view lib-src/make-msgfile.c @ 3767:6b2ef948e140

[xemacs-hg @ 2006-12-29 18:09:38 by aidan] etc/ChangeLog addition: 2006-12-21 Aidan Kehoe <kehoea@parhasard.net> * unicode/unicode-consortium/8859-7.TXT: Update the mapping to the 2003 version of ISO 8859-7. lisp/ChangeLog addition: 2006-12-21 Aidan Kehoe <kehoea@parhasard.net> * mule/cyrillic.el: * mule/cyrillic.el (iso-8859-5): * mule/cyrillic.el (cyrillic-koi8-r-encode-table): Add syntax, case support for Cyrillic; make some parentheses more Lispy. * mule/european.el: Content moved to latin.el, file deleted. * mule/general-late.el: If Unicode tables are to be loaded at dump time, do it here, not in loadup.el. * mule/greek.el: Add syntax, case support for Greek. * mule/latin.el: Move the content of european.el here. Change the case table mappings to use hexadecimal codes, to make cross reference to the standards easier. In all cases, take character syntax from similar characters in Latin-1 , rather than deciding separately what syntax they should take. Add (incomplete) support for case with Turkish. Remove description of the character sets used from the language environments' doc strings, since now that we create variant language environments on the fly, such descriptions will often be inaccurate. Set the native-coding-system language info property while setting the other coding-system properties of the language. * mule/misc-lang.el (ipa): Remove the language environment. The International Phonetic _Alphabet_ is not a language, it's inane to have a corresponding language environment in XEmacs. * mule/mule-cmds.el (create-variant-language-environment): Also modify the coding-priority when creating a new language environment; document that. * mule/mule-cmds.el (get-language-environment-from-locale): Recognise that the 'native-coding-system language-info property can be a list, interpret it correctly when it is one. 2006-12-21 Aidan Kehoe <kehoea@parhasard.net> * coding.el (coding-system-category): Use the new 'unicode-type property for finding what sort of Unicode coding system subtype a coding system is, instead of the overshadowed 'type property. * dumped-lisp.el (preloaded-file-list): mule/european.el has been removed. * loadup.el (really-early-error-handler): Unicode tables loaded at dump time are now in mule/general-late.el. * simple.el (count-lines): Add some backslashes to to parentheses in docstrings to help fontification along. * simple.el (what-cursor-position): Wrap a line to fit in 80 characters. * unicode.el: Use the 'unicode-type property, not 'type, for setting the Unicode coding-system subtype. src/ChangeLog addition: 2006-12-21 Aidan Kehoe <kehoea@parhasard.net> * file-coding.c: Update the make-coding-system docstring to reflect unicode-type * general-slots.h: New symbol, unicode-type, since 'type was being overridden when accessing a coding system's Unicode subtype. * intl-win32.c: Backslash a few parentheses, to help fontification along. * intl-win32.c (complex_vars_of_intl_win32): Use the 'unicode-type symbol, not 'type, when creating the Microsoft Unicode coding system. * unicode.c (unicode_putprop): * unicode.c (unicode_getprop): * unicode.c (unicode_print): Using 'type as the property name when working out what Unicode subtype a given coding system is was broken, since there's a general coding system property called 'type. Change the former to use 'unicode-type instead.
author aidan
date Fri, 29 Dec 2006 18:09:51 +0000
parents ecf1ebac70d8
children
line wrap: on
line source

/* #### Old code!  Replaced with make-msgfile.lex. */


/* Scan specified C and Lisp files, extracting the following messages:

     C files:
	GETTEXT (...)
	DEFER_GETTEXT (...)
	DEFUN interactive prompts
     Lisp files:
	(gettext ...)
	(dgettext "domain-name" ...)
	(defer-gettext ...)
	(interactive ...)

  The arguments given to this program are all the C and Lisp source files
  of XEmacs.  .el and .c files are allowed.  There is no support for .elc
  files at this time, but they may be specified; the corresponding .el file
  will be used.  Similarly, .o files can also be specified, and the corresponding
  .c file will be used.  This helps the makefile pass the correct list of files.

  The results, which go to standard output or to a file specified with -a or -o
  (-a to append, -o to start from nothing), are quoted strings wrapped in
  gettext(...).  The results can be passed to xgettext to produce a .po message
  file.
*/

#include <stdio.h>
#include <string.h>

#define LINESIZE 256
#define GET_LINE	fgets (line, LINESIZE, infile)
#define CHECK_EOL(p)	if (*(p) == '\0')  (p) = GET_LINE
#define SKIP_BLANKS(p)	while ((*p) == ' ' || (*p) == '\t')  (p)++

enum filetype { C_FILE, LISP_FILE, INVALID_FILE };
/* some brain-dead headers define this ... */
#undef FALSE
#undef TRUE
enum boolean { FALSE, TRUE };

FILE *infile;
FILE *outfile;
char line[LINESIZE];


void scan_file (char *filename);
void process_C_file (void);
void process_Lisp_file (void);
char *copy_up_to_paren (register char *p);
char *copy_quoted_string (register char *p);
enum boolean no_interactive_prompt (register char *q);
char *skip_blanks (register char *p);


main (int argc, char *argv[])
{
  register int i;

  outfile = stdout;

  /* If first two args are -o FILE, output to FILE. */
  i = 1;
  if (argc > i + 1 && strcmp (argv[i], "-o") == 0) {
    outfile = fopen (argv[++i], "w");
    ++i;
  }
  /* ...Or if args are -a FILE, append to FILE. */
  if (argc > i + 1 && strcmp (argv[i], "-a") == 0) {
    outfile = fopen (argv[++i], "a");
    ++i;
  }
  if (!outfile) {
    fprintf (stderr, "Unable to open output file %s\n", argv[--i]);
    return;
  }

  for (; i < argc; i++)
    scan_file (argv[i]);

  return 0;
}


void scan_file (char *filename)
{
  enum filetype type = INVALID_FILE;
  register char *p = filename + strlen (filename);

  if (strcmp (p - 4, ".elc") == 0) {
    *--p = '\0';				/* Use .el file instead */
    type = LISP_FILE;
  } else if (strcmp (p - 3, ".el") == 0)
    type = LISP_FILE;
  else if (strcmp (p - 2, ".o") == 0) {
    *--p = 'c';					/* Use .c file instead */
    type = C_FILE;
  } else if (strcmp (p - 2, ".c") == 0)
    type = C_FILE;

  if (type == INVALID_FILE) {
    fprintf (stderr, "File %s being ignored\n", filename);
    return;
  }
  infile = fopen (filename, "r");
  if (!infile) {
    fprintf (stderr, "Unable to open input file %s\n", filename);
    return;
  }

  fprintf (outfile, "/* %s */\n", filename);
  if (type == C_FILE)
    process_C_file ();
  else
    process_Lisp_file ();
  fputc ('\n', outfile);

  fclose (infile);
}


void process_C_file (void)
{
  register char *p;
  char *gettext, *defun;

  while (p = GET_LINE) {
    gettext = strstr (p, "GETTEXT");
    defun = strstr (p, "DEFUN");
    if (gettext || defun) {
      if (gettext) {
	p = gettext;
	p += 7;			/* Skip over "GETTEXT" */
      }
      else if (defun) {
	p = defun;
	p += 5;			/* Skip over "DEFUN" */
      }

      p = skip_blanks (p);
      if (*p++ != '(')
	continue;

      if (defun) {
	register int i;

	for (i = 0; i < 5; i++)	/* Skip over commas to doc string */
	  while (*p++ != ',')
	    CHECK_EOL (p);
	if (*p == '\n')
	  p = GET_LINE;
      }

      p = skip_blanks (p);
      if (*p != '\"')		/* Make sure there is a quoted string */
	continue;

      if (defun && no_interactive_prompt (p))
	continue;

      fprintf (outfile, "gettext(");
      if (gettext)
	p = copy_up_to_paren (p);
      else
	p = copy_quoted_string (p);
      fprintf (outfile, ")\n");
    }
  }
}


void process_Lisp_file (void)
{
  register char *p;
  char *gettext, *interactive;
  enum boolean dgettext = FALSE;

  while (p = GET_LINE) {
    gettext = strstr (p, "gettext");
    interactive = strstr (p, "(interactive");
    if (gettext || interactive) {
      if (!interactive)
	p = gettext;
      else if (!gettext)
	p = interactive;
      else if (gettext < interactive) {
	p = gettext;
	interactive = NULL;
      } else {
	p = interactive;
	gettext = NULL;
      }

      if (gettext) {
	if (p > line && *(p-1) == 'd')
	  dgettext = TRUE;
	p += 7;		/* Skip over "gettext" */
      } else
	p += 12;	/* Skip over "(interactive" */

      p = skip_blanks (p);
      if (*p != '\"')		/* Make sure there is a quoted string */
	continue;

      if (dgettext) {		/* Skip first quoted string (domain name) */
	while (*++p != '"')
	  ;  /* null statement */
	++p;
	p = skip_blanks (p);
	if (*p != '\"')		/* Check for second quoted string (message) */
	  continue;
      }

      if (interactive && no_interactive_prompt (p))
	continue;

      fprintf (outfile, "gettext(");
      p = copy_up_to_paren (p);
      fprintf (outfile, ")\n");
    }
  }
}


/* Assuming p points to some character beyond an opening parenthesis, copy
   everything to outfile up to but not including the closing parenthesis.
*/
char *copy_up_to_paren (register char *p)
{
  for (;;) {
    SKIP_BLANKS (p);	/* We don't call skip_blanks() in order to */
    CHECK_EOL (p);	/* preserve blanks at the beginning of the line */
    if (*p == ')')
      break;

    if (*p == '\"')
      p = copy_quoted_string (p);
    else
      fputc (*p++, outfile);
  }
  return p;
}


/* Assuming p points to a quote character, copy the quoted string to outfile.
*/
char *copy_quoted_string (register char *p)
{
  do {
    if (*p == '\\')
      fputc (*p++, outfile);
    fputc (*p++, outfile);
    CHECK_EOL (p);
  } while (*p != '\"');

  fputc (*p++, outfile);
  return p;
}


/* Return TRUE if the interactive specification consists only
   of code letters and no prompt.
*/
enum boolean no_interactive_prompt (register char *q)
{
  while (++q, *q == '*' || *q == '@')
    ; /* null statement */
  if (*q == '\"')
    return TRUE;
 skip_code_letter:
  if (*++q == '\"')
    return TRUE;
  if (*q == '\\' && *++q == 'n') {
    ++q;
    goto skip_code_letter;
  }
  return FALSE;
}


char *skip_blanks (register char *p)
{
  while (*p == ' ' || *p == '\t' || *p == '\n') {
    p++;
    CHECK_EOL (p);
  }
  return p;
}