(string(p.getField(i)), 10, 64)
|
|
||||||
p.setError(err)
|
|
||||||
return int(x)
|
|
||||||
}
|
|
||||||
|
|
||||||
// Uint parses and returns field i as an unsigned integer value.
|
|
||||||
func (p *Parser) Uint(i int) uint {
|
|
||||||
x, err := strconv.ParseUint(string(p.getField(i)), 10, 64)
|
|
||||||
p.setError(err)
|
|
||||||
return uint(x)
|
|
||||||
}
|
|
||||||
|
|
||||||
// Float parses and returns field i as a decimal value.
|
|
||||||
func (p *Parser) Float(i int) float64 {
|
|
||||||
x, err := strconv.ParseFloat(string(p.getField(i)), 64)
|
|
||||||
p.setError(err)
|
|
||||||
return x
|
|
||||||
}
|
|
||||||
|
|
||||||
// String parses and returns field i as a string value.
|
|
||||||
func (p *Parser) String(i int) string {
|
|
||||||
return string(p.getField(i))
|
|
||||||
}
|
|
||||||
|
|
||||||
// Strings parses and returns field i as a space-separated list of strings.
|
|
||||||
func (p *Parser) Strings(i int) []string {
|
|
||||||
ss := strings.Split(string(p.getField(i)), " ")
|
|
||||||
for i, s := range ss {
|
|
||||||
ss[i] = strings.TrimSpace(s)
|
|
||||||
}
|
|
||||||
return ss
|
|
||||||
}
|
|
||||||
|
|
||||||
// Comment returns the comments for the current line.
|
|
||||||
func (p *Parser) Comment() string {
|
|
||||||
return string(p.comment)
|
|
||||||
}
|
|
||||||
|
|
||||||
var errUndefinedEnum = errors.New("ucd: undefined enum value")
|
|
||||||
|
|
||||||
// Enum interprets and returns field i as a value that must be one of the values
|
|
||||||
// in enum.
|
|
||||||
func (p *Parser) Enum(i int, enum ...string) string {
|
|
||||||
b := p.getField(i)
|
|
||||||
for _, s := range enum {
|
|
||||||
if bstrEq(b, s) {
|
|
||||||
return s
|
|
||||||
}
|
|
||||||
}
|
|
||||||
p.setError(errUndefinedEnum)
|
|
||||||
return ""
|
|
||||||
}
|
|
||||||
|
|
||||||
func bstrEq(b []byte, s string) bool {
|
|
||||||
if len(b) != len(s) {
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
for i, c := range b {
|
|
||||||
if c != s[i] {
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return true
|
|
||||||
}
|
|
|
@ -1,133 +0,0 @@
|
||||||
// Copyright 2015 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
package main
|
|
||||||
|
|
||||||
import (
|
|
||||||
"flag"
|
|
||||||
"log"
|
|
||||||
|
|
||||||
"golang.org/x/text/internal/gen"
|
|
||||||
"golang.org/x/text/internal/triegen"
|
|
||||||
"golang.org/x/text/internal/ucd"
|
|
||||||
)
|
|
||||||
|
|
||||||
var outputFile = flag.String("out", "tables.go", "output file")
|
|
||||||
|
|
||||||
func main() {
|
|
||||||
gen.Init()
|
|
||||||
gen.Repackage("gen_trieval.go", "trieval.go", "bidi")
|
|
||||||
gen.Repackage("gen_ranges.go", "ranges_test.go", "bidi")
|
|
||||||
|
|
||||||
genTables()
|
|
||||||
}
|
|
||||||
|
|
||||||
// bidiClass names and codes taken from class "bc" in
|
|
||||||
// http://www.unicode.org/Public/8.0.0/ucd/PropertyValueAliases.txt
|
|
||||||
var bidiClass = map[string]Class{
|
|
||||||
"AL": AL, // ArabicLetter
|
|
||||||
"AN": AN, // ArabicNumber
|
|
||||||
"B": B, // ParagraphSeparator
|
|
||||||
"BN": BN, // BoundaryNeutral
|
|
||||||
"CS": CS, // CommonSeparator
|
|
||||||
"EN": EN, // EuropeanNumber
|
|
||||||
"ES": ES, // EuropeanSeparator
|
|
||||||
"ET": ET, // EuropeanTerminator
|
|
||||||
"L": L, // LeftToRight
|
|
||||||
"NSM": NSM, // NonspacingMark
|
|
||||||
"ON": ON, // OtherNeutral
|
|
||||||
"R": R, // RightToLeft
|
|
||||||
"S": S, // SegmentSeparator
|
|
||||||
"WS": WS, // WhiteSpace
|
|
||||||
|
|
||||||
"FSI": Control,
|
|
||||||
"PDF": Control,
|
|
||||||
"PDI": Control,
|
|
||||||
"LRE": Control,
|
|
||||||
"LRI": Control,
|
|
||||||
"LRO": Control,
|
|
||||||
"RLE": Control,
|
|
||||||
"RLI": Control,
|
|
||||||
"RLO": Control,
|
|
||||||
}
|
|
||||||
|
|
||||||
func genTables() {
|
|
||||||
if numClass > 0x0F {
|
|
||||||
log.Fatalf("Too many Class constants (%#x > 0x0F).", numClass)
|
|
||||||
}
|
|
||||||
w := gen.NewCodeWriter()
|
|
||||||
defer w.WriteGoFile(*outputFile, "bidi")
|
|
||||||
|
|
||||||
gen.WriteUnicodeVersion(w)
|
|
||||||
|
|
||||||
t := triegen.NewTrie("bidi")
|
|
||||||
|
|
||||||
// Build data about bracket mapping. These bits need to be or-ed with
|
|
||||||
// any other bits.
|
|
||||||
orMask := map[rune]uint64{}
|
|
||||||
|
|
||||||
xorMap := map[rune]int{}
|
|
||||||
xorMasks := []rune{0} // First value is no-op.
|
|
||||||
|
|
||||||
ucd.Parse(gen.OpenUCDFile("BidiBrackets.txt"), func(p *ucd.Parser) {
|
|
||||||
r1 := p.Rune(0)
|
|
||||||
r2 := p.Rune(1)
|
|
||||||
xor := r1 ^ r2
|
|
||||||
if _, ok := xorMap[xor]; !ok {
|
|
||||||
xorMap[xor] = len(xorMasks)
|
|
||||||
xorMasks = append(xorMasks, xor)
|
|
||||||
}
|
|
||||||
entry := uint64(xorMap[xor]) << xorMaskShift
|
|
||||||
switch p.String(2) {
|
|
||||||
case "o":
|
|
||||||
entry |= openMask
|
|
||||||
case "c", "n":
|
|
||||||
default:
|
|
||||||
log.Fatalf("Unknown bracket class %q.", p.String(2))
|
|
||||||
}
|
|
||||||
orMask[r1] = entry
|
|
||||||
})
|
|
||||||
|
|
||||||
w.WriteComment(`
|
|
||||||
xorMasks contains masks to be xor-ed with brackets to get the reverse
|
|
||||||
version.`)
|
|
||||||
w.WriteVar("xorMasks", xorMasks)
|
|
||||||
|
|
||||||
done := map[rune]bool{}
|
|
||||||
|
|
||||||
insert := func(r rune, c Class) {
|
|
||||||
if !done[r] {
|
|
||||||
t.Insert(r, orMask[r]|uint64(c))
|
|
||||||
done[r] = true
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Insert the derived BiDi properties.
|
|
||||||
ucd.Parse(gen.OpenUCDFile("extracted/DerivedBidiClass.txt"), func(p *ucd.Parser) {
|
|
||||||
r := p.Rune(0)
|
|
||||||
class, ok := bidiClass[p.String(1)]
|
|
||||||
if !ok {
|
|
||||||
log.Fatalf("%U: Unknown BiDi class %q", r, p.String(1))
|
|
||||||
}
|
|
||||||
insert(r, class)
|
|
||||||
})
|
|
||||||
visitDefaults(insert)
|
|
||||||
|
|
||||||
// TODO: use sparse blocks. This would reduce table size considerably
|
|
||||||
// from the looks of it.
|
|
||||||
|
|
||||||
sz, err := t.Gen(w)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
w.Size += sz
|
|
||||||
}
|
|
||||||
|
|
||||||
// dummy values to make methods in gen_common compile. The real versions
|
|
||||||
// will be generated by this file to tables.go.
|
|
||||||
var (
|
|
||||||
xorMasks []rune
|
|
||||||
)
|
|
|
@ -1,57 +0,0 @@
|
||||||
// Copyright 2015 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
package main
|
|
||||||
|
|
||||||
import (
|
|
||||||
"unicode"
|
|
||||||
|
|
||||||
"golang.org/x/text/internal/gen"
|
|
||||||
"golang.org/x/text/internal/ucd"
|
|
||||||
"golang.org/x/text/unicode/rangetable"
|
|
||||||
)
|
|
||||||
|
|
||||||
// These tables are hand-extracted from:
|
|
||||||
// http://www.unicode.org/Public/8.0.0/ucd/extracted/DerivedBidiClass.txt
|
|
||||||
func visitDefaults(fn func(r rune, c Class)) {
|
|
||||||
// first write default values for ranges listed above.
|
|
||||||
visitRunes(fn, AL, []rune{
|
|
||||||
0x0600, 0x07BF, // Arabic
|
|
||||||
0x08A0, 0x08FF, // Arabic Extended-A
|
|
||||||
0xFB50, 0xFDCF, // Arabic Presentation Forms
|
|
||||||
0xFDF0, 0xFDFF,
|
|
||||||
0xFE70, 0xFEFF,
|
|
||||||
0x0001EE00, 0x0001EEFF, // Arabic Mathematical Alpha Symbols
|
|
||||||
})
|
|
||||||
visitRunes(fn, R, []rune{
|
|
||||||
0x0590, 0x05FF, // Hebrew
|
|
||||||
0x07C0, 0x089F, // Nko et al.
|
|
||||||
0xFB1D, 0xFB4F,
|
|
||||||
0x00010800, 0x00010FFF, // Cypriot Syllabary et. al.
|
|
||||||
0x0001E800, 0x0001EDFF,
|
|
||||||
0x0001EF00, 0x0001EFFF,
|
|
||||||
})
|
|
||||||
visitRunes(fn, ET, []rune{ // European Terminator
|
|
||||||
0x20A0, 0x20Cf, // Currency symbols
|
|
||||||
})
|
|
||||||
rangetable.Visit(unicode.Noncharacter_Code_Point, func(r rune) {
|
|
||||||
fn(r, BN) // Boundary Neutral
|
|
||||||
})
|
|
||||||
ucd.Parse(gen.OpenUCDFile("DerivedCoreProperties.txt"), func(p *ucd.Parser) {
|
|
||||||
if p.String(1) == "Default_Ignorable_Code_Point" {
|
|
||||||
fn(p.Rune(0), BN) // Boundary Neutral
|
|
||||||
}
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
func visitRunes(fn func(r rune, c Class), c Class, runes []rune) {
|
|
||||||
for i := 0; i < len(runes); i += 2 {
|
|
||||||
lo, hi := runes[i], runes[i+1]
|
|
||||||
for j := lo; j <= hi; j++ {
|
|
||||||
fn(j, c)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
|
@ -1,64 +0,0 @@
|
||||||
// Copyright 2015 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
package main
|
|
||||||
|
|
||||||
// Class is the Unicode BiDi class. Each rune has a single class.
|
|
||||||
type Class uint
|
|
||||||
|
|
||||||
const (
|
|
||||||
L Class = iota // LeftToRight
|
|
||||||
R // RightToLeft
|
|
||||||
EN // EuropeanNumber
|
|
||||||
ES // EuropeanSeparator
|
|
||||||
ET // EuropeanTerminator
|
|
||||||
AN // ArabicNumber
|
|
||||||
CS // CommonSeparator
|
|
||||||
B // ParagraphSeparator
|
|
||||||
S // SegmentSeparator
|
|
||||||
WS // WhiteSpace
|
|
||||||
ON // OtherNeutral
|
|
||||||
BN // BoundaryNeutral
|
|
||||||
NSM // NonspacingMark
|
|
||||||
AL // ArabicLetter
|
|
||||||
Control // Control LRO - PDI
|
|
||||||
|
|
||||||
numClass
|
|
||||||
|
|
||||||
LRO // LeftToRightOverride
|
|
||||||
RLO // RightToLeftOverride
|
|
||||||
LRE // LeftToRightEmbedding
|
|
||||||
RLE // RightToLeftEmbedding
|
|
||||||
PDF // PopDirectionalFormat
|
|
||||||
LRI // LeftToRightIsolate
|
|
||||||
RLI // RightToLeftIsolate
|
|
||||||
FSI // FirstStrongIsolate
|
|
||||||
PDI // PopDirectionalIsolate
|
|
||||||
|
|
||||||
unknownClass = ^Class(0)
|
|
||||||
)
|
|
||||||
|
|
||||||
var controlToClass = map[rune]Class{
|
|
||||||
0x202D: LRO, // LeftToRightOverride,
|
|
||||||
0x202E: RLO, // RightToLeftOverride,
|
|
||||||
0x202A: LRE, // LeftToRightEmbedding,
|
|
||||||
0x202B: RLE, // RightToLeftEmbedding,
|
|
||||||
0x202C: PDF, // PopDirectionalFormat,
|
|
||||||
0x2066: LRI, // LeftToRightIsolate,
|
|
||||||
0x2067: RLI, // RightToLeftIsolate,
|
|
||||||
0x2068: FSI, // FirstStrongIsolate,
|
|
||||||
0x2069: PDI, // PopDirectionalIsolate,
|
|
||||||
}
|
|
||||||
|
|
||||||
// A trie entry has the following bits:
|
|
||||||
// 7..5 XOR mask for brackets
|
|
||||||
// 4 1: Bracket open, 0: Bracket close
|
|
||||||
// 3..0 Class type
|
|
||||||
|
|
||||||
const (
|
|
||||||
openMask = 0x10
|
|
||||||
xorMaskShift = 5
|
|
||||||
)
|
|
|
@ -1,100 +0,0 @@
|
||||||
// Copyright 2013 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
package cldr
|
|
||||||
|
|
||||||
import (
|
|
||||||
"encoding/xml"
|
|
||||||
"regexp"
|
|
||||||
"strconv"
|
|
||||||
)
|
|
||||||
|
|
||||||
// Elem is implemented by every XML element.
|
|
||||||
type Elem interface {
|
|
||||||
setEnclosing(Elem)
|
|
||||||
setName(string)
|
|
||||||
enclosing() Elem
|
|
||||||
|
|
||||||
GetCommon() *Common
|
|
||||||
}
|
|
||||||
|
|
||||||
type hidden struct {
|
|
||||||
CharData string `xml:",chardata"`
|
|
||||||
Alias *struct {
|
|
||||||
Common
|
|
||||||
Source string `xml:"source,attr"`
|
|
||||||
Path string `xml:"path,attr"`
|
|
||||||
} `xml:"alias"`
|
|
||||||
Def *struct {
|
|
||||||
Common
|
|
||||||
Choice string `xml:"choice,attr,omitempty"`
|
|
||||||
Type string `xml:"type,attr,omitempty"`
|
|
||||||
} `xml:"default"`
|
|
||||||
}
|
|
||||||
|
|
||||||
// Common holds several of the most common attributes and sub elements
|
|
||||||
// of an XML element.
|
|
||||||
type Common struct {
|
|
||||||
XMLName xml.Name
|
|
||||||
name string
|
|
||||||
enclElem Elem
|
|
||||||
Type string `xml:"type,attr,omitempty"`
|
|
||||||
Reference string `xml:"reference,attr,omitempty"`
|
|
||||||
Alt string `xml:"alt,attr,omitempty"`
|
|
||||||
ValidSubLocales string `xml:"validSubLocales,attr,omitempty"`
|
|
||||||
Draft string `xml:"draft,attr,omitempty"`
|
|
||||||
hidden
|
|
||||||
}
|
|
||||||
|
|
||||||
// Default returns the default type to select from the enclosed list
|
|
||||||
// or "" if no default value is specified.
|
|
||||||
func (e *Common) Default() string {
|
|
||||||
if e.Def == nil {
|
|
||||||
return ""
|
|
||||||
}
|
|
||||||
if e.Def.Choice != "" {
|
|
||||||
return e.Def.Choice
|
|
||||||
} else if e.Def.Type != "" {
|
|
||||||
// Type is still used by the default element in collation.
|
|
||||||
return e.Def.Type
|
|
||||||
}
|
|
||||||
return ""
|
|
||||||
}
|
|
||||||
|
|
||||||
// GetCommon returns e. It is provided such that Common implements Elem.
|
|
||||||
func (e *Common) GetCommon() *Common {
|
|
||||||
return e
|
|
||||||
}
|
|
||||||
|
|
||||||
// Data returns the character data accumulated for this element.
|
|
||||||
func (e *Common) Data() string {
|
|
||||||
e.CharData = charRe.ReplaceAllStringFunc(e.CharData, replaceUnicode)
|
|
||||||
return e.CharData
|
|
||||||
}
|
|
||||||
|
|
||||||
func (e *Common) setName(s string) {
|
|
||||||
e.name = s
|
|
||||||
}
|
|
||||||
|
|
||||||
func (e *Common) enclosing() Elem {
|
|
||||||
return e.enclElem
|
|
||||||
}
|
|
||||||
|
|
||||||
func (e *Common) setEnclosing(en Elem) {
|
|
||||||
e.enclElem = en
|
|
||||||
}
|
|
||||||
|
|
||||||
// Escape characters that can be escaped without further escaping the string.
|
|
||||||
var charRe = regexp.MustCompile(`&#x[0-9a-fA-F]*;|\\u[0-9a-fA-F]{4}|\\U[0-9a-fA-F]{8}|\\x[0-9a-fA-F]{2}|\\[0-7]{3}|\\[abtnvfr]`)
|
|
||||||
|
|
||||||
// replaceUnicode converts hexadecimal Unicode codepoint notations to a one-rune string.
|
|
||||||
// It assumes the input string is correctly formatted.
|
|
||||||
func replaceUnicode(s string) string {
|
|
||||||
if s[1] == '#' {
|
|
||||||
r, _ := strconv.ParseInt(s[3:len(s)-1], 16, 32)
|
|
||||||
return string(r)
|
|
||||||
}
|
|
||||||
r, _, _, _ := strconv.UnquoteChar(s, 0)
|
|
||||||
return string(r)
|
|
||||||
}
|
|
|
@ -1,130 +0,0 @@
|
||||||
// Copyright 2013 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
//go:generate go run makexml.go -output xml.go
|
|
||||||
|
|
||||||
// Package cldr provides a parser for LDML and related XML formats.
|
|
||||||
// This package is intended to be used by the table generation tools
|
|
||||||
// for the various internationalization-related packages.
|
|
||||||
// As the XML types are generated from the CLDR DTD, and as the CLDR standard
|
|
||||||
// is periodically amended, this package may change considerably over time.
|
|
||||||
// This mostly means that data may appear and disappear between versions.
|
|
||||||
// That is, old code should keep compiling for newer versions, but data
|
|
||||||
// may have moved or changed.
|
|
||||||
// CLDR version 22 is the first version supported by this package.
|
|
||||||
// Older versions may not work.
|
|
||||||
package cldr // import "golang.org/x/text/unicode/cldr"
|
|
||||||
|
|
||||||
import (
|
|
||||||
"fmt"
|
|
||||||
"sort"
|
|
||||||
)
|
|
||||||
|
|
||||||
// CLDR provides access to parsed data of the Unicode Common Locale Data Repository.
|
|
||||||
type CLDR struct {
|
|
||||||
parent map[string][]string
|
|
||||||
locale map[string]*LDML
|
|
||||||
resolved map[string]*LDML
|
|
||||||
bcp47 *LDMLBCP47
|
|
||||||
supp *SupplementalData
|
|
||||||
}
|
|
||||||
|
|
||||||
func makeCLDR() *CLDR {
|
|
||||||
return &CLDR{
|
|
||||||
parent: make(map[string][]string),
|
|
||||||
locale: make(map[string]*LDML),
|
|
||||||
resolved: make(map[string]*LDML),
|
|
||||||
bcp47: &LDMLBCP47{},
|
|
||||||
supp: &SupplementalData{},
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// BCP47 returns the parsed BCP47 LDML data. If no such data was parsed, nil is returned.
|
|
||||||
func (cldr *CLDR) BCP47() *LDMLBCP47 {
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
|
|
||||||
// Draft indicates the draft level of an element.
|
|
||||||
type Draft int
|
|
||||||
|
|
||||||
const (
|
|
||||||
Approved Draft = iota
|
|
||||||
Contributed
|
|
||||||
Provisional
|
|
||||||
Unconfirmed
|
|
||||||
)
|
|
||||||
|
|
||||||
var drafts = []string{"unconfirmed", "provisional", "contributed", "approved", ""}
|
|
||||||
|
|
||||||
// ParseDraft returns the Draft value corresponding to the given string. The
|
|
||||||
// empty string corresponds to Approved.
|
|
||||||
func ParseDraft(level string) (Draft, error) {
|
|
||||||
if level == "" {
|
|
||||||
return Approved, nil
|
|
||||||
}
|
|
||||||
for i, s := range drafts {
|
|
||||||
if level == s {
|
|
||||||
return Unconfirmed - Draft(i), nil
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return Approved, fmt.Errorf("cldr: unknown draft level %q", level)
|
|
||||||
}
|
|
||||||
|
|
||||||
func (d Draft) String() string {
|
|
||||||
return drafts[len(drafts)-1-int(d)]
|
|
||||||
}
|
|
||||||
|
|
||||||
// SetDraftLevel sets which draft levels to include in the evaluated LDML.
|
|
||||||
// Any draft element for which the draft level is higher than lev will be excluded.
|
|
||||||
// If multiple draft levels are available for a single element, the one with the
|
|
||||||
// lowest draft level will be selected, unless preferDraft is true, in which case
|
|
||||||
// the highest draft will be chosen.
|
|
||||||
// It is assumed that the underlying LDML is canonicalized.
|
|
||||||
func (cldr *CLDR) SetDraftLevel(lev Draft, preferDraft bool) {
|
|
||||||
// TODO: implement
|
|
||||||
cldr.resolved = make(map[string]*LDML)
|
|
||||||
}
|
|
||||||
|
|
||||||
// RawLDML returns the LDML XML for id in unresolved form.
|
|
||||||
// id must be one of the strings returned by Locales.
|
|
||||||
func (cldr *CLDR) RawLDML(loc string) *LDML {
|
|
||||||
return cldr.locale[loc]
|
|
||||||
}
|
|
||||||
|
|
||||||
// LDML returns the fully resolved LDML XML for loc, which must be one of
|
|
||||||
// the strings returned by Locales.
|
|
||||||
func (cldr *CLDR) LDML(loc string) (*LDML, error) {
|
|
||||||
return cldr.resolve(loc)
|
|
||||||
}
|
|
||||||
|
|
||||||
// Supplemental returns the parsed supplemental data. If no such data was parsed,
|
|
||||||
// nil is returned.
|
|
||||||
func (cldr *CLDR) Supplemental() *SupplementalData {
|
|
||||||
return cldr.supp
|
|
||||||
}
|
|
||||||
|
|
||||||
// Locales returns the locales for which there exist files.
|
|
||||||
// Valid sublocales for which there is no file are not included.
|
|
||||||
// The root locale is always sorted first.
|
|
||||||
func (cldr *CLDR) Locales() []string {
|
|
||||||
loc := []string{"root"}
|
|
||||||
hasRoot := false
|
|
||||||
for l, _ := range cldr.locale {
|
|
||||||
if l == "root" {
|
|
||||||
hasRoot = true
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
loc = append(loc, l)
|
|
||||||
}
|
|
||||||
sort.Strings(loc[1:])
|
|
||||||
if !hasRoot {
|
|
||||||
return loc[1:]
|
|
||||||
}
|
|
||||||
return loc
|
|
||||||
}
|
|
||||||
|
|
||||||
// Get fills in the fields of x based on the XPath path.
|
|
||||||
func Get(e Elem, path string) (res Elem, err error) {
|
|
||||||
return walkXPath(e, path)
|
|
||||||
}
|
|
|
@ -1,359 +0,0 @@
|
||||||
// Copyright 2013 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
package cldr
|
|
||||||
|
|
||||||
import (
|
|
||||||
"bufio"
|
|
||||||
"encoding/xml"
|
|
||||||
"errors"
|
|
||||||
"fmt"
|
|
||||||
"strconv"
|
|
||||||
"strings"
|
|
||||||
"unicode"
|
|
||||||
"unicode/utf8"
|
|
||||||
)
|
|
||||||
|
|
||||||
// RuleProcessor can be passed to Collator's Process method, which
|
|
||||||
// parses the rules and calls the respective method for each rule found.
|
|
||||||
type RuleProcessor interface {
|
|
||||||
Reset(anchor string, before int) error
|
|
||||||
Insert(level int, str, context, extend string) error
|
|
||||||
Index(id string)
|
|
||||||
}
|
|
||||||
|
|
||||||
const (
|
|
||||||
// cldrIndex is a Unicode-reserved sentinel value used to mark the start
|
|
||||||
// of a grouping within an index.
|
|
||||||
// We ignore any rule that starts with this rune.
|
|
||||||
// See http://unicode.org/reports/tr35/#Collation_Elements for details.
|
|
||||||
cldrIndex = "\uFDD0"
|
|
||||||
|
|
||||||
// specialAnchor is the format in which to represent logical reset positions,
|
|
||||||
// such as "first tertiary ignorable".
|
|
||||||
specialAnchor = "<%s/>"
|
|
||||||
)
|
|
||||||
|
|
||||||
// Process parses the rules for the tailorings of this collation
|
|
||||||
// and calls the respective methods of p for each rule found.
|
|
||||||
func (c Collation) Process(p RuleProcessor) (err error) {
|
|
||||||
if len(c.Cr) > 0 {
|
|
||||||
if len(c.Cr) > 1 {
|
|
||||||
return fmt.Errorf("multiple cr elements, want 0 or 1")
|
|
||||||
}
|
|
||||||
return processRules(p, c.Cr[0].Data())
|
|
||||||
}
|
|
||||||
if c.Rules.Any != nil {
|
|
||||||
return c.processXML(p)
|
|
||||||
}
|
|
||||||
return errors.New("no tailoring data")
|
|
||||||
}
|
|
||||||
|
|
||||||
// processRules parses rules in the Collation Rule Syntax defined in
|
|
||||||
// http://www.unicode.org/reports/tr35/tr35-collation.html#Collation_Tailorings.
|
|
||||||
func processRules(p RuleProcessor, s string) (err error) {
|
|
||||||
chk := func(s string, e error) string {
|
|
||||||
if err == nil {
|
|
||||||
err = e
|
|
||||||
}
|
|
||||||
return s
|
|
||||||
}
|
|
||||||
i := 0 // Save the line number for use after the loop.
|
|
||||||
scanner := bufio.NewScanner(strings.NewReader(s))
|
|
||||||
for ; scanner.Scan() && err == nil; i++ {
|
|
||||||
for s := skipSpace(scanner.Text()); s != "" && s[0] != '#'; s = skipSpace(s) {
|
|
||||||
level := 5
|
|
||||||
var ch byte
|
|
||||||
switch ch, s = s[0], s[1:]; ch {
|
|
||||||
case '&': // followed by <anchor> or '[' <key> ']'
|
|
||||||
if s = skipSpace(s); consume(&s, '[') {
|
|
||||||
s = chk(parseSpecialAnchor(p, s))
|
|
||||||
} else {
|
|
||||||
s = chk(parseAnchor(p, 0, s))
|
|
||||||
}
|
|
||||||
case '<': // sort relation '<'{1,4}, optionally followed by '*'.
|
|
||||||
for level = 1; consume(&s, '<'); level++ {
|
|
||||||
}
|
|
||||||
if level > 4 {
|
|
||||||
err = fmt.Errorf("level %d > 4", level)
|
|
||||||
}
|
|
||||||
fallthrough
|
|
||||||
case '=': // identity relation, optionally followed by *.
|
|
||||||
if consume(&s, '*') {
|
|
||||||
s = chk(parseSequence(p, level, s))
|
|
||||||
} else {
|
|
||||||
s = chk(parseOrder(p, level, s))
|
|
||||||
}
|
|
||||||
default:
|
|
||||||
chk("", fmt.Errorf("illegal operator %q", ch))
|
|
||||||
break
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
if chk("", scanner.Err()); err != nil {
|
|
||||||
return fmt.Errorf("%d: %v", i, err)
|
|
||||||
}
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
|
|
||||||
// parseSpecialAnchor parses the anchor syntax which is either of the form
|
|
||||||
// ['before' <level>] <anchor>
|
|
||||||
// or
|
|
||||||
// [<label>]
|
|
||||||
// The starting should already be consumed.
|
|
||||||
func parseSpecialAnchor(p RuleProcessor, s string) (tail string, err error) {
|
|
||||||
i := strings.IndexByte(s, ']')
|
|
||||||
if i == -1 {
|
|
||||||
return "", errors.New("unmatched bracket")
|
|
||||||
}
|
|
||||||
a := strings.TrimSpace(s[:i])
|
|
||||||
s = s[i+1:]
|
|
||||||
if strings.HasPrefix(a, "before ") {
|
|
||||||
l, err := strconv.ParseUint(skipSpace(a[len("before "):]), 10, 3)
|
|
||||||
if err != nil {
|
|
||||||
return s, err
|
|
||||||
}
|
|
||||||
return parseAnchor(p, int(l), s)
|
|
||||||
}
|
|
||||||
return s, p.Reset(fmt.Sprintf(specialAnchor, a), 0)
|
|
||||||
}
|
|
||||||
|
|
||||||
func parseAnchor(p RuleProcessor, level int, s string) (tail string, err error) {
|
|
||||||
anchor, s, err := scanString(s)
|
|
||||||
if err != nil {
|
|
||||||
return s, err
|
|
||||||
}
|
|
||||||
return s, p.Reset(anchor, level)
|
|
||||||
}
|
|
||||||
|
|
||||||
func parseOrder(p RuleProcessor, level int, s string) (tail string, err error) {
|
|
||||||
var value, context, extend string
|
|
||||||
if value, s, err = scanString(s); err != nil {
|
|
||||||
return s, err
|
|
||||||
}
|
|
||||||
if strings.HasPrefix(value, cldrIndex) {
|
|
||||||
p.Index(value[len(cldrIndex):])
|
|
||||||
return
|
|
||||||
}
|
|
||||||
if consume(&s, '|') {
|
|
||||||
if context, s, err = scanString(s); err != nil {
|
|
||||||
return s, errors.New("missing string after context")
|
|
||||||
}
|
|
||||||
}
|
|
||||||
if consume(&s, '/') {
|
|
||||||
if extend, s, err = scanString(s); err != nil {
|
|
||||||
return s, errors.New("missing string after extension")
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return s, p.Insert(level, value, context, extend)
|
|
||||||
}
|
|
||||||
|
|
||||||
// scanString scans a single input string.
|
|
||||||
func scanString(s string) (str, tail string, err error) {
|
|
||||||
if s = skipSpace(s); s == "" {
|
|
||||||
return s, s, errors.New("missing string")
|
|
||||||
}
|
|
||||||
buf := [16]byte{} // small but enough to hold most cases.
|
|
||||||
value := buf[:0]
|
|
||||||
for s != "" {
|
|
||||||
if consume(&s, '\'') {
|
|
||||||
i := strings.IndexByte(s, '\'')
|
|
||||||
if i == -1 {
|
|
||||||
return "", "", errors.New(`unmatched single quote`)
|
|
||||||
}
|
|
||||||
if i == 0 {
|
|
||||||
value = append(value, '\'')
|
|
||||||
} else {
|
|
||||||
value = append(value, s[:i]...)
|
|
||||||
}
|
|
||||||
s = s[i+1:]
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
r, sz := utf8.DecodeRuneInString(s)
|
|
||||||
if unicode.IsSpace(r) || strings.ContainsRune("&<=#", r) {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
value = append(value, s[:sz]...)
|
|
||||||
s = s[sz:]
|
|
||||||
}
|
|
||||||
return string(value), skipSpace(s), nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func parseSequence(p RuleProcessor, level int, s string) (tail string, err error) {
|
|
||||||
if s = skipSpace(s); s == "" {
|
|
||||||
return s, errors.New("empty sequence")
|
|
||||||
}
|
|
||||||
last := rune(0)
|
|
||||||
for s != "" {
|
|
||||||
r, sz := utf8.DecodeRuneInString(s)
|
|
||||||
s = s[sz:]
|
|
||||||
|
|
||||||
if r == '-' {
|
|
||||||
// We have a range. The first element was already written.
|
|
||||||
if last == 0 {
|
|
||||||
return s, errors.New("range without starter value")
|
|
||||||
}
|
|
||||||
r, sz = utf8.DecodeRuneInString(s)
|
|
||||||
s = s[sz:]
|
|
||||||
if r == utf8.RuneError || r < last {
|
|
||||||
return s, fmt.Errorf("invalid range %q-%q", last, r)
|
|
||||||
}
|
|
||||||
for i := last + 1; i <= r; i++ {
|
|
||||||
if err := p.Insert(level, string(i), "", ""); err != nil {
|
|
||||||
return s, err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
last = 0
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
|
|
||||||
if unicode.IsSpace(r) || unicode.IsPunct(r) {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
|
|
||||||
// normal case
|
|
||||||
if err := p.Insert(level, string(r), "", ""); err != nil {
|
|
||||||
return s, err
|
|
||||||
}
|
|
||||||
last = r
|
|
||||||
}
|
|
||||||
return s, nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func skipSpace(s string) string {
|
|
||||||
return strings.TrimLeftFunc(s, unicode.IsSpace)
|
|
||||||
}
|
|
||||||
|
|
||||||
// consumes returns whether the next byte is ch. If so, it gobbles it by
|
|
||||||
// updating s.
|
|
||||||
func consume(s *string, ch byte) (ok bool) {
|
|
||||||
if *s == "" || (*s)[0] != ch {
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
*s = (*s)[1:]
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
|
|
||||||
// The following code parses Collation rules of CLDR version 24 and before.
|
|
||||||
|
|
||||||
var lmap = map[byte]int{
|
|
||||||
'p': 1,
|
|
||||||
's': 2,
|
|
||||||
't': 3,
|
|
||||||
'i': 5,
|
|
||||||
}
|
|
||||||
|
|
||||||
type rulesElem struct {
|
|
||||||
Rules struct {
|
|
||||||
Common
|
|
||||||
Any []*struct {
|
|
||||||
XMLName xml.Name
|
|
||||||
rule
|
|
||||||
} `xml:",any"`
|
|
||||||
} `xml:"rules"`
|
|
||||||
}
|
|
||||||
|
|
||||||
type rule struct {
|
|
||||||
Value string `xml:",chardata"`
|
|
||||||
Before string `xml:"before,attr"`
|
|
||||||
Any []*struct {
|
|
||||||
XMLName xml.Name
|
|
||||||
rule
|
|
||||||
} `xml:",any"`
|
|
||||||
}
|
|
||||||
|
|
||||||
var emptyValueError = errors.New("cldr: empty rule value")
|
|
||||||
|
|
||||||
func (r *rule) value() (string, error) {
|
|
||||||
// Convert hexadecimal Unicode codepoint notation to a string.
|
|
||||||
s := charRe.ReplaceAllStringFunc(r.Value, replaceUnicode)
|
|
||||||
r.Value = s
|
|
||||||
if s == "" {
|
|
||||||
if len(r.Any) != 1 {
|
|
||||||
return "", emptyValueError
|
|
||||||
}
|
|
||||||
r.Value = fmt.Sprintf(specialAnchor, r.Any[0].XMLName.Local)
|
|
||||||
r.Any = nil
|
|
||||||
} else if len(r.Any) != 0 {
|
|
||||||
return "", fmt.Errorf("cldr: XML elements found in collation rule: %v", r.Any)
|
|
||||||
}
|
|
||||||
return r.Value, nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func (r rule) process(p RuleProcessor, name, context, extend string) error {
|
|
||||||
v, err := r.value()
|
|
||||||
if err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
switch name {
|
|
||||||
case "p", "s", "t", "i":
|
|
||||||
if strings.HasPrefix(v, cldrIndex) {
|
|
||||||
p.Index(v[len(cldrIndex):])
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
if err := p.Insert(lmap[name[0]], v, context, extend); err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
case "pc", "sc", "tc", "ic":
|
|
||||||
level := lmap[name[0]]
|
|
||||||
for _, s := range v {
|
|
||||||
if err := p.Insert(level, string(s), context, extend); err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
default:
|
|
||||||
return fmt.Errorf("cldr: unsupported tag: %q", name)
|
|
||||||
}
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
|
|
||||||
// processXML parses the format of CLDR versions 24 and older.
|
|
||||||
func (c Collation) processXML(p RuleProcessor) (err error) {
|
|
||||||
// Collation is generated and defined in xml.go.
|
|
||||||
var v string
|
|
||||||
for _, r := range c.Rules.Any {
|
|
||||||
switch r.XMLName.Local {
|
|
||||||
case "reset":
|
|
||||||
level := 0
|
|
||||||
switch r.Before {
|
|
||||||
case "primary", "1":
|
|
||||||
level = 1
|
|
||||||
case "secondary", "2":
|
|
||||||
level = 2
|
|
||||||
case "tertiary", "3":
|
|
||||||
level = 3
|
|
||||||
case "":
|
|
||||||
default:
|
|
||||||
return fmt.Errorf("cldr: unknown level %q", r.Before)
|
|
||||||
}
|
|
||||||
v, err = r.value()
|
|
||||||
if err == nil {
|
|
||||||
err = p.Reset(v, level)
|
|
||||||
}
|
|
||||||
case "x":
|
|
||||||
var context, extend string
|
|
||||||
for _, r1 := range r.Any {
|
|
||||||
v, err = r1.value()
|
|
||||||
switch r1.XMLName.Local {
|
|
||||||
case "context":
|
|
||||||
context = v
|
|
||||||
case "extend":
|
|
||||||
extend = v
|
|
||||||
}
|
|
||||||
}
|
|
||||||
for _, r1 := range r.Any {
|
|
||||||
if t := r1.XMLName.Local; t == "context" || t == "extend" {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
r1.rule.process(p, r1.XMLName.Local, context, extend)
|
|
||||||
}
|
|
||||||
default:
|
|
||||||
err = r.rule.process(p, r.XMLName.Local, "", "")
|
|
||||||
}
|
|
||||||
if err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return nil
|
|
||||||
}
|
|
|
@ -1,171 +0,0 @@
|
||||||
// Copyright 2013 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
package cldr
|
|
||||||
|
|
||||||
import (
|
|
||||||
"archive/zip"
|
|
||||||
"bytes"
|
|
||||||
"encoding/xml"
|
|
||||||
"fmt"
|
|
||||||
"io"
|
|
||||||
"io/ioutil"
|
|
||||||
"log"
|
|
||||||
"os"
|
|
||||||
"path/filepath"
|
|
||||||
"regexp"
|
|
||||||
)
|
|
||||||
|
|
||||||
// A Decoder loads an archive of CLDR data.
|
|
||||||
type Decoder struct {
|
|
||||||
dirFilter []string
|
|
||||||
sectionFilter []string
|
|
||||||
loader Loader
|
|
||||||
cldr *CLDR
|
|
||||||
curLocale string
|
|
||||||
}
|
|
||||||
|
|
||||||
// SetSectionFilter takes a list top-level LDML element names to which
|
|
||||||
// evaluation of LDML should be limited. It automatically calls SetDirFilter.
|
|
||||||
func (d *Decoder) SetSectionFilter(filter ...string) {
|
|
||||||
d.sectionFilter = filter
|
|
||||||
// TODO: automatically set dir filter
|
|
||||||
}
|
|
||||||
|
|
||||||
// SetDirFilter limits the loading of LDML XML files of the specied directories.
|
|
||||||
// Note that sections may be split across directories differently for different CLDR versions.
|
|
||||||
// For more robust code, use SetSectionFilter.
|
|
||||||
func (d *Decoder) SetDirFilter(dir ...string) {
|
|
||||||
d.dirFilter = dir
|
|
||||||
}
|
|
||||||
|
|
||||||
// A Loader provides access to the files of a CLDR archive.
|
|
||||||
type Loader interface {
|
|
||||||
Len() int
|
|
||||||
Path(i int) string
|
|
||||||
Reader(i int) (io.ReadCloser, error)
|
|
||||||
}
|
|
||||||
|
|
||||||
var fileRe = regexp.MustCompile(".*/(.*)/(.*)\\.xml")
|
|
||||||
|
|
||||||
// Decode loads and decodes the files represented by l.
|
|
||||||
func (d *Decoder) Decode(l Loader) (cldr *CLDR, err error) {
|
|
||||||
d.cldr = makeCLDR()
|
|
||||||
for i := 0; i < l.Len(); i++ {
|
|
||||||
fname := l.Path(i)
|
|
||||||
if m := fileRe.FindStringSubmatch(fname); m != nil {
|
|
||||||
if len(d.dirFilter) > 0 && !in(d.dirFilter, m[1]) {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
var r io.Reader
|
|
||||||
if r, err = l.Reader(i); err == nil {
|
|
||||||
err = d.decode(m[1], m[2], r)
|
|
||||||
}
|
|
||||||
if err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
d.cldr.finalize(d.sectionFilter)
|
|
||||||
return d.cldr, nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func (d *Decoder) decode(dir, id string, r io.Reader) error {
|
|
||||||
var v interface{}
|
|
||||||
var l *LDML
|
|
||||||
cldr := d.cldr
|
|
||||||
switch {
|
|
||||||
case dir == "supplemental":
|
|
||||||
v = cldr.supp
|
|
||||||
case dir == "transforms":
|
|
||||||
return nil
|
|
||||||
case dir == "bcp47":
|
|
||||||
v = cldr.bcp47
|
|
||||||
case dir == "validity":
|
|
||||||
return nil
|
|
||||||
default:
|
|
||||||
ok := false
|
|
||||||
if v, ok = cldr.locale[id]; !ok {
|
|
||||||
l = &LDML{}
|
|
||||||
v, cldr.locale[id] = l, l
|
|
||||||
}
|
|
||||||
}
|
|
||||||
x := xml.NewDecoder(r)
|
|
||||||
if err := x.Decode(v); err != nil {
|
|
||||||
log.Printf("%s/%s: %v", dir, id, err)
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
if l != nil {
|
|
||||||
if l.Identity == nil {
|
|
||||||
return fmt.Errorf("%s/%s: missing identity element", dir, id)
|
|
||||||
}
|
|
||||||
// TODO: verify when CLDR bug http://unicode.org/cldr/trac/ticket/8970
|
|
||||||
// is resolved.
|
|
||||||
// path := strings.Split(id, "_")
|
|
||||||
// if lang := l.Identity.Language.Type; lang != path[0] {
|
|
||||||
// return fmt.Errorf("%s/%s: language was %s; want %s", dir, id, lang, path[0])
|
|
||||||
// }
|
|
||||||
}
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
|
|
||||||
type pathLoader []string
|
|
||||||
|
|
||||||
func makePathLoader(path string) (pl pathLoader, err error) {
|
|
||||||
err = filepath.Walk(path, func(path string, _ os.FileInfo, err error) error {
|
|
||||||
pl = append(pl, path)
|
|
||||||
return err
|
|
||||||
})
|
|
||||||
return pl, err
|
|
||||||
}
|
|
||||||
|
|
||||||
func (pl pathLoader) Len() int {
|
|
||||||
return len(pl)
|
|
||||||
}
|
|
||||||
|
|
||||||
func (pl pathLoader) Path(i int) string {
|
|
||||||
return pl[i]
|
|
||||||
}
|
|
||||||
|
|
||||||
func (pl pathLoader) Reader(i int) (io.ReadCloser, error) {
|
|
||||||
return os.Open(pl[i])
|
|
||||||
}
|
|
||||||
|
|
||||||
// DecodePath loads CLDR data from the given path.
|
|
||||||
func (d *Decoder) DecodePath(path string) (cldr *CLDR, err error) {
|
|
||||||
loader, err := makePathLoader(path)
|
|
||||||
if err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
return d.Decode(loader)
|
|
||||||
}
|
|
||||||
|
|
||||||
type zipLoader struct {
|
|
||||||
r *zip.Reader
|
|
||||||
}
|
|
||||||
|
|
||||||
func (zl zipLoader) Len() int {
|
|
||||||
return len(zl.r.File)
|
|
||||||
}
|
|
||||||
|
|
||||||
func (zl zipLoader) Path(i int) string {
|
|
||||||
return zl.r.File[i].Name
|
|
||||||
}
|
|
||||||
|
|
||||||
func (zl zipLoader) Reader(i int) (io.ReadCloser, error) {
|
|
||||||
return zl.r.File[i].Open()
|
|
||||||
}
|
|
||||||
|
|
||||||
// DecodeZip loads CLDR data from the zip archive for which r is the source.
|
|
||||||
func (d *Decoder) DecodeZip(r io.Reader) (cldr *CLDR, err error) {
|
|
||||||
buffer, err := ioutil.ReadAll(r)
|
|
||||||
if err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
archive, err := zip.NewReader(bytes.NewReader(buffer), int64(len(buffer)))
|
|
||||||
if err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
return d.Decode(zipLoader{archive})
|
|
||||||
}
|
|
|
@ -1,400 +0,0 @@
|
||||||
// Copyright 2013 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
// This tool generates types for the various XML formats of CLDR.
|
|
||||||
package main
|
|
||||||
|
|
||||||
import (
|
|
||||||
"archive/zip"
|
|
||||||
"bytes"
|
|
||||||
"encoding/xml"
|
|
||||||
"flag"
|
|
||||||
"fmt"
|
|
||||||
"io"
|
|
||||||
"io/ioutil"
|
|
||||||
"log"
|
|
||||||
"os"
|
|
||||||
"regexp"
|
|
||||||
"strings"
|
|
||||||
|
|
||||||
"golang.org/x/text/internal/gen"
|
|
||||||
)
|
|
||||||
|
|
||||||
var outputFile = flag.String("output", "xml.go", "output file name")
|
|
||||||
|
|
||||||
func main() {
|
|
||||||
flag.Parse()
|
|
||||||
|
|
||||||
r := gen.OpenCLDRCoreZip()
|
|
||||||
buffer, err := ioutil.ReadAll(r)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal("Could not read zip file")
|
|
||||||
}
|
|
||||||
r.Close()
|
|
||||||
z, err := zip.NewReader(bytes.NewReader(buffer), int64(len(buffer)))
|
|
||||||
if err != nil {
|
|
||||||
log.Fatalf("Could not read zip archive: %v", err)
|
|
||||||
}
|
|
||||||
|
|
||||||
var buf bytes.Buffer
|
|
||||||
|
|
||||||
version := gen.CLDRVersion()
|
|
||||||
|
|
||||||
for _, dtd := range files {
|
|
||||||
for _, f := range z.File {
|
|
||||||
if strings.HasSuffix(f.Name, dtd.file+".dtd") {
|
|
||||||
r, err := f.Open()
|
|
||||||
failOnError(err)
|
|
||||||
|
|
||||||
b := makeBuilder(&buf, dtd)
|
|
||||||
b.parseDTD(r)
|
|
||||||
b.resolve(b.index[dtd.top[0]])
|
|
||||||
b.write()
|
|
||||||
if b.version != "" && version != b.version {
|
|
||||||
println(f.Name)
|
|
||||||
log.Fatalf("main: inconsistent versions: found %s; want %s", b.version, version)
|
|
||||||
}
|
|
||||||
break
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
fmt.Fprintln(&buf, "// Version is the version of CLDR from which the XML definitions are generated.")
|
|
||||||
fmt.Fprintf(&buf, "const Version = %q\n", version)
|
|
||||||
|
|
||||||
gen.WriteGoFile(*outputFile, "cldr", buf.Bytes())
|
|
||||||
}
|
|
||||||
|
|
||||||
func failOnError(err error) {
|
|
||||||
if err != nil {
|
|
||||||
log.New(os.Stderr, "", log.Lshortfile).Output(2, err.Error())
|
|
||||||
os.Exit(1)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// configuration data per DTD type
|
|
||||||
type dtd struct {
|
|
||||||
file string // base file name
|
|
||||||
root string // Go name of the root XML element
|
|
||||||
top []string // create a different type for this section
|
|
||||||
|
|
||||||
skipElem []string // hard-coded or deprecated elements
|
|
||||||
skipAttr []string // attributes to exclude
|
|
||||||
predefined []string // hard-coded elements exist of the form <name>Elem
|
|
||||||
forceRepeat []string // elements to make slices despite DTD
|
|
||||||
}
|
|
||||||
|
|
||||||
var files = []dtd{
|
|
||||||
{
|
|
||||||
file: "ldmlBCP47",
|
|
||||||
root: "LDMLBCP47",
|
|
||||||
top: []string{"ldmlBCP47"},
|
|
||||||
skipElem: []string{
|
|
||||||
"cldrVersion", // deprecated, not used
|
|
||||||
},
|
|
||||||
},
|
|
||||||
{
|
|
||||||
file: "ldmlSupplemental",
|
|
||||||
root: "SupplementalData",
|
|
||||||
top: []string{"supplementalData"},
|
|
||||||
skipElem: []string{
|
|
||||||
"cldrVersion", // deprecated, not used
|
|
||||||
},
|
|
||||||
forceRepeat: []string{
|
|
||||||
"plurals", // data defined in plurals.xml and ordinals.xml
|
|
||||||
},
|
|
||||||
},
|
|
||||||
{
|
|
||||||
file: "ldml",
|
|
||||||
root: "LDML",
|
|
||||||
top: []string{
|
|
||||||
"ldml", "collation", "calendar", "timeZoneNames", "localeDisplayNames", "numbers",
|
|
||||||
},
|
|
||||||
skipElem: []string{
|
|
||||||
"cp", // not used anywhere
|
|
||||||
"special", // not used anywhere
|
|
||||||
"fallback", // deprecated, not used
|
|
||||||
"alias", // in Common
|
|
||||||
"default", // in Common
|
|
||||||
},
|
|
||||||
skipAttr: []string{
|
|
||||||
"hiraganaQuarternary", // typo in DTD, correct version included as well
|
|
||||||
},
|
|
||||||
predefined: []string{"rules"},
|
|
||||||
},
|
|
||||||
}
|
|
||||||
|
|
||||||
var comments = map[string]string{
|
|
||||||
"ldmlBCP47": `
|
|
||||||
// LDMLBCP47 holds information on allowable values for various variables in LDML.
|
|
||||||
`,
|
|
||||||
"supplementalData": `
|
|
||||||
// SupplementalData holds information relevant for internationalization
|
|
||||||
// and proper use of CLDR, but that is not contained in the locale hierarchy.
|
|
||||||
`,
|
|
||||||
"ldml": `
|
|
||||||
// LDML is the top-level type for locale-specific data.
|
|
||||||
`,
|
|
||||||
"collation": `
|
|
||||||
// Collation contains rules that specify a certain sort-order,
|
|
||||||
// as a tailoring of the root order.
|
|
||||||
// The parsed rules are obtained by passing a RuleProcessor to Collation's
|
|
||||||
// Process method.
|
|
||||||
`,
|
|
||||||
"calendar": `
|
|
||||||
// Calendar specifies the fields used for formatting and parsing dates and times.
|
|
||||||
// The month and quarter names are identified numerically, starting at 1.
|
|
||||||
// The day (of the week) names are identified with short strings, since there is
|
|
||||||
// no universally-accepted numeric designation.
|
|
||||||
`,
|
|
||||||
"dates": `
|
|
||||||
// Dates contains information regarding the format and parsing of dates and times.
|
|
||||||
`,
|
|
||||||
"localeDisplayNames": `
|
|
||||||
// LocaleDisplayNames specifies localized display names for for scripts, languages,
|
|
||||||
// countries, currencies, and variants.
|
|
||||||
`,
|
|
||||||
"numbers": `
|
|
||||||
// Numbers supplies information for formatting and parsing numbers and currencies.
|
|
||||||
`,
|
|
||||||
}
|
|
||||||
|
|
||||||
type element struct {
|
|
||||||
name string // XML element name
|
|
||||||
category string // elements contained by this element
|
|
||||||
signature string // category + attrKey*
|
|
||||||
|
|
||||||
attr []*attribute // attributes supported by this element.
|
|
||||||
sub []struct { // parsed and evaluated sub elements of this element.
|
|
||||||
e *element
|
|
||||||
repeat bool // true if the element needs to be a slice
|
|
||||||
}
|
|
||||||
|
|
||||||
resolved bool // prevent multiple resolutions of this element.
|
|
||||||
}
|
|
||||||
|
|
||||||
type attribute struct {
|
|
||||||
name string
|
|
||||||
key string
|
|
||||||
list []string
|
|
||||||
|
|
||||||
tag string // Go tag
|
|
||||||
}
|
|
||||||
|
|
||||||
var (
|
|
||||||
reHead = regexp.MustCompile(` *(\w+) +([\w\-]+)`)
|
|
||||||
reAttr = regexp.MustCompile(` *(\w+) *(?:(\w+)|\(([\w\- \|]+)\)) *(?:#([A-Z]*) *(?:\"([\.\d+])\")?)? *("[\w\-:]*")?`)
|
|
||||||
reElem = regexp.MustCompile(`^ *(EMPTY|ANY|\(.*\)[\*\+\?]?) *$`)
|
|
||||||
reToken = regexp.MustCompile(`\w\-`)
|
|
||||||
)
|
|
||||||
|
|
||||||
// builder is used to read in the DTD files from CLDR and generate Go code
|
|
||||||
// to be used with the encoding/xml package.
|
|
||||||
type builder struct {
|
|
||||||
w io.Writer
|
|
||||||
index map[string]*element
|
|
||||||
elem []*element
|
|
||||||
info dtd
|
|
||||||
version string
|
|
||||||
}
|
|
||||||
|
|
||||||
func makeBuilder(w io.Writer, d dtd) builder {
|
|
||||||
return builder{
|
|
||||||
w: w,
|
|
||||||
index: make(map[string]*element),
|
|
||||||
elem: []*element{},
|
|
||||||
info: d,
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// parseDTD parses a DTD file.
|
|
||||||
func (b *builder) parseDTD(r io.Reader) {
|
|
||||||
for d := xml.NewDecoder(r); ; {
|
|
||||||
t, err := d.Token()
|
|
||||||
if t == nil {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
failOnError(err)
|
|
||||||
dir, ok := t.(xml.Directive)
|
|
||||||
if !ok {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
m := reHead.FindSubmatch(dir)
|
|
||||||
dir = dir[len(m[0]):]
|
|
||||||
ename := string(m[2])
|
|
||||||
el, elementFound := b.index[ename]
|
|
||||||
switch string(m[1]) {
|
|
||||||
case "ELEMENT":
|
|
||||||
if elementFound {
|
|
||||||
log.Fatal("parseDTD: duplicate entry for element %q", ename)
|
|
||||||
}
|
|
||||||
m := reElem.FindSubmatch(dir)
|
|
||||||
if m == nil {
|
|
||||||
log.Fatalf("parseDTD: invalid element %q", string(dir))
|
|
||||||
}
|
|
||||||
if len(m[0]) != len(dir) {
|
|
||||||
log.Fatal("parseDTD: invalid element %q", string(dir), len(dir), len(m[0]), string(m[0]))
|
|
||||||
}
|
|
||||||
s := string(m[1])
|
|
||||||
el = &element{
|
|
||||||
name: ename,
|
|
||||||
category: s,
|
|
||||||
}
|
|
||||||
b.index[ename] = el
|
|
||||||
case "ATTLIST":
|
|
||||||
if !elementFound {
|
|
||||||
log.Fatalf("parseDTD: unknown element %q", ename)
|
|
||||||
}
|
|
||||||
s := string(dir)
|
|
||||||
m := reAttr.FindStringSubmatch(s)
|
|
||||||
if m == nil {
|
|
||||||
log.Fatal(fmt.Errorf("parseDTD: invalid attribute %q", string(dir)))
|
|
||||||
}
|
|
||||||
if m[4] == "FIXED" {
|
|
||||||
b.version = m[5]
|
|
||||||
} else {
|
|
||||||
switch m[1] {
|
|
||||||
case "draft", "references", "alt", "validSubLocales", "standard" /* in Common */ :
|
|
||||||
case "type", "choice":
|
|
||||||
default:
|
|
||||||
el.attr = append(el.attr, &attribute{
|
|
||||||
name: m[1],
|
|
||||||
key: s,
|
|
||||||
list: reToken.FindAllString(m[3], -1),
|
|
||||||
})
|
|
||||||
el.signature = fmt.Sprintf("%s=%s+%s", el.signature, m[1], m[2])
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
var reCat = regexp.MustCompile(`[ ,\|]*(?:(\(|\)|\#?[\w_-]+)([\*\+\?]?))?`)
|
|
||||||
|
|
||||||
// resolve takes a parsed element and converts it into structured data
|
|
||||||
// that can be used to generate the XML code.
|
|
||||||
func (b *builder) resolve(e *element) {
|
|
||||||
if e.resolved {
|
|
||||||
return
|
|
||||||
}
|
|
||||||
b.elem = append(b.elem, e)
|
|
||||||
e.resolved = true
|
|
||||||
s := e.category
|
|
||||||
found := make(map[string]bool)
|
|
||||||
sequenceStart := []int{}
|
|
||||||
for len(s) > 0 {
|
|
||||||
m := reCat.FindStringSubmatch(s)
|
|
||||||
if m == nil {
|
|
||||||
log.Fatalf("%s: invalid category string %q", e.name, s)
|
|
||||||
}
|
|
||||||
repeat := m[2] == "*" || m[2] == "+" || in(b.info.forceRepeat, m[1])
|
|
||||||
switch m[1] {
|
|
||||||
case "":
|
|
||||||
case "(":
|
|
||||||
sequenceStart = append(sequenceStart, len(e.sub))
|
|
||||||
case ")":
|
|
||||||
if len(sequenceStart) == 0 {
|
|
||||||
log.Fatalf("%s: unmatched closing parenthesis", e.name)
|
|
||||||
}
|
|
||||||
for i := sequenceStart[len(sequenceStart)-1]; i < len(e.sub); i++ {
|
|
||||||
e.sub[i].repeat = e.sub[i].repeat || repeat
|
|
||||||
}
|
|
||||||
sequenceStart = sequenceStart[:len(sequenceStart)-1]
|
|
||||||
default:
|
|
||||||
if in(b.info.skipElem, m[1]) {
|
|
||||||
} else if sub, ok := b.index[m[1]]; ok {
|
|
||||||
if !found[sub.name] {
|
|
||||||
e.sub = append(e.sub, struct {
|
|
||||||
e *element
|
|
||||||
repeat bool
|
|
||||||
}{sub, repeat})
|
|
||||||
found[sub.name] = true
|
|
||||||
b.resolve(sub)
|
|
||||||
}
|
|
||||||
} else if m[1] == "#PCDATA" || m[1] == "ANY" {
|
|
||||||
} else if m[1] != "EMPTY" {
|
|
||||||
log.Fatalf("resolve:%s: element %q not found", e.name, m[1])
|
|
||||||
}
|
|
||||||
}
|
|
||||||
s = s[len(m[0]):]
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// return true if s is contained in set.
|
|
||||||
func in(set []string, s string) bool {
|
|
||||||
for _, v := range set {
|
|
||||||
if v == s {
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
|
|
||||||
var repl = strings.NewReplacer("-", " ", "_", " ")
|
|
||||||
|
|
||||||
// title puts the first character or each character following '_' in title case and
|
|
||||||
// removes all occurrences of '_'.
|
|
||||||
func title(s string) string {
|
|
||||||
return strings.Replace(strings.Title(repl.Replace(s)), " ", "", -1)
|
|
||||||
}
|
|
||||||
|
|
||||||
// writeElem generates Go code for a single element, recursively.
|
|
||||||
func (b *builder) writeElem(tab int, e *element) {
|
|
||||||
p := func(f string, x ...interface{}) {
|
|
||||||
f = strings.Replace(f, "\n", "\n"+strings.Repeat("\t", tab), -1)
|
|
||||||
fmt.Fprintf(b.w, f, x...)
|
|
||||||
}
|
|
||||||
if len(e.sub) == 0 && len(e.attr) == 0 {
|
|
||||||
p("Common")
|
|
||||||
return
|
|
||||||
}
|
|
||||||
p("struct {")
|
|
||||||
tab++
|
|
||||||
p("\nCommon")
|
|
||||||
for _, attr := range e.attr {
|
|
||||||
if !in(b.info.skipAttr, attr.name) {
|
|
||||||
p("\n%s string `xml:\"%s,attr\"`", title(attr.name), attr.name)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
for _, sub := range e.sub {
|
|
||||||
if in(b.info.predefined, sub.e.name) {
|
|
||||||
p("\n%sElem", sub.e.name)
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
if in(b.info.skipElem, sub.e.name) {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
p("\n%s ", title(sub.e.name))
|
|
||||||
if sub.repeat {
|
|
||||||
p("[]")
|
|
||||||
}
|
|
||||||
p("*")
|
|
||||||
if in(b.info.top, sub.e.name) {
|
|
||||||
p(title(sub.e.name))
|
|
||||||
} else {
|
|
||||||
b.writeElem(tab, sub.e)
|
|
||||||
}
|
|
||||||
p(" `xml:\"%s\"`", sub.e.name)
|
|
||||||
}
|
|
||||||
tab--
|
|
||||||
p("\n}")
|
|
||||||
}
|
|
||||||
|
|
||||||
// write generates the Go XML code.
|
|
||||||
func (b *builder) write() {
|
|
||||||
for i, name := range b.info.top {
|
|
||||||
e := b.index[name]
|
|
||||||
if e != nil {
|
|
||||||
fmt.Fprintf(b.w, comments[name])
|
|
||||||
name := title(e.name)
|
|
||||||
if i == 0 {
|
|
||||||
name = b.info.root
|
|
||||||
}
|
|
||||||
fmt.Fprintf(b.w, "type %s ", name)
|
|
||||||
b.writeElem(0, e)
|
|
||||||
fmt.Fprint(b.w, "\n")
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
|
@ -1,602 +0,0 @@
|
||||||
// Copyright 2013 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
package cldr
|
|
||||||
|
|
||||||
// This file implements the various inheritance constructs defined by LDML.
|
|
||||||
// See http://www.unicode.org/reports/tr35/#Inheritance_and_Validity
|
|
||||||
// for more details.
|
|
||||||
|
|
||||||
import (
|
|
||||||
"fmt"
|
|
||||||
"log"
|
|
||||||
"reflect"
|
|
||||||
"regexp"
|
|
||||||
"sort"
|
|
||||||
"strings"
|
|
||||||
)
|
|
||||||
|
|
||||||
// fieldIter iterates over fields in a struct. It includes
|
|
||||||
// fields of embedded structs.
|
|
||||||
type fieldIter struct {
|
|
||||||
v reflect.Value
|
|
||||||
index, n []int
|
|
||||||
}
|
|
||||||
|
|
||||||
func iter(v reflect.Value) fieldIter {
|
|
||||||
if v.Kind() != reflect.Struct {
|
|
||||||
log.Panicf("value %v must be a struct", v)
|
|
||||||
}
|
|
||||||
i := fieldIter{
|
|
||||||
v: v,
|
|
||||||
index: []int{0},
|
|
||||||
n: []int{v.NumField()},
|
|
||||||
}
|
|
||||||
i.descent()
|
|
||||||
return i
|
|
||||||
}
|
|
||||||
|
|
||||||
func (i *fieldIter) descent() {
|
|
||||||
for f := i.field(); f.Anonymous && f.Type.NumField() > 0; f = i.field() {
|
|
||||||
i.index = append(i.index, 0)
|
|
||||||
i.n = append(i.n, f.Type.NumField())
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
func (i *fieldIter) done() bool {
|
|
||||||
return len(i.index) == 1 && i.index[0] >= i.n[0]
|
|
||||||
}
|
|
||||||
|
|
||||||
func skip(f reflect.StructField) bool {
|
|
||||||
return !f.Anonymous && (f.Name[0] < 'A' || f.Name[0] > 'Z')
|
|
||||||
}
|
|
||||||
|
|
||||||
func (i *fieldIter) next() {
|
|
||||||
for {
|
|
||||||
k := len(i.index) - 1
|
|
||||||
i.index[k]++
|
|
||||||
if i.index[k] < i.n[k] {
|
|
||||||
if !skip(i.field()) {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
} else {
|
|
||||||
if k == 0 {
|
|
||||||
return
|
|
||||||
}
|
|
||||||
i.index = i.index[:k]
|
|
||||||
i.n = i.n[:k]
|
|
||||||
}
|
|
||||||
}
|
|
||||||
i.descent()
|
|
||||||
}
|
|
||||||
|
|
||||||
func (i *fieldIter) value() reflect.Value {
|
|
||||||
return i.v.FieldByIndex(i.index)
|
|
||||||
}
|
|
||||||
|
|
||||||
func (i *fieldIter) field() reflect.StructField {
|
|
||||||
return i.v.Type().FieldByIndex(i.index)
|
|
||||||
}
|
|
||||||
|
|
||||||
type visitor func(v reflect.Value) error
|
|
||||||
|
|
||||||
var stopDescent = fmt.Errorf("do not recurse")
|
|
||||||
|
|
||||||
func (f visitor) visit(x interface{}) error {
|
|
||||||
return f.visitRec(reflect.ValueOf(x))
|
|
||||||
}
|
|
||||||
|
|
||||||
// visit recursively calls f on all nodes in v.
|
|
||||||
func (f visitor) visitRec(v reflect.Value) error {
|
|
||||||
if v.Kind() == reflect.Ptr {
|
|
||||||
if v.IsNil() {
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
return f.visitRec(v.Elem())
|
|
||||||
}
|
|
||||||
if err := f(v); err != nil {
|
|
||||||
if err == stopDescent {
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
switch v.Kind() {
|
|
||||||
case reflect.Struct:
|
|
||||||
for i := iter(v); !i.done(); i.next() {
|
|
||||||
if err := f.visitRec(i.value()); err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
case reflect.Slice:
|
|
||||||
for i := 0; i < v.Len(); i++ {
|
|
||||||
if err := f.visitRec(v.Index(i)); err != nil {
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
|
|
||||||
// getPath is used for error reporting purposes only.
|
|
||||||
func getPath(e Elem) string {
|
|
||||||
if e == nil {
|
|
||||||
return "<nil>"
|
|
||||||
}
|
|
||||||
if e.enclosing() == nil {
|
|
||||||
return e.GetCommon().name
|
|
||||||
}
|
|
||||||
if e.GetCommon().Type == "" {
|
|
||||||
return fmt.Sprintf("%s.%s", getPath(e.enclosing()), e.GetCommon().name)
|
|
||||||
}
|
|
||||||
return fmt.Sprintf("%s.%s[type=%s]", getPath(e.enclosing()), e.GetCommon().name, e.GetCommon().Type)
|
|
||||||
}
|
|
||||||
|
|
||||||
// xmlName returns the xml name of the element or attribute
|
|
||||||
func xmlName(f reflect.StructField) (name string, attr bool) {
|
|
||||||
tags := strings.Split(f.Tag.Get("xml"), ",")
|
|
||||||
for _, s := range tags {
|
|
||||||
attr = attr || s == "attr"
|
|
||||||
}
|
|
||||||
return tags[0], attr
|
|
||||||
}
|
|
||||||
|
|
||||||
func findField(v reflect.Value, key string) (reflect.Value, error) {
|
|
||||||
v = reflect.Indirect(v)
|
|
||||||
for i := iter(v); !i.done(); i.next() {
|
|
||||||
if n, _ := xmlName(i.field()); n == key {
|
|
||||||
return i.value(), nil
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return reflect.Value{}, fmt.Errorf("cldr: no field %q in element %#v", key, v.Interface())
|
|
||||||
}
|
|
||||||
|
|
||||||
var xpathPart = regexp.MustCompile(`(\pL+)(?:\[@(\pL+)='([\w-]+)'\])?`)
|
|
||||||
|
|
||||||
func walkXPath(e Elem, path string) (res Elem, err error) {
|
|
||||||
for _, c := range strings.Split(path, "/") {
|
|
||||||
if c == ".." {
|
|
||||||
if e = e.enclosing(); e == nil {
|
|
||||||
panic("path ..")
|
|
||||||
return nil, fmt.Errorf(`cldr: ".." moves past root in path %q`, path)
|
|
||||||
}
|
|
||||||
continue
|
|
||||||
} else if c == "" {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
m := xpathPart.FindStringSubmatch(c)
|
|
||||||
if len(m) == 0 || len(m[0]) != len(c) {
|
|
||||||
return nil, fmt.Errorf("cldr: syntax error in path component %q", c)
|
|
||||||
}
|
|
||||||
v, err := findField(reflect.ValueOf(e), m[1])
|
|
||||||
if err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
switch v.Kind() {
|
|
||||||
case reflect.Slice:
|
|
||||||
i := 0
|
|
||||||
if m[2] != "" || v.Len() > 1 {
|
|
||||||
if m[2] == "" {
|
|
||||||
m[2] = "type"
|
|
||||||
if m[3] = e.GetCommon().Default(); m[3] == "" {
|
|
||||||
return nil, fmt.Errorf("cldr: type selector or default value needed for element %s", m[1])
|
|
||||||
}
|
|
||||||
}
|
|
||||||
for ; i < v.Len(); i++ {
|
|
||||||
vi := v.Index(i)
|
|
||||||
key, err := findField(vi.Elem(), m[2])
|
|
||||||
if err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
key = reflect.Indirect(key)
|
|
||||||
if key.Kind() == reflect.String && key.String() == m[3] {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
if i == v.Len() || v.Index(i).IsNil() {
|
|
||||||
return nil, fmt.Errorf("no %s found with %s==%s", m[1], m[2], m[3])
|
|
||||||
}
|
|
||||||
e = v.Index(i).Interface().(Elem)
|
|
||||||
case reflect.Ptr:
|
|
||||||
if v.IsNil() {
|
|
||||||
return nil, fmt.Errorf("cldr: element %q not found within element %q", m[1], e.GetCommon().name)
|
|
||||||
}
|
|
||||||
var ok bool
|
|
||||||
if e, ok = v.Interface().(Elem); !ok {
|
|
||||||
return nil, fmt.Errorf("cldr: %q is not an XML element", m[1])
|
|
||||||
} else if m[2] != "" || m[3] != "" {
|
|
||||||
return nil, fmt.Errorf("cldr: no type selector allowed for element %s", m[1])
|
|
||||||
}
|
|
||||||
default:
|
|
||||||
return nil, fmt.Errorf("cldr: %q is not an XML element", m[1])
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return e, nil
|
|
||||||
}
|
|
||||||
|
|
||||||
const absPrefix = "//ldml/"
|
|
||||||
|
|
||||||
func (cldr *CLDR) resolveAlias(e Elem, src, path string) (res Elem, err error) {
|
|
||||||
if src != "locale" {
|
|
||||||
if !strings.HasPrefix(path, absPrefix) {
|
|
||||||
return nil, fmt.Errorf("cldr: expected absolute path, found %q", path)
|
|
||||||
}
|
|
||||||
path = path[len(absPrefix):]
|
|
||||||
if e, err = cldr.resolve(src); err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return walkXPath(e, path)
|
|
||||||
}
|
|
||||||
|
|
||||||
func (cldr *CLDR) resolveAndMergeAlias(e Elem) error {
|
|
||||||
alias := e.GetCommon().Alias
|
|
||||||
if alias == nil {
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
a, err := cldr.resolveAlias(e, alias.Source, alias.Path)
|
|
||||||
if err != nil {
|
|
||||||
return fmt.Errorf("%v: error evaluating path %q: %v", getPath(e), alias.Path, err)
|
|
||||||
}
|
|
||||||
// Ensure alias node was already evaluated. TODO: avoid double evaluation.
|
|
||||||
err = cldr.resolveAndMergeAlias(a)
|
|
||||||
v := reflect.ValueOf(e).Elem()
|
|
||||||
for i := iter(reflect.ValueOf(a).Elem()); !i.done(); i.next() {
|
|
||||||
if vv := i.value(); vv.Kind() != reflect.Ptr || !vv.IsNil() {
|
|
||||||
if _, attr := xmlName(i.field()); !attr {
|
|
||||||
v.FieldByIndex(i.index).Set(vv)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
|
|
||||||
func (cldr *CLDR) aliasResolver() visitor {
|
|
||||||
return func(v reflect.Value) (err error) {
|
|
||||||
if e, ok := v.Addr().Interface().(Elem); ok {
|
|
||||||
err = cldr.resolveAndMergeAlias(e)
|
|
||||||
if err == nil && blocking[e.GetCommon().name] {
|
|
||||||
return stopDescent
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// elements within blocking elements do not inherit.
|
|
||||||
// Taken from CLDR's supplementalMetaData.xml.
|
|
||||||
var blocking = map[string]bool{
|
|
||||||
"identity": true,
|
|
||||||
"supplementalData": true,
|
|
||||||
"cldrTest": true,
|
|
||||||
"collation": true,
|
|
||||||
"transform": true,
|
|
||||||
}
|
|
||||||
|
|
||||||
// Distinguishing attributes affect inheritance; two elements with different
|
|
||||||
// distinguishing attributes are treated as different for purposes of inheritance,
|
|
||||||
// except when such attributes occur in the indicated elements.
|
|
||||||
// Taken from CLDR's supplementalMetaData.xml.
|
|
||||||
var distinguishing = map[string][]string{
|
|
||||||
"key": nil,
|
|
||||||
"request_id": nil,
|
|
||||||
"id": nil,
|
|
||||||
"registry": nil,
|
|
||||||
"alt": nil,
|
|
||||||
"iso4217": nil,
|
|
||||||
"iso3166": nil,
|
|
||||||
"mzone": nil,
|
|
||||||
"from": nil,
|
|
||||||
"to": nil,
|
|
||||||
"type": []string{
|
|
||||||
"abbreviationFallback",
|
|
||||||
"default",
|
|
||||||
"mapping",
|
|
||||||
"measurementSystem",
|
|
||||||
"preferenceOrdering",
|
|
||||||
},
|
|
||||||
"numberSystem": nil,
|
|
||||||
}
|
|
||||||
|
|
||||||
func in(set []string, s string) bool {
|
|
||||||
for _, v := range set {
|
|
||||||
if v == s {
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
|
|
||||||
// attrKey computes a key based on the distinguishable attributes of
|
|
||||||
// an element and it's values.
|
|
||||||
func attrKey(v reflect.Value, exclude ...string) string {
|
|
||||||
parts := []string{}
|
|
||||||
ename := v.Interface().(Elem).GetCommon().name
|
|
||||||
v = v.Elem()
|
|
||||||
for i := iter(v); !i.done(); i.next() {
|
|
||||||
if name, attr := xmlName(i.field()); attr {
|
|
||||||
if except, ok := distinguishing[name]; ok && !in(exclude, name) && !in(except, ename) {
|
|
||||||
v := i.value()
|
|
||||||
if v.Kind() == reflect.Ptr {
|
|
||||||
v = v.Elem()
|
|
||||||
}
|
|
||||||
if v.IsValid() {
|
|
||||||
parts = append(parts, fmt.Sprintf("%s=%s", name, v.String()))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
sort.Strings(parts)
|
|
||||||
return strings.Join(parts, ";")
|
|
||||||
}
|
|
||||||
|
|
||||||
// Key returns a key for e derived from all distinguishing attributes
|
|
||||||
// except those specified by exclude.
|
|
||||||
func Key(e Elem, exclude ...string) string {
|
|
||||||
return attrKey(reflect.ValueOf(e), exclude...)
|
|
||||||
}
|
|
||||||
|
|
||||||
// linkEnclosing sets the enclosing element as well as the name
|
|
||||||
// for all sub-elements of child, recursively.
|
|
||||||
func linkEnclosing(parent, child Elem) {
|
|
||||||
child.setEnclosing(parent)
|
|
||||||
v := reflect.ValueOf(child).Elem()
|
|
||||||
for i := iter(v); !i.done(); i.next() {
|
|
||||||
vf := i.value()
|
|
||||||
if vf.Kind() == reflect.Slice {
|
|
||||||
for j := 0; j < vf.Len(); j++ {
|
|
||||||
linkEnclosing(child, vf.Index(j).Interface().(Elem))
|
|
||||||
}
|
|
||||||
} else if vf.Kind() == reflect.Ptr && !vf.IsNil() && vf.Elem().Kind() == reflect.Struct {
|
|
||||||
linkEnclosing(child, vf.Interface().(Elem))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
func setNames(e Elem, name string) {
|
|
||||||
e.setName(name)
|
|
||||||
v := reflect.ValueOf(e).Elem()
|
|
||||||
for i := iter(v); !i.done(); i.next() {
|
|
||||||
vf := i.value()
|
|
||||||
name, _ = xmlName(i.field())
|
|
||||||
if vf.Kind() == reflect.Slice {
|
|
||||||
for j := 0; j < vf.Len(); j++ {
|
|
||||||
setNames(vf.Index(j).Interface().(Elem), name)
|
|
||||||
}
|
|
||||||
} else if vf.Kind() == reflect.Ptr && !vf.IsNil() && vf.Elem().Kind() == reflect.Struct {
|
|
||||||
setNames(vf.Interface().(Elem), name)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// deepCopy copies elements of v recursively. All elements of v that may
|
|
||||||
// be modified by inheritance are explicitly copied.
|
|
||||||
func deepCopy(v reflect.Value) reflect.Value {
|
|
||||||
switch v.Kind() {
|
|
||||||
case reflect.Ptr:
|
|
||||||
if v.IsNil() || v.Elem().Kind() != reflect.Struct {
|
|
||||||
return v
|
|
||||||
}
|
|
||||||
nv := reflect.New(v.Elem().Type())
|
|
||||||
nv.Elem().Set(v.Elem())
|
|
||||||
deepCopyRec(nv.Elem(), v.Elem())
|
|
||||||
return nv
|
|
||||||
case reflect.Slice:
|
|
||||||
nv := reflect.MakeSlice(v.Type(), v.Len(), v.Len())
|
|
||||||
for i := 0; i < v.Len(); i++ {
|
|
||||||
deepCopyRec(nv.Index(i), v.Index(i))
|
|
||||||
}
|
|
||||||
return nv
|
|
||||||
}
|
|
||||||
panic("deepCopy: must be called with pointer or slice")
|
|
||||||
}
|
|
||||||
|
|
||||||
// deepCopyRec is only called by deepCopy.
|
|
||||||
func deepCopyRec(nv, v reflect.Value) {
|
|
||||||
if v.Kind() == reflect.Struct {
|
|
||||||
t := v.Type()
|
|
||||||
for i := 0; i < v.NumField(); i++ {
|
|
||||||
if name, attr := xmlName(t.Field(i)); name != "" && !attr {
|
|
||||||
deepCopyRec(nv.Field(i), v.Field(i))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
} else {
|
|
||||||
nv.Set(deepCopy(v))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// newNode is used to insert a missing node during inheritance.
|
|
||||||
func (cldr *CLDR) newNode(v, enc reflect.Value) reflect.Value {
|
|
||||||
n := reflect.New(v.Type())
|
|
||||||
for i := iter(v); !i.done(); i.next() {
|
|
||||||
if name, attr := xmlName(i.field()); name == "" || attr {
|
|
||||||
n.Elem().FieldByIndex(i.index).Set(i.value())
|
|
||||||
}
|
|
||||||
}
|
|
||||||
n.Interface().(Elem).GetCommon().setEnclosing(enc.Addr().Interface().(Elem))
|
|
||||||
return n
|
|
||||||
}
|
|
||||||
|
|
||||||
// v, parent must be pointers to struct
|
|
||||||
func (cldr *CLDR) inheritFields(v, parent reflect.Value) (res reflect.Value, err error) {
|
|
||||||
t := v.Type()
|
|
||||||
nv := reflect.New(t)
|
|
||||||
nv.Elem().Set(v)
|
|
||||||
for i := iter(v); !i.done(); i.next() {
|
|
||||||
vf := i.value()
|
|
||||||
f := i.field()
|
|
||||||
name, attr := xmlName(f)
|
|
||||||
if name == "" || attr {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
pf := parent.FieldByIndex(i.index)
|
|
||||||
if blocking[name] {
|
|
||||||
if vf.IsNil() {
|
|
||||||
vf = pf
|
|
||||||
}
|
|
||||||
nv.Elem().FieldByIndex(i.index).Set(deepCopy(vf))
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
switch f.Type.Kind() {
|
|
||||||
case reflect.Ptr:
|
|
||||||
if f.Type.Elem().Kind() == reflect.Struct {
|
|
||||||
if !vf.IsNil() {
|
|
||||||
if vf, err = cldr.inheritStructPtr(vf, pf); err != nil {
|
|
||||||
return reflect.Value{}, err
|
|
||||||
}
|
|
||||||
vf.Interface().(Elem).setEnclosing(nv.Interface().(Elem))
|
|
||||||
nv.Elem().FieldByIndex(i.index).Set(vf)
|
|
||||||
} else if !pf.IsNil() {
|
|
||||||
n := cldr.newNode(pf.Elem(), v)
|
|
||||||
if vf, err = cldr.inheritStructPtr(n, pf); err != nil {
|
|
||||||
return reflect.Value{}, err
|
|
||||||
}
|
|
||||||
vf.Interface().(Elem).setEnclosing(nv.Interface().(Elem))
|
|
||||||
nv.Elem().FieldByIndex(i.index).Set(vf)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
case reflect.Slice:
|
|
||||||
vf, err := cldr.inheritSlice(nv.Elem(), vf, pf)
|
|
||||||
if err != nil {
|
|
||||||
return reflect.Zero(t), err
|
|
||||||
}
|
|
||||||
nv.Elem().FieldByIndex(i.index).Set(vf)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return nv, nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func root(e Elem) *LDML {
|
|
||||||
for ; e.enclosing() != nil; e = e.enclosing() {
|
|
||||||
}
|
|
||||||
return e.(*LDML)
|
|
||||||
}
|
|
||||||
|
|
||||||
// inheritStructPtr first merges possible aliases in with v and then inherits
|
|
||||||
// any underspecified elements from parent.
|
|
||||||
func (cldr *CLDR) inheritStructPtr(v, parent reflect.Value) (r reflect.Value, err error) {
|
|
||||||
if !v.IsNil() {
|
|
||||||
e := v.Interface().(Elem).GetCommon()
|
|
||||||
alias := e.Alias
|
|
||||||
if alias == nil && !parent.IsNil() {
|
|
||||||
alias = parent.Interface().(Elem).GetCommon().Alias
|
|
||||||
}
|
|
||||||
if alias != nil {
|
|
||||||
a, err := cldr.resolveAlias(v.Interface().(Elem), alias.Source, alias.Path)
|
|
||||||
if a != nil {
|
|
||||||
if v, err = cldr.inheritFields(v.Elem(), reflect.ValueOf(a).Elem()); err != nil {
|
|
||||||
return reflect.Value{}, err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
if !parent.IsNil() {
|
|
||||||
return cldr.inheritFields(v.Elem(), parent.Elem())
|
|
||||||
}
|
|
||||||
} else if parent.IsNil() {
|
|
||||||
panic("should not reach here")
|
|
||||||
}
|
|
||||||
return v, nil
|
|
||||||
}
|
|
||||||
|
|
||||||
// Must be slice of struct pointers.
|
|
||||||
func (cldr *CLDR) inheritSlice(enc, v, parent reflect.Value) (res reflect.Value, err error) {
|
|
||||||
t := v.Type()
|
|
||||||
index := make(map[string]reflect.Value)
|
|
||||||
if !v.IsNil() {
|
|
||||||
for i := 0; i < v.Len(); i++ {
|
|
||||||
vi := v.Index(i)
|
|
||||||
key := attrKey(vi)
|
|
||||||
index[key] = vi
|
|
||||||
}
|
|
||||||
}
|
|
||||||
if !parent.IsNil() {
|
|
||||||
for i := 0; i < parent.Len(); i++ {
|
|
||||||
vi := parent.Index(i)
|
|
||||||
key := attrKey(vi)
|
|
||||||
if w, ok := index[key]; ok {
|
|
||||||
index[key], err = cldr.inheritStructPtr(w, vi)
|
|
||||||
} else {
|
|
||||||
n := cldr.newNode(vi.Elem(), enc)
|
|
||||||
index[key], err = cldr.inheritStructPtr(n, vi)
|
|
||||||
}
|
|
||||||
index[key].Interface().(Elem).setEnclosing(enc.Addr().Interface().(Elem))
|
|
||||||
if err != nil {
|
|
||||||
return v, err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
keys := make([]string, 0, len(index))
|
|
||||||
for k, _ := range index {
|
|
||||||
keys = append(keys, k)
|
|
||||||
}
|
|
||||||
sort.Strings(keys)
|
|
||||||
sl := reflect.MakeSlice(t, len(index), len(index))
|
|
||||||
for i, k := range keys {
|
|
||||||
sl.Index(i).Set(index[k])
|
|
||||||
}
|
|
||||||
return sl, nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func parentLocale(loc string) string {
|
|
||||||
parts := strings.Split(loc, "_")
|
|
||||||
if len(parts) == 1 {
|
|
||||||
return "root"
|
|
||||||
}
|
|
||||||
parts = parts[:len(parts)-1]
|
|
||||||
key := strings.Join(parts, "_")
|
|
||||||
return key
|
|
||||||
}
|
|
||||||
|
|
||||||
func (cldr *CLDR) resolve(loc string) (res *LDML, err error) {
|
|
||||||
if r := cldr.resolved[loc]; r != nil {
|
|
||||||
return r, nil
|
|
||||||
}
|
|
||||||
x := cldr.RawLDML(loc)
|
|
||||||
if x == nil {
|
|
||||||
return nil, fmt.Errorf("cldr: unknown locale %q", loc)
|
|
||||||
}
|
|
||||||
var v reflect.Value
|
|
||||||
if loc == "root" {
|
|
||||||
x = deepCopy(reflect.ValueOf(x)).Interface().(*LDML)
|
|
||||||
linkEnclosing(nil, x)
|
|
||||||
err = cldr.aliasResolver().visit(x)
|
|
||||||
} else {
|
|
||||||
key := parentLocale(loc)
|
|
||||||
var parent *LDML
|
|
||||||
for ; cldr.locale[key] == nil; key = parentLocale(key) {
|
|
||||||
}
|
|
||||||
if parent, err = cldr.resolve(key); err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
v, err = cldr.inheritFields(reflect.ValueOf(x).Elem(), reflect.ValueOf(parent).Elem())
|
|
||||||
x = v.Interface().(*LDML)
|
|
||||||
linkEnclosing(nil, x)
|
|
||||||
}
|
|
||||||
if err != nil {
|
|
||||||
return nil, err
|
|
||||||
}
|
|
||||||
cldr.resolved[loc] = x
|
|
||||||
return x, err
|
|
||||||
}
|
|
||||||
|
|
||||||
// finalize finalizes the initialization of the raw LDML structs. It also
|
|
||||||
// removed unwanted fields, as specified by filter, so that they will not
|
|
||||||
// be unnecessarily evaluated.
|
|
||||||
func (cldr *CLDR) finalize(filter []string) {
|
|
||||||
for _, x := range cldr.locale {
|
|
||||||
if filter != nil {
|
|
||||||
v := reflect.ValueOf(x).Elem()
|
|
||||||
t := v.Type()
|
|
||||||
for i := 0; i < v.NumField(); i++ {
|
|
||||||
f := t.Field(i)
|
|
||||||
name, _ := xmlName(f)
|
|
||||||
if name != "" && name != "identity" && !in(filter, name) {
|
|
||||||
v.Field(i).Set(reflect.Zero(f.Type))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
linkEnclosing(nil, x) // for resolving aliases and paths
|
|
||||||
setNames(x, "ldml")
|
|
||||||
}
|
|
||||||
}
|
|
|
@ -1,144 +0,0 @@
|
||||||
// Copyright 2013 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
package cldr
|
|
||||||
|
|
||||||
import (
|
|
||||||
"fmt"
|
|
||||||
"reflect"
|
|
||||||
"sort"
|
|
||||||
)
|
|
||||||
|
|
||||||
// Slice provides utilities for modifying slices of elements.
|
|
||||||
// It can be wrapped around any slice of which the element type implements
|
|
||||||
// interface Elem.
|
|
||||||
type Slice struct {
|
|
||||||
ptr reflect.Value
|
|
||||||
typ reflect.Type
|
|
||||||
}
|
|
||||||
|
|
||||||
// Value returns the reflect.Value of the underlying slice.
|
|
||||||
func (s *Slice) Value() reflect.Value {
|
|
||||||
return s.ptr.Elem()
|
|
||||||
}
|
|
||||||
|
|
||||||
// MakeSlice wraps a pointer to a slice of Elems.
|
|
||||||
// It replaces the array pointed to by the slice so that subsequent modifications
|
|
||||||
// do not alter the data in a CLDR type.
|
|
||||||
// It panics if an incorrect type is passed.
|
|
||||||
func MakeSlice(slicePtr interface{}) Slice {
|
|
||||||
ptr := reflect.ValueOf(slicePtr)
|
|
||||||
if ptr.Kind() != reflect.Ptr {
|
|
||||||
panic(fmt.Sprintf("MakeSlice: argument must be pointer to slice, found %v", ptr.Type()))
|
|
||||||
}
|
|
||||||
sl := ptr.Elem()
|
|
||||||
if sl.Kind() != reflect.Slice {
|
|
||||||
panic(fmt.Sprintf("MakeSlice: argument must point to a slice, found %v", sl.Type()))
|
|
||||||
}
|
|
||||||
intf := reflect.TypeOf((*Elem)(nil)).Elem()
|
|
||||||
if !sl.Type().Elem().Implements(intf) {
|
|
||||||
panic(fmt.Sprintf("MakeSlice: element type of slice (%v) does not implement Elem", sl.Type().Elem()))
|
|
||||||
}
|
|
||||||
nsl := reflect.MakeSlice(sl.Type(), sl.Len(), sl.Len())
|
|
||||||
reflect.Copy(nsl, sl)
|
|
||||||
sl.Set(nsl)
|
|
||||||
return Slice{
|
|
||||||
ptr: ptr,
|
|
||||||
typ: sl.Type().Elem().Elem(),
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
func (s Slice) indexForAttr(a string) []int {
|
|
||||||
for i := iter(reflect.Zero(s.typ)); !i.done(); i.next() {
|
|
||||||
if n, _ := xmlName(i.field()); n == a {
|
|
||||||
return i.index
|
|
||||||
}
|
|
||||||
}
|
|
||||||
panic(fmt.Sprintf("MakeSlice: no attribute %q for type %v", a, s.typ))
|
|
||||||
}
|
|
||||||
|
|
||||||
// Filter filters s to only include elements for which fn returns true.
|
|
||||||
func (s Slice) Filter(fn func(e Elem) bool) {
|
|
||||||
k := 0
|
|
||||||
sl := s.Value()
|
|
||||||
for i := 0; i < sl.Len(); i++ {
|
|
||||||
vi := sl.Index(i)
|
|
||||||
if fn(vi.Interface().(Elem)) {
|
|
||||||
sl.Index(k).Set(vi)
|
|
||||||
k++
|
|
||||||
}
|
|
||||||
}
|
|
||||||
sl.Set(sl.Slice(0, k))
|
|
||||||
}
|
|
||||||
|
|
||||||
// Group finds elements in s for which fn returns the same value and groups
|
|
||||||
// them in a new Slice.
|
|
||||||
func (s Slice) Group(fn func(e Elem) string) []Slice {
|
|
||||||
m := make(map[string][]reflect.Value)
|
|
||||||
sl := s.Value()
|
|
||||||
for i := 0; i < sl.Len(); i++ {
|
|
||||||
vi := sl.Index(i)
|
|
||||||
key := fn(vi.Interface().(Elem))
|
|
||||||
m[key] = append(m[key], vi)
|
|
||||||
}
|
|
||||||
keys := []string{}
|
|
||||||
for k, _ := range m {
|
|
||||||
keys = append(keys, k)
|
|
||||||
}
|
|
||||||
sort.Strings(keys)
|
|
||||||
res := []Slice{}
|
|
||||||
for _, k := range keys {
|
|
||||||
nsl := reflect.New(sl.Type())
|
|
||||||
nsl.Elem().Set(reflect.Append(nsl.Elem(), m[k]...))
|
|
||||||
res = append(res, MakeSlice(nsl.Interface()))
|
|
||||||
}
|
|
||||||
return res
|
|
||||||
}
|
|
||||||
|
|
||||||
// SelectAnyOf filters s to contain only elements for which attr matches
|
|
||||||
// any of the values.
|
|
||||||
func (s Slice) SelectAnyOf(attr string, values ...string) {
|
|
||||||
index := s.indexForAttr(attr)
|
|
||||||
s.Filter(func(e Elem) bool {
|
|
||||||
vf := reflect.ValueOf(e).Elem().FieldByIndex(index)
|
|
||||||
return in(values, vf.String())
|
|
||||||
})
|
|
||||||
}
|
|
||||||
|
|
||||||
// SelectOnePerGroup filters s to include at most one element e per group of
|
|
||||||
// elements matching Key(attr), where e has an attribute a that matches any
|
|
||||||
// the values in v.
|
|
||||||
// If more than one element in a group matches a value in v preference
|
|
||||||
// is given to the element that matches the first value in v.
|
|
||||||
func (s Slice) SelectOnePerGroup(a string, v []string) {
|
|
||||||
index := s.indexForAttr(a)
|
|
||||||
grouped := s.Group(func(e Elem) string { return Key(e, a) })
|
|
||||||
sl := s.Value()
|
|
||||||
sl.Set(sl.Slice(0, 0))
|
|
||||||
for _, g := range grouped {
|
|
||||||
e := reflect.Value{}
|
|
||||||
found := len(v)
|
|
||||||
gsl := g.Value()
|
|
||||||
for i := 0; i < gsl.Len(); i++ {
|
|
||||||
vi := gsl.Index(i).Elem().FieldByIndex(index)
|
|
||||||
j := 0
|
|
||||||
for ; j < len(v) && v[j] != vi.String(); j++ {
|
|
||||||
}
|
|
||||||
if j < found {
|
|
||||||
found = j
|
|
||||||
e = gsl.Index(i)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
if found < len(v) {
|
|
||||||
sl.Set(reflect.Append(sl, e))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// SelectDraft drops all elements from the list with a draft level smaller than d
|
|
||||||
// and selects the highest draft level of the remaining.
|
|
||||||
// This method assumes that the input CLDR is canonicalized.
|
|
||||||
func (s Slice) SelectDraft(d Draft) {
|
|
||||||
s.SelectOnePerGroup("draft", drafts[len(drafts)-2-int(d):])
|
|
||||||
}
|
|
File diff suppressed because it is too large
Load Diff
|
@ -1,978 +0,0 @@
|
||||||
// Copyright 2011 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
// Normalization table generator.
|
|
||||||
// Data read from the web.
|
|
||||||
// See forminfo.go for a description of the trie values associated with each rune.
|
|
||||||
|
|
||||||
package main
|
|
||||||
|
|
||||||
import (
|
|
||||||
"bytes"
|
|
||||||
"flag"
|
|
||||||
"fmt"
|
|
||||||
"io"
|
|
||||||
"log"
|
|
||||||
"sort"
|
|
||||||
"strconv"
|
|
||||||
"strings"
|
|
||||||
|
|
||||||
"golang.org/x/text/internal/gen"
|
|
||||||
"golang.org/x/text/internal/triegen"
|
|
||||||
"golang.org/x/text/internal/ucd"
|
|
||||||
)
|
|
||||||
|
|
||||||
func main() {
|
|
||||||
gen.Init()
|
|
||||||
loadUnicodeData()
|
|
||||||
compactCCC()
|
|
||||||
loadCompositionExclusions()
|
|
||||||
completeCharFields(FCanonical)
|
|
||||||
completeCharFields(FCompatibility)
|
|
||||||
computeNonStarterCounts()
|
|
||||||
verifyComputed()
|
|
||||||
printChars()
|
|
||||||
if *test {
|
|
||||||
testDerived()
|
|
||||||
printTestdata()
|
|
||||||
} else {
|
|
||||||
makeTables()
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
var (
|
|
||||||
tablelist = flag.String("tables",
|
|
||||||
"all",
|
|
||||||
"comma-separated list of which tables to generate; "+
|
|
||||||
"can be 'decomp', 'recomp', 'info' and 'all'")
|
|
||||||
test = flag.Bool("test",
|
|
||||||
false,
|
|
||||||
"test existing tables against DerivedNormalizationProps and generate test data for regression testing")
|
|
||||||
verbose = flag.Bool("verbose",
|
|
||||||
false,
|
|
||||||
"write data to stdout as it is parsed")
|
|
||||||
)
|
|
||||||
|
|
||||||
const MaxChar = 0x10FFFF // anything above this shouldn't exist
|
|
||||||
|
|
||||||
// Quick Check properties of runes allow us to quickly
|
|
||||||
// determine whether a rune may occur in a normal form.
|
|
||||||
// For a given normal form, a rune may be guaranteed to occur
|
|
||||||
// verbatim (QC=Yes), may or may not combine with another
|
|
||||||
// rune (QC=Maybe), or may not occur (QC=No).
|
|
||||||
type QCResult int
|
|
||||||
|
|
||||||
const (
|
|
||||||
QCUnknown QCResult = iota
|
|
||||||
QCYes
|
|
||||||
QCNo
|
|
||||||
QCMaybe
|
|
||||||
)
|
|
||||||
|
|
||||||
func (r QCResult) String() string {
|
|
||||||
switch r {
|
|
||||||
case QCYes:
|
|
||||||
return "Yes"
|
|
||||||
case QCNo:
|
|
||||||
return "No"
|
|
||||||
case QCMaybe:
|
|
||||||
return "Maybe"
|
|
||||||
}
|
|
||||||
return "***UNKNOWN***"
|
|
||||||
}
|
|
||||||
|
|
||||||
const (
|
|
||||||
FCanonical = iota // NFC or NFD
|
|
||||||
FCompatibility // NFKC or NFKD
|
|
||||||
FNumberOfFormTypes
|
|
||||||
)
|
|
||||||
|
|
||||||
const (
|
|
||||||
MComposed = iota // NFC or NFKC
|
|
||||||
MDecomposed // NFD or NFKD
|
|
||||||
MNumberOfModes
|
|
||||||
)
|
|
||||||
|
|
||||||
// This contains only the properties we're interested in.
|
|
||||||
type Char struct {
|
|
||||||
name string
|
|
||||||
codePoint rune // if zero, this index is not a valid code point.
|
|
||||||
ccc uint8 // canonical combining class
|
|
||||||
origCCC uint8
|
|
||||||
excludeInComp bool // from CompositionExclusions.txt
|
|
||||||
compatDecomp bool // it has a compatibility expansion
|
|
||||||
|
|
||||||
nTrailingNonStarters uint8
|
|
||||||
nLeadingNonStarters uint8 // must be equal to trailing if non-zero
|
|
||||||
|
|
||||||
forms [FNumberOfFormTypes]FormInfo // For FCanonical and FCompatibility
|
|
||||||
|
|
||||||
state State
|
|
||||||
}
|
|
||||||
|
|
||||||
var chars = make([]Char, MaxChar+1)
|
|
||||||
var cccMap = make(map[uint8]uint8)
|
|
||||||
|
|
||||||
func (c Char) String() string {
|
|
||||||
buf := new(bytes.Buffer)
|
|
||||||
|
|
||||||
fmt.Fprintf(buf, "%U [%s]:\n", c.codePoint, c.name)
|
|
||||||
fmt.Fprintf(buf, " ccc: %v\n", c.ccc)
|
|
||||||
fmt.Fprintf(buf, " excludeInComp: %v\n", c.excludeInComp)
|
|
||||||
fmt.Fprintf(buf, " compatDecomp: %v\n", c.compatDecomp)
|
|
||||||
fmt.Fprintf(buf, " state: %v\n", c.state)
|
|
||||||
fmt.Fprintf(buf, " NFC:\n")
|
|
||||||
fmt.Fprint(buf, c.forms[FCanonical])
|
|
||||||
fmt.Fprintf(buf, " NFKC:\n")
|
|
||||||
fmt.Fprint(buf, c.forms[FCompatibility])
|
|
||||||
|
|
||||||
return buf.String()
|
|
||||||
}
|
|
||||||
|
|
||||||
// In UnicodeData.txt, some ranges are marked like this:
|
|
||||||
// 3400;<CJK Ideograph Extension A, First>;Lo;0;L;;;;;N;;;;;
|
|
||||||
// 4DB5;<CJK Ideograph Extension A, Last>;Lo;0;L;;;;;N;;;;;
|
|
||||||
// parseCharacter keeps a state variable indicating the weirdness.
|
|
||||||
type State int
|
|
||||||
|
|
||||||
const (
|
|
||||||
SNormal State = iota // known to be zero for the type
|
|
||||||
SFirst
|
|
||||||
SLast
|
|
||||||
SMissing
|
|
||||||
)
|
|
||||||
|
|
||||||
var lastChar = rune('\u0000')
|
|
||||||
|
|
||||||
func (c Char) isValid() bool {
|
|
||||||
return c.codePoint != 0 && c.state != SMissing
|
|
||||||
}
|
|
||||||
|
|
||||||
type FormInfo struct {
|
|
||||||
quickCheck [MNumberOfModes]QCResult // index: MComposed or MDecomposed
|
|
||||||
verified [MNumberOfModes]bool // index: MComposed or MDecomposed
|
|
||||||
|
|
||||||
combinesForward bool // May combine with rune on the right
|
|
||||||
combinesBackward bool // May combine with rune on the left
|
|
||||||
isOneWay bool // Never appears in result
|
|
||||||
inDecomp bool // Some decompositions result in this char.
|
|
||||||
decomp Decomposition
|
|
||||||
expandedDecomp Decomposition
|
|
||||||
}
|
|
||||||
|
|
||||||
func (f FormInfo) String() string {
|
|
||||||
buf := bytes.NewBuffer(make([]byte, 0))
|
|
||||||
|
|
||||||
fmt.Fprintf(buf, " quickCheck[C]: %v\n", f.quickCheck[MComposed])
|
|
||||||
fmt.Fprintf(buf, " quickCheck[D]: %v\n", f.quickCheck[MDecomposed])
|
|
||||||
fmt.Fprintf(buf, " cmbForward: %v\n", f.combinesForward)
|
|
||||||
fmt.Fprintf(buf, " cmbBackward: %v\n", f.combinesBackward)
|
|
||||||
fmt.Fprintf(buf, " isOneWay: %v\n", f.isOneWay)
|
|
||||||
fmt.Fprintf(buf, " inDecomp: %v\n", f.inDecomp)
|
|
||||||
fmt.Fprintf(buf, " decomposition: %X\n", f.decomp)
|
|
||||||
fmt.Fprintf(buf, " expandedDecomp: %X\n", f.expandedDecomp)
|
|
||||||
|
|
||||||
return buf.String()
|
|
||||||
}
|
|
||||||
|
|
||||||
type Decomposition []rune
|
|
||||||
|
|
||||||
func parseDecomposition(s string, skipfirst bool) (a []rune, err error) {
|
|
||||||
decomp := strings.Split(s, " ")
|
|
||||||
if len(decomp) > 0 && skipfirst {
|
|
||||||
decomp = decomp[1:]
|
|
||||||
}
|
|
||||||
for _, d := range decomp {
|
|
||||||
point, err := strconv.ParseUint(d, 16, 64)
|
|
||||||
if err != nil {
|
|
||||||
return a, err
|
|
||||||
}
|
|
||||||
a = append(a, rune(point))
|
|
||||||
}
|
|
||||||
return a, nil
|
|
||||||
}
|
|
||||||
|
|
||||||
func loadUnicodeData() {
|
|
||||||
f := gen.OpenUCDFile("UnicodeData.txt")
|
|
||||||
defer f.Close()
|
|
||||||
p := ucd.New(f)
|
|
||||||
for p.Next() {
|
|
||||||
r := p.Rune(ucd.CodePoint)
|
|
||||||
char := &chars[r]
|
|
||||||
|
|
||||||
char.ccc = uint8(p.Uint(ucd.CanonicalCombiningClass))
|
|
||||||
decmap := p.String(ucd.DecompMapping)
|
|
||||||
|
|
||||||
exp, err := parseDecomposition(decmap, false)
|
|
||||||
isCompat := false
|
|
||||||
if err != nil {
|
|
||||||
if len(decmap) > 0 {
|
|
||||||
exp, err = parseDecomposition(decmap, true)
|
|
||||||
if err != nil {
|
|
||||||
log.Fatalf(`%U: bad decomp |%v|: "%s"`, r, decmap, err)
|
|
||||||
}
|
|
||||||
isCompat = true
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
char.name = p.String(ucd.Name)
|
|
||||||
char.codePoint = r
|
|
||||||
char.forms[FCompatibility].decomp = exp
|
|
||||||
if !isCompat {
|
|
||||||
char.forms[FCanonical].decomp = exp
|
|
||||||
} else {
|
|
||||||
char.compatDecomp = true
|
|
||||||
}
|
|
||||||
if len(decmap) > 0 {
|
|
||||||
char.forms[FCompatibility].decomp = exp
|
|
||||||
}
|
|
||||||
}
|
|
||||||
if err := p.Err(); err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// compactCCC converts the sparse set of CCC values to a continguous one,
|
|
||||||
// reducing the number of bits needed from 8 to 6.
|
|
||||||
func compactCCC() {
|
|
||||||
m := make(map[uint8]uint8)
|
|
||||||
for i := range chars {
|
|
||||||
c := &chars[i]
|
|
||||||
m[c.ccc] = 0
|
|
||||||
}
|
|
||||||
cccs := []int{}
|
|
||||||
for v, _ := range m {
|
|
||||||
cccs = append(cccs, int(v))
|
|
||||||
}
|
|
||||||
sort.Ints(cccs)
|
|
||||||
for i, c := range cccs {
|
|
||||||
cccMap[uint8(i)] = uint8(c)
|
|
||||||
m[uint8(c)] = uint8(i)
|
|
||||||
}
|
|
||||||
for i := range chars {
|
|
||||||
c := &chars[i]
|
|
||||||
c.origCCC = c.ccc
|
|
||||||
c.ccc = m[c.ccc]
|
|
||||||
}
|
|
||||||
if len(m) >= 1<<6 {
|
|
||||||
log.Fatalf("too many difference CCC values: %d >= 64", len(m))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// CompositionExclusions.txt has form:
|
|
||||||
// 0958 # ...
|
|
||||||
// See http://unicode.org/reports/tr44/ for full explanation
|
|
||||||
func loadCompositionExclusions() {
|
|
||||||
f := gen.OpenUCDFile("CompositionExclusions.txt")
|
|
||||||
defer f.Close()
|
|
||||||
p := ucd.New(f)
|
|
||||||
for p.Next() {
|
|
||||||
c := &chars[p.Rune(0)]
|
|
||||||
if c.excludeInComp {
|
|
||||||
log.Fatalf("%U: Duplicate entry in exclusions.", c.codePoint)
|
|
||||||
}
|
|
||||||
c.excludeInComp = true
|
|
||||||
}
|
|
||||||
if e := p.Err(); e != nil {
|
|
||||||
log.Fatal(e)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// hasCompatDecomp returns true if any of the recursive
|
|
||||||
// decompositions contains a compatibility expansion.
|
|
||||||
// In this case, the character may not occur in NFK*.
|
|
||||||
func hasCompatDecomp(r rune) bool {
|
|
||||||
c := &chars[r]
|
|
||||||
if c.compatDecomp {
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
for _, d := range c.forms[FCompatibility].decomp {
|
|
||||||
if hasCompatDecomp(d) {
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
|
|
||||||
// Hangul related constants.
|
|
||||||
const (
|
|
||||||
HangulBase = 0xAC00
|
|
||||||
HangulEnd = 0xD7A4 // hangulBase + Jamo combinations (19 * 21 * 28)
|
|
||||||
|
|
||||||
JamoLBase = 0x1100
|
|
||||||
JamoLEnd = 0x1113
|
|
||||||
JamoVBase = 0x1161
|
|
||||||
JamoVEnd = 0x1176
|
|
||||||
JamoTBase = 0x11A8
|
|
||||||
JamoTEnd = 0x11C3
|
|
||||||
|
|
||||||
JamoLVTCount = 19 * 21 * 28
|
|
||||||
JamoTCount = 28
|
|
||||||
)
|
|
||||||
|
|
||||||
func isHangul(r rune) bool {
|
|
||||||
return HangulBase <= r && r < HangulEnd
|
|
||||||
}
|
|
||||||
|
|
||||||
func isHangulWithoutJamoT(r rune) bool {
|
|
||||||
if !isHangul(r) {
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
r -= HangulBase
|
|
||||||
return r < JamoLVTCount && r%JamoTCount == 0
|
|
||||||
}
|
|
||||||
|
|
||||||
func ccc(r rune) uint8 {
|
|
||||||
return chars[r].ccc
|
|
||||||
}
|
|
||||||
|
|
||||||
// Insert a rune in a buffer, ordered by Canonical Combining Class.
|
|
||||||
func insertOrdered(b Decomposition, r rune) Decomposition {
|
|
||||||
n := len(b)
|
|
||||||
b = append(b, 0)
|
|
||||||
cc := ccc(r)
|
|
||||||
if cc > 0 {
|
|
||||||
// Use bubble sort.
|
|
||||||
for ; n > 0; n-- {
|
|
||||||
if ccc(b[n-1]) <= cc {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
b[n] = b[n-1]
|
|
||||||
}
|
|
||||||
}
|
|
||||||
b[n] = r
|
|
||||||
return b
|
|
||||||
}
|
|
||||||
|
|
||||||
// Recursively decompose.
|
|
||||||
func decomposeRecursive(form int, r rune, d Decomposition) Decomposition {
|
|
||||||
dcomp := chars[r].forms[form].decomp
|
|
||||||
if len(dcomp) == 0 {
|
|
||||||
return insertOrdered(d, r)
|
|
||||||
}
|
|
||||||
for _, c := range dcomp {
|
|
||||||
d = decomposeRecursive(form, c, d)
|
|
||||||
}
|
|
||||||
return d
|
|
||||||
}
|
|
||||||
|
|
||||||
func completeCharFields(form int) {
|
|
||||||
// Phase 0: pre-expand decomposition.
|
|
||||||
for i := range chars {
|
|
||||||
f := &chars[i].forms[form]
|
|
||||||
if len(f.decomp) == 0 {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
exp := make(Decomposition, 0)
|
|
||||||
for _, c := range f.decomp {
|
|
||||||
exp = decomposeRecursive(form, c, exp)
|
|
||||||
}
|
|
||||||
f.expandedDecomp = exp
|
|
||||||
}
|
|
||||||
|
|
||||||
// Phase 1: composition exclusion, mark decomposition.
|
|
||||||
for i := range chars {
|
|
||||||
c := &chars[i]
|
|
||||||
f := &c.forms[form]
|
|
||||||
|
|
||||||
// Marks script-specific exclusions and version restricted.
|
|
||||||
f.isOneWay = c.excludeInComp
|
|
||||||
|
|
||||||
// Singletons
|
|
||||||
f.isOneWay = f.isOneWay || len(f.decomp) == 1
|
|
||||||
|
|
||||||
// Non-starter decompositions
|
|
||||||
if len(f.decomp) > 1 {
|
|
||||||
chk := c.ccc != 0 || chars[f.decomp[0]].ccc != 0
|
|
||||||
f.isOneWay = f.isOneWay || chk
|
|
||||||
}
|
|
||||||
|
|
||||||
// Runes that decompose into more than two runes.
|
|
||||||
f.isOneWay = f.isOneWay || len(f.decomp) > 2
|
|
||||||
|
|
||||||
if form == FCompatibility {
|
|
||||||
f.isOneWay = f.isOneWay || hasCompatDecomp(c.codePoint)
|
|
||||||
}
|
|
||||||
|
|
||||||
for _, r := range f.decomp {
|
|
||||||
chars[r].forms[form].inDecomp = true
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Phase 2: forward and backward combining.
|
|
||||||
for i := range chars {
|
|
||||||
c := &chars[i]
|
|
||||||
f := &c.forms[form]
|
|
||||||
|
|
||||||
if !f.isOneWay && len(f.decomp) == 2 {
|
|
||||||
f0 := &chars[f.decomp[0]].forms[form]
|
|
||||||
f1 := &chars[f.decomp[1]].forms[form]
|
|
||||||
if !f0.isOneWay {
|
|
||||||
f0.combinesForward = true
|
|
||||||
}
|
|
||||||
if !f1.isOneWay {
|
|
||||||
f1.combinesBackward = true
|
|
||||||
}
|
|
||||||
}
|
|
||||||
if isHangulWithoutJamoT(rune(i)) {
|
|
||||||
f.combinesForward = true
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Phase 3: quick check values.
|
|
||||||
for i := range chars {
|
|
||||||
c := &chars[i]
|
|
||||||
f := &c.forms[form]
|
|
||||||
|
|
||||||
switch {
|
|
||||||
case len(f.decomp) > 0:
|
|
||||||
f.quickCheck[MDecomposed] = QCNo
|
|
||||||
case isHangul(rune(i)):
|
|
||||||
f.quickCheck[MDecomposed] = QCNo
|
|
||||||
default:
|
|
||||||
f.quickCheck[MDecomposed] = QCYes
|
|
||||||
}
|
|
||||||
switch {
|
|
||||||
case f.isOneWay:
|
|
||||||
f.quickCheck[MComposed] = QCNo
|
|
||||||
case (i & 0xffff00) == JamoLBase:
|
|
||||||
f.quickCheck[MComposed] = QCYes
|
|
||||||
if JamoLBase <= i && i < JamoLEnd {
|
|
||||||
f.combinesForward = true
|
|
||||||
}
|
|
||||||
if JamoVBase <= i && i < JamoVEnd {
|
|
||||||
f.quickCheck[MComposed] = QCMaybe
|
|
||||||
f.combinesBackward = true
|
|
||||||
f.combinesForward = true
|
|
||||||
}
|
|
||||||
if JamoTBase <= i && i < JamoTEnd {
|
|
||||||
f.quickCheck[MComposed] = QCMaybe
|
|
||||||
f.combinesBackward = true
|
|
||||||
}
|
|
||||||
case !f.combinesBackward:
|
|
||||||
f.quickCheck[MComposed] = QCYes
|
|
||||||
default:
|
|
||||||
f.quickCheck[MComposed] = QCMaybe
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
func computeNonStarterCounts() {
|
|
||||||
// Phase 4: leading and trailing non-starter count
|
|
||||||
for i := range chars {
|
|
||||||
c := &chars[i]
|
|
||||||
|
|
||||||
runes := []rune{rune(i)}
|
|
||||||
// We always use FCompatibility so that the CGJ insertion points do not
|
|
||||||
// change for repeated normalizations with different forms.
|
|
||||||
if exp := c.forms[FCompatibility].expandedDecomp; len(exp) > 0 {
|
|
||||||
runes = exp
|
|
||||||
}
|
|
||||||
// We consider runes that combine backwards to be non-starters for the
|
|
||||||
// purpose of Stream-Safe Text Processing.
|
|
||||||
for _, r := range runes {
|
|
||||||
if cr := &chars[r]; cr.ccc == 0 && !cr.forms[FCompatibility].combinesBackward {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
c.nLeadingNonStarters++
|
|
||||||
}
|
|
||||||
for i := len(runes) - 1; i >= 0; i-- {
|
|
||||||
if cr := &chars[runes[i]]; cr.ccc == 0 && !cr.forms[FCompatibility].combinesBackward {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
c.nTrailingNonStarters++
|
|
||||||
}
|
|
||||||
if c.nTrailingNonStarters > 3 {
|
|
||||||
log.Fatalf("%U: Decomposition with more than 3 (%d) trailing modifiers (%U)", i, c.nTrailingNonStarters, runes)
|
|
||||||
}
|
|
||||||
|
|
||||||
if isHangul(rune(i)) {
|
|
||||||
c.nTrailingNonStarters = 2
|
|
||||||
if isHangulWithoutJamoT(rune(i)) {
|
|
||||||
c.nTrailingNonStarters = 1
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if l, t := c.nLeadingNonStarters, c.nTrailingNonStarters; l > 0 && l != t {
|
|
||||||
log.Fatalf("%U: number of leading and trailing non-starters should be equal (%d vs %d)", i, l, t)
|
|
||||||
}
|
|
||||||
if t := c.nTrailingNonStarters; t > 3 {
|
|
||||||
log.Fatalf("%U: number of trailing non-starters is %d > 3", t)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
func printBytes(w io.Writer, b []byte, name string) {
|
|
||||||
fmt.Fprintf(w, "// %s: %d bytes\n", name, len(b))
|
|
||||||
fmt.Fprintf(w, "var %s = [...]byte {", name)
|
|
||||||
for i, c := range b {
|
|
||||||
switch {
|
|
||||||
case i%64 == 0:
|
|
||||||
fmt.Fprintf(w, "\n// Bytes %x - %x\n", i, i+63)
|
|
||||||
case i%8 == 0:
|
|
||||||
fmt.Fprintf(w, "\n")
|
|
||||||
}
|
|
||||||
fmt.Fprintf(w, "0x%.2X, ", c)
|
|
||||||
}
|
|
||||||
fmt.Fprint(w, "\n}\n\n")
|
|
||||||
}
|
|
||||||
|
|
||||||
// See forminfo.go for format.
|
|
||||||
func makeEntry(f *FormInfo, c *Char) uint16 {
|
|
||||||
e := uint16(0)
|
|
||||||
if r := c.codePoint; HangulBase <= r && r < HangulEnd {
|
|
||||||
e |= 0x40
|
|
||||||
}
|
|
||||||
if f.combinesForward {
|
|
||||||
e |= 0x20
|
|
||||||
}
|
|
||||||
if f.quickCheck[MDecomposed] == QCNo {
|
|
||||||
e |= 0x4
|
|
||||||
}
|
|
||||||
switch f.quickCheck[MComposed] {
|
|
||||||
case QCYes:
|
|
||||||
case QCNo:
|
|
||||||
e |= 0x10
|
|
||||||
case QCMaybe:
|
|
||||||
e |= 0x18
|
|
||||||
default:
|
|
||||||
log.Fatalf("Illegal quickcheck value %v.", f.quickCheck[MComposed])
|
|
||||||
}
|
|
||||||
e |= uint16(c.nTrailingNonStarters)
|
|
||||||
return e
|
|
||||||
}
|
|
||||||
|
|
||||||
// decompSet keeps track of unique decompositions, grouped by whether
|
|
||||||
// the decomposition is followed by a trailing and/or leading CCC.
|
|
||||||
type decompSet [7]map[string]bool
|
|
||||||
|
|
||||||
const (
|
|
||||||
normalDecomp = iota
|
|
||||||
firstMulti
|
|
||||||
firstCCC
|
|
||||||
endMulti
|
|
||||||
firstLeadingCCC
|
|
||||||
firstCCCZeroExcept
|
|
||||||
firstStarterWithNLead
|
|
||||||
lastDecomp
|
|
||||||
)
|
|
||||||
|
|
||||||
var cname = []string{"firstMulti", "firstCCC", "endMulti", "firstLeadingCCC", "firstCCCZeroExcept", "firstStarterWithNLead", "lastDecomp"}
|
|
||||||
|
|
||||||
func makeDecompSet() decompSet {
|
|
||||||
m := decompSet{}
|
|
||||||
for i := range m {
|
|
||||||
m[i] = make(map[string]bool)
|
|
||||||
}
|
|
||||||
return m
|
|
||||||
}
|
|
||||||
func (m *decompSet) insert(key int, s string) {
|
|
||||||
m[key][s] = true
|
|
||||||
}
|
|
||||||
|
|
||||||
func printCharInfoTables(w io.Writer) int {
|
|
||||||
mkstr := func(r rune, f *FormInfo) (int, string) {
|
|
||||||
d := f.expandedDecomp
|
|
||||||
s := string([]rune(d))
|
|
||||||
if max := 1 << 6; len(s) >= max {
|
|
||||||
const msg = "%U: too many bytes in decomposition: %d >= %d"
|
|
||||||
log.Fatalf(msg, r, len(s), max)
|
|
||||||
}
|
|
||||||
head := uint8(len(s))
|
|
||||||
if f.quickCheck[MComposed] != QCYes {
|
|
||||||
head |= 0x40
|
|
||||||
}
|
|
||||||
if f.combinesForward {
|
|
||||||
head |= 0x80
|
|
||||||
}
|
|
||||||
s = string([]byte{head}) + s
|
|
||||||
|
|
||||||
lccc := ccc(d[0])
|
|
||||||
tccc := ccc(d[len(d)-1])
|
|
||||||
cc := ccc(r)
|
|
||||||
if cc != 0 && lccc == 0 && tccc == 0 {
|
|
||||||
log.Fatalf("%U: trailing and leading ccc are 0 for non-zero ccc %d", r, cc)
|
|
||||||
}
|
|
||||||
if tccc < lccc && lccc != 0 {
|
|
||||||
const msg = "%U: lccc (%d) must be <= tcc (%d)"
|
|
||||||
log.Fatalf(msg, r, lccc, tccc)
|
|
||||||
}
|
|
||||||
index := normalDecomp
|
|
||||||
nTrail := chars[r].nTrailingNonStarters
|
|
||||||
if tccc > 0 || lccc > 0 || nTrail > 0 {
|
|
||||||
tccc <<= 2
|
|
||||||
tccc |= nTrail
|
|
||||||
s += string([]byte{tccc})
|
|
||||||
index = endMulti
|
|
||||||
for _, r := range d[1:] {
|
|
||||||
if ccc(r) == 0 {
|
|
||||||
index = firstCCC
|
|
||||||
}
|
|
||||||
}
|
|
||||||
if lccc > 0 {
|
|
||||||
s += string([]byte{lccc})
|
|
||||||
if index == firstCCC {
|
|
||||||
log.Fatalf("%U: multi-segment decomposition not supported for decompositions with leading CCC != 0", r)
|
|
||||||
}
|
|
||||||
index = firstLeadingCCC
|
|
||||||
}
|
|
||||||
if cc != lccc {
|
|
||||||
if cc != 0 {
|
|
||||||
log.Fatalf("%U: for lccc != ccc, expected ccc to be 0; was %d", r, cc)
|
|
||||||
}
|
|
||||||
index = firstCCCZeroExcept
|
|
||||||
}
|
|
||||||
} else if len(d) > 1 {
|
|
||||||
index = firstMulti
|
|
||||||
}
|
|
||||||
return index, s
|
|
||||||
}
|
|
||||||
|
|
||||||
decompSet := makeDecompSet()
|
|
||||||
const nLeadStr = "\x00\x01" // 0-byte length and tccc with nTrail.
|
|
||||||
decompSet.insert(firstStarterWithNLead, nLeadStr)
|
|
||||||
|
|
||||||
// Store the uniqued decompositions in a byte buffer,
|
|
||||||
// preceded by their byte length.
|
|
||||||
for _, c := range chars {
|
|
||||||
for _, f := range c.forms {
|
|
||||||
if len(f.expandedDecomp) == 0 {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
if f.combinesBackward {
|
|
||||||
log.Fatalf("%U: combinesBackward and decompose", c.codePoint)
|
|
||||||
}
|
|
||||||
index, s := mkstr(c.codePoint, &f)
|
|
||||||
decompSet.insert(index, s)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
decompositions := bytes.NewBuffer(make([]byte, 0, 10000))
|
|
||||||
size := 0
|
|
||||||
positionMap := make(map[string]uint16)
|
|
||||||
decompositions.WriteString("\000")
|
|
||||||
fmt.Fprintln(w, "const (")
|
|
||||||
for i, m := range decompSet {
|
|
||||||
sa := []string{}
|
|
||||||
for s := range m {
|
|
||||||
sa = append(sa, s)
|
|
||||||
}
|
|
||||||
sort.Strings(sa)
|
|
||||||
for _, s := range sa {
|
|
||||||
p := decompositions.Len()
|
|
||||||
decompositions.WriteString(s)
|
|
||||||
positionMap[s] = uint16(p)
|
|
||||||
}
|
|
||||||
if cname[i] != "" {
|
|
||||||
fmt.Fprintf(w, "%s = 0x%X\n", cname[i], decompositions.Len())
|
|
||||||
}
|
|
||||||
}
|
|
||||||
fmt.Fprintln(w, "maxDecomp = 0x8000")
|
|
||||||
fmt.Fprintln(w, ")")
|
|
||||||
b := decompositions.Bytes()
|
|
||||||
printBytes(w, b, "decomps")
|
|
||||||
size += len(b)
|
|
||||||
|
|
||||||
varnames := []string{"nfc", "nfkc"}
|
|
||||||
for i := 0; i < FNumberOfFormTypes; i++ {
|
|
||||||
trie := triegen.NewTrie(varnames[i])
|
|
||||||
|
|
||||||
for r, c := range chars {
|
|
||||||
f := c.forms[i]
|
|
||||||
d := f.expandedDecomp
|
|
||||||
if len(d) != 0 {
|
|
||||||
_, key := mkstr(c.codePoint, &f)
|
|
||||||
trie.Insert(rune(r), uint64(positionMap[key]))
|
|
||||||
if c.ccc != ccc(d[0]) {
|
|
||||||
// We assume the lead ccc of a decomposition !=0 in this case.
|
|
||||||
if ccc(d[0]) == 0 {
|
|
||||||
log.Fatalf("Expected leading CCC to be non-zero; ccc is %d", c.ccc)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
} else if c.nLeadingNonStarters > 0 && len(f.expandedDecomp) == 0 && c.ccc == 0 && !f.combinesBackward {
|
|
||||||
// Handle cases where it can't be detected that the nLead should be equal
|
|
||||||
// to nTrail.
|
|
||||||
trie.Insert(c.codePoint, uint64(positionMap[nLeadStr]))
|
|
||||||
} else if v := makeEntry(&f, &c)<<8 | uint16(c.ccc); v != 0 {
|
|
||||||
trie.Insert(c.codePoint, uint64(0x8000|v))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
sz, err := trie.Gen(w, triegen.Compact(&normCompacter{name: varnames[i]}))
|
|
||||||
if err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
size += sz
|
|
||||||
}
|
|
||||||
return size
|
|
||||||
}
|
|
||||||
|
|
||||||
func contains(sa []string, s string) bool {
|
|
||||||
for _, a := range sa {
|
|
||||||
if a == s {
|
|
||||||
return true
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return false
|
|
||||||
}
|
|
||||||
|
|
||||||
func makeTables() {
|
|
||||||
w := &bytes.Buffer{}
|
|
||||||
|
|
||||||
size := 0
|
|
||||||
if *tablelist == "" {
|
|
||||||
return
|
|
||||||
}
|
|
||||||
list := strings.Split(*tablelist, ",")
|
|
||||||
if *tablelist == "all" {
|
|
||||||
list = []string{"recomp", "info"}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Compute maximum decomposition size.
|
|
||||||
max := 0
|
|
||||||
for _, c := range chars {
|
|
||||||
if n := len(string(c.forms[FCompatibility].expandedDecomp)); n > max {
|
|
||||||
max = n
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
fmt.Fprintln(w, "const (")
|
|
||||||
fmt.Fprintln(w, "\t// Version is the Unicode edition from which the tables are derived.")
|
|
||||||
fmt.Fprintf(w, "\tVersion = %q\n", gen.UnicodeVersion())
|
|
||||||
fmt.Fprintln(w)
|
|
||||||
fmt.Fprintln(w, "\t// MaxTransformChunkSize indicates the maximum number of bytes that Transform")
|
|
||||||
fmt.Fprintln(w, "\t// may need to write atomically for any Form. Making a destination buffer at")
|
|
||||||
fmt.Fprintln(w, "\t// least this size ensures that Transform can always make progress and that")
|
|
||||||
fmt.Fprintln(w, "\t// the user does not need to grow the buffer on an ErrShortDst.")
|
|
||||||
fmt.Fprintf(w, "\tMaxTransformChunkSize = %d+maxNonStarters*4\n", len(string(0x034F))+max)
|
|
||||||
fmt.Fprintln(w, ")\n")
|
|
||||||
|
|
||||||
// Print the CCC remap table.
|
|
||||||
size += len(cccMap)
|
|
||||||
fmt.Fprintf(w, "var ccc = [%d]uint8{", len(cccMap))
|
|
||||||
for i := 0; i < len(cccMap); i++ {
|
|
||||||
if i%8 == 0 {
|
|
||||||
fmt.Fprintln(w)
|
|
||||||
}
|
|
||||||
fmt.Fprintf(w, "%3d, ", cccMap[uint8(i)])
|
|
||||||
}
|
|
||||||
fmt.Fprintln(w, "\n}\n")
|
|
||||||
|
|
||||||
if contains(list, "info") {
|
|
||||||
size += printCharInfoTables(w)
|
|
||||||
}
|
|
||||||
|
|
||||||
if contains(list, "recomp") {
|
|
||||||
// Note that we use 32 bit keys, instead of 64 bit.
|
|
||||||
// This clips the bits of three entries, but we know
|
|
||||||
// this won't cause a collision. The compiler will catch
|
|
||||||
// any changes made to UnicodeData.txt that introduces
|
|
||||||
// a collision.
|
|
||||||
// Note that the recomposition map for NFC and NFKC
|
|
||||||
// are identical.
|
|
||||||
|
|
||||||
// Recomposition map
|
|
||||||
nrentries := 0
|
|
||||||
for _, c := range chars {
|
|
||||||
f := c.forms[FCanonical]
|
|
||||||
if !f.isOneWay && len(f.decomp) > 0 {
|
|
||||||
nrentries++
|
|
||||||
}
|
|
||||||
}
|
|
||||||
sz := nrentries * 8
|
|
||||||
size += sz
|
|
||||||
fmt.Fprintf(w, "// recompMap: %d bytes (entries only)\n", sz)
|
|
||||||
fmt.Fprintln(w, "var recompMap = map[uint32]rune{")
|
|
||||||
for i, c := range chars {
|
|
||||||
f := c.forms[FCanonical]
|
|
||||||
d := f.decomp
|
|
||||||
if !f.isOneWay && len(d) > 0 {
|
|
||||||
key := uint32(uint16(d[0]))<<16 + uint32(uint16(d[1]))
|
|
||||||
fmt.Fprintf(w, "0x%.8X: 0x%.4X,\n", key, i)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
fmt.Fprintf(w, "}\n\n")
|
|
||||||
}
|
|
||||||
|
|
||||||
fmt.Fprintf(w, "// Total size of tables: %dKB (%d bytes)\n", (size+512)/1024, size)
|
|
||||||
gen.WriteGoFile("tables.go", "norm", w.Bytes())
|
|
||||||
}
|
|
||||||
|
|
||||||
func printChars() {
|
|
||||||
if *verbose {
|
|
||||||
for _, c := range chars {
|
|
||||||
if !c.isValid() || c.state == SMissing {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
fmt.Println(c)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// verifyComputed does various consistency tests.
|
|
||||||
func verifyComputed() {
|
|
||||||
for i, c := range chars {
|
|
||||||
for _, f := range c.forms {
|
|
||||||
isNo := (f.quickCheck[MDecomposed] == QCNo)
|
|
||||||
if (len(f.decomp) > 0) != isNo && !isHangul(rune(i)) {
|
|
||||||
log.Fatalf("%U: NF*D QC must be No if rune decomposes", i)
|
|
||||||
}
|
|
||||||
|
|
||||||
isMaybe := f.quickCheck[MComposed] == QCMaybe
|
|
||||||
if f.combinesBackward != isMaybe {
|
|
||||||
log.Fatalf("%U: NF*C QC must be Maybe if combinesBackward", i)
|
|
||||||
}
|
|
||||||
if len(f.decomp) > 0 && f.combinesForward && isMaybe {
|
|
||||||
log.Fatalf("%U: NF*C QC must be Yes or No if combinesForward and decomposes", i)
|
|
||||||
}
|
|
||||||
|
|
||||||
if len(f.expandedDecomp) != 0 {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
if a, b := c.nLeadingNonStarters > 0, (c.ccc > 0 || f.combinesBackward); a != b {
|
|
||||||
// We accept these runes to be treated differently (it only affects
|
|
||||||
// segment breaking in iteration, most likely on improper use), but
|
|
||||||
// reconsider if more characters are added.
|
|
||||||
// U+FF9E HALFWIDTH KATAKANA VOICED SOUND MARK;Lm;0;L;<narrow> 3099;;;;N;;;;;
|
|
||||||
// U+FF9F HALFWIDTH KATAKANA SEMI-VOICED SOUND MARK;Lm;0;L;<narrow> 309A;;;;N;;;;;
|
|
||||||
// U+3133 HANGUL LETTER KIYEOK-SIOS;Lo;0;L;<compat> 11AA;;;;N;HANGUL LETTER GIYEOG SIOS;;;;
|
|
||||||
// U+318E HANGUL LETTER ARAEAE;Lo;0;L;<compat> 11A1;;;;N;HANGUL LETTER ALAE AE;;;;
|
|
||||||
// U+FFA3 HALFWIDTH HANGUL LETTER KIYEOK-SIOS;Lo;0;L;<narrow> 3133;;;;N;HALFWIDTH HANGUL LETTER GIYEOG SIOS;;;;
|
|
||||||
// U+FFDC HALFWIDTH HANGUL LETTER I;Lo;0;L;<narrow> 3163;;;;N;;;;;
|
|
||||||
if i != 0xFF9E && i != 0xFF9F && !(0x3133 <= i && i <= 0x318E) && !(0xFFA3 <= i && i <= 0xFFDC) {
|
|
||||||
log.Fatalf("%U: nLead was %v; want %v", i, a, b)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
nfc := c.forms[FCanonical]
|
|
||||||
nfkc := c.forms[FCompatibility]
|
|
||||||
if nfc.combinesBackward != nfkc.combinesBackward {
|
|
||||||
log.Fatalf("%U: Cannot combine combinesBackward\n", c.codePoint)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Use values in DerivedNormalizationProps.txt to compare against the
|
|
||||||
// values we computed.
|
|
||||||
// DerivedNormalizationProps.txt has form:
|
|
||||||
// 00C0..00C5 ; NFD_QC; N # ...
|
|
||||||
// 0374 ; NFD_QC; N # ...
|
|
||||||
// See http://unicode.org/reports/tr44/ for full explanation
|
|
||||||
func testDerived() {
|
|
||||||
f := gen.OpenUCDFile("DerivedNormalizationProps.txt")
|
|
||||||
defer f.Close()
|
|
||||||
p := ucd.New(f)
|
|
||||||
for p.Next() {
|
|
||||||
r := p.Rune(0)
|
|
||||||
c := &chars[r]
|
|
||||||
|
|
||||||
var ftype, mode int
|
|
||||||
qt := p.String(1)
|
|
||||||
switch qt {
|
|
||||||
case "NFC_QC":
|
|
||||||
ftype, mode = FCanonical, MComposed
|
|
||||||
case "NFD_QC":
|
|
||||||
ftype, mode = FCanonical, MDecomposed
|
|
||||||
case "NFKC_QC":
|
|
||||||
ftype, mode = FCompatibility, MComposed
|
|
||||||
case "NFKD_QC":
|
|
||||||
ftype, mode = FCompatibility, MDecomposed
|
|
||||||
default:
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
var qr QCResult
|
|
||||||
switch p.String(2) {
|
|
||||||
case "Y":
|
|
||||||
qr = QCYes
|
|
||||||
case "N":
|
|
||||||
qr = QCNo
|
|
||||||
case "M":
|
|
||||||
qr = QCMaybe
|
|
||||||
default:
|
|
||||||
log.Fatalf(`Unexpected quick check value "%s"`, p.String(2))
|
|
||||||
}
|
|
||||||
if got := c.forms[ftype].quickCheck[mode]; got != qr {
|
|
||||||
log.Printf("%U: FAILED %s (was %v need %v)\n", r, qt, got, qr)
|
|
||||||
}
|
|
||||||
c.forms[ftype].verified[mode] = true
|
|
||||||
}
|
|
||||||
if err := p.Err(); err != nil {
|
|
||||||
log.Fatal(err)
|
|
||||||
}
|
|
||||||
// Any unspecified value must be QCYes. Verify this.
|
|
||||||
for i, c := range chars {
|
|
||||||
for j, fd := range c.forms {
|
|
||||||
for k, qr := range fd.quickCheck {
|
|
||||||
if !fd.verified[k] && qr != QCYes {
|
|
||||||
m := "%U: FAIL F:%d M:%d (was %v need Yes) %s\n"
|
|
||||||
log.Printf(m, i, j, k, qr, c.name)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
var testHeader = `const (
|
|
||||||
Yes = iota
|
|
||||||
No
|
|
||||||
Maybe
|
|
||||||
)
|
|
||||||
|
|
||||||
type formData struct {
|
|
||||||
qc uint8
|
|
||||||
combinesForward bool
|
|
||||||
decomposition string
|
|
||||||
}
|
|
||||||
|
|
||||||
type runeData struct {
|
|
||||||
r rune
|
|
||||||
ccc uint8
|
|
||||||
nLead uint8
|
|
||||||
nTrail uint8
|
|
||||||
f [2]formData // 0: canonical; 1: compatibility
|
|
||||||
}
|
|
||||||
|
|
||||||
func f(qc uint8, cf bool, dec string) [2]formData {
|
|
||||||
return [2]formData{{qc, cf, dec}, {qc, cf, dec}}
|
|
||||||
}
|
|
||||||
|
|
||||||
func g(qc, qck uint8, cf, cfk bool, d, dk string) [2]formData {
|
|
||||||
return [2]formData{{qc, cf, d}, {qck, cfk, dk}}
|
|
||||||
}
|
|
||||||
|
|
||||||
var testData = []runeData{
|
|
||||||
`
|
|
||||||
|
|
||||||
func printTestdata() {
|
|
||||||
type lastInfo struct {
|
|
||||||
ccc uint8
|
|
||||||
nLead uint8
|
|
||||||
nTrail uint8
|
|
||||||
f string
|
|
||||||
}
|
|
||||||
|
|
||||||
last := lastInfo{}
|
|
||||||
w := &bytes.Buffer{}
|
|
||||||
fmt.Fprintf(w, testHeader)
|
|
||||||
for r, c := range chars {
|
|
||||||
f := c.forms[FCanonical]
|
|
||||||
qc, cf, d := f.quickCheck[MComposed], f.combinesForward, string(f.expandedDecomp)
|
|
||||||
f = c.forms[FCompatibility]
|
|
||||||
qck, cfk, dk := f.quickCheck[MComposed], f.combinesForward, string(f.expandedDecomp)
|
|
||||||
s := ""
|
|
||||||
if d == dk && qc == qck && cf == cfk {
|
|
||||||
s = fmt.Sprintf("f(%s, %v, %q)", qc, cf, d)
|
|
||||||
} else {
|
|
||||||
s = fmt.Sprintf("g(%s, %s, %v, %v, %q, %q)", qc, qck, cf, cfk, d, dk)
|
|
||||||
}
|
|
||||||
current := lastInfo{c.ccc, c.nLeadingNonStarters, c.nTrailingNonStarters, s}
|
|
||||||
if last != current {
|
|
||||||
fmt.Fprintf(w, "\t{0x%x, %d, %d, %d, %s},\n", r, c.origCCC, c.nLeadingNonStarters, c.nTrailingNonStarters, s)
|
|
||||||
last = current
|
|
||||||
}
|
|
||||||
}
|
|
||||||
fmt.Fprintln(w, "}")
|
|
||||||
gen.WriteGoFile("data_test.go", "norm", w.Bytes())
|
|
||||||
}
|
|
|
@ -1,117 +0,0 @@
|
||||||
// Copyright 2011 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
// Trie table generator.
|
|
||||||
// Used by make*tables tools to generate a go file with trie data structures
|
|
||||||
// for mapping UTF-8 to a 16-bit value. All but the last byte in a UTF-8 byte
|
|
||||||
// sequence are used to lookup offsets in the index table to be used for the
|
|
||||||
// next byte. The last byte is used to index into a table with 16-bit values.
|
|
||||||
|
|
||||||
package main
|
|
||||||
|
|
||||||
import (
|
|
||||||
"fmt"
|
|
||||||
"io"
|
|
||||||
)
|
|
||||||
|
|
||||||
const maxSparseEntries = 16
|
|
||||||
|
|
||||||
type normCompacter struct {
|
|
||||||
sparseBlocks [][]uint64
|
|
||||||
sparseOffset []uint16
|
|
||||||
sparseCount int
|
|
||||||
name string
|
|
||||||
}
|
|
||||||
|
|
||||||
func mostFrequentStride(a []uint64) int {
|
|
||||||
counts := make(map[int]int)
|
|
||||||
var v int
|
|
||||||
for _, x := range a {
|
|
||||||
if stride := int(x) - v; v != 0 && stride >= 0 {
|
|
||||||
counts[stride]++
|
|
||||||
}
|
|
||||||
v = int(x)
|
|
||||||
}
|
|
||||||
var maxs, maxc int
|
|
||||||
for stride, cnt := range counts {
|
|
||||||
if cnt > maxc || (cnt == maxc && stride < maxs) {
|
|
||||||
maxs, maxc = stride, cnt
|
|
||||||
}
|
|
||||||
}
|
|
||||||
return maxs
|
|
||||||
}
|
|
||||||
|
|
||||||
func countSparseEntries(a []uint64) int {
|
|
||||||
stride := mostFrequentStride(a)
|
|
||||||
var v, count int
|
|
||||||
for _, tv := range a {
|
|
||||||
if int(tv)-v != stride {
|
|
||||||
if tv != 0 {
|
|
||||||
count++
|
|
||||||
}
|
|
||||||
}
|
|
||||||
v = int(tv)
|
|
||||||
}
|
|
||||||
return count
|
|
||||||
}
|
|
||||||
|
|
||||||
func (c *normCompacter) Size(v []uint64) (sz int, ok bool) {
|
|
||||||
if n := countSparseEntries(v); n <= maxSparseEntries {
|
|
||||||
return (n+1)*4 + 2, true
|
|
||||||
}
|
|
||||||
return 0, false
|
|
||||||
}
|
|
||||||
|
|
||||||
func (c *normCompacter) Store(v []uint64) uint32 {
|
|
||||||
h := uint32(len(c.sparseOffset))
|
|
||||||
c.sparseBlocks = append(c.sparseBlocks, v)
|
|
||||||
c.sparseOffset = append(c.sparseOffset, uint16(c.sparseCount))
|
|
||||||
c.sparseCount += countSparseEntries(v) + 1
|
|
||||||
return h
|
|
||||||
}
|
|
||||||
|
|
||||||
func (c *normCompacter) Handler() string {
|
|
||||||
return c.name + "Sparse.lookup"
|
|
||||||
}
|
|
||||||
|
|
||||||
func (c *normCompacter) Print(w io.Writer) (retErr error) {
|
|
||||||
p := func(f string, x ...interface{}) {
|
|
||||||
if _, err := fmt.Fprintf(w, f, x...); retErr == nil && err != nil {
|
|
||||||
retErr = err
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
ls := len(c.sparseBlocks)
|
|
||||||
p("// %sSparseOffset: %d entries, %d bytes\n", c.name, ls, ls*2)
|
|
||||||
p("var %sSparseOffset = %#v\n\n", c.name, c.sparseOffset)
|
|
||||||
|
|
||||||
ns := c.sparseCount
|
|
||||||
p("// %sSparseValues: %d entries, %d bytes\n", c.name, ns, ns*4)
|
|
||||||
p("var %sSparseValues = [%d]valueRange {", c.name, ns)
|
|
||||||
for i, b := range c.sparseBlocks {
|
|
||||||
p("\n// Block %#x, offset %#x", i, c.sparseOffset[i])
|
|
||||||
var v int
|
|
||||||
stride := mostFrequentStride(b)
|
|
||||||
n := countSparseEntries(b)
|
|
||||||
p("\n{value:%#04x,lo:%#02x},", stride, uint8(n))
|
|
||||||
for i, nv := range b {
|
|
||||||
if int(nv)-v != stride {
|
|
||||||
if v != 0 {
|
|
||||||
p(",hi:%#02x},", 0x80+i-1)
|
|
||||||
}
|
|
||||||
if nv != 0 {
|
|
||||||
p("\n{value:%#04x,lo:%#02x", nv, 0x80+i)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
v = int(nv)
|
|
||||||
}
|
|
||||||
if v != 0 {
|
|
||||||
p(",hi:%#02x},", 0x80+len(b)-1)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
p("\n}\n\n")
|
|
||||||
return
|
|
||||||
}
|
|
|
@ -1,113 +0,0 @@
|
||||||
// Copyright 2015 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// +build ignore
|
|
||||||
|
|
||||||
package main
|
|
||||||
|
|
||||||
import (
|
|
||||||
"bytes"
|
|
||||||
"flag"
|
|
||||||
"fmt"
|
|
||||||
"io"
|
|
||||||
"log"
|
|
||||||
"reflect"
|
|
||||||
"sort"
|
|
||||||
"strings"
|
|
||||||
"unicode"
|
|
||||||
|
|
||||||
"golang.org/x/text/internal/gen"
|
|
||||||
"golang.org/x/text/internal/ucd"
|
|
||||||
"golang.org/x/text/unicode/rangetable"
|
|
||||||
)
|
|
||||||
|
|
||||||
var versionList = flag.String("versions", "",
|
|
||||||
"list of versions for which to generate RangeTables")
|
|
||||||
|
|
||||||
const bootstrapMessage = `No versions specified.
|
|
||||||
To bootstrap the code generation, run:
|
|
||||||
go run gen.go --versions=4.1.0,5.0.0,6.0.0,6.1.0,6.2.0,6.3.0,7.0.0
|
|
||||||
|
|
||||||
and ensure that the latest versions are included by checking:
|
|
||||||
http://www.unicode.org/Public/`
|
|
||||||
|
|
||||||
func getVersions() []string {
|
|
||||||
if *versionList == "" {
|
|
||||||
log.Fatal(bootstrapMessage)
|
|
||||||
}
|
|
||||||
|
|
||||||
versions := strings.Split(*versionList, ",")
|
|
||||||
sort.Strings(versions)
|
|
||||||
|
|
||||||
// Ensure that at least the current version is included.
|
|
||||||
for _, v := range versions {
|
|
||||||
if v == gen.UnicodeVersion() {
|
|
||||||
return versions
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
versions = append(versions, gen.UnicodeVersion())
|
|
||||||
sort.Strings(versions)
|
|
||||||
return versions
|
|
||||||
}
|
|
||||||
|
|
||||||
func main() {
|
|
||||||
gen.Init()
|
|
||||||
|
|
||||||
versions := getVersions()
|
|
||||||
|
|
||||||
w := &bytes.Buffer{}
|
|
||||||
|
|
||||||
fmt.Fprintf(w, "//go:generate go run gen.go --versions=%s\n\n", strings.Join(versions, ","))
|
|
||||||
fmt.Fprintf(w, "import \"unicode\"\n\n")
|
|
||||||
|
|
||||||
vstr := func(s string) string { return strings.Replace(s, ".", "_", -1) }
|
|
||||||
|
|
||||||
fmt.Fprintf(w, "var assigned = map[string]*unicode.RangeTable{\n")
|
|
||||||
for _, v := range versions {
|
|
||||||
fmt.Fprintf(w, "\t%q: assigned%s,\n", v, vstr(v))
|
|
||||||
}
|
|
||||||
fmt.Fprintf(w, "}\n\n")
|
|
||||||
|
|
||||||
var size int
|
|
||||||
for _, v := range versions {
|
|
||||||
assigned := []rune{}
|
|
||||||
|
|
||||||
r := gen.Open("http://www.unicode.org/Public/", "", v+"/ucd/UnicodeData.txt")
|
|
||||||
ucd.Parse(r, func(p *ucd.Parser) {
|
|
||||||
assigned = append(assigned, p.Rune(0))
|
|
||||||
})
|
|
||||||
|
|
||||||
rt := rangetable.New(assigned...)
|
|
||||||
sz := int(reflect.TypeOf(unicode.RangeTable{}).Size())
|
|
||||||
sz += int(reflect.TypeOf(unicode.Range16{}).Size()) * len(rt.R16)
|
|
||||||
sz += int(reflect.TypeOf(unicode.Range32{}).Size()) * len(rt.R32)
|
|
||||||
|
|
||||||
fmt.Fprintf(w, "// size %d bytes (%d KiB)\n", sz, sz/1024)
|
|
||||||
fmt.Fprintf(w, "var assigned%s = ", vstr(v))
|
|
||||||
print(w, rt)
|
|
||||||
|
|
||||||
size += sz
|
|
||||||
}
|
|
||||||
|
|
||||||
fmt.Fprintf(w, "// Total size %d bytes (%d KiB)\n", size, size/1024)
|
|
||||||
|
|
||||||
gen.WriteGoFile("tables.go", "rangetable", w.Bytes())
|
|
||||||
}
|
|
||||||
|
|
||||||
func print(w io.Writer, rt *unicode.RangeTable) {
|
|
||||||
fmt.Fprintln(w, "&unicode.RangeTable{")
|
|
||||||
fmt.Fprintln(w, "\tR16: []unicode.Range16{")
|
|
||||||
for _, r := range rt.R16 {
|
|
||||||
fmt.Fprintf(w, "\t\t{%#04x, %#04x, %d},\n", r.Lo, r.Hi, r.Stride)
|
|
||||||
}
|
|
||||||
fmt.Fprintln(w, "\t},")
|
|
||||||
fmt.Fprintln(w, "\tR32: []unicode.Range32{")
|
|
||||||
for _, r := range rt.R32 {
|
|
||||||
fmt.Fprintf(w, "\t\t{%#08x, %#08x, %d},\n", r.Lo, r.Hi, r.Stride)
|
|
||||||
}
|
|
||||||
fmt.Fprintln(w, "\t},")
|
|
||||||
fmt.Fprintf(w, "\tLatinOffset: %d,\n", rt.LatinOffset)
|
|
||||||
fmt.Fprintf(w, "}\n\n")
|
|
||||||
}
|
|
|
@ -1,260 +0,0 @@
|
||||||
// Copyright 2015 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
package rangetable
|
|
||||||
|
|
||||||
import (
|
|
||||||
"unicode"
|
|
||||||
)
|
|
||||||
|
|
||||||
// atEnd is used to mark a completed iteration.
|
|
||||||
const atEnd = unicode.MaxRune + 1
|
|
||||||
|
|
||||||
// Merge returns a new RangeTable that is the union of the given tables.
|
|
||||||
// It can also be used to compact user-created RangeTables. The entries in
|
|
||||||
// R16 and R32 for any given RangeTable should be sorted and non-overlapping.
|
|
||||||
//
|
|
||||||
// A lookup in the resulting table can be several times faster than using In
|
|
||||||
// directly on the ranges. Merge is an expensive operation, however, and only
|
|
||||||
// makes sense if one intends to use the result for more than a couple of
|
|
||||||
// hundred lookups.
|
|
||||||
func Merge(ranges ...*unicode.RangeTable) *unicode.RangeTable {
|
|
||||||
rt := &unicode.RangeTable{}
|
|
||||||
if len(ranges) == 0 {
|
|
||||||
return rt
|
|
||||||
}
|
|
||||||
|
|
||||||
iter := tablesIter(make([]tableIndex, len(ranges)))
|
|
||||||
|
|
||||||
for i, t := range ranges {
|
|
||||||
iter[i] = tableIndex{t, 0, atEnd}
|
|
||||||
if len(t.R16) > 0 {
|
|
||||||
iter[i].next = rune(t.R16[0].Lo)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if r0 := iter.next16(); r0.Stride != 0 {
|
|
||||||
for {
|
|
||||||
r1 := iter.next16()
|
|
||||||
if r1.Stride == 0 {
|
|
||||||
rt.R16 = append(rt.R16, r0)
|
|
||||||
break
|
|
||||||
}
|
|
||||||
stride := r1.Lo - r0.Hi
|
|
||||||
if (r1.Lo == r1.Hi || stride == r1.Stride) && (r0.Lo == r0.Hi || stride == r0.Stride) {
|
|
||||||
// Fully merge the next range into the previous one.
|
|
||||||
r0.Hi, r0.Stride = r1.Hi, stride
|
|
||||||
continue
|
|
||||||
} else if stride == r0.Stride {
|
|
||||||
// Move the first element of r1 to r0. This may eliminate an
|
|
||||||
// entry.
|
|
||||||
r0.Hi = r1.Lo
|
|
||||||
r0.Stride = stride
|
|
||||||
r1.Lo = r1.Lo + r1.Stride
|
|
||||||
if r1.Lo > r1.Hi {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
}
|
|
||||||
rt.R16 = append(rt.R16, r0)
|
|
||||||
r0 = r1
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
for i, t := range ranges {
|
|
||||||
iter[i] = tableIndex{t, 0, atEnd}
|
|
||||||
if len(t.R32) > 0 {
|
|
||||||
iter[i].next = rune(t.R32[0].Lo)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if r0 := iter.next32(); r0.Stride != 0 {
|
|
||||||
for {
|
|
||||||
r1 := iter.next32()
|
|
||||||
if r1.Stride == 0 {
|
|
||||||
rt.R32 = append(rt.R32, r0)
|
|
||||||
break
|
|
||||||
}
|
|
||||||
stride := r1.Lo - r0.Hi
|
|
||||||
if (r1.Lo == r1.Hi || stride == r1.Stride) && (r0.Lo == r0.Hi || stride == r0.Stride) {
|
|
||||||
// Fully merge the next range into the previous one.
|
|
||||||
r0.Hi, r0.Stride = r1.Hi, stride
|
|
||||||
continue
|
|
||||||
} else if stride == r0.Stride {
|
|
||||||
// Move the first element of r1 to r0. This may eliminate an
|
|
||||||
// entry.
|
|
||||||
r0.Hi = r1.Lo
|
|
||||||
r1.Lo = r1.Lo + r1.Stride
|
|
||||||
if r1.Lo > r1.Hi {
|
|
||||||
continue
|
|
||||||
}
|
|
||||||
}
|
|
||||||
rt.R32 = append(rt.R32, r0)
|
|
||||||
r0 = r1
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
for i := 0; i < len(rt.R16) && rt.R16[i].Hi <= unicode.MaxLatin1; i++ {
|
|
||||||
rt.LatinOffset = i + 1
|
|
||||||
}
|
|
||||||
|
|
||||||
return rt
|
|
||||||
}
|
|
||||||
|
|
||||||
type tableIndex struct {
|
|
||||||
t *unicode.RangeTable
|
|
||||||
p uint32
|
|
||||||
next rune
|
|
||||||
}
|
|
||||||
|
|
||||||
type tablesIter []tableIndex
|
|
||||||
|
|
||||||
// sortIter does an insertion sort using the next field of tableIndex. Insertion
|
|
||||||
// sort is a good sorting algorithm for this case.
|
|
||||||
func sortIter(t []tableIndex) {
|
|
||||||
for i := range t {
|
|
||||||
for j := i; j > 0 && t[j-1].next > t[j].next; j-- {
|
|
||||||
t[j], t[j-1] = t[j-1], t[j]
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// next16 finds the ranged to be added to the table. If ranges overlap between
|
|
||||||
// multiple tables it clips the result to a non-overlapping range if the
|
|
||||||
// elements are not fully subsumed. It returns a zero range if there are no more
|
|
||||||
// ranges.
|
|
||||||
func (ti tablesIter) next16() unicode.Range16 {
|
|
||||||
sortIter(ti)
|
|
||||||
|
|
||||||
t0 := ti[0]
|
|
||||||
if t0.next == atEnd {
|
|
||||||
return unicode.Range16{}
|
|
||||||
}
|
|
||||||
r0 := t0.t.R16[t0.p]
|
|
||||||
r0.Lo = uint16(t0.next)
|
|
||||||
|
|
||||||
// We restrict the Hi of the current range if it overlaps with another range.
|
|
||||||
for i := range ti {
|
|
||||||
tn := ti[i]
|
|
||||||
// Since our tableIndices are sorted by next, we can break if the there
|
|
||||||
// is no overlap. The first value of a next range can always be merged
|
|
||||||
// into the current one, so we can break in case of equality as well.
|
|
||||||
if rune(r0.Hi) <= tn.next {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
rn := tn.t.R16[tn.p]
|
|
||||||
rn.Lo = uint16(tn.next)
|
|
||||||
|
|
||||||
// Limit r0.Hi based on next ranges in list, but allow it to overlap
|
|
||||||
// with ranges as long as it subsumes it.
|
|
||||||
m := (rn.Lo - r0.Lo) % r0.Stride
|
|
||||||
if m == 0 && (rn.Stride == r0.Stride || rn.Lo == rn.Hi) {
|
|
||||||
// Overlap, take the min of the two Hi values: for simplicity's sake
|
|
||||||
// we only process one range at a time.
|
|
||||||
if r0.Hi > rn.Hi {
|
|
||||||
r0.Hi = rn.Hi
|
|
||||||
}
|
|
||||||
} else {
|
|
||||||
// Not a compatible stride. Set to the last possible value before
|
|
||||||
// rn.Lo, but ensure there is at least one value.
|
|
||||||
if x := rn.Lo - m; r0.Lo <= x {
|
|
||||||
r0.Hi = x
|
|
||||||
}
|
|
||||||
break
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Update the next values for each table.
|
|
||||||
for i := range ti {
|
|
||||||
tn := &ti[i]
|
|
||||||
if rune(r0.Hi) < tn.next {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
rn := tn.t.R16[tn.p]
|
|
||||||
stride := rune(rn.Stride)
|
|
||||||
tn.next += stride * (1 + ((rune(r0.Hi) - tn.next) / stride))
|
|
||||||
if rune(rn.Hi) < tn.next {
|
|
||||||
if tn.p++; int(tn.p) == len(tn.t.R16) {
|
|
||||||
tn.next = atEnd
|
|
||||||
} else {
|
|
||||||
tn.next = rune(tn.t.R16[tn.p].Lo)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if r0.Lo == r0.Hi {
|
|
||||||
r0.Stride = 1
|
|
||||||
}
|
|
||||||
|
|
||||||
return r0
|
|
||||||
}
|
|
||||||
|
|
||||||
// next32 finds the ranged to be added to the table. If ranges overlap between
|
|
||||||
// multiple tables it clips the result to a non-overlapping range if the
|
|
||||||
// elements are not fully subsumed. It returns a zero range if there are no more
|
|
||||||
// ranges.
|
|
||||||
func (ti tablesIter) next32() unicode.Range32 {
|
|
||||||
sortIter(ti)
|
|
||||||
|
|
||||||
t0 := ti[0]
|
|
||||||
if t0.next == atEnd {
|
|
||||||
return unicode.Range32{}
|
|
||||||
}
|
|
||||||
r0 := t0.t.R32[t0.p]
|
|
||||||
r0.Lo = uint32(t0.next)
|
|
||||||
|
|
||||||
// We restrict the Hi of the current range if it overlaps with another range.
|
|
||||||
for i := range ti {
|
|
||||||
tn := ti[i]
|
|
||||||
// Since our tableIndices are sorted by next, we can break if the there
|
|
||||||
// is no overlap. The first value of a next range can always be merged
|
|
||||||
// into the current one, so we can break in case of equality as well.
|
|
||||||
if rune(r0.Hi) <= tn.next {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
rn := tn.t.R32[tn.p]
|
|
||||||
rn.Lo = uint32(tn.next)
|
|
||||||
|
|
||||||
// Limit r0.Hi based on next ranges in list, but allow it to overlap
|
|
||||||
// with ranges as long as it subsumes it.
|
|
||||||
m := (rn.Lo - r0.Lo) % r0.Stride
|
|
||||||
if m == 0 && (rn.Stride == r0.Stride || rn.Lo == rn.Hi) {
|
|
||||||
// Overlap, take the min of the two Hi values: for simplicity's sake
|
|
||||||
// we only process one range at a time.
|
|
||||||
if r0.Hi > rn.Hi {
|
|
||||||
r0.Hi = rn.Hi
|
|
||||||
}
|
|
||||||
} else {
|
|
||||||
// Not a compatible stride. Set to the last possible value before
|
|
||||||
// rn.Lo, but ensure there is at least one value.
|
|
||||||
if x := rn.Lo - m; r0.Lo <= x {
|
|
||||||
r0.Hi = x
|
|
||||||
}
|
|
||||||
break
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Update the next values for each table.
|
|
||||||
for i := range ti {
|
|
||||||
tn := &ti[i]
|
|
||||||
if rune(r0.Hi) < tn.next {
|
|
||||||
break
|
|
||||||
}
|
|
||||||
rn := tn.t.R32[tn.p]
|
|
||||||
stride := rune(rn.Stride)
|
|
||||||
tn.next += stride * (1 + ((rune(r0.Hi) - tn.next) / stride))
|
|
||||||
if rune(rn.Hi) < tn.next {
|
|
||||||
if tn.p++; int(tn.p) == len(tn.t.R32) {
|
|
||||||
tn.next = atEnd
|
|
||||||
} else {
|
|
||||||
tn.next = rune(tn.t.R32[tn.p].Lo)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
if r0.Lo == r0.Hi {
|
|
||||||
r0.Stride = 1
|
|
||||||
}
|
|
||||||
|
|
||||||
return r0
|
|
||||||
}
|
|
|
@ -1,70 +0,0 @@
|
||||||
// Copyright 2015 The Go Authors. All rights reserved.
|
|
||||||
// Use of this source code is governed by a BSD-style
|
|
||||||
// license that can be found in the LICENSE file.
|
|
||||||
|
|
||||||
// Package rangetable provides utilities for creating and inspecting
|
|
||||||
// unicode.RangeTables.
|
|
||||||
package rangetable
|
|
||||||
|
|
||||||
import (
|
|
||||||
"sort"
|
|
||||||
"unicode"
|
|
||||||
)
|
|
||||||
|
|
||||||
// New creates a RangeTable from the given runes, which may contain duplicates.
|
|
||||||
func New(r ...rune) *unicode.RangeTable {
|
|
||||||
if len(r) == 0 {
|
|
||||||
return &unicode.RangeTable{}
|
|
||||||
}
|
|
||||||
|
|
||||||
sort.Sort(byRune(r))
|
|
||||||
|
|
||||||
// Remove duplicates.
|
|
||||||
k := 1
|
|
||||||
for i := 1; i < len(r); i++ {
|
|
||||||
if r[k-1] != r[i] {
|
|
||||||
r[k] = r[i]
|
|
||||||
k++
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
var rt unicode.RangeTable
|
|
||||||
for _, r := range r[:k] {
|
|
||||||
if r <= 0xFFFF {
|
|
||||||
rt.R16 = append(rt.R16, unicode.Range16{Lo: uint16(r), Hi: uint16(r), Stride: 1})
|
|
||||||
} else {
|
|
||||||
rt.R32 = append(rt.R32, unicode.Range32{Lo: uint32(r), Hi: uint32(r), Stride: 1})
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Optimize RangeTable.
|
|
||||||
return Merge(&rt)
|
|
||||||
}
|
|
||||||
|
|
||||||
type byRune []rune
|
|
||||||
|
|
||||||
func (r byRune) Len() int { return len(r) }
|
|
||||||
func (r byRune) Swap(i, j int) { r[i], r[j] = r[j], r[i] }
|
|
||||||
func (r byRune) Less(i, j int) bool { return r[i] < r[j] }
|
|
||||||
|
|
||||||
// Visit visits all runes in the given RangeTable in order, calling fn for each.
|
|
||||||
func Visit(rt *unicode.RangeTable, fn func(rune)) {
|
|
||||||
for _, r16 := range rt.R16 {
|
|
||||||
for r := rune(r16.Lo); r <= rune(r16.Hi); r += rune(r16.Stride) {
|
|
||||||
fn(r)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
for _, r32 := range rt.R32 {
|
|
||||||
for r := rune(r32.Lo); r <= rune(r32.Hi); r += rune(r32.Stride) {
|
|
||||||
fn(r)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Assigned returns a RangeTable with all assigned code points for a given
|
|
||||||
// Unicode version. This includes graphic, format, control, and private-use
|
|
||||||
// characters. It returns nil if the data for the given version is not
|
|
||||||
// available.
|
|
||||||
func Assigned(version string) *unicode.RangeTable {
|
|
||||||
return assigned[version]
|
|
||||||
}
|
|
File diff suppressed because it is too large
Load Diff
Loading…
Reference in New Issue