Tcl Source Code

compat/zlib/win32/zdll.lib
compat/zlib/win32/zlib1.dll
compat/zlib/win64/zdll.lib
compat/zlib/win64/zlib1.dll
compat/zlib/win64/libz.dll.a
compat/zlib/zlib.3.pdf
tools/empty.zip
*.bmp
*.gif
*.png

libtommath/*.vcproj
tools/tcl.hpj.in
tools/tcl.wse.in
win/buildall.vc.bat
win/coffbase.txt
win/makefile.vc
win/rules.vc
win/rules-ext.vc
win/targets.vc
win/tcl.dsp
win/tcl.dsw
win/tcl.hpj.in

libtommath/*.vcproj
tools/tcl.hpj.in
tools/tcl.wse.in
win/buildall.vc.bat
win/coffbase.txt
win/makefile.vc
win/rules.vc
win/rules-ext.vc
win/targets.vc
win/tcl.dsp
win/tcl.dsw
win/tcl.hpj.in

unix/dltest.marker
unix/tcl.pc
unix/tclIndex
unix/pkgs/*
win/Debug*
win/Release*
win/pkgs/*
win/coffbase.txt
win/tcl.hpj
win/nmhlp-out.txt

Important Note
==========
Please do not file issues with Tcl on Github. They are unlikely to be noticed in a timely fashion. Tcl issues are hosted in the [tcl fossil repository on core.tcl.tk](https://core.tcl.tk/tcl/tktnew); please post them there.

Important Note
==========
Please do not file pull requests with Tcl on Github. They are unlikely to be noticed in a timely fashion. Tcl issues (including patches) are hosted in the [tcl fossil repository on core.tcl.tk](https://core.tcl.tk/tcl/tktnew); please post them there.

<?xml version="1.0" encoding="UTF-8"?>
<projectDescription>
	<name>tcl9</name>
	<name>tcl8</name>
	<comment></comment>
	<projects>
	</projects>
	<buildSpec>
	</buildSpec>
	<natures>
	</natures>

README:  Tcl
    This is the Tcl 9.0a0 source distribution.
    This is the Tcl 8.7a2 source distribution.
	http://sourceforge.net/projects/tcl/files/Tcl/
    You can get any source release of Tcl from the URL above.

Contents
--------
    1. Introduction
    2. Documentation

2013-05-06 (platform support) Cygwin64 (nijtmans)

2013-05-15 (enhancement) Improved [list {*}...] compile (fellows)

2013-05-16 (platform support) mingw-4.0 (nijtmans)

2013-05-19 (platform support) FreeBSD updates (gahr)
2013-05-19 (platform support) FreeBSD updates (cerutti)

2013-05-20 (bug fix)[3613567] access error temp file creation (keene)

2013-05-20 (bug fix)[3613569] temp file open fail can crash [load] (keene)

2013-05-22 (bug fix)[3613609] [lsort -nocase] failed on non-ASCII (fellows)

2016-05-13 (bug)[3154ea] Mem corruption in assembler exceptions (tkob,kenny)

2016-05-13 (bug) registry package support any Unicode env (nijtmans)
=> registry 1.3.2

2016-05-21 (bug)[f7d4e] [namespace delete] performance (fellows)

2016-06-02 (TIP 447) execution time verbosity option (gahr)
2016-06-02 (TIP 447) execution time verbosity option (cerutti)
=> tcltest 2.4.0

2016-06-16 (bug)[16828b] crash due to [vwait] trace undo fail (dah,porter)

2016-06-16 (enhancement)[4b61af] good [info frame] from more cases (beric)

2016-06-21 (bug)[c383eb] crash in [glob -path a] (oehlmann,porter)

2017-07-06 (bug)[adb198] Plug memleak in TclJoinPath (sebres,porter)

2017-07-17 (bug)[fb2208] Repeatable tclIndex generation (wiedemann,nijtmans)

--- Released 8.6.7, August 9, 2017 --- http://core.tcl.tk/tcl/ for details

2017-08-10 [array names -regexp] supports backrefs (goth)

2017-08-10 Fix gcc build failures due to #pragma placement (cassoff,fellows)

2017-08-29 (bug)[b50fb2] exec redir append stdout and stderr to file (coulter)

2017-08-31 (bug)[2a9465] http state 100 continue handling broken (oehlmann)
=> http 2.8.12

2017-09-02 (bug)[0e4d88] replace command, delete trace kills namespace (porter)

2017-10-19 (bug)[1a5655] [info * methods] includes mixins (fellows)

2017-10-23 tzdata updated to Olson's tzdata2017c (jima)

2017-10-24 (bug)[fc1409] segfault in method cloning, oo-15.15 (coulter,fellows)

2017-11-03 (bug)[6f2f83] More robust [load] for ReactOS (werner)

2017-11-08 (bug)[3298012] Stop crash when hash tables overflow 32 bits (porter)

2017-11-14 (bug)[5d6de6] Close failing case of [package prefer stable] (kupries)

2017-11-17 (bug)[fab924] Fix misleading [load] message on Windows (oehlmann)

2017-12-05 (bug)[4f6a1e] Crash when ensemble map and list are same (sebres)

2017-12-06 (bug)[ce3a21] file normalize failure when tail is empty (porter)

2017-12-08 (new)[TIP 477] nmake build system reform (nadkarni)

2017-12-19 (bug)[586e71] EvalObjv exception handling at level #0 (sebres,porter)

--- Released 8.6.8, December 22, 2017 --- http://core.tcl.tk/tcl/ for details

Changes to 8.7a1 include all changes to the 8.6 line through 8.6.7,
plus the following, which focuses on the high-level feature changes
in this changeset (new minor version) rather than bug fixes:

2016-03-17 (bug)[0b8c38] socket accept callbacks always in global ns (porter)
        *** POTENTIAL INCOMPATIBILITY ***

2016-07-01 Hack accommodations for legacy Itcl 3 disabled (porter)

2016-07-12 Make TCL_HASH_TYPE build-time configurable (nijtmans)

2017-06-22 (TIP 470) Tcl_GetDefineContextObject();[oo::define [self]] (fellows)
=> TclOO 1.2.0

2017-06-23 (TIP 472) Support 0d as prefix of decimal numbers (iyer,griffin)

2017-08-31 (bug)[2a9465] http state 100 continue handling broken (oehlmann)
=> http 2.8.12

2017-09-02 (bug)[0e4d88] replace command, delete trace kills namespace (porter)

--- Released 8.7a1, September 8, 2017 --- http://core.tcl.tk/tcl/ for details

2018-03-12 (TIP 490) add oo support for msgcat => msgcat 1.7.0 (oehlmann)

2018-03-12 (TIP 499) custom locale preference list (oehlmann)
=> msgcat 1.7.0

/*
 * float.h --
 *
 *	This is a dummy header file to #include in Tcl when there
 *	is no float.h in /usr/include.  Right now this file is empty:
 *	Tcl contains #ifdefs to deal with the lack of definitions;
 *	all it needs is for the #include statement to work.
 *
 * Copyright (c) 1993 The Regents of the University of California.
 * Copyright (c) 1994 Sun Microsystems, Inc.
 *
 * See the file "license.terms" for information on usage and redistribution
 * of this file, and for a DISCLAIMER OF ALL WARRANTIES.
 */

         Modifications of Unzip for Zip64
         Copyright (C) 2007-2008 Even Rouault

         Modifications for Zip64 support on both zip and unzip
         Copyright (C) 2009-2010 Mathias Svensson ( http://result42.com )
*/


#if (!defined(_WIN32)) && (!defined(WIN32)) && (!defined(__APPLE__))
        #ifndef __USE_FILE_OFFSET64
                #define __USE_FILE_OFFSET64
        #endif
        #ifndef __USE_LARGEFILE64
                #define __USE_LARGEFILE64
        #endif

#define FSEEKO_FUNC(stream, offset, origin) fseeko(stream, offset, origin)
#else
#define FOPEN_FUNC(filename, mode) fopen64(filename, mode)
#define FTELLO_FUNC(stream) ftello64(stream)
#define FSEEKO_FUNC(stream, offset, origin) fseeko64(stream, offset, origin)
#endif


#include "tinydir.h"

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <errno.h>
#include <fcntl.h>

    printf("MiniZip 1.1, demo of zLib + MiniZip64 package, written by Gilles Vollant\n");
    printf("more info on MiniZip at http://www.winimage.com/zLibDll/minizip.html\n\n");
}

void do_help()
{
    printf("Usage : minizip [-o] [-a] [-0 to -9] [-p password] [-j] file.zip [files_to_add]\n\n" \
           "  -r  Scan directories recursively\n" \
           "  -o  Overwrite existing file.zip\n" \
           "  -a  Append to existing file.zip\n" \
           "  -0  Store only\n" \
           "  -1  Compress faster\n" \
           "  -9  Compress better\n\n" \
           "  -j  exclude path. store only the file name.\n\n");
}

     largeFile = 1;

                fclose(pFile);
  }

 return largeFile;
}

void addFileToZip(zipFile zf, const char *filenameinzip, const char *password, int opt_exclude_path,int opt_compress_level) {
    FILE * fin;
    int size_read;
    const char *savefilenameinzip;
    zip_fileinfo zi;
    unsigned long crcFile=0;
    int zip64 = 0;
    int err=0;
    int size_buf=WRITEBUFFERSIZE;
    unsigned char buf[WRITEBUFFERSIZE];
    zi.tmz_date.tm_sec = zi.tmz_date.tm_min = zi.tmz_date.tm_hour =
    zi.tmz_date.tm_mday = zi.tmz_date.tm_mon = zi.tmz_date.tm_year = 0;
    zi.dosDate = 0;
    zi.internal_fa = 0;
    zi.external_fa = 0;
    filetime(filenameinzip,&zi.tmz_date,&zi.dosDate);

/*
    err = zipOpenNewFileInZip(zf,filenameinzip,&zi,
                     NULL,0,NULL,0,NULL / * comment * /,
                     (opt_compress_level != 0) ? Z_DEFLATED : 0,
                     opt_compress_level);
*/
    if ((password != NULL) && (err==ZIP_OK))
        err = getFileCrc(filenameinzip,buf,size_buf,&crcFile);

    zip64 = isLargeFile(filenameinzip);

   /* The path name saved, should not include a leading slash. */
   /*if it did, windows/xp and dynazip couldn't read the zip file. */
     savefilenameinzip = filenameinzip;
     while( savefilenameinzip[0] == '\\' || savefilenameinzip[0] == '/' )
     {
         savefilenameinzip++;
     }

     /*should the zip file contain any path at all?*/
     if( opt_exclude_path )
     {
         const char *tmpptr;
         const char *lastslash = 0;
         for( tmpptr = savefilenameinzip; *tmpptr; tmpptr++)
         {
             if( *tmpptr == '\\' || *tmpptr == '/')
             {
                 lastslash = tmpptr;
             }
         }
         if( lastslash != NULL )
         {
             savefilenameinzip = lastslash+1; // base filename follows last slash.
         }
     }

     /**/
    err = zipOpenNewFileInZip3_64(zf,savefilenameinzip,&zi,
                     NULL,0,NULL,0,NULL /* comment*/,
                     (opt_compress_level != 0) ? Z_DEFLATED : 0,
                     opt_compress_level,0,
                     /* -MAX_WBITS, DEF_MEM_LEVEL, Z_DEFAULT_STRATEGY, */
                     -MAX_WBITS, DEF_MEM_LEVEL, Z_DEFAULT_STRATEGY,
                     password,crcFile, zip64);

    if (err != ZIP_OK)
        printf("error in opening %s in zipfile\n",filenameinzip);
    else
    {
        fin = FOPEN_FUNC(filenameinzip,"rb");
        if (fin==NULL)
        {
            err=ZIP_ERRNO;
            printf("error in opening %s for reading\n",filenameinzip);
        }
    }

    if (err == ZIP_OK)
        do
        {
            err = ZIP_OK;
            size_read = (int)fread(buf,1,size_buf,fin);
            if (size_read < size_buf)
                if (feof(fin)==0)
            {
                printf("error in reading %s\n",filenameinzip);
                err = ZIP_ERRNO;
            }

            if (size_read>0)
            {
                err = zipWriteInFileInZip (zf,buf,size_read);
                if (err<0)
                {
                    printf("error in writing %s in the zipfile\n",
                                     filenameinzip);
                }

            }
        } while ((err == ZIP_OK) && (size_read>0));

    if (fin)
        fclose(fin);

    if (err<0)
        err=ZIP_ERRNO;
    else
    {
        err = zipCloseFileInZip(zf);
        if (err!=ZIP_OK)
            printf("error in closing %s in the zipfile\n",
                        filenameinzip);
    }
}


void addPathToZip(zipFile zf, const char *filenameinzip, const char *password, int opt_exclude_path,int opt_compress_level) {
    tinydir_dir dir;
    int i;
    char *newname[512];

    tinydir_open_sorted(&dir, filenameinzip);

    for (i = 0; i < dir.n_files; i++)
    {
        tinydir_file file;
        tinydir_readfile_n(&dir, &file, i);
        if(strcmp(file.name,".")==0) continue;
        if(strcmp(file.name,"..")==0) continue;
        sprintf(newname,"%s/%s",dir.path,file.name);
        if (file.is_dir)
        {
            addPathToZip(zf,newname,password,opt_exclude_path,opt_compress_level);
        } else {
            addFileToZip(zf,newname,password,opt_exclude_path,opt_compress_level);
        }
    }

    tinydir_close(&dir);
}


int main(argc,argv)
    int argc;
    char *argv[];
{
    int i;
    int opt_recursive=0;
    int opt_overwrite=0;
    int opt_overwrite=1;
    int opt_compress_level=Z_DEFAULT_COMPRESSION;
    int opt_exclude_path=0;
    int zipfilenamearg = 0;
    char filename_try[MAXFILENAME+16];
    int zipok;
    int err=0;
    int size_buf=0;

                        opt_overwrite = 1;
                    if ((c=='a') || (c=='A'))
                        opt_overwrite = 2;
                    if ((c>='0') && (c<='9'))
                        opt_compress_level = c-'0';
                    if ((c=='j') || (c=='J'))
                        opt_exclude_path = 1;

                    if ((c=='r') || (c=='R'))
                        opt_recursive = 1;
                    if (((c=='p') || (c=='P')) && (i+1<argc))
                    {
                        password=argv[i+1];
                        i++;
                    }
                }
            }

        for (i=zipfilenamearg+1;(i<argc) && (err==ZIP_OK);i++)
        {
            if (!((((*(argv[i]))=='-') || ((*(argv[i]))=='/')) &&
                  ((argv[i][1]=='o') || (argv[i][1]=='O') ||
                   (argv[i][1]=='a') || (argv[i][1]=='A') ||
                   (argv[i][1]=='p') || (argv[i][1]=='P') ||
                   (argv[i][1]=='r') || (argv[i][1]=='R') ||
                   ((argv[i][1]>='0') || (argv[i][1]<='9'))) &&
                  (strlen(argv[i]) == 2)))
            {
                FILE * fin;
                int size_read;
                const char* filenameinzip = argv[i];
                const char *savefilenameinzip;
                zip_fileinfo zi;
                unsigned long crcFile=0;
                int zip64 = 0;

                if(opt_recursive) {
                zi.tmz_date.tm_sec = zi.tmz_date.tm_min = zi.tmz_date.tm_hour =
                zi.tmz_date.tm_mday = zi.tmz_date.tm_mon = zi.tmz_date.tm_year = 0;
                zi.dosDate = 0;
                zi.internal_fa = 0;
                zi.external_fa = 0;
                filetime(filenameinzip,&zi.tmz_date,&zi.dosDate);

/*
                err = zipOpenNewFileInZip(zf,filenameinzip,&zi,
                                 NULL,0,NULL,0,NULL / * comment * /,
                                 (opt_compress_level != 0) ? Z_DEFLATED : 0,
                                 opt_compress_level);
                    addPathToZip(zf,argv[i],password,opt_exclude_path,opt_compress_level);
*/
                if ((password != NULL) && (err==ZIP_OK))
                    err = getFileCrc(filenameinzip,buf,size_buf,&crcFile);

                zip64 = isLargeFile(filenameinzip);

                                                         /* The path name saved, should not include a leading slash. */
               /*if it did, windows/xp and dynazip couldn't read the zip file. */
                 savefilenameinzip = filenameinzip;
                 while( savefilenameinzip[0] == '\\' || savefilenameinzip[0] == '/' )
                 {
                     savefilenameinzip++;
                 }

                 /*should the zip file contain any path at all?*/
                 if( opt_exclude_path )
                 {
                     const char *tmpptr;
                     const char *lastslash = 0;
                     for( tmpptr = savefilenameinzip; *tmpptr; tmpptr++)
                     {
                         if( *tmpptr == '\\' || *tmpptr == '/')
                         {
                             lastslash = tmpptr;
                         }
                     }
                     if( lastslash != NULL )
                     {
                         savefilenameinzip = lastslash+1; // base filename follows last slash.
                     }
                 }

                 /**/
                err = zipOpenNewFileInZip3_64(zf,savefilenameinzip,&zi,
                                 NULL,0,NULL,0,NULL /* comment*/,
                                 (opt_compress_level != 0) ? Z_DEFLATED : 0,
                                 opt_compress_level,0,
                                 /* -MAX_WBITS, DEF_MEM_LEVEL, Z_DEFAULT_STRATEGY, */
                                 -MAX_WBITS, DEF_MEM_LEVEL, Z_DEFAULT_STRATEGY,
                                 password,crcFile, zip64);

                if (err != ZIP_OK)
                    printf("error in opening %s in zipfile\n",filenameinzip);
                else
                } else {
                {
                    fin = FOPEN_FUNC(filenameinzip,"rb");
                    if (fin==NULL)
                    {
                        err=ZIP_ERRNO;
                        printf("error in opening %s for reading\n",filenameinzip);
                    }
                }

                    addFileToZip(zf,argv[i],password,opt_exclude_path,opt_compress_level);
                if (err == ZIP_OK)
                    do
                    {
                        err = ZIP_OK;
                        size_read = (int)fread(buf,1,size_buf,fin);
                        if (size_read < size_buf)
                            if (feof(fin)==0)
                        {
                            printf("error in reading %s\n",filenameinzip);
                            err = ZIP_ERRNO;
                        }

                        if (size_read>0)
                        {
                            err = zipWriteInFileInZip (zf,buf,size_read);
                            if (err<0)
                            {
                                printf("error in writing %s in the zipfile\n",
                                                 filenameinzip);
                            }

                        }
                    } while ((err == ZIP_OK) && (size_read>0));

                if (fin)
                    fclose(fin);

                if (err<0)
                    err=ZIP_ERRNO;
                else
                {
                    err = zipCloseFileInZip(zf);
                    if (err!=ZIP_OK)
                        printf("error in closing %s in the zipfile\n",
                                    filenameinzip);
                }
            }
        }
        errclose = zipClose(zf,NULL);
        if (errclose != ZIP_OK)
            printf("error in closing %s\n",filename_try);
    }

/*
Copyright (c) 2013-2017, tinydir authors:
- Cong Xu
- Lautis Sun
- Baudouin Feildel
- Andargor <[email protected]>
All rights reserved.

Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:

1. Redistributions of source code must retain the above copyright notice, this
   list of conditions and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice,
   this list of conditions and the following disclaimer in the documentation
   and/or other materials provided with the distribution.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
#ifndef TINYDIR_H
#define TINYDIR_H

#ifdef __cplusplus
extern "C" {
#endif

#if ((defined _UNICODE) && !(defined UNICODE))
#define UNICODE
#endif

#if ((defined UNICODE) && !(defined _UNICODE))
#define _UNICODE
#endif

#include <errno.h>
#include <stdlib.h>
#include <string.h>
#ifdef _MSC_VER
# define WIN32_LEAN_AND_MEAN
# include <windows.h>
# include <tchar.h>
# pragma warning(push)
# pragma warning (disable : 4996)
#else
# include <dirent.h>
# include <libgen.h>
# include <sys/stat.h>
# include <stddef.h>
#endif
#ifdef __MINGW32__
# include <tchar.h>
#endif


/* types */

/* Windows UNICODE wide character support */
#if defined _MSC_VER || defined __MINGW32__
# define _tinydir_char_t TCHAR
# define TINYDIR_STRING(s) _TEXT(s)
# define _tinydir_strlen _tcslen
# define _tinydir_strcpy _tcscpy
# define _tinydir_strcat _tcscat
# define _tinydir_strcmp _tcscmp
# define _tinydir_strrchr _tcsrchr
# define _tinydir_strncmp _tcsncmp
#else
# define _tinydir_char_t char
# define TINYDIR_STRING(s) s
# define _tinydir_strlen strlen
# define _tinydir_strcpy strcpy
# define _tinydir_strcat strcat
# define _tinydir_strcmp strcmp
# define _tinydir_strrchr strrchr
# define _tinydir_strncmp strncmp
#endif

#if (defined _MSC_VER || defined __MINGW32__)
# include <windows.h>
# define _TINYDIR_PATH_MAX MAX_PATH
#elif defined  __linux__
# include <limits.h>
# define _TINYDIR_PATH_MAX PATH_MAX
#elif defined(__unix__) || (defined(__APPLE__) && defined(__MACH__))
# include <sys/param.h>
# if defined(BSD)
#  include <limits.h>
#  define _TINYDIR_PATH_MAX PATH_MAX
# endif
#endif

#ifndef _TINYDIR_PATH_MAX
#define _TINYDIR_PATH_MAX 4096
#endif

#ifdef _MSC_VER
/* extra chars for the "\\*" mask */
# define _TINYDIR_PATH_EXTRA 2
#else
# define _TINYDIR_PATH_EXTRA 0
#endif

#define _TINYDIR_FILENAME_MAX 256

#if (defined _MSC_VER || defined __MINGW32__)
#define _TINYDIR_DRIVE_MAX 3
#endif

#ifdef _MSC_VER
# define _TINYDIR_FUNC static __inline
#elif !defined __STDC_VERSION__ || __STDC_VERSION__ < 199901L
# define _TINYDIR_FUNC static __inline__
#else
# define _TINYDIR_FUNC static inline
#endif

/* readdir_r usage; define TINYDIR_USE_READDIR_R to use it (if supported) */
#ifdef TINYDIR_USE_READDIR_R

/* readdir_r is a POSIX-only function, and may not be available under various
 * environments/settings, e.g. MinGW. Use readdir fallback */
#if _POSIX_C_SOURCE >= 1 || _XOPEN_SOURCE || _BSD_SOURCE || _SVID_SOURCE ||\
	_POSIX_SOURCE
# define _TINYDIR_HAS_READDIR_R
#endif
#if _POSIX_C_SOURCE >= 200112L
# define _TINYDIR_HAS_FPATHCONF
# include <unistd.h>
#endif
#if _BSD_SOURCE || _SVID_SOURCE || \
	(_POSIX_C_SOURCE >= 200809L || _XOPEN_SOURCE >= 700)
# define _TINYDIR_HAS_DIRFD
# include <sys/types.h>
#endif
#if defined _TINYDIR_HAS_FPATHCONF && defined _TINYDIR_HAS_DIRFD &&\
	defined _PC_NAME_MAX
# define _TINYDIR_USE_FPATHCONF
#endif
#if defined __MINGW32__ || !defined _TINYDIR_HAS_READDIR_R ||\
	!(defined _TINYDIR_USE_FPATHCONF || defined NAME_MAX)
# define _TINYDIR_USE_READDIR
#endif

/* Use readdir by default */
#else
# define _TINYDIR_USE_READDIR
#endif

/* MINGW32 has two versions of dirent, ASCII and UNICODE*/
#ifndef _MSC_VER
#if (defined __MINGW32__) && (defined _UNICODE)
#define _TINYDIR_DIR _WDIR
#define _tinydir_dirent _wdirent
#define _tinydir_opendir _wopendir
#define _tinydir_readdir _wreaddir
#define _tinydir_closedir _wclosedir
#else
#define _TINYDIR_DIR DIR
#define _tinydir_dirent dirent
#define _tinydir_opendir opendir
#define _tinydir_readdir readdir
#define _tinydir_closedir closedir
#endif
#endif

/* Allow user to use a custom allocator by defining _TINYDIR_MALLOC and _TINYDIR_FREE. */
#if    defined(_TINYDIR_MALLOC) &&  defined(_TINYDIR_FREE)
#elif !defined(_TINYDIR_MALLOC) && !defined(_TINYDIR_FREE)
#else
#error "Either define both alloc and free or none of them!"
#endif

#if !defined(_TINYDIR_MALLOC)
	#define _TINYDIR_MALLOC(_size) malloc(_size)
	#define _TINYDIR_FREE(_ptr)    free(_ptr)
#endif /* !defined(_TINYDIR_MALLOC) */

typedef struct tinydir_file
{
	_tinydir_char_t path[_TINYDIR_PATH_MAX];
	_tinydir_char_t name[_TINYDIR_FILENAME_MAX];
	_tinydir_char_t *extension;
	int is_dir;
	int is_reg;

#ifndef _MSC_VER
#ifdef __MINGW32__
	struct _stat _s;
#else
	struct stat _s;
#endif
#endif
} tinydir_file;

typedef struct tinydir_dir
{
	_tinydir_char_t path[_TINYDIR_PATH_MAX];
	int has_next;
	size_t n_files;

	tinydir_file *_files;
#ifdef _MSC_VER
	HANDLE _h;
	WIN32_FIND_DATA _f;
#else
	_TINYDIR_DIR *_d;
	struct _tinydir_dirent *_e;
#ifndef _TINYDIR_USE_READDIR
	struct _tinydir_dirent *_ep;
#endif
#endif
} tinydir_dir;


/* declarations */

_TINYDIR_FUNC
int tinydir_open(tinydir_dir *dir, const _tinydir_char_t *path);
_TINYDIR_FUNC
int tinydir_open_sorted(tinydir_dir *dir, const _tinydir_char_t *path);
_TINYDIR_FUNC
void tinydir_close(tinydir_dir *dir);

_TINYDIR_FUNC
int tinydir_next(tinydir_dir *dir);
_TINYDIR_FUNC
int tinydir_readfile(const tinydir_dir *dir, tinydir_file *file);
_TINYDIR_FUNC
int tinydir_readfile_n(const tinydir_dir *dir, tinydir_file *file, size_t i);
_TINYDIR_FUNC
int tinydir_open_subdir_n(tinydir_dir *dir, size_t i);

_TINYDIR_FUNC
int tinydir_file_open(tinydir_file *file, const _tinydir_char_t *path);
_TINYDIR_FUNC
void _tinydir_get_ext(tinydir_file *file);
_TINYDIR_FUNC
int _tinydir_file_cmp(const void *a, const void *b);
#ifndef _MSC_VER
#ifndef _TINYDIR_USE_READDIR
_TINYDIR_FUNC
size_t _tinydir_dirent_buf_size(_TINYDIR_DIR *dirp);
#endif
#endif


/* definitions*/

_TINYDIR_FUNC
int tinydir_open(tinydir_dir *dir, const _tinydir_char_t *path)
{
#ifndef _MSC_VER
#ifndef _TINYDIR_USE_READDIR
	int error;
	int size;	/* using int size */
#endif
#else
	_tinydir_char_t path_buf[_TINYDIR_PATH_MAX];
#endif
	_tinydir_char_t *pathp;

	if (dir == NULL || path == NULL || _tinydir_strlen(path) == 0)
	{
		errno = EINVAL;
		return -1;
	}
	if (_tinydir_strlen(path) + _TINYDIR_PATH_EXTRA >= _TINYDIR_PATH_MAX)
	{
		errno = ENAMETOOLONG;
		return -1;
	}

	/* initialise dir */
	dir->_files = NULL;
#ifdef _MSC_VER
	dir->_h = INVALID_HANDLE_VALUE;
#else
	dir->_d = NULL;
#ifndef _TINYDIR_USE_READDIR
	dir->_ep = NULL;
#endif
#endif
	tinydir_close(dir);

	_tinydir_strcpy(dir->path, path);
	/* Remove trailing slashes */
	pathp = &dir->path[_tinydir_strlen(dir->path) - 1];
	while (pathp != dir->path && (*pathp == TINYDIR_STRING('\\') || *pathp == TINYDIR_STRING('/')))
	{
		*pathp = TINYDIR_STRING('\0');
		pathp++;
	}
#ifdef _MSC_VER
	_tinydir_strcpy(path_buf, dir->path);
	_tinydir_strcat(path_buf, TINYDIR_STRING("\\*"));
#if (defined WINAPI_FAMILY) && (WINAPI_FAMILY != WINAPI_FAMILY_DESKTOP_APP)
	dir->_h = FindFirstFileEx(path_buf, FindExInfoStandard, &dir->_f, FindExSearchNameMatch, NULL, 0);
#else
	dir->_h = FindFirstFile(path_buf, &dir->_f);
#endif
	if (dir->_h == INVALID_HANDLE_VALUE)
	{
		errno = ENOENT;
#else
	dir->_d = _tinydir_opendir(path);
	if (dir->_d == NULL)
	{
#endif
		goto bail;
	}

	/* read first file */
	dir->has_next = 1;
#ifndef _MSC_VER
#ifdef _TINYDIR_USE_READDIR
	dir->_e = _tinydir_readdir(dir->_d);
#else
	/* allocate dirent buffer for readdir_r */
	size = _tinydir_dirent_buf_size(dir->_d); /* conversion to int */
	if (size == -1) return -1;
	dir->_ep = (struct _tinydir_dirent*)_TINYDIR_MALLOC(size);
	if (dir->_ep == NULL) return -1;

	error = readdir_r(dir->_d, dir->_ep, &dir->_e);
	if (error != 0) return -1;
#endif
	if (dir->_e == NULL)
	{
		dir->has_next = 0;
	}
#endif

	return 0;

bail:
	tinydir_close(dir);
	return -1;
}

_TINYDIR_FUNC
int tinydir_open_sorted(tinydir_dir *dir, const _tinydir_char_t *path)
{
	/* Count the number of files first, to pre-allocate the files array */
	size_t n_files = 0;
	if (tinydir_open(dir, path) == -1)
	{
		return -1;
	}
	while (dir->has_next)
	{
		n_files++;
		if (tinydir_next(dir) == -1)
		{
			goto bail;
		}
	}
	tinydir_close(dir);

	if (tinydir_open(dir, path) == -1)
	{
		return -1;
	}

	dir->n_files = 0;
	dir->_files = (tinydir_file *)_TINYDIR_MALLOC(sizeof *dir->_files * n_files);
	if (dir->_files == NULL)
	{
		goto bail;
	}
	while (dir->has_next)
	{
		tinydir_file *p_file;
		dir->n_files++;

		p_file = &dir->_files[dir->n_files - 1];
		if (tinydir_readfile(dir, p_file) == -1)
		{
			goto bail;
		}

		if (tinydir_next(dir) == -1)
		{
			goto bail;
		}

		/* Just in case the number of files has changed between the first and
		second reads, terminate without writing into unallocated memory */
		if (dir->n_files == n_files)
		{
			break;
		}
	}

	qsort(dir->_files, dir->n_files, sizeof(tinydir_file), _tinydir_file_cmp);

	return 0;

bail:
	tinydir_close(dir);
	return -1;
}

_TINYDIR_FUNC
void tinydir_close(tinydir_dir *dir)
{
	if (dir == NULL)
	{
		return;
	}

	memset(dir->path, 0, sizeof(dir->path));
	dir->has_next = 0;
	dir->n_files = 0;
	_TINYDIR_FREE(dir->_files);
	dir->_files = NULL;
#ifdef _MSC_VER
	if (dir->_h != INVALID_HANDLE_VALUE)
	{
		FindClose(dir->_h);
	}
	dir->_h = INVALID_HANDLE_VALUE;
#else
	if (dir->_d)
	{
		_tinydir_closedir(dir->_d);
	}
	dir->_d = NULL;
	dir->_e = NULL;
#ifndef _TINYDIR_USE_READDIR
	_TINYDIR_FREE(dir->_ep);
	dir->_ep = NULL;
#endif
#endif
}

_TINYDIR_FUNC
int tinydir_next(tinydir_dir *dir)
{
	if (dir == NULL)
	{
		errno = EINVAL;
		return -1;
	}
	if (!dir->has_next)
	{
		errno = ENOENT;
		return -1;
	}

#ifdef _MSC_VER
	if (FindNextFile(dir->_h, &dir->_f) == 0)
#else
#ifdef _TINYDIR_USE_READDIR
	dir->_e = _tinydir_readdir(dir->_d);
#else
	if (dir->_ep == NULL)
	{
		return -1;
	}
	if (readdir_r(dir->_d, dir->_ep, &dir->_e) != 0)
	{
		return -1;
	}
#endif
	if (dir->_e == NULL)
#endif
	{
		dir->has_next = 0;
#ifdef _MSC_VER
		if (GetLastError() != ERROR_SUCCESS &&
			GetLastError() != ERROR_NO_MORE_FILES)
		{
			tinydir_close(dir);
			errno = EIO;
			return -1;
		}
#endif
	}

	return 0;
}

_TINYDIR_FUNC
int tinydir_readfile(const tinydir_dir *dir, tinydir_file *file)
{
	if (dir == NULL || file == NULL)
	{
		errno = EINVAL;
		return -1;
	}
#ifdef _MSC_VER
	if (dir->_h == INVALID_HANDLE_VALUE)
#else
	if (dir->_e == NULL)
#endif
	{
		errno = ENOENT;
		return -1;
	}
	if (_tinydir_strlen(dir->path) +
		_tinydir_strlen(
#ifdef _MSC_VER
			dir->_f.cFileName
#else
			dir->_e->d_name
#endif
		) + 1 + _TINYDIR_PATH_EXTRA >=
		_TINYDIR_PATH_MAX)
	{
		/* the path for the file will be too long */
		errno = ENAMETOOLONG;
		return -1;
	}
	if (_tinydir_strlen(
#ifdef _MSC_VER
			dir->_f.cFileName
#else
			dir->_e->d_name
#endif
		) >= _TINYDIR_FILENAME_MAX)
	{
		errno = ENAMETOOLONG;
		return -1;
	}

	_tinydir_strcpy(file->path, dir->path);
	_tinydir_strcat(file->path, TINYDIR_STRING("/"));
	_tinydir_strcpy(file->name,
#ifdef _MSC_VER
		dir->_f.cFileName
#else
		dir->_e->d_name
#endif
	);
	_tinydir_strcat(file->path, file->name);
#ifndef _MSC_VER
#ifdef __MINGW32__
	if (_tstat(
#else
	if (stat(
#endif
		file->path, &file->_s) == -1)
	{
		return -1;
	}
#endif
	_tinydir_get_ext(file);

	file->is_dir =
#ifdef _MSC_VER
		!!(dir->_f.dwFileAttributes & FILE_ATTRIBUTE_DIRECTORY);
#else
		S_ISDIR(file->_s.st_mode);
#endif
	file->is_reg =
#ifdef _MSC_VER
		!!(dir->_f.dwFileAttributes & FILE_ATTRIBUTE_NORMAL) ||
		(
			!(dir->_f.dwFileAttributes & FILE_ATTRIBUTE_DEVICE) &&
			!(dir->_f.dwFileAttributes & FILE_ATTRIBUTE_DIRECTORY) &&
			!(dir->_f.dwFileAttributes & FILE_ATTRIBUTE_ENCRYPTED) &&
#ifdef FILE_ATTRIBUTE_INTEGRITY_STREAM
			!(dir->_f.dwFileAttributes & FILE_ATTRIBUTE_INTEGRITY_STREAM) &&
#endif
#ifdef FILE_ATTRIBUTE_NO_SCRUB_DATA
			!(dir->_f.dwFileAttributes & FILE_ATTRIBUTE_NO_SCRUB_DATA) &&
#endif
			!(dir->_f.dwFileAttributes & FILE_ATTRIBUTE_OFFLINE) &&
			!(dir->_f.dwFileAttributes & FILE_ATTRIBUTE_TEMPORARY));
#else
		S_ISREG(file->_s.st_mode);
#endif

	return 0;
}

_TINYDIR_FUNC
int tinydir_readfile_n(const tinydir_dir *dir, tinydir_file *file, size_t i)
{
	if (dir == NULL || file == NULL)
	{
		errno = EINVAL;
		return -1;
	}
	if (i >= dir->n_files)
	{
		errno = ENOENT;
		return -1;
	}

	memcpy(file, &dir->_files[i], sizeof(tinydir_file));
	_tinydir_get_ext(file);

	return 0;
}

_TINYDIR_FUNC
int tinydir_open_subdir_n(tinydir_dir *dir, size_t i)
{
	_tinydir_char_t path[_TINYDIR_PATH_MAX];
	if (dir == NULL)
	{
		errno = EINVAL;
		return -1;
	}
	if (i >= dir->n_files || !dir->_files[i].is_dir)
	{
		errno = ENOENT;
		return -1;
	}

	_tinydir_strcpy(path, dir->_files[i].path);
	tinydir_close(dir);
	if (tinydir_open_sorted(dir, path) == -1)
	{
		return -1;
	}

	return 0;
}

/* Open a single file given its path */
_TINYDIR_FUNC
int tinydir_file_open(tinydir_file *file, const _tinydir_char_t *path)
{
	tinydir_dir dir;
	int result = 0;
	int found = 0;
	_tinydir_char_t dir_name_buf[_TINYDIR_PATH_MAX];
	_tinydir_char_t file_name_buf[_TINYDIR_FILENAME_MAX];
	_tinydir_char_t *dir_name;
	_tinydir_char_t *base_name;
#if (defined _MSC_VER || defined __MINGW32__)
	_tinydir_char_t drive_buf[_TINYDIR_PATH_MAX];
	_tinydir_char_t ext_buf[_TINYDIR_FILENAME_MAX];
#endif

	if (file == NULL || path == NULL || _tinydir_strlen(path) == 0)
	{
		errno = EINVAL;
		return -1;
	}
	if (_tinydir_strlen(path) + _TINYDIR_PATH_EXTRA >= _TINYDIR_PATH_MAX)
	{
		errno = ENAMETOOLONG;
		return -1;
	}

	/* Get the parent path */
#if (defined _MSC_VER || defined __MINGW32__)
#if ((defined _MSC_VER) && (_MSC_VER >= 1400))
		_tsplitpath_s(
			path,
			drive_buf, _TINYDIR_DRIVE_MAX,
			dir_name_buf, _TINYDIR_FILENAME_MAX,
			file_name_buf, _TINYDIR_FILENAME_MAX,
			ext_buf, _TINYDIR_FILENAME_MAX);
#else
		_tsplitpath(
			path,
			drive_buf,
			dir_name_buf,
			file_name_buf,
			ext_buf);
#endif

/* _splitpath_s not work fine with only filename and widechar support */
#ifdef _UNICODE
		if (drive_buf[0] == L'\xFEFE')
			drive_buf[0] = '\0';
		if (dir_name_buf[0] == L'\xFEFE')
			dir_name_buf[0] = '\0';
#endif

	if (errno)
	{
		errno = EINVAL;
		return -1;
	}
	/* Emulate the behavior of dirname by returning "." for dir name if it's
	empty */
	if (drive_buf[0] == '\0' && dir_name_buf[0] == '\0')
	{
		_tinydir_strcpy(dir_name_buf, TINYDIR_STRING("."));
	}
	/* Concatenate the drive letter and dir name to form full dir name */
	_tinydir_strcat(drive_buf, dir_name_buf);
	dir_name = drive_buf;
	/* Concatenate the file name and extension to form base name */
	_tinydir_strcat(file_name_buf, ext_buf);
	base_name = file_name_buf;
#else
	_tinydir_strcpy(dir_name_buf, path);
	dir_name = dirname(dir_name_buf);
	_tinydir_strcpy(file_name_buf, path);
	base_name =basename(file_name_buf);
#endif

	/* Open the parent directory */
	if (tinydir_open(&dir, dir_name) == -1)
	{
		return -1;
	}

	/* Read through the parent directory and look for the file */
	while (dir.has_next)
	{
		if (tinydir_readfile(&dir, file) == -1)
		{
			result = -1;
			goto bail;
		}
		if (_tinydir_strcmp(file->name, base_name) == 0)
		{
			/* File found */
			found = 1;
			break;
		}
		tinydir_next(&dir);
	}
	if (!found)
	{
		result = -1;
		errno = ENOENT;
	}

bail:
	tinydir_close(&dir);
	return result;
}

_TINYDIR_FUNC
void _tinydir_get_ext(tinydir_file *file)
{
	_tinydir_char_t *period = _tinydir_strrchr(file->name, TINYDIR_STRING('.'));
	if (period == NULL)
	{
		file->extension = &(file->name[_tinydir_strlen(file->name)]);
	}
	else
	{
		file->extension = period + 1;
	}
}

_TINYDIR_FUNC
int _tinydir_file_cmp(const void *a, const void *b)
{
	const tinydir_file *fa = (const tinydir_file *)a;
	const tinydir_file *fb = (const tinydir_file *)b;
	if (fa->is_dir != fb->is_dir)
	{
		return -(fa->is_dir - fb->is_dir);
	}
	return _tinydir_strncmp(fa->name, fb->name, _TINYDIR_FILENAME_MAX);
}

#ifndef _MSC_VER
#ifndef _TINYDIR_USE_READDIR
/*
The following authored by Ben Hutchings <[email protected]>
from https://womble.decadent.org.uk/readdir_r-advisory.html
*/
/* Calculate the required buffer size (in bytes) for directory      *
* entries read from the given directory handle.  Return -1 if this  *
* this cannot be done.                                              *
*                                                                   *
* This code does not trust values of NAME_MAX that are less than    *
* 255, since some systems (including at least HP-UX) incorrectly    *
* define it to be a smaller value.                                  */
_TINYDIR_FUNC
size_t _tinydir_dirent_buf_size(_TINYDIR_DIR *dirp)
{
	long name_max;
	size_t name_end;
	/* parameter may be unused */
	(void)dirp;

#if defined _TINYDIR_USE_FPATHCONF
	name_max = fpathconf(dirfd(dirp), _PC_NAME_MAX);
	if (name_max == -1)
#if defined(NAME_MAX)
		name_max = (NAME_MAX > 255) ? NAME_MAX : 255;
#else
		return (size_t)(-1);
#endif
#elif defined(NAME_MAX)
 	name_max = (NAME_MAX > 255) ? NAME_MAX : 255;
#else
#error "buffer size for readdir_r cannot be determined"
#endif
	name_end = (size_t)offsetof(struct _tinydir_dirent, d_name) + name_max + 1;
	return (name_end > sizeof(struct _tinydir_dirent) ?
		name_end : sizeof(struct _tinydir_dirent));
}
#endif
#endif

#ifdef __cplusplus
}
#endif

# if defined (_MSC_VER)
# pragma warning(pop)
# endif

#endif

\fBTcl_ExternalToUtf\fR.
.PP
\fBTcl_WinUtfToTChar\fR and \fBTcl_WinTCharToUtf\fR are
Windows-only convenience
functions for converting between UTF-8 and Windows strings
based on the TCHAR type which is by convention
a Unicode character on Windows NT.
These functions are essentially wrappers around
\fBTcl_UtfToExternalDString\fR and
\fBTcl_ExternalToUtfDString\fR that convert to and from the
Unicode encoding.
.PP
\fBTcl_GetEncodingName\fR is roughly the inverse of \fBTcl_GetEncoding\fR.
Given an \fIencoding\fR, the return value is the \fIname\fR argument that
was used to create the encoding.  The string returned by
\fBTcl_GetEncodingName\fR is only guaranteed to persist until the
\fIencoding\fR is deleted.  The caller must not modify this string.
.PP

\fBTCL_EVAL_DIRECT\fR flag is useful in situations where the
contents of a value are going to change immediately, so the
bytecodes will not be reused in a future execution.  In this case,
it is faster to execute the script directly.
.TP 23
\fBTCL_EVAL_GLOBAL\fR
.
If this flag is set, the script is processed at global level.  This
If this flag is set, the script is evaluated in the global namespace instead of
means that it is evaluated in the global namespace and its variable
context consists of global variables only (it ignores any Tcl
procedures that are active).
the current namespace and its variable context consists of global variables
only (it ignores any Tcl procedures that are active).
.\" TODO: document TCL_EVAL_INVOKE and TCL_EVAL_NOERR.

.SH "MISCELLANEOUS DETAILS"
.PP
During the processing of a Tcl command it is legal to make nested
calls to evaluate other commands (this is how procedures and
some control structures are implemented).
If a code other than \fBTCL_OK\fR is returned

.QW \fB0d\fR
then \fIsrc\fR is expected to be in decimal form; otherwise,
if the first such characters are
.QW \fB0o\fR
then \fIsrc\fR is expected to be in octal form;  otherwise,
if the first such characters are
.QW \fB0b\fR
then \fIsrc\fR is expected to be in binary form;  otherwise,
if the first such character is
.QW \fB0\fR
then \fIsrc\fR
is expected to be in binary form;  otherwise, \fIsrc\fR is
expected to be in decimal form.
is expected to be in octal form;  otherwise, \fIsrc\fR
is expected to be in decimal form.
.PP
\fBTcl_GetDouble\fR expects \fIsrc\fR to consist of a floating-point
number, which is:  white space;  a sign; a sequence of digits;  a
decimal point
.QW \fB.\fR ;
a sequence of digits;  the letter
.QW \fBe\fR ;

with which values might be exchanged.  The C integral types for which Tcl
provides value exchange routines are \fBint\fR, \fBlong int\fR,
\fBTcl_WideInt\fR, and \fBmp_int\fR.  The \fBint\fR and \fBlong int\fR types
are provided by the C language standard.  The \fBTcl_WideInt\fR type is a
typedef defined to be whatever signed integral type covers at least the
64-bit integer range (-9223372036854775808 to 9223372036854775807).  Depending
on the platform and the C compiler, the actual type might be
\fBlong int\fR, \fBlong long int\fR, \fB__int64\fR, or something else.
\fBlong long int\fR, \fB__int64\fR, or something else.
The \fBmp_int\fR type is a multiple-precision integer type defined
by the LibTomMath multiple-precision integer library.
.PP
The \fBTcl_NewIntObj\fR, \fBTcl_NewLongObj\fR, \fBTcl_NewWideIntObj\fR,
and \fBTcl_NewBignumObj\fR routines each create and return a new
Tcl value initialized to the integral value of the argument.  The
returned Tcl value is unshared.

structure.  Callers of \fBTcl_CreateInterp\fR should use this pointer
as an opaque token, suitable for nothing other than passing back to
other routines in the Tcl interface.  Accessing fields directly through
the pointer as described below is no longer supported.  The supported
public routines \fBTcl_SetResult\fR, \fBTcl_GetResult\fR,
\fBTcl_SetErrorLine\fR, \fBTcl_GetErrorLine\fR must be used instead.
.PP
For legacy programs and extensions no longer being maintained, compiles
against the Tcl 8.6 header files are only possible with the compiler
directives
.CS
#define USE_INTERP_RESULT
.CE
and/or
.CS
#define USE_INTERP_ERRORLINE
.CE
depending on which fields of the \fBTcl_Interp\fR struct are accessed.
These directives may be embedded in code or supplied via compiler options.
.PP
The \fIresult\fR and \fIfreeProc\fR fields are used to return
results or error messages from commands.
This information is returned by command procedures back to \fBTcl_Eval\fR,
and by \fBTcl_Eval\fR back to its callers.
The \fIresult\fR field points to the string that represents the
result or error message, and the \fIfreeProc\fR field tells how
to dispose of the storage for the string when it is not needed anymore.

As part of processing each command, \fBTcl_Eval\fR initializes
\fIinterp->result\fR
and \fIinterp->freeProc\fR just before calling the command procedure for
the command.  The \fIfreeProc\fR field will be initialized to zero,
and \fIinterp->result\fR will point to an empty string.  Commands that
do not return any value can simply leave the fields alone.
Furthermore, the empty string pointed to by \fIresult\fR is actually
part of an array of \fBTCL_RESULT_SIZE\fR characters (approximately 200).
part of an array of approximately 200 characters.
If a command wishes to return a short string, it can simply copy
it to the area pointed to by \fIinterp->result\fR.  Or, it can use
the sprintf procedure to generate a short result string at the location
pointed to by \fIinterp->result\fR.
.PP
It is a general convention in Tcl-based applications that the result
of an interpreter is normally in the initialized state described

'\"
'\" Copyright (c) 2007 Donal K. Fellows
'\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH Tcl_Method 3 0.1 TclOO "TclOO Library Functions"
.so man.macros
.BS
'\" Note:  do not modify the .SH NAME line immediately below!
.SH NAME
Tcl_ClassSetConstructor, Tcl_ClassSetDestructor, Tcl_MethodDeclarerClass, Tcl_MethodDeclarerObject, Tcl_MethodIsPublic, Tcl_MethodIsType, Tcl_MethodName, Tcl_NewInstanceMethod, Tcl_NewMethod, Tcl_ObjectContextInvokeNext, Tcl_ObjectContextIsFiltering, Tcl_ObjectContextMethod, Tcl_ObjectContextObject, Tcl_ObjectContextSkippedArgs \- manipulate methods and method-call contexts
Tcl_ClassSetConstructor, Tcl_ClassSetDestructor, Tcl_MethodDeclarerClass, Tcl_MethodDeclarerObject, Tcl_MethodIsPublic, Tcl_MethodIsPrivate, Tcl_MethodIsType, Tcl_MethodName, Tcl_NewInstanceMethod, Tcl_NewMethod, Tcl_ObjectContextInvokeNext, Tcl_ObjectContextIsFiltering, Tcl_ObjectContextMethod, Tcl_ObjectContextObject, Tcl_ObjectContextSkippedArgs \- manipulate methods and method-call contexts
.SH SYNOPSIS
.nf
\fB#include <tclOO.h>\fR
.sp
Tcl_Method
\fBTcl_NewMethod\fR(\fIinterp, class, nameObj, isPublic,
              methodTypePtr, clientData\fR)
\fBTcl_NewMethod\fR(\fIinterp, class, nameObj, flags, methodTypePtr,
              clientData\fR)
.sp
Tcl_Method
\fBTcl_NewInstanceMethod\fR(\fIinterp, object, nameObj, isPublic,
                      methodTypePtr, clientData\fR)
\fBTcl_NewInstanceMethod\fR(\fIinterp, object, nameObj, flags, methodTypePtr,
                      clientData\fR)
.sp
\fBTcl_ClassSetConstructor\fR(\fIinterp, class, method\fR)
.sp
\fBTcl_ClassSetDestructor\fR(\fIinterp, class, method\fR)
.sp
Tcl_Class
\fBTcl_MethodDeclarerClass\fR(\fImethod\fR)
.sp
Tcl_Object
\fBTcl_MethodDeclarerObject\fR(\fImethod\fR)
.sp
Tcl_Obj *
\fBTcl_MethodName\fR(\fImethod\fR)
.sp
.VS TIP500
int
\fBTcl_MethodIsPublic\fR(\fImethod\fR)
.VE TIP500
.sp
int
\fBTcl_MethodIsPrivate\fR(\fImethod\fR)
.sp
int
\fBTcl_MethodIsType\fR(\fImethod, methodTypePtr, clientDataPtr\fR)
.sp
int
\fBTcl_ObjectContextInvokeNext\fR(\fIinterp, context, objc, objv, skip\fR)
.sp

.AP Tcl_Object object in
The object to create the method in.
.AP Tcl_Class class in
The class to create the method in.
.AP Tcl_Obj *nameObj in
The name of the method to create. Should not be NULL unless creating
constructors or destructors.
.AP int isPublic in
A flag saying what the visibility of the method is. The only supported public
values of this flag are 0 for a non-exported method, and 1 for an exported
method.
.AP int flags in
A flag saying (currently) what the visibility of the method is. The supported
public values of this flag are \fBTCL_OO_METHOD_PUBLIC\fR (which is fixed at 1
for backward compatibility) for an exported method,
\fBTCL_OO_METHOD_UNEXPORTED\fR (which is fixed at 0 for backward
compatibility) for a non-exported method,
.VS TIP500
and \fBTCL_OO_METHOD_PRIVATE\fR for a private method.
.VE TIP500
.AP Tcl_MethodType *methodTypePtr in
A description of the type of the method to create, or the type of method to
compare against.
.AP ClientData clientData in
A piece of data that is passed to the implementation of the method without
interpretation.
.AP ClientData *clientDataPtr out

that class.
.PP
Given a method, the entity that declared it can be found using
\fBTcl_MethodDeclarerClass\fR which returns the class that the method is
attached to (or NULL if the method is not attached to any class) and
\fBTcl_MethodDeclarerObject\fR which returns the object that the method is
attached to (or NULL if the method is not attached to an object). The name of
the method can be retrieved with \fBTcl_MethodName\fR and whether the method
is exported is retrieved with \fBTcl_MethodIsPublic\fR. The type of the method
the method can be retrieved with \fBTcl_MethodName\fR, whether the method
is exported is retrieved with \fBTcl_MethodIsPublic\fR,
.VS TIP500
and whether the method is private is retrieved with \fBTcl_MethodIsPrivate\fR.
.VE TIP500
The type of the method
can also be introspected upon to a limited degree; the function
\fBTcl_MethodIsType\fR returns whether a method is of a particular type,
assigning the per-method \fIclientData\fR to the variable pointed to by
\fIclientDataPtr\fR if (that is non-NULL) if the type is matched.
.SS "METHOD CREATION"
.PP
Methods are created by \fBTcl_NewMethod\fR and \fBTcl_NewInstanceMethod\fR,
which
create a method attached to a class or an object respectively. In both cases,
the \fInameObj\fR argument gives the name of the method to create, the
\fIisPublic\fR argument states whether the method should be exported
initially, the \fImethodTypePtr\fR argument describes the implementation of
\fIflags\fR argument states whether the method should be exported
initially
.VS TIP500
or be marked as a private method,
.VE TIP500
the \fImethodTypePtr\fR argument describes the implementation of
the method (see the \fBMETHOD TYPES\fR section below) and the \fIclientData\fR
argument gives some implementation-specific data that is passed on to the
implementation of the method when it is called.
.PP
When the \fInameObj\fR argument to \fBTcl_NewMethod\fR is NULL, an
unnamed method is created, which is used for constructors and destructors.
Constructors should be installed into their class using the

.\"
.\" Copyright (c) 2008 by Kevin B. Kenny.
.\" Copyright (c) 2018 by Nathan Coulter.
.\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH NRE 3 8.6 Tcl "Tcl Library Procedures"
.so man.macros
.BS

.sp
void
\fBTcl_NRAddCallback\fR(\fIinterp, postProcPtr, data0, data1, data2, data3\fR)
.fi
.SH ARGUMENTS
.AS Tcl_CmdDeleteProc *interp in
.AP Tcl_Interp *interp in
Interpreter in which to create or evaluate a command.
The relevant Interpreter.
.AP char *cmdName in
Name of a new command to create.
Name of the command to create.
.AP Tcl_ObjCmdProc *proc in
Implementation of a command that will be called whenever \fIcmdName\fR
is invoked as a command in the unoptimized way.
Called in order to evaluate a command.  Is often just a small wrapper that uses
\fBTcl_NRCallObjProc\fR to call \fInreProc\fR using a new trampoline.  Behaves
in the same way as the \fIproc\fR argument to \fBTcl_CreateObjCommand\fR(3)
(\fIq.v.\fR).
.AP Tcl_ObjCmdProc *nreProc in
Implementation of a command that will be called whenever \fIcmdName\fR
is invoked and requested to conserve the C stack.
Called instead of \fIproc\fR when a trampoline is already in use.
.AP ClientData clientData in
Arbitrary one-word value that will be passed to \fIproc\fR, \fInreProc\fR,
\fIdeleteProc\fR and \fIobjProc\fR.
Arbitrary one-word value passed to \fIproc\fR, \fInreProc\fR, \fIdeleteProc\fR
and \fIobjProc\fR.
.AP Tcl_CmdDeleteProc *deleteProc in/out
Procedure to call before \fIcmdName\fR is deleted from the interpreter.
This procedure allows for command-specific cleanup. If \fIdeleteProc\fR
Called before \fIcmdName\fR is deleted from the interpreter, allowing for
command-specific cleanup. May be NULL.
is \fBNULL\fR, then no procedure is called before the command is deleted.
.AP int objc in
Count of parameters provided to the implementation of a command.
Number of items in \fIobjv\fR.
.AP Tcl_Obj **objv in
Pointer to an array of Tcl values. Each value holds the value of a
single word in the command to execute.
Words in the command.
.AP Tcl_Obj *objPtr in
Pointer to a Tcl_Obj whose value is a script or expression to execute.
A script or expression to evaluate.
.AP int flags in
As described for \fITcl_EvalObjv\fR.
ORed combination of flag bits that specify additional options.
.PP
\fBTCL_EVAL_GLOBAL\fR is the only flag that is currently supported.
.\" TODO: This is a lie. But kbk didn't grasp TCL_EVAL_INVOKE and
.\"       TCL_EVAL_NOERR well enough to document them.
.AP Tcl_Command cmd in
Token for a command that is to be used instead of the currently
executing command.
Token to use instead of one derived from the first word of \fIobjv\fR in order
to evaluate a command.
.AP Tcl_Obj *resultPtr out
Pointer to an unshared Tcl_Obj where the result of expression
evaluation is written.
Pointer to an unshared Tcl_Obj where the result of the evaluation is stored if
the return code is TCL_OK.
.AP Tcl_NRPostProc *postProcPtr in
Pointer to a function that will be invoked when the command currently
A function to push.
executing in the interpreter designated by \fIinterp\fR completes.
.AP ClientData data0 in
.AP ClientData data1 in
.AP ClientData data2 in
.AP ClientData data3 in
\fIdata0\fR through \fIdata3\fR are four one-word values that will be passed
to the function designated by \fIpostProcPtr\fR when it is invoked.
.BE
.SH DESCRIPTION
.PP
This series of C functions provides an interface whereby commands that
are implemented in C can be evaluated, and invoke Tcl commands scripts
and scripts, without consuming space on the C stack. The non-recursive
evaluation is done by installing a \fItrampoline\fR, a small piece of
These functions provide an interface to the function stack that an interpreter
iterates through to evaluate commands.  The routine behind a command is
implemented by an initial function and any additional functions that the
routine pushes onto the stack as it progresses.  The interpreter itself pushes
functions onto the stack to react to the end of a routine and to exercise other
forms of control such as switching between in-progress stacks and the
evaluation of other scripts at additional levels without adding frames to the C
stack.  To execute a routine, the initial function for the routine is called
and then a small bit of code called a \fItrampoline\fR iteratively takes
code that invokes a command or script, and then executes a series of
callbacks when the command or script returns.
functions off the stack and calls them, using the value of the last call as the
value of the routine.
.PP
The \fBTcl_NRCreateCommand\fR function creates a Tcl command in the
interpreter designated by \fIinterp\fR that is prepared to handle
nonrecursive evaluation with a trampoline. The \fIcmdName\fR argument
gives the name of the new command. If \fIcmdName\fR contains any
namespace qualifiers, then the new command is added to the specified
namespace; otherwise, it is added to the global namespace. \fIproc\fR
gives the procedure that will be called when the interpreter wishes to
evaluate the command in an unoptimized manner, and \fInreProc\fR is
\fBTcl_NRCallObjProc\fR calls \fInreProc\fR using a new trampoline.
the procedure that will be called when the interpreter wishes to
evaluate the command using a trampoline. \fIdeleteProc\fR is a
function that will be called before the command is deleted from the
interpreter. When any of the three functions is invoked, it is passed
the \fIclientData\fR parameter.
.PP
\fBTcl_NRCreateCommand\fR deletes any existing command
\fBTcl_NRCreateCommand\fR, an alternative to \fBTcl_CreateObjCommand\fR,
\fIname\fR already associated with the interpreter
(however see below for an exception where the existing command
is not deleted).
It returns a token that may be used to refer
to the command in subsequent calls to \fBTcl_GetCommandName\fR.
resolves \fIcmdName\fR, which may contain namespace qualifiers, relative to the
current namespace, creates a command by that name, and returns a token for the
command which may be used in subsequent calls to \fBTcl_GetCommandName\fR.
If \fBTcl_NRCreateCommand\fR is called for an interpreter that is in
the process of being deleted, then it does not create a new command,
does not delete any existing command of the same name, and returns NULL.
Except for a few cases noted below any existing command by the same name is
.PP
The \fIproc\fR and \fInreProc\fR function are expected to conform to
all the rules set forth for the \fIproc\fR argument to
\fBTcl_CreateObjCommand\fR(3) (\fIq.v.\fR).
.PP
When a command that is written to cope with evaluation via trampoline
is invoked without a trampoline on the stack, it will usually respond
to the invocation by creating a trampoline and calling the
trampoline-enabled implementation of the same command. This call is done by
means of \fBTcl_NRCallObjProc\fR. In the call to
\fBTcl_NRCallObjProc\fR, the \fIinterp\fR, \fIclientData\fR,
\fIobjc\fR and \fIobjv\fR parameters should be the same ones that were
passed to \fIproc\fR. The \fInreProc\fR parameter should designate the
trampoline-enabled implementation of the command.
.PP
\fBTcl_NREvalObj\fR arranges for the script contained in \fIobjPtr\fR
to be evaluated in the interpreter designated by \fIinterp\fR after
the current command (which must be trampoline-enabled) returns. It is
the method by which a command may invoke a script without consuming
space on the C stack. Similarly, \fBTcl_NREvalObjv\fR arranges to
first deleted.  If \fIinterp\fR is in the process of being deleted
\fBTcl_NRCreateCommand\fR does not create any command, does not delete any
command, and returns NULL.
.PP
\fBTcl_NREvalObj\fR pushes a function that is like \fBTcl_EvalObjEx\fR but
consumes no space on the C stack.
invoke a single Tcl command whose words have already been separated
and substituted. The \fIobjc\fR and \fIobjv\fR parameters give the
words of the command to be evaluated when execution reaches the
trampoline.
.PP
\fBTcl_NRCmdSwap\fR allows for trampoline evaluation of a command whose
resolution is already known.  The \fIcmd\fR parameter gives a
\fBTcl_Command\fR token (returned from \fBTcl_CreateObjCommand\fR or
\fBTcl_GetCommandFromObj\fR) identifying the command to be invoked in
the trampoline; this command must match the word in \fIobjv[0]\fR.
The remaining arguments are as for \fBTcl_NREvalObjv\fR.
\fBTcl_NREvalObjv\fR pushes a function that is like \fBTcl_EvalObjv\fR but
consumes no space on the C stack.
.PP
\fBTcl_NREvalObj\fR, \fBTcl_NREvalObjv\fR and \fBTcl_NRCmdSwap\fR
\fBTcl_NRCmdSwap\fR is like \fBTcl_NREvalObjv\fR, but uses \fIcmd\fR, a token
all accept a \fIflags\fR parameter, which is an OR-ed-together set of
bits to control evaluation. At the present time, the only supported flag
available to callers is \fBTCL_EVAL_GLOBAL\fR.
previously returned by \fBTcl_CreateObjCommand\fR or
.\" TODO: Again, this is a lie. Do we want to explain TCL_EVAL_INVOKE
.\"       and TCL_EVAL_NOERR?
If the \fBTCL_EVAL_GLOBAL\fR flag is set, the script or command is
\fBTcl_GetCommandFromObj\fR, instead of resolving the first word of \fIobjv\fR.
.  The name of this command must be the same as \fIobjv[0]\fR.
evaluated in the global namespace. If it is not set, it is evaluated
in the current namespace.
.PP
\fBTcl_NRExprObj\fR arranges for the expression contained in \fIobjPtr\fR
\fBTcl_NRExprObj\fR pushes a function that evaluates \fIobjPtr\fR as an
to be evaluated in the interpreter designated by \fIinterp\fR after
the current command (which must be trampoline-enabled) returns. It is
the method by which a command may evaluate a Tcl expression without consuming
space on the C stack.  The argument \fIresultPtr\fR is a pointer to an
expression in the same manner as \fBTcl_ExprObj\fR but without consuming space
on the C stack.
unshared Tcl_Obj where the result of expression evaluation is to be written.
If expression evaluation returns any code other than TCL_OK, the
\fIresultPtr\fR value is left untouched.
.PP
All of the routines return \fBTCL_OK\fR if command or expression invocation
has been scheduled successfully. If for any reason the scheduling cannot
All of the functions return \fBTCL_OK\fR if the evaluation of the script,
command, or expression has been scheduled successfully.  Otherwise (for example
be completed (for example, if the interpreter is unable to find
the requested command), they return \fBTCL_ERROR\fR with an
appropriate message left in the interpreter's result.
if the command name cannot be resolved), they return \fBTCL_ERROR\fR and store
a message as the interpreter's result.
.PP
\fBTcl_NRAddCallback\fR arranges to have a C function called when the
\fBTcl_NRAddCallback\fR pushes \fIpostProcPtr\fR.  The signature for
current trampoline-enabled command in the Tcl interpreter designated
by \fIinterp\fR returns.  The \fIpostProcPtr\fR argument is a pointer
to the callback function, which must have arguments and return value
consistent with the \fBTcl_NRPostProc\fR data type:
\fBTcl_NRPostProc\fR is:
.PP
.CS
typedef int
\fBTcl_NRPostProc\fR(
        \fBClientData\fR \fIdata\fR[],
        \fBTcl_Interp\fR *\fIinterp\fR,
        int \fIresult\fR);
.CE
.PP
When the trampoline invokes the callback function, the \fIdata\fR
parameter will point to an array containing the four one-word
quantities that were passed to \fBTcl_NRAddCallback\fR in the
\fIdata0\fR through \fIdata3\fR parameters. The Tcl interpreter will
\fIdata\fR is a pointer to an array containing \fIdata0\fR through \fIdata3\fR.
be designated by the \fIinterp\fR parameter, and the \fIresult\fR
parameter will contain the result (\fBTCL_OK\fR, \fBTCL_ERROR\fR,
\fBTCL_RETURN\fR, \fBTCL_BREAK\fR or \fBTCL_CONTINUE\fR) that was
returned by the command evaluation. The callback function is expected,
\fIresult\fR is the value returned by the previous function implementing part
in turn, either to return a \fIresult\fR to control further evaluation.
.PP
Multiple \fBTcl_NRAddCallback\fR invocations may request multiple
callbacks, which may be to the same or different callback
the routine.
functions. If multiple callbacks are requested, they are executed in
last-in, first-out order, that is, the most recently requested
callback is executed first.
.SH EXAMPLE
.PP
The usual pattern for Tcl commands that invoke other Tcl commands
The following command uses \fBTcl_EvalObjEx\fR, which consumes space on the C
is something like:
stack, to evalute a script:
.PP
.CS
int
\fITheCmdOldObjProc\fR(
    ClientData clientData,
    Tcl_Interp *interp,
    int objc,

    return result;
}
\fBTcl_CreateObjCommand\fR(interp, "theCommand",
        \fITheCmdOldObjProc\fR, clientData, TheCmdDeleteProc);
.CE
.PP
To enable a command like this one for trampoline-based evaluation,
it must be split into three pieces:
To avoid consuming space on the C stack, \fITheCmdOldObjProc\fR is renamed to
\fITheCmdNRObjProc\fR and the postprocessing step is split into a separate
.IP \(bu
A non-trampoline implementation, \fITheCmdNewObjProc\fR,
which will simply create a trampoline
and invoke the trampoline-based implementation.
function, \fITheCmdPostProc\fR, which is pushed onto the function stack.
\fITcl_EvalObjEx\fR is replaced with \fITcl_NREvalObj\fR, which uses a
trampoline instead of consuming space on the C stack.  A new version of
.IP \(bu
A trampoline-enabled implementation, \fITheCmdNRObjProc\fR.  This
\fITheCmdOldObjProc\fR is just a a wrapper that uses \fBTcl_NRCallObjProc\fR to
call \fITheCmdNRObjProc\fR:
function will perform the initialization, request that the trampoline
call the postprocessing routine after command evaluation, and finally,
request that the trampoline call the inner command.
.IP \(bu
A postprocessing routine, \fITheCmdPostProc\fR. This function will
perform the postprocessing formerly done after the return from the
inner command in \fITheCmdObjProc\fR.
.PP
The non-trampoline implementation is simple and stylized, containing
a single statement:
.PP
.CS
int
\fITheCmdNewObjProc\fR(
\fITheCmdOldObjProc\fR(
    ClientData clientData,
    Tcl_Interp *interp,
    int objc,
    Tcl_Obj *const objv[])
{
    return \fBTcl_NRCallObjProc\fR(interp, \fITheCmdNRObjProc\fR,
            clientData, objc, objv);
}
.CE
.PP
The trampoline-enabled implementation requests postprocessing,
and returns to the trampoline requesting command evaluation.
.PP
.CS
int
\fITheCmdNRObjProc\fR
    ClientData clientData,
    Tcl_Interp *interp,
    int objc,
    Tcl_Obj *const objv[])

    /* \fIdata0 .. data3\fR are up to four one-word items to
     * pass to the postprocessing procedure */

    return \fBTcl_NREvalObj\fR(interp, objPtr, 0);
}
.CE
.PP
The postprocessing procedure does whatever the original command did
upon return from the inner evaluation.
.PP
.CS
int
\fITheCmdNRPostProc\fR(
    ClientData data[],
    Tcl_Interp *interp,
    int result)
{
    /* \fIdata[0] .. data[3]\fR are the four words of data
     * passed to \fBTcl_NRAddCallback\fR */

    \fI... postprocessing ...\fR

    return result;
}
.CE
.PP
If \fItheCommand\fR is a command that results in multiple commands or
Any function comprising a routine can push other functions, making it possible
scripts being evaluated, its postprocessing routine may schedule
additional postprocessing and then request another command evaluation
by means of \fBTcl_NREvalObj\fR or one of the other evaluation
routines. Looping and sequencing constructs may be implemented in this way.
implement looping and sequencing constructs using the function stack.
.PP
Finally, to install a trampoline-enabled command in the interpreter,
\fBTcl_NRCreateCommand\fR is used in place of
\fBTcl_CreateObjCommand\fR.  It accepts two command procedures instead
of one. The first is for use when no trampoline is yet on the stack,
and the second is for use when there is already a trampoline in place.
.PP
.CS
\fBTcl_NRCreateCommand\fR(interp, "theCommand",
        \fITheCmdNewObjProc\fR, \fITheCmdNRObjProc\fR, clientData,
        TheCmdDeleteProc);
.CE
.SH "SEE ALSO"
Tcl_CreateCommand(3), Tcl_CreateObjCommand(3), Tcl_EvalObjEx(3), Tcl_GetCommandFromObj(3), Tcl_ExprObj(3)
.SH KEYWORDS
stackless, nonrecursive, execute, command, global, value, result, script
.SH COPYRIGHT
Copyright (c) 2008 by Kevin B. Kenny
Copyright (c) 2008 by Kevin B. Kenny.
Copyright (c) 2018 by Nathan Coulter.

\fBincr x\fR
.CE
.PP
The \fBincr\fR command first gets an integer from \fIx\fR's value
by calling \fBTcl_GetIntFromObj\fR.
This procedure checks whether the value is already an integer value.
Since it is not, it converts the value
by setting the value's \fIinternalRep.longValue\fR member
by setting the value's internal representation
to the integer \fB123\fR
and setting the value's \fItypePtr\fR
to point to the integer Tcl_ObjType structure.
Both representations are now valid.
\fBincr\fR increments the value's integer internal representation
then invalidates its string representation
(by calling \fBTcl_InvalidateStringRep\fR)

'\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH Tcl_Panic 3 8.4 Tcl "Tcl Library Procedures"
.so man.macros
.BS
'\"  Note:  do not modify the .SH NAME line immediately below!
.SH NAME
Tcl_Panic, Tcl_PanicVA, Tcl_SetPanicProc \- report fatal error and abort
Tcl_Panic, Tcl_PanicVA, Tcl_SetPanicProc, Tcl_ConsolePanic \- report fatal error and abort
.SH SYNOPSIS
.nf
\fB#include <tcl.h>\fR
.sp
void
\fBTcl_Panic\fR(\fIformat\fR, \fIarg\fR, \fIarg\fR, \fI...\fR)
.sp
void
\fBTcl_PanicVA\fR(\fIformat\fR, \fIargList\fR)
.sp
void
\fBTcl_SetPanicProc\fR(\fIpanicProc\fR)
.sp
void
\fBTcl_ConsolePanic\fR(\fIformat\fR, \fIarg\fR, \fIarg\fR, \fI...\fR)
.sp
.SH ARGUMENTS
.AS Tcl_PanicProc *panicProc
.AP "const char*" format in
A printf-style format string.
.AP "" arg in
Arguments matching the format string.
.AP va_list argList in

In a freshly loaded Tcl library, \fBTcl_Panic\fR prints the formatted
error message to the standard error file of the process, and then
calls \fBabort\fR to terminate the process.  \fBTcl_Panic\fR does not
return. On Windows, when a debugger is running, the formatted error
message is sent to the debugger in stead. If the windows executable
does not have a stderr channel (e.g. \fBwish.exe\fR), then a
system dialog box is used to display the panic message.
.PP
If your application doesn't use \fBTcl_Main\fR or \fBTk_Main\fR
and you want to implicitly use the stderr channel of your
application's C runtime (in stead of the stderr channel of the
C runtime used by Tcl), you can call \fBTcl_SetPanicProc\fR
with \fBTcl_ConsolePanic\fR as its argument. On platforms which
only have one C runtime (almost all platforms except Windows)
\fBTcl_ConsolePanic\fR is equivalent to NULL.
.PP
\fBTcl_SetPanicProc\fR may be used to modify the behavior of
\fBTcl_Panic\fR.  The \fIpanicProc\fR argument should match the
type \fBTcl_PanicProc\fR:
.PP
.CS
typedef void \fBTcl_PanicProc\fR(

'\"
'\" Copyright (c) 1997 by Sun Microsystems, Inc.
'\" Contributions from Don Porter, NIST, 2004. (not subject to US copyright)
'\" Copyright (c) 2018 Nathan Coulter.
'\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH Tcl_SaveResult 3 8.1 Tcl "Tcl Library Procedures"
.so man.macros
.BS
.SH NAME
Tcl_SaveInterpState, Tcl_RestoreInterpState, Tcl_DiscardInterpState, Tcl_SaveResult, Tcl_RestoreResult, Tcl_DiscardResult \- save and restore an interpreter's state
Tcl_SaveInterpState, Tcl_RestoreInterpState, Tcl_DiscardInterpState,
Tcl_SaveResult, Tcl_RestoreResult, Tcl_DiscardResult \- Save and restore the
state of an an interpreter.
.SH SYNOPSIS
.nf
\fB#include <tcl.h>\fR
.sp
Tcl_InterpState
\fBTcl_SaveInterpState\fR(\fIinterp, status\fR)
.sp
int
\fBTcl_RestoreInterpState\fR(\fIinterp, state\fR)
.sp
\fBTcl_DiscardInterpState\fR(\fIstate\fR)
.sp
\fBTcl_SaveResult\fR(\fIinterp, savedPtr\fR)
.sp
\fBTcl_RestoreResult\fR(\fIinterp, savedPtr\fR)
.sp
\fBTcl_DiscardResult\fR(\fIsavedPtr\fR)
.SH ARGUMENTS
.AS Tcl_InterpState savedPtr
.AP Tcl_Interp *interp in
Interpreter for which state should be saved.
The interpreter for the operation.
.AP int status in
Return code value to save as part of interpreter state.
The return code for the state.
.AP Tcl_InterpState state in
Saved state token to be restored or discarded.
A token for saved state.
.AP Tcl_SavedResult *savedPtr in
Pointer to location where interpreter result should be saved or restored.
A pointer to storage for saved state.
.BE
.SH DESCRIPTION
.PP
These routines allows a C procedure to take a snapshot of the current
state of an interpreter so that it can be restored after a call
to \fBTcl_Eval\fR or some other routine that modifies the interpreter
These routines save the state of an interpreter before a call to a routine such
as \fBTcl_Eval\fR, and restore the state afterwards.
state.  There are two triplets of routines meant to work together.
.PP
The first triplet stores the snapshot of interpreter state in
an opaque token returned by \fBTcl_SaveInterpState\fR.  That token
\fBTcl_SaveInterpState\fR saves the parts of \fIinterp\fR that comprise the
value may then be passed back to one of \fBTcl_RestoreInterpState\fR
or \fBTcl_DiscardInterpState\fR, depending on whether the interp
state is to be restored.  So long as one of the latter two routines
result of a script, including the resulting value, the return code passed as
\fIstatus\fR, and any options such as \fB\-errorinfo\fR and \fB\-errorcode\fR.
It returns a token for the saved state.  The interpreter result is not reset
and no interpreter state is changed.
is called, Tcl will take care of memory management.
.PP
The second triplet stores the snapshot of only the interpreter
\fBTcl_RestoreInterpState\fR restores the state indicated by \fIstate\fR and
result (not its complete state) in memory allocated by the caller.
These routines are passed a pointer to \fBTcl_SavedResult\fR
returns the \fIstatus\fR originally passed in the corresponding call to
that is used to store enough information to restore the interpreter result.
\fBTcl_SavedResult\fR can be allocated on the stack of the calling
\fBTcl_SaveInterpState\fR.
procedure.  These routines do not save the state of any error
information in the interpreter (e.g. the \fB\-errorcode\fR or
\fB\-errorinfo\fR return options, when an error is in progress).
.PP
Because the routines \fBTcl_SaveInterpState\fR,
\fBTcl_RestoreInterpState\fR, and \fBTcl_DiscardInterpState\fR perform
If a saved state is not restored, \fBTcl_DiscardInterpState\fR must be called
a superset of the functions provided by the other routines,
any new code should only make use of the more powerful routines.
The older, weaker routines \fBTcl_SaveResult\fR, \fBTcl_RestoreResult\fR,
and \fBTcl_DiscardResult\fR continue to exist only for the sake
of existing programs that may already be using them.
to release it.  A token used to discard or restore state must not be used
again.
.PP
\fBTcl_SaveInterpState\fR takes a snapshot of those portions of
\fBTcl_SaveResult\fR, \fBTcl_RestoreResult\fR, and \fBTcl_DiscardResult\fR are
interpreter state that make up the full result of script evaluation.
This include the interpreter result, the return code (passed in
as the \fIstatus\fR argument, and any return options, including
\fB\-errorinfo\fR and \fB\-errorcode\fR when an error is in progress.
This snapshot is returned as an opaque token of type \fBTcl_InterpState\fR.
The call to \fBTcl_SaveInterpState\fR does not itself change the
state of the interpreter.  Unlike \fBTcl_SaveResult\fR, it does
deprecated.  Instead use \fBTcl_SaveInterpState\fR,
\fBTcl_RestoreInterpState\fR, and \fBTcl_DiscardInterpState\fR, which are more
not reset the interpreter.
capable.
.PP
\fBTcl_RestoreInterpState\fR accepts a \fBTcl_InterpState\fR token
previously returned by \fBTcl_SaveInterpState\fR and restores the
state of the interp to the state held in that snapshot.  The return
value of \fBTcl_RestoreInterpState\fR is the status value originally
passed to \fBTcl_SaveInterpState\fR when the snapshot token was
created.
.PP
\fBTcl_DiscardInterpState\fR is called to release a \fBTcl_InterpState\fR
token previously returned by \fBTcl_SaveInterpState\fR when that
snapshot is not to be restored to an interp.
.PP
The \fBTcl_InterpState\fR token returned by \fBTcl_SaveInterpState\fR
must eventually be passed to either \fBTcl_RestoreInterpState\fR
or \fBTcl_DiscardInterpState\fR to avoid a memory leak.  Once
the \fBTcl_InterpState\fR token is passed to one of them, the
token is no longer valid and should not be used anymore.
.PP
\fBTcl_SaveResult\fR moves the string and value results
of \fIinterp\fR into the location specified by \fIstatePtr\fR.
\fBTcl_SaveResult\fR moves the result of \fIinterp\fR to the location
\fIstatePtr\fR points to and returns the interpreter result to its initial
\fBTcl_SaveResult\fR clears the result for \fIinterp\fR and
leaves the result in its normal empty initialized state.
state.  It does not save options such as \fB\-errorcode\fR or
\fB\-errorinfo\fR.
.PP
\fBTcl_RestoreResult\fR moves the string and value results from
\fIstatePtr\fR back into \fIinterp\fR.  Any result or error that was
\fBTcl_RestoreResult\fR clears any existing result or error in \fIinterp\fR and
moves the result from \fIstatePtr\fR back to \fIinterp\fR.  \fIstatePtr\fR is
already in the interpreter will be cleared.  The \fIstatePtr\fR is left
in an uninitialized state and cannot be used until another call to
then in an undefined state and must not be used until passed again to
\fBTcl_SaveResult\fR.
.PP
\fBTcl_DiscardResult\fR releases the saved interpreter state
\fBTcl_DiscardResult\fR releases the state stored at \fBstatePtr\fR, which is
stored at \fBstatePtr\fR.  The state structure is left in an
uninitialized state and cannot be used until another call to
then in an undefined state and must not be used until passed again to
\fBTcl_SaveResult\fR.
.PP
Once \fBTcl_SaveResult\fR is called to save the interpreter
result, either \fBTcl_RestoreResult\fR or
\fBTcl_DiscardResult\fR must be called to properly clean up the
If a saved result is not restored, \fBTcl_DiscardResult\fR must be called to
memory associated with the saved state.
release it.
.SH KEYWORDS
result, state, interp

\fBTcl_FreeResult\fR performs part of the work
of \fBTcl_ResetResult\fR.
It frees up the memory associated with \fIinterp\fR's result.
It also sets \fIinterp->freeProc\fR to zero, but does not
change \fIinterp->result\fR or clear error state.
\fBTcl_FreeResult\fR is most commonly used when a procedure
is about to replace one result value with another.
.SS "DIRECT ACCESS TO INTERP->RESULT"
.PP
It used to be legal for programs to
directly read and write \fIinterp->result\fR
to manipulate the interpreter result.  The Tcl headers no longer
permit this access by default, and C code still doing this must
be updated to use supported routines \fBTcl_GetObjResult\fR,
\fBTcl_GetStringResult\fR, \fBTcl_SetObjResult\fR, and \fBTcl_SetResult\fR.
As a migration aid, access can be restored with the compiler directive
.CS
#define USE_INTERP_RESULT
.CE
but this is meant only to offer life support to otherwise dead code.
.SH "THE TCL_FREEPROC ARGUMENT TO TCL_SETRESULT"
.PP
\fBTcl_SetResult\fR's \fIfreeProc\fR argument specifies how
the Tcl system is to manage the storage for the \fIresult\fR argument.
If \fBTcl_SetResult\fR or \fBTcl_SetObjResult\fR are called
at a time when \fIinterp\fR holds a string result,
they do whatever is necessary to dispose of the old string result

.sp
Tcl_UniChar *
\fBTcl_GetUnicodeFromObj\fR(\fIobjPtr, lengthPtr\fR)
.sp
Tcl_UniChar *
\fBTcl_GetUnicode\fR(\fIobjPtr\fR)
.sp
Tcl_UniChar
int
\fBTcl_GetUniChar\fR(\fIobjPtr, index\fR)
.sp
int
\fBTcl_GetCharLength\fR(\fIobjPtr\fR)
.sp
Tcl_Obj *
\fBTcl_GetRange\fR(\fIobjPtr, first, last\fR)

\fIlengthPtr\fR if it is non-NULL.  The storage referenced by the returned
byte pointer is owned by the value manager and should not be modified by
the caller.  The procedure \fBTcl_GetUnicode\fR is used in the common case
where the caller does not need the length of the unicode string
representation.
.PP
\fBTcl_GetUniChar\fR returns the \fIindex\fR'th character in the
value's Unicode representation.
value's Unicode representation. If the index is out of range or
it references a low surrogate preceded by a high surrogate, it returns -1;
.PP
\fBTcl_GetRange\fR returns a newly created value comprised of the
characters between \fIfirst\fR and \fIlast\fR (inclusive) in the
value's Unicode representation.  If the value's Unicode
representation is invalid, the Unicode representation is regenerated
from the value's string representation.
.PP

.BS
.SH NAME
Tcl_UniCharToUpper, Tcl_UniCharToLower, Tcl_UniCharToTitle, Tcl_UtfToUpper, Tcl_UtfToLower, Tcl_UtfToTitle \- routines for manipulating the case of Unicode characters and UTF-8 strings
.SH SYNOPSIS
.nf
\fB#include <tcl.h>\fR
.sp
Tcl_UniChar
int
\fBTcl_UniCharToUpper\fR(\fIch\fR)
.sp
Tcl_UniChar
int
\fBTcl_UniCharToLower\fR(\fIch\fR)
.sp
Tcl_UniChar
int
\fBTcl_UniCharToTitle\fR(\fIch\fR)
.sp
int
\fBTcl_UtfToUpper\fR(\fIstr\fR)
.sp
int
\fBTcl_UtfToLower\fR(\fIstr\fR)
.sp
int
\fBTcl_UtfToTitle\fR(\fIstr\fR)
.SH ARGUMENTS
.AS char *str in/out
.AP int ch in
The Tcl_UniChar to be converted.
The Unicode character to be converted.
.AP char *str in/out
Pointer to UTF-8 string to be converted in place.
.BE

.SH DESCRIPTION
.PP
The first three routines convert the case of individual Unicode characters:

\fBTcl_UniCharIsUpper\fR(\fIch\fR)
.sp
int
\fBTcl_UniCharIsWordChar\fR(\fIch\fR)
.SH ARGUMENTS
.AS int ch
.AP int ch in
The Tcl_UniChar to be examined.
The Unicode character to be examined.
.BE

.SH DESCRIPTION
.PP
All of the routines described examine Tcl_UniChars and return a
All of the routines described examine Unicode characters and return a
boolean value. A non-zero return value means that the character does
belong to the character class associated with the called routine. The
rest of this document just describes the character classes associated
with the various routines.
.PP
Note: A Tcl_UniChar is a Unicode character represented as an unsigned,
fixed-size quantity.

.SH "CHARACTER CLASSES"
.PP
\fBTcl_UniCharIsAlnum\fR tests if the character is an alphanumeric Unicode character.
.PP
\fBTcl_UniCharIsAlpha\fR tests if the character is an alphabetic Unicode character.
.PP

.sp
const char *
\fBTcl_UtfNext\fR(\fIsrc\fR)
.sp
const char *
\fBTcl_UtfPrev\fR(\fIsrc, start\fR)
.sp
Tcl_UniChar
int
\fBTcl_UniCharAtIndex\fR(\fIsrc, index\fR)
.sp
const char *
\fBTcl_UtfAtIndex\fR(\fIsrc, index\fR)
.sp
int
\fBTcl_UtfBackslash\fR(\fIsrc, readPtr, dst\fR)
.SH ARGUMENTS
.AS "const Tcl_UniChar" *uniPattern in/out
.AP char *buf out
Buffer in which the UTF-8 representation of the Tcl_UniChar is stored.  At most
\fBTCL_UTF_MAX\fR bytes are stored in the buffer.
.AP int ch in
The Tcl_UniChar to be converted or examined.
The Unicode character to be converted or examined.
.AP Tcl_UniChar *chPtr out
Filled with the Tcl_UniChar represented by the head of the UTF-8 string.
.AP "const char" *src in
Pointer to a UTF-8 string.
.AP "const char" *cs in
Pointer to a UTF-8 string.
.AP "const char" *ct in

.AP int nocase in
Specifies whether the match should be done case-sensitive (0) or
case-insensitive (1).
.BE

.SH DESCRIPTION
.PP
These routines convert between UTF-8 strings and Tcl_UniChars.  A
Tcl_UniChar is a Unicode character represented as an unsigned, fixed-size
These routines convert between UTF-8 strings and Unicode characters.  An
Unicode character represented as an unsigned, fixed-size
quantity.  A UTF-8 character is a Unicode character represented as
a varying-length sequence of up to \fBTCL_UTF_MAX\fR bytes.  A multibyte UTF-8
sequence consists of a lead byte followed by some number of trail bytes.
.PP
\fBTCL_UTF_MAX\fR is the maximum number of bytes that it takes to
represent one Unicode character in the UTF-8 representation.
.PP
\fBTcl_UniCharToUtf\fR stores the Tcl_UniChar \fIch\fR as a UTF-8 string
\fBTcl_UniCharToUtf\fR stores the character \fIch\fR as a UTF-8 string
in starting at \fIbuf\fR.  The return value is the number of bytes stored
in \fIbuf\fR.
in \fIbuf\fR. If ch is an upper surrogate (range U+D800 - U+DBFF), then
the return value will be 0 and nothing will be stored. If you still
want to produce UTF-8 output for it (even though knowing it's an illegal
code-point on its own), just call \fBTcl_UniCharToUtf\fR again using ch = -1.
.PP
\fBTcl_UtfToUniChar\fR reads one UTF-8 character starting at \fIsrc\fR
and stores it as a Tcl_UniChar in \fI*chPtr\fR.  The return value is the
number of bytes read from \fIsrc\fR.  The caller must ensure that the
source buffer is long enough such that this routine does not run off the
end and dereference non-existent or random memory; if the source buffer
is known to be null-terminated, this will not happen.  If the input is
a byte in the range 0x80 - 0x9F, \fBTcl_UtfToUniChar\fR assumes the
cp1252 encoding, stores the corresponding Tcl_UniChar in \fI*chPtr\fR
and returns 1. If the input is otherwise
not in proper UTF-8 format, \fBTcl_UtfToUniChar\fR will store the first
byte of \fIsrc\fR in \fI*chPtr\fR as a Tcl_UniChar between 0x0000 and
0x00ff and return 1.
.PP
\fBTcl_UniCharToUtfDString\fR converts the given Unicode string
to UTF-8, storing the result in a previously initialized \fBTcl_DString\fR.
You must specify \fIuniLength\fR, the length of the given Unicode string.

characters.
.PP
\fBTcl_UtfCharComplete\fR returns 1 if the source UTF-8 string \fIsrc\fR
of \fIlength\fR bytes is long enough to be decoded by
\fBTcl_UtfToUniChar\fR, or 0 otherwise.  This function does not guarantee
that the UTF-8 string is properly formed.  This routine is used by
procedures that are operating on a byte at a time and need to know if a
full Tcl_UniChar has been seen.
full Unicode character has been seen.
.PP
\fBTcl_NumUtfChars\fR corresponds to \fBstrlen\fR for UTF-8 strings.  It
returns the number of Tcl_UniChars that are represented by the UTF-8 string
\fIsrc\fR.  The length of the source string is \fIlength\fR bytes.  If the
length is negative, all bytes up to the first null byte are used.
.PP
\fBTcl_UtfFindFirst\fR corresponds to \fBstrchr\fR for UTF-8 strings.  It
returns a pointer to the first occurrence of the Tcl_UniChar \fIch\fR
returns a pointer to the first occurrence of the Unicode character \fIch\fR
in the null-terminated UTF-8 string \fIsrc\fR.  The null terminator is
considered part of the UTF-8 string.
.PP
\fBTcl_UtfFindLast\fR corresponds to \fBstrrchr\fR for UTF-8 strings.  It
returns a pointer to the last occurrence of the Tcl_UniChar \fIch\fR
returns a pointer to the last occurrence of the Unicode character \fIch\fR
in the null-terminated UTF-8 string \fIsrc\fR.  The null terminator is
considered part of the UTF-8 string.
.PP
Given \fIsrc\fR, a pointer to some location in a UTF-8 string,
\fBTcl_UtfNext\fR returns a pointer to the next UTF-8 character in the
string.  The caller must not ask for the next character after the last
character in the string if the string is not terminated by a null

'\"
'\" Copyright (c) 2018 Donal K. Fellows
'\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH abstract n 0.3 TclOO "TclOO Commands"
.so man.macros
.BS
'\" Note:  do not modify the .SH NAME line immediately below!
.SH NAME
oo::abstract \- a class that does not allow direct instances of itself
.SH SYNOPSIS
.nf
package require TclOO

\fBoo::abstract\fI method \fR?\fIarg ...\fR?
.fi
.SH "CLASS HIERARCHY"
.nf
\fBoo::object\fR
   \(-> \fBoo::class\fR
       \(-> \fBoo::abstract\fR
.fi
.BE
.SH DESCRIPTION
Abstract classes are classes that can contain definitions, but which cannot be
directly manufactured; they are intended to only ever be inherited from and
instantiated indirectly. The characteristic methods of \fBoo::class\fR
(\fBcreate\fR and \fBnew\fR) are not exported by an instance of
\fBoo::abstract\fR.
.PP
Note that \fBoo::abstract\fR is not itself an instance of \fBoo::abstract\fR.
.SS CONSTRUCTOR
The \fBoo::abstract\fR class does not define an explicit constructor; this
means that it is effectively the same as the constructor of the
\fBoo::class\fR class.
.SS DESTRUCTOR
The \fBoo::abstract\fR class does not define an explicit destructor;
destroying an instance of it is just like destroying an ordinary class (and
will destroy all its subclasses).
.SS "EXPORTED METHODS"
The \fBoo::abstract\fR class defines no new exported methods.
.SS "NON-EXPORTED METHODS"
The \fBoo::abstract\fR class explicitly states that \fBcreate\fR,
\fBcreateWithNamespace\fR, and \fBnew\fR are unexported.
.SH EXAMPLES
.PP
This example defines a simple class hierarchy and creates a new instance of
it. It then invokes a method of the object before destroying the hierarchy and
showing that the destruction is transitive.
.PP
.CS
\fBoo::abstract\fR create fruit {
    method eat {} {
        puts "yummy!"
    }
}
oo::class create banana {
    superclass fruit
    method peel {} {
        puts "skin now off"
    }
}
set b [banana \fBnew\fR]
$b peel              \fI\(-> prints 'skin now off'\fR
$b eat               \fI\(-> prints 'yummy!'\fR
set f [fruit new]    \fI\(-> error 'unknown method "new"...'\fR
.CE
.SH "SEE ALSO"
oo::define(n), oo::object(n)
.SH KEYWORDS
abstract class, class, metaclass, object
.\" Local variables:
.\" mode: nroff
.\" fill-column: 78
.\" End:

been the return value from a previous invocation of
\fBarray startsearch\fR.  Returns an empty string.
.TP
\fBarray exists \fIarrayName\fR
Returns 1 if \fIarrayName\fR is an array variable, 0 if there
is no variable by that name or if it is a scalar variable.
.TP
\fBarray for {\fIkeyVariable valueVariable\fB} \fIarrayName body\fP
The first argument is a two element list of variable names for the
key and value of each entry in the array.  The second argument is the
array name to iterate over.  The third argument is the body to execute
for each key and value returned.
The ordering of the returned keys is undefined.
If an array element is deleted or a new array element is inserted during
the \fIarray for\fP process, the command will terminate with an error.
.TP
\fBarray get \fIarrayName\fR ?\fIpattern\fR?
Returns a list containing pairs of elements.  The first
element in each pair is the name of an element in \fIarrayName\fR
and the second element of each pair is the value of the
array element.  The order of the pairs is undefined.
If \fIpattern\fR is not specified, then all of the elements of the
array are included in the result.

'\"
'\" Copyright (c) 2018 Donal K. Fellows
'\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH callback n 0.3 TclOO "TclOO Commands"
.so man.macros
.BS
'\" Note:  do not modify the .SH NAME line immediately below!
.SH NAME
callback, mymethod \- generate callbacks to methods
.SH SYNOPSIS
.nf
package require TclOO

\fBcallback\fR \fImethodName\fR ?\fIarg ...\fR?
\fBmymethod\fR \fImethodName\fR ?\fIarg ...\fR?
.fi
.BE
.SH DESCRIPTION
The \fBcallback\fR command,
'\" Based on notes in the tcllib docs, we know the provenance of mymethod
also called \fBmymethod\fR for compatibility with the ooutil and snit packages
of Tcllib,
and which should only be used from within the context of a call to a method
(i.e. inside a method, constructor or destructor body) is used to generate a
script fragment that will invoke the method, \fImethodName\fR, on the current
object (as reported by \fBself\fR) when executed. Any additional arguments
provided will be provided as leading arguments to the callback. The resulting
script fragment shall be a proper list.
.PP
Note that it is up to the caller to ensure that the current object is able to
handle the call of \fImethodName\fR; this command does not check that.
\fImethodName\fR may refer to any exported or unexported method, but may not
refer to a private method as those can only be invoked directly from within
methods. If there is no such method present at the point when the callback is
invoked, the standard \fBunknown\fR method handler will be called.
.SH EXAMPLE
This is a simple echo server class. The \fBcallback\fR command is used in two
places, to arrange for the incoming socket connections to be handled by the
\fIAccept\fR method, and to arrange for the incoming bytes on those
connections to be handled by the \fIReceive\fR method.
.PP
.CS
oo::class create EchoServer {
    variable server clients
    constructor {port} {
        set server [socket -server [\fBcallback\fR Accept] $port]
        set clients {}
    }
    destructor {
        chan close $server
        foreach client [dict keys $clients] {
            chan close $client
        }
    }

    method Accept {channel clientAddress clientPort} {
        dict set clients $channel [dict create \e
                address $clientAddress port $clientPort]
        chan event $channel readable [\fBcallback\fR Receive $channel]
    }
    method Receive {channel} {
        if {[chan gets $channel line] >= 0} {
            my echo $channel $line
        } else {
            chan close $channel
            dict unset clients $channel
        }
    }

    method echo {channel line} {
        dict with clients $channel {
            chan puts $channel \e
                    [format {[%s:%d] %s} $address $port $line]
        }
    }
}
.CE
.SH "SEE ALSO"
chan(n), fileevent(n), my(n), self(n), socket(n), trace(n)
.SH KEYWORDS
callback, object
.\" Local Variables:
.\" mode: nroff
.\" fill-column: 78
.\" End:

'\"
'\" Copyright (c) 2011-2015 Andreas Kupries
'\" Copyright (c) 2018 Donal K. Fellows
'\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH classvariable n 0.3 TclOO "TclOO Commands"
.so man.macros
.BS
'\" Note:  do not modify the .SH NAME line immediately below!
.SH NAME
classvariable \- create link from local variable to variable in class
.SH SYNOPSIS
.nf
package require TclOO

\fBclassvariable\fR \fIvariableName\fR ?\fI...\fR?
.fi
.BE
.SH DESCRIPTION
The \fBclassvariable\fR command is available within methods. It takes a series
of one or more variable names and makes them available in the method's scope;
those variable names must not be qualified and must not refer to array
elements. The originating scope for the variables is the namespace of the
class that the method was defined by. In other words, the referenced variables
are shared between all instances of that class.
.PP
Note: This command is equivalent to the command \fBtypevariable\fR provided by
the snit package in tcllib for approximately the same purpose. If used in a
method defined directly on a class instance (e.g., through the
\fBoo::objdefine\fR \fBmethod\fR definition) this is very much like just
using:
.PP
.CS
namespace upvar [namespace current] $var $var
.CE
.PP
for each variable listed to \fBclassvariable\fR.
.SH EXAMPLE
This class counts how many instances of it have been made.
.PP
.CS
oo::class create Counted {
    initialise {
        variable count 0
    }

    variable number
    constructor {} {
        \fBclassvariable\fR count
        set number [incr count]
    }

    method report {} {
        \fBclassvariable\fR count
        puts "This is instance $number of $count"
    }
}

set a [Counted new]
set b [Counted new]
$a report
        \fI\(-> This is instance 1 of 2\fR
set c [Counted new]
$b report
        \fI\(-> This is instance 2 of 3\fR
$c report
        \fI\(-> This is instance 3 of 3\fR
.CE
.SH "SEE ALSO"
global(n), namespace(n), oo::class(n), oo::define(n), upvar(n), variable(n)
.SH KEYWORDS
class, class variable, variable
.\" Local Variables:
.\" mode: nroff
.\" fill-column: 78
.\" End:

If a format string lacks a \fB%z\fR or \fB%Z\fR format group,
it is possible for the time to be ambiguous because it appears twice
in the same day, once without and once with Daylight Saving Time.
If this situation occurs, the first occurrence of the time is chosen.
(For this reason, it is wise to have the input string contain the
time zone when converting local times.  This caveat does not apply to
UTC times.)
.PP
If the interpretation of the groups yields an impossible time because
a field is out of range, enough of that field's unit will be added to
or subtracted from the time to bring it in range. Thus, if attempting to
scan or format day 0 of the month, one day will be subtracted from day
1 of the month, yielding the last day of the previous month.
.PP
If the interpretation of the groups yields an impossible time because
a Daylight Saving Time change skips over that time, or an ambiguous
time because a Daylight Saving Time change skips back so that the clock
observes the given time twice, and no time zone specifier (\fB%z\fR
or \fB%Z\fR) is present in the format, the time is interpreted as
if the clock had not changed.
.SH "FORMAT GROUPS"
.PP
The following format groups are recognized by the \fBclock scan\fR and
\fBclock format\fR commands.
.TP
\fB%a\fR
On output, receives an abbreviation (\fIe.g.,\fR \fBMon\fR) for the day

'\"
'\" Copyright (c) 2007 Donal K. Fellows
'\" Copyright (c) 2007-2018 Donal K. Fellows
'\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH define n 0.3 TclOO "TclOO Commands"
.so man.macros
.BS

\fIarg\fR arguments; when the second is present, it is exactly as if all the
arguments from \fIsubcommand\fR onwards are made into a list and that list is
used as the \fIdefScript\fR argument.
.SS "CONFIGURING CLASSES"
.PP
The following commands are supported in the \fIdefScript\fR for
\fBoo::define\fR, each of which may also be used in the \fIsubcommand\fR form:
.TP
\fBclassmethod\fI name\fR ?\fIargList bodyScrip\fR?
.VS TIP478
This creates a class method, or (if \fIargList\fR and \fIbodyScript\fR are
omitted) promotes an existing method on the class object to be a class
method. The \fIname\fR, \fIargList\fR and \fIbodyScript\fR arguments are as in
the \fBmethod\fR definition, below.
.RS
.PP
Class methods can be called on either the class itself or on the instances of
that class. When they are called, the current object (see the \fBself\R and
\fBmy\fR commands) is the class on which they are called or the class of the
instance on which they are called, depending on whether they are called on the
class or an instance of the class, respectively. If called on a subclass or
instance of the subclass, the current object is the subclass.
.PP
In a private definition context, the methods as invoked on classes are
\fInot\fR private, but the methods as invoked on instances of classes are
private.
.RE
.VE TIP478
.TP
\fBconstructor\fI argList bodyScript\fR
.
This creates or updates the constructor for a class. The formal arguments to
the constructor (defined using the same format as for the Tcl \fBproc\fR
command) will be \fIargList\fR, and the body of the constructor will be
\fIbodyScript\fR. When the body of the constructor is evaluated, the current
namespace of the constructor will be a namespace that is unique to the object
being constructed. Within the constructor, the \fBnext\fR command should be
used to call the superclasses' constructors. If \fIbodyScript\fR is the empty
string, the constructor will be deleted.
.TP
\fBdeletemethod\fI name\fR ?\fIname ...\fR
\fBdeletemethod\fI name\fR ?\fIname ...\fR?
.
This deletes each of the methods called \fIname\fR from a class. The methods
must have previously existed in that class. Does not affect the superclasses
of the class, nor does it affect the subclasses or instances of the class
(except when they have a call chain through the class being modified).
.TP
\fBdestructor\fI bodyScript\fR

This arranges for each of the named methods, \fIname\fR, to be exported
(i.e. usable outside an instance through the instance object's command) by the
class being defined. Note that the methods themselves may be actually defined
by a superclass; subclass exports override superclass visibility, and may in
turn be overridden by instances.
.TP
\fBfilter\fR ?\fI\-slotOperation\fR? ?\fImethodName ...\fR?
.VS
.
This slot (see \fBSLOTTED DEFINITIONS\fR below)
.VE
sets or updates the list of method names that are used to guard whether
method call to instances of the class may be called and what the method's
results are. Each \fImethodName\fR names a single filtering method (which may
be exposed or not exposed); it is not an error for a non-existent method to be
named since they may be defined by subclasses.
.VS
By default, this slot works by appending.
.VE
.TP
\fBforward\fI name cmdName \fR?\fIarg ...\fR?
.
This creates or updates a forwarded method called \fIname\fR. The method is
defined be forwarded to the command called \fIcmdName\fR, with additional
arguments, \fIarg\fR etc., added before those arguments specified by the
caller of the method. The \fIcmdName\fR will always be resolved using the
rules of the invoking objects' namespaces, i.e., when \fIcmdName\fR is not
fully-qualified, the command will be searched for in each object's namespace,
using the instances' namespace's path, or by looking in the global namespace.
The method will be exported if \fIname\fR starts with a lower-case letter, and
non-exported otherwise.
.RS
.PP
.VS TIP500
If in a private definition context (see the \fBprivate\fR definition command,
below), this command creates private forwarded methods.
.VE TIP500
.RE
.TP
\fBinitialise\fI script\fR
.TP
\fBinitialize\fI script\fR
.VS TIP478
This evaluates \fIscript\fR in a context which supports local variables and
where the current namespace is the instance namespace of the class object
itself. This is useful for setting up, e.g., class-scoped variables.
.VE TIP478
.TP
\fBmethod\fI name argList bodyScript\fR
.
This creates or updates a method that is implemented as a procedure-like
script. The name of the method is \fIname\fR, the formal arguments to the
method (defined using the same format as for the Tcl \fBproc\fR command) will
be \fIargList\fR, and the body of the method will be \fIbodyScript\fR. When
the body of the method is evaluated, the current namespace of the method will
be a namespace that is unique to the current object. The method will be
exported if \fIname\fR starts with a lower-case letter, and non-exported
otherwise; this behavior can be overridden via \fBexport\fR and
\fBunexport\fR.
.RS
.PP
.VS TIP500
If in a private definition context (see the \fBprivate\fR definition command,
below), this command creates private procedure-like methods.
.VE TIP500
.RE
.TP
\fBmixin\fR ?\fI\-slotOperation\fR? ?\fIclassName ...\fR?
.VS
.
This slot (see \fBSLOTTED DEFINITIONS\fR below)
.VE
sets or updates the list of additional classes that are to be mixed into
all the instances of the class being defined. Each \fIclassName\fR argument
names a single class that is to be mixed in.
.VS
By default, this slot works by replacement.
.TP
\fBprivate \fIcmd arg...\fR
.TP
\fBprivate \fIscript\fR
.
.VS TIP500
This evaluates the \fIscript\fR (or the list of command and arguments given by
\fIcmd\fR and \fIarg\fRs) in a context where the definitions made on the
current class will be private definitions.
.RS
.PP
The following class definition commands are affected by \fBprivate\fR:
\fBforward\fR, \fBmethod\fR, \fBself\fR, and \fBvariable\fR. Nesting
\fBprivate\fR inside \fBprivate\fR has no cumulative effect; the innermost
definition context is just a private definition context. All other definition
commands have no difference in behavior when used in a private definition
context.
.RE
.VE
.VE TIP500
.TP
\fBrenamemethod\fI fromName toName\fR
.
This renames the method called \fIfromName\fR in a class to \fItoName\fR. The
method must have previously existed in the class, and \fItoName\fR must not
previously refer to a method in that class. Does not affect the superclasses
of the class, nor does it affect the subclasses or instances of the class

.QW "\fBoo::objdefine \fIcls subcommand ...\fR" .
.RS
.PP
.VS TIP470
If no arguments at all are used, this gives the name of the class currently
being configured.
.VE TIP470
.VS TIP500
If in a private definition context (see the \fBprivate\fR definition command,
below), the definitions on the class object will also be made in a private
definition context.
.VE TIP500
.RE
.TP
\fBsuperclass\fR ?\fI\-slotOperation\fR? ?\fIclassName ...\fR?
.VS
.
This slot (see \fBSLOTTED DEFINITIONS\fR below)
.VE
allows the alteration of the superclasses of the class being defined.
Each \fIclassName\fR argument names one class that is to be a superclass of
the defined class. Note that objects must not be changed from being classes to
being non-classes or vice-versa, that an empty parent class is equivalent to
\fBoo::object\fR, and that the parent classes of \fBoo::object\fR and
\fBoo::class\fR may not be modified.
.VS
By default, this slot works by replacement.
.VE
.TP
\fBunexport\fI name \fR?\fIname ...\fR?
.
This arranges for each of the named methods, \fIname\fR, to be not exported
(i.e. not usable outside the instance through the instance object's command,
but instead just through the \fBmy\fR command visible in each object's
context) by the class being defined. Note that the methods themselves may be
actually defined by a superclass; subclass unexports override superclass
visibility, and may be overridden by instance unexports.
.TP
\fBvariable\fR ?\fI\-slotOperation\fR? ?\fIname ...\fR?
.VS
.
This slot (see \fBSLOTTED DEFINITIONS\fR below) arranges for each of the named
variables to be automatically made
available in the methods, constructor and destructor declared by the class
being defined. Each variable name must not have any namespace
separators and must not look like an array access. All variables will be
actually present in the namespace of the instance object on which the method
actually present in the instance object on which the method is executed. Note
is executed. Note
that the variable lists declared by a superclass or subclass are completely
disjoint, as are variable lists declared by instances; the list of variable
names is just for methods (and constructors and destructors) declared by this
class. By default, this slot works by appending.
.RS
.PP
.VS TIP500
If in a private definition context (see the \fBprivate\fR definition command,
below), this slot manipulates the list of private variable bindings for this
class. In a private variable binding, the name of the variable within the
instance object is different to the name given in the definition; the name
used in the definition is the name that you use to access the variable within
the methods of this class, and the name of the variable in the instance
namespace has a unique prefix that makes accidental use from other classes
extremely unlikely.
.VE
.VE TIP500
.RE
.SS "CONFIGURING OBJECTS"
.PP
The following commands are supported in the \fIdefScript\fR for
\fBoo::objdefine\fR, each of which may also be used in the \fIsubcommand\fR
form:
.TP
\fBclass\fI className\fR

.
This arranges for each of the named methods, \fIname\fR, to be exported
(i.e. usable outside the object through the object's command) by the object
being defined. Note that the methods themselves may be actually defined by a
class or superclass; object exports override class visibility.
.TP
\fBfilter\fR ?\fI\-slotOperation\fR? ?\fImethodName ...\fR?
.VS
.
This slot (see \fBSLOTTED DEFINITIONS\fR below)
.VE
sets or updates the list of method names that are used to guard whether a
method call to the object may be called and what the method's results are.
Each \fImethodName\fR names a single filtering method (which may be exposed or
not exposed); it is not an error for a non-existent method to be named. Note
that the actual list of filters also depends on the filters set upon any
classes that the object is an instance of.
.VS
By default, this slot works by appending.
.VE
.TP
\fBforward\fI name cmdName \fR?\fIarg ...\fR?
.
This creates or updates a forwarded object method called \fIname\fR. The
method is defined be forwarded to the command called \fIcmdName\fR, with
additional arguments, \fIarg\fR etc., added before those arguments specified
by the caller of the method. Forwarded methods should be deleted using the
\fBmethod\fR subcommand. The method will be exported if \fIname\fR starts with
a lower-case letter, and non-exported otherwise.
.RS
.PP
.VS TIP500
If in a private definition context (see the \fBprivate\fR definition command,
below), this command creates private forwarded methods.
.VE TIP500
.RE
.TP
\fBmethod\fI name argList bodyScript\fR
.
This creates, updates or deletes an object method. The name of the method is
\fIname\fR, the formal arguments to the method (defined using the same format
as for the Tcl \fBproc\fR command) will be \fIargList\fR, and the body of the
method will be \fIbodyScript\fR. When the body of the method is evaluated, the
current namespace of the method will be a namespace that is unique to the
object. The method will be exported if \fIname\fR starts with a lower-case
letter, and non-exported otherwise.
.RS
.PP
.VS TIP500
If in a private definition context (see the \fBprivate\fR definition command,
below), this command creates private procedure-like methods.
.VE TIP500
.RE
.TP
\fBmixin\fR ?\fI\-slotOperation\fR? ?\fIclassName ...\fR?
.VS
.
This slot (see \fBSLOTTED DEFINITIONS\fR below)
.VE
sets or updates a per-object list of additional classes that are to be
mixed into the object. Each argument, \fIclassName\fR, names a single class
that is to be mixed in.
.VS
By default, this slot works by replacement.
.TP
\fBprivate \fIcmd arg...\fR
.TP
\fBprivate \fIscript\fR
.VS TIP500
This evaluates the \fIscript\fR (or the list of command and arguments given by
\fIcmd\fR and \fIarg\fRs) in a context where the definitions made on the
current object will be private definitions.
.RS
.PP
The following class definition commands are affected by \fBprivate\fR:
\fBforward\fR, \fBmethod\fR, and \fBvariable\fR. Nesting \fBprivate\fR inside
\fBprivate\fR has no cumulative effect; the innermost definition context is
just a private definition context. All other definition commands have no
difference in behavior when used in a private definition context.
.RE
.VE
.VE TIP500
.TP
\fBrenamemethod\fI fromName toName\fR
.
This renames the method called \fIfromName\fR in an object to \fItoName\fR.
The method must have previously existed in the object, and \fItoName\fR must
not previously refer to a method in that object. Does not affect the classes
that the object is an instance of. Does not change the export status of the
method; if it was exported before, it will be afterwards.
.TP
\fBself \fR
.
.VS TIP470
This gives the name of the object currently being configured.
.VE TIP470
.TP
\fBunexport\fI name \fR?\fIname ...\fR?
.
This arranges for each of the named methods, \fIname\fR, to be not exported
(i.e. not usable outside the object through the object's command, but instead
just through the \fBmy\fR command visible in the object's context) by the
object being defined. Note that the methods themselves may be actually defined
by a class; instance unexports override class visibility.
.TP
\fBvariable\fR ?\fI\-slotOperation\fR? ?\fIname ...\fR?
.VS
.
This slot (see \fBSLOTTED DEFINITIONS\fR below) arranges for each of the named
variables to be automatically made available in the methods declared by the
object being defined.  Each variable name must not have any namespace
separators and must not look like an array access. All variables will be
actually present in the namespace of the object on which the method is
actually present in the object on which the method is executed. Note that the
executed. Note that the
variable lists declared by the classes and mixins of which the object is an
instance are completely disjoint; the list of variable names is just for
methods declared by this object. By default, this slot works by appending.
.RS
.PP
.VS TIP500
If in a private definition context (see the \fBprivate\fR definition command,
below), this slot manipulates the list of private variable bindings for this
object.  In a private variable binding, the name of the variable within the
instance object is different to the name given in the definition; the name
used in the definition is the name that you use to access the variable within
the methods of this instance object, and the name of the variable in the
instance namespace has a unique prefix that makes accidental use from
superclass methods extremely unlikely.
.VE TIP500
.RE
.SH "PRIVATE METHODS"
.VS TIP500
When a class or instance has a private method, that private method can only be
invoked from within methods of that class or instance. Other callers of the
object's methods \fIcannot\fR invoke private methods, it is as if the private
methods do not exist. However, a private method of a class \fIcan\fR be
invoked from the class's methods when those methods are being used on another
instance object; this means that a class can use them to coordinate behaviour
between several instances of itself without interfering with how other
classes (especially either subclasses or superclasses) interact. Private
methods precede all mixed in classes in the method call order (as reported by
\fBself call\fR).
.VE TIP500
.SH "SLOTTED DEFINITIONS"
Some of the configurable definitions of a class or object are \fIslotted
definitions\fR. This means that the configuration is implemented by a slot
object, that is an instance of the class \fBoo::Slot\fR, which manages a list
of values (class names, variable names, etc.) that comprises the contents of
the slot. The class defines three operations (as methods) that may be done on
the slot:
.VE
.TP
\fIslot\fR \fB\-append\fR ?\fImember ...\fR?
.VS
.
This appends the given \fImember\fR elements to the slot definition.
.VE
.TP
\fIslot\fR \fB\-clear\fR
.VS
.
This sets the slot definition to the empty list.
.VE
.TP
\fIslot\fR \fB\-set\fR ?\fImember ...\fR?
.VS
.
This replaces the slot definition with the given \fImember\fR elements.
.PP
A consequence of this is that any use of a slot's default operation where the
first member argument begins with a hyphen will be an error. One of the above
operations should be used explicitly in those circumstances.
.SS "SLOT IMPLEMENTATION"
Internally, slot objects also define a method \fB\-\-default\-operation\fR
which is forwarded to the default operation of the slot (thus, for the class
.QW \fBvariable\fR
slot, this is forwarded to
.QW "\fBmy \-append\fR" ),
and these methods which provide the implementation interface:
.VE
.TP
\fIslot\fR \fBGet\fR
.VS
.
Returns a list that is the current contents of the slot. This method must
always be called from a stack frame created by a call to \fBoo::define\fR or
\fBoo::objdefine\fR.
.VE
.TP
\fIslot\fR \fBSet \fIelementList\fR
.VS
.
Sets the contents of the slot to the list \fIelementList\fR and returns the
empty string. This method must always be called from a stack frame created by
a call to \fBoo::define\fR or \fBoo::objdefine\fR.
.PP
The implementation of these methods is slot-dependent (and responsible for
accessing the correct part of the class or object definition). Slots also have
an unknown method handler to tie all these pieces together, and they hide
their \fBdestroy\fR method so that it is not invoked inadvertently. It is
\fIrecommended\fR that any user changes to the slot mechanism be restricted to
defining new operations whose names start with a hyphen.
.VE
.SH EXAMPLES
This example demonstrates how to use both forms of the \fBoo::define\fR and
\fBoo::objdefine\fR commands (they work in the same way), as well as
illustrating four of the subcommands of them.
.PP
.CS
oo::class create c

}
\fBoo::objdefine\fR inst {
    \fBmixin -append\fR B
}
inst m1              \fI\(-> prints "red brick"\fR
inst m2              \fI\(-> prints "blue brick"\fR
.CE
.PP
.VS TIP478
This example shows how to create and use class variables. It is a class that
counts how many instances of itself have been made.
.PP
.CS
oo::class create Counted
\fBoo::define\fR Counted {
    \fBinitialise\fR {
        variable count 0
    }

    \fBvariable\fR number
    \fBconstructor\fR {} {
        classvariable count
        set number [incr count]
    }

    \fBmethod\fR report {} {
        classvariable count
        puts "This is instance $number of $count"
    }
}

set a [Counted new]
set b [Counted new]
$a report
        \fI\(-> This is instance 1 of 2\fR
set c [Counted new]
$b report
        \fI\(-> This is instance 2 of 3\fR
$c report
        \fI\(-> This is instance 3 of 3\fR
.CE
.PP
This example demonstrates how to use class methods. (Note that the constructor
for \fBoo::class\fR calls \fBoo::define\fR on the class.)
.PP
.CS
oo::class create DBTable {
    \fBclassmethod\fR find {description} {
        puts "DB: locate row from [self] matching $description"
        return [my new]
    }
    \fBclassmethod\fR insert {description} {
        puts "DB: create row in [self] matching $description"
        return [my new]
    }
    \fBmethod\fR update {description} {
        puts "DB: update row [self] with $description"
    }
    \fBmethod\fR delete {} {
        puts "DB: delete row [self]"
        my destroy; # Just delete the object, not the DB row
    }
}

oo::class create Users {
    \fBsuperclass\fR DBTable
}
oo::class create Groups {
    \fBsuperclass\fR DBTable
}

set u1 [Users insert "username=abc"]
        \fI\(-> DB: create row from ::Users matching username=abc\fR
set u2 [Users insert "username=def"]
        \fI\(-> DB: create row from ::Users matching username=def\fR
$u2 update "group=NULL"
        \fI\(-> DB: update row ::oo::Obj124 with group=NULL\fR
$u1 delete
        \fI\(-> DB: delete row ::oo::Obj123\fR
set g [Group find "groupname=webadmins"]
        \fI\(-> DB: locate row ::Group with groupname=webadmins\fR
$g update "emailaddress=admins"
        \fI\(-> DB: update row ::oo::Obj125 with emailaddress=admins\fR
.CE
.VE TIP478
.SH "SEE ALSO"
next(n), oo::class(n), oo::object(n)
.SH KEYWORDS
class, definition, method, object, slot
.\" Local variables:
.\" mode: nroff
.\" fill-column: 78
.\" End:

.QW \fB@\0\fIfileId\fR
notation, does not work.  When reading from a socket, a 16-bit DOS
application will hang and a 32-bit application will return immediately with
end-of-file.  When either type of application writes to a socket, the
information is instead sent to the console, if one is present, or is
discarded.
.RS
.PP
Note that the current escape resp. quoting of arguments for windows works only
with executables using CommandLineToArgv, CRT-library or similar, as well as
with the windows batch files (excepting the newline, see below).
Although it is the common escape algorithm, but, in fact, the way how the
executable parses the command-line (resp. splits it into single arguments)
is decisive.
.PP
Unfortunately, there is currently no way to supply newline character within
an argument to the batch files (\fB.cmd\fR or \fB.bat\fR) or to the command
processor (\fBcmd.exe /c\fR), because this causes truncation of command-line
(also the argument chain) on the first newline character.
But it works properly with an executable (using CommandLineToArgv, etc).
.PP
The Tk console text widget does not provide real standard IO capabilities.
Under Tk, when redirecting from standard input, all applications will see an
immediate end-of-file; information redirected to standard output or standard
error will be discarded.
.PP
Either forward or backward slashes are accepted as path separators for

.CS
\fBexec\fR cmp.bat somefile.c -o somefile
.CE
.PP
With the file \fIcmp.bat\fR looking something like:
.PP
.CS
@gcc %*
.CE
or like another variant using single parameters:
.CS
@gcc %1 %2 %3 %4 %5 %6 %7 %8 %9
.CE
.SS "WORKING WITH COMMAND BUILT-INS"
.PP
Sometimes you need to be careful, as different programs may have the
same name and be in the path. It can then happen that typing a command
at the DOS prompt finds \fIa different program\fR than the same

An expression consists of a combination of operands, operators, parentheses and
commas, possibly with whitespace between any of these elements, which is
ignored.
An integer operand may be specified in decimal (the normal case, the optional
first two characters are \fB0d\fR), binary
(the first two characters are \fB0b\fR), octal
(the first two characters are \fB0o\fR), or hexadecimal
(the first two characters are \fB0x\fR) form.
(the first two characters are \fB0x\fR) form.  For
compatibility with older Tcl releases, an operand that begins with \fB0\fR is
interpreted as an octal integer even if the second character is not \fBo\fR.
A floating-point number may be specified in any of several
common decimal formats, and may use the decimal point \fB.\fR,
\fBe\fR or \fBE\fR for scientific notation, and
the sign characters \fB+\fR and \fB\-\fR.  The
following are all valid floating-point numbers:  2.1, 3., 6e4, 7.91e+16.
The strings \fBInf\fR
and \fBNaN\fR, in any combination of case, are also recognized as floating point

number if the first character is not a sign.
.TP 10
\fB0\fR
Specifies that the number should be padded on the left with
zeroes instead of spaces.
.TP 10
\fB#\fR
Requests an alternate output form. For \fBo\fR and \fBO\fR
conversions it guarantees that the first digit is always \fB0\fR.
For \fBx\fR or \fBX\fR conversions, \fB0x\fR or \fB0X\fR (respectively)
Requests an alternate output form. For \fBo\fR conversions,
\fB0o\fR will be added to the beginning of the result unless
it is zero. For \fBx\fR or \fBX\fR conversions, \fB0x\fR
will be added to the beginning of the result unless it is zero.
For \fBb\fR conversions, \fB0b\fR
will be added to the beginning of the result unless it is zero.
For \fBd\fR conversions, \fB0d\fR will be added to the beginning
of the result unless it is zero.
For \fBd\fR conversions, \fB0d\fR there is no effect unless
the \fB0\fR specifier is used as well: In that case, \fB0d\fR
will be added to the beginning.
For all floating-point conversions (\fBe\fR, \fBE\fR, \fBf\fR,
\fBg\fR, and \fBG\fR) it guarantees that the result always
has a decimal point.
For \fBg\fR and \fBG\fR conversions it specifies that
trailing zeroes should not be removed.
.SS "OPTIONAL FIELD WIDTH"
.PP

printed; if the string is longer than this then the trailing characters will be dropped.
If the precision is specified with \fB*\fR rather than a number
then the next argument to the \fBformat\fR command determines the precision;
it must be a numeric string.
.SS "OPTIONAL SIZE MODIFIER"
.PP
The fifth part of a conversion specifier is a size modifier,
which must be \fBll\fR, \fBh\fR, or \fBl\fR.
which must be \fBll\fR, \fBh\fR, \fBl\fR, or \fBL\fR.
If it is \fBll\fR it specifies that an integer value is taken
without truncation for conversion to a formatted substring.
If it is \fBh\fR it specifies that an integer value is
truncated to a 16-bit range before converting.  This option is rarely useful.
If it is \fBl\fR it specifies that the integer value is
truncated to the same range as that produced by the \fBwide()\fR
function of the \fBexpr\fR command (at least a 64-bit range).
If it is \fBL\fR it specifies that an integer or double value is taken
without truncation for conversion to a formatted substring.
If neither \fBh\fR nor \fBl\fR are present, the integer value is
If neither \fBh\fR nor \fBl\fR nor \fBL\fR are present, the integer value is
truncated to the same range as that produced by the \fBint()\fR
function of the \fBexpr\fR command (at least a 32-bit range, but
determined by the value of the \fBwordSize\fR element of the
\fBtcl_platform\fR array).
.SS "MANDATORY CONVERSION TYPE"
.PP
The last thing in a conversion specifier is an alphabetic character

Convert integer to unsigned hexadecimal string, using digits
.QW 0123456789abcdef
for \fBx\fR and
.QW 0123456789ABCDEF
for \fBX\fR).
.TP 10
\fBb\fR
Convert integer to binary string, using digits 0 and 1.
Convert integer to unsigned binary string, using digits 0 and 1.
.TP 10
\fBc\fR
Convert integer to the Unicode character it represents.
.TP 10
\fBs\fR
No conversion; just insert string.
.TP 10

\fBg\fR or \fBG\fR
If the exponent is less than \-4 or greater than or equal to the
precision, then convert number as for \fB%e\fR or
\fB%E\fR.
Otherwise convert as for \fB%f\fR.
Trailing zeroes and a trailing decimal point are omitted.
.TP 10
\fBa\fR or \fBA\fR
Convert double to hexadecimal notation in the form
\fI0x1.yyy\fBp\(+-\fIzz\fR, where the number of \fIy\fR's is
determined by the precision (default: 13).
If the \fBA\fR form is used then the hex characters
are printed in uppercase.
.TP 10
\fB%\fR
No conversion: just insert \fB%\fR.
.TP 10
\fBp\fR
Shorthand form for \fB0x%zx\fR, so it outputs the integer in
hexadecimal form with \fB0x\fR prefix.
.SH "DIFFERENCES FROM ANSI SPRINTF"
.PP
The behavior of the format command is the same as the
ANSI C \fBsprintf\fR procedure except for the following
differences:
.IP [1]
Tcl guarantees that it will be working with UNICODE characters.
.IP [2]
\fB%p\fR and \fB%n\fR specifiers are not supported.
\fB%n\fR specifier is not supported.
.IP [3]
For \fB%c\fR conversions the argument must be an integer value,
which will then be converted to the corresponding character value.
.IP [4]
The size modifiers are ignored when formatting floating-point values.
The \fBll\fR modifier has no \fBsprintf\fR counterpart.
The \fBb\fR specifier has no \fBsprintf\fR counterpart.
.SH EXAMPLES
.PP
Convert the numeric value of a UNICODE character to the character
itself:
.PP
.CS

'\"
'\" Copyright (c) 1995-1997 Sun Microsystems, Inc.
'\" Copyright (c) 1998-2000 by Ajuba Solutions.
'\" Copyright (c) 2004 ActiveState Corporation.
'\"
'\" See the file "license.terms" for information on usage and redistribution
'\" of this file, and for a DISCLAIMER OF ALL WARRANTIES.
'\"
.TH "http" n 2.7 http "Tcl Bundled Packages"
.TH "http" n 2.9 http "Tcl Bundled Packages"
.so man.macros
.BS
'\" Note:  do not modify the .SH NAME line immediately below!
.SH NAME
http \- Client-side implementation of the HTTP/1.1 protocol
.SH SYNOPSIS
\fBpackage require http ?2.7?\fR
\fBpackage require http ?2.8?\fR
.\" See Also -useragent option documentation in body!
.sp
\fB::http::config ?\fI\-option value\fR ...?
.sp
\fB::http::geturl \fIurl\fR ?\fI\-option value\fR ...?
.sp
\fB::http::formatQuery\fR \fIkey value\fR ?\fIkey value\fR ...?
.sp
\fB::http::quoteString\fR \fIvalue\fR
.sp
\fB::http::quoteString\fR \fIvalue\fR
\fB::http::reset\fR \fItoken\fR ?\fIwhy\fR?
.sp
\fB::http::wait \fItoken\fR
.sp
\fB::http::status \fItoken\fR
.sp
\fB::http::size \fItoken\fR

.sp
\fB::http::error \fItoken\fR
.sp
\fB::http::cleanup \fItoken\fR
.sp
\fB::http::register \fIproto port command\fR
.sp
\fB::http::registerError \fIport\fR ?\fImessage\fR?
.sp
\fB::http::unregister \fIproto\fR
.BE
.SH DESCRIPTION
.PP
The \fBhttp\fR package provides the client side of the HTTP/1.1
protocol, as defined in RFC 2616.
protocol, as defined in RFC 7230 to RFC 7235, which supersede RFC 2616.
The package implements the GET, POST, and HEAD operations
of HTTP/1.1.  It allows configuration of a proxy host to get through
firewalls.  The package is compatible with the \fBSafesock\fR security
policy, so it can be used by untrusted applets to do URL fetching from
a restricted set of hosts. This package can be extended to support
additional HTTP transport protocols, such as HTTPS, by providing
a custom \fBsocket\fR command, via \fB::http::register\fR.

\fB\-accept\fR \fImimetypes\fR
.
The Accept header of the request.  The default is */*, which means that
all types of documents are accepted.  Otherwise you can supply a
comma-separated list of mime type patterns that you are
willing to receive.  For example,
.QW "image/gif, image/jpeg, text/*" .
.TP
\fB\-pipeline\fR \fIboolean\fR
.
Specifies whether HTTP/1.1 transactions on a persistent socket will be
pipelined.  See the \fBPERSISTENT SOCKETS\fR section for details. The default
is 1.
.TP
\fB\-postfresh\fR \fIboolean\fR
.
Specifies whether requests that use the \fBPOST\fR method will always use a
fresh socket, overriding the \fB-keepalive\fR option of
command \fBhttp::geturl\fR.  See the \fBPERSISTENT SOCKETS\fR section for details.
The default is 0.
.TP
\fB\-proxyhost\fR \fIhostname\fR
.
The name of the proxy host, if any.  If this value is the
empty string, the URL host is contacted directly.
.TP
\fB\-proxyport\fR \fInumber\fR

to determine if a proxy is required for a given host.  One argument, a
host name, is added to \fIcommand\fR when it is invoked.  If a proxy
is required, the callback should return a two-element list containing
the proxy server and proxy port.  Otherwise the filter should return
an empty list.  The default filter returns the values of the
\fB\-proxyhost\fR and \fB\-proxyport\fR settings if they are
non-empty.
.TP
\fB\-repost\fR \fIboolean\fR
.
Specifies what to do if a POST request over a persistent connection fails
because the server has half-closed the connection.  If boolean \fBtrue\fR, the
request
will be automatically retried; if boolean \fBfalse\fR it will not, and the
application
that uses \fBhttp::geturl\fR is expected to seek user confirmation before
retrying the POST.  The value \fBtrue\fR should be used only under certain
conditions. See the \fBPERSISTENT SOCKETS\fR section for details. The
default is 0.
.TP
\fB\-urlencoding\fR \fIencoding\fR
.
The \fIencoding\fR used for creating the x-url-encoded URLs with
\fB::http::formatQuery\fR and \fB::http::quoteString\fR.
\fB::http::formatQuery\fR.  The default is \fButf-8\fR, as specified by RFC
The default is \fButf-8\fR, as specified by RFC
2718.  Prior to http 2.5 this was unspecified, and that behavior can be
returned by specifying the empty string (\fB{}\fR), although
\fIiso8859-1\fR is recommended to restore similar behavior but without the
\fB::http::formatQuery\fR throwing an error processing non-latin-1
characters.
\fB::http::formatQuery\fR or \fB::http::quoteString\fR
throwing an error processing non-latin-1 characters.
.TP
\fB\-useragent\fR \fIstring\fR
.
The value of the User-Agent header in the HTTP request.  The default is
.QW "\fBTcl http client package 2.7\fR" .
The value of the User-Agent header in the HTTP request.  In an unsafe
interpreter, the default value depends upon the operating system, and
the version numbers of \fBhttp\fR and \fBTcl\fR, and is (for example)
.QW "\fBMozilla/5.0 (Windows; U; Windows NT 10.0) http/2.8.12 Tcl/8.6.8\fR" .
A safe interpreter cannot determine its operating system, and so the default
in a safe interpreter is to use a Windows 10 value with the current version
numbers of \fBhttp\fR and \fBTcl\fR.
.TP
\fB\-zip\fR \fIboolean\fR
.
If the value is boolean \fBtrue\fR, then by default requests will send a header
.QW "\fBAccept-Encoding: gzip,deflate,compress\fR" .
If the value is boolean \fBfalse\fR, then by default this header will not be sent.
In either case the default can be overridden for an individual request by
supplying a custom \fBAccept-Encoding\fR header in the \fB-headers\fR option
of \fBhttp::geturl\fR. The default is 1.
.RE
.TP
\fB::http::geturl\fR \fIurl\fR ?\fIoptions\fR?
.
The \fB::http::geturl\fR command is the main procedure in the package.
The \fB\-query\fR option causes a POST operation and
the \fB\-validate\fR option causes a HEAD operation;

.CS
Pragma: no-cache
.CE
.RE
.TP
\fB\-keepalive\fR \fIboolean\fR
.
If true, attempt to keep the connection open for servicing
If boolean \fBtrue\fR, attempt to keep the connection open for servicing
multiple requests.  Default is 0.
.TP
\fB\-method\fR \fItype\fR
.
Force the HTTP request method to \fItype\fR. \fB::http::geturl\fR will
auto-select GET, POST or HEAD based on other options, but this option
enables choices like PUT and DELETE for webdav support.

\fB::http::formatQuery\fR \fIkey value\fR ?\fIkey value\fR ...?
.
This procedure does x-url-encoding of query data.  It takes an even
number of arguments that are the keys and values of the query.  It
encodes the keys and values, and generates one string that has the
proper & and = separators.  The result is suitable for the
\fB\-query\fR value passed to \fB::http::geturl\fR.
.TP
\fB::http::quoteString\fR \fIvalue\fR
.
This procedure does x-url-encoding of string.  It takes a single argument and
encodes it.
.TP
\fB::http::reset\fR \fItoken\fR ?\fIwhy\fR?
.
This command resets the HTTP transaction identified by \fItoken\fR, if any.
This sets the \fBstate(status)\fR value to \fIwhy\fR, which defaults to
\fBreset\fR, and then calls the registered \fB\-command\fR callback.
.TP

package require tls

::http::register https 443 ::tls::socket

set token [::http::geturl https://my.secure.site/]
.CE
.RE
.TP
\fB::http::registerError\fR \fIport\fR ?\fImessage\fR?
.
This procedure allows a registered protocol handler to deliver an error
message for use by \fBhttp\fR.  Calling this command does not raise an
error. The command is useful when a registered protocol detects an problem
(for example, an invalid TLS certificate) that will cause an error to
propagate to \fBhttp\fR.  The command allows \fBhttp\fR to provide a
precise error message rather than a general one.  The command returns the
value provided by the last call with argument \fImessage\fR, or the empty
string if no such call has been made.
.TP
\fB::http::unregister\fR \fIproto\fR
.
This procedure unregisters a protocol handler that was previously
registered via \fB::http::register\fR, returning a two-item list of
the default port and handler command that was previously installed
(via \fB::http::register\fR) if there was such a handler, and an error if

Once the data associated with the URL is no longer needed, the state
array should be unset to free up storage.
The \fB::http::cleanup\fR procedure is provided for that purpose.
The following elements of
the array are supported:
.RS
.TP
\fBbinary\fR
.
This is boolean \fBtrue\fR if (after decoding any compression specified
by the
.QW "Content-Encoding"
response header) the HTTP response is binary.  It is boolean \fBfalse\fR
if the HTTP response is text.
.TP
\fBbody\fR
.
The contents of the URL.  This will be empty if the \fB\-channel\fR
option has been specified.  This value is returned by the \fB::http::data\fR command.
.TP
\fBcharset\fR
.

.
A copy of the \fBContent-Type\fR meta-data value.
.TP
\fBurl\fR
.
The requested URL.
.RE
.SH "PERSISTENT CONNECTIONS"
.PP
.SS "BASICS"
.PP
See RFC 7230 Sec 6, which supersedes RFC 2616 Sec 8.1.
.PP
A persistent connection allows multiple HTTP/1.1 transactions to be
carried over the same TCP connection.  Pipelining allows a
client to make multiple requests over a persistent connection without
waiting for each response.  The server sends responses in the same order
that the requests were received.
.PP
If a POST request fails to complete, typically user confirmation is
needed before sending the request again.  The user may wish to verify
whether the server was modified by the failed POST request, before
sending the same request again.
.PP
A HTTP request will use a persistent socket if the call to
\fBhttp::geturl\fR has the option \fB-keepalive true\fR. It will use
pipelining where permitted if the \fBhttp::config\fR option
\fB-pipeline\fR is boolean \fBtrue\fR (its default value).
.PP
The http package maintains no more than one persistent connection to each
server (i.e. each value of
.QW "domain:port" ).
If \fBhttp::geturl\fR is called to make a request over a persistent
connection while the connection is busy with another request, the new
request will be held in a queue until the connection is free.
.PP
The http package does not support HTTP/1.0 persistent connections
controlled by the \fBKeep-Alive\fR header.
.SS "SPECIAL CASES"
.PP
This subsection discusses issues related to closure of the
persistent connection by the server, automatic retry of failed requests,
the special treatment necessary for POST requests, and the options for
dealing with these cases.
.PP
In accordance with RFC 7230, \fBhttp::geturl\fR does not pipeline
requests that use the POST method.  If a POST uses a persistent
connection and is not the first request on that connection,
\fBhttp::geturl\fR waits until it has received the response for the previous
request; or (if \fBhttp::config\fR option \fB-postfresh\fR is boolean \fBtrue\fR) it
uses a new connection for each POST.
.PP
If the server is processing a number of pipelined requests, and sends a
response header
.QW "\fBConnection: close\fR"
with one of the responses (other than the last), then subsequent responses
are unfulfilled. \fBhttp::geturl\fR will send the unfulfilled requests again
over a new connection.
.PP
A difficulty arises when a HTTP client sends a request over a persistent
connection that has been idle for a while.  The HTTP server may
half-close an apparently idle connection while the client is sending a
request, but before the request arrives at the server: in this case (an
.QW "asynchronous close event" )
the request will fail.  The difficulty arises because the client cannot
be certain whether the POST modified the state of the server.  For HEAD or
GET requests, \fBhttp::geturl\fR opens another connection and retransmits
the failed request. However, if the request was a POST, RFC 7230 forbids
automatic retry by default, suggesting either user confirmation, or
confirmation by user-agent software that has semantic understanding of
the application.  The \fBhttp::config\fR option \fB-repost\fR allows for
either possibility.
.PP
Asynchronous close events can occur only in a short interval of time.  The
\fBhttp\fR package monitors each persistent connection for closure by the
server.  Upon detection, the connection is also closed at the client end,
and subsequent requests will use a fresh connection.
.PP
If the \fBhttp::geturl\fR command is called with option \fB-keepalive true\fR,
then it will both try to use an existing persistent connection
(if one is available), and it will send the server a
.QW "\fBConnection: keep-alive\fR"
request header asking to keep the connection open for future requests.
.PP
The \fBhttp::config\fR options \fB-pipeline\fR, \fB-postfresh\fR, and
\fB-repost\fR relate to persistent connections.
.PP
Option \fB-pipeline\fR, if boolean \fBtrue\fR, will pipeline GET and HEAD requests
made
over a persistent connection.  POST requests will not be pipelined - if the
POST is not the first transaction on the connection, its request will not
be sent until the previous response has finished.  GET and HEAD requests
made after a POST will not be sent until the POST response has been
delivered, and will not be sent if the POST fails.
.PP
Option \fB-postfresh\fR, if boolean \fBtrue\fR, will override the \fBhttp::geturl\fR option
\fB-keepalive\fR, and always open a fresh connection for a POST request.
.PP
Option \fB-repost\fR, if \fBtrue\fR, permits automatic retry of a POST request
that fails because it uses a persistent connection that the server has
half-closed (an
.QW "asynchronous close event" ).
Subsequent GET and HEAD requests in a failed pipeline will also be retried.
\fIThe -repost option should be used only if the application understands
that the retry is appropriate\fR - specifically, the application must know
that if the failed POST successfully modified the state of the server, a repeat POST
would have no adverse effect.
.SH EXAMPLE
.PP
This example creates a procedure to copy a URL to a file while printing a
progress meter, and prints the meta-data associated with the URL.
.PP
.CS
proc httpcopy { url file {chunk 4096} } {