summary refs log tree commit diff
path: root/intl/dcgettext.c
diff options
context:
space:
mode:
authorRoland McGrath <roland@gnu.org>1995-09-28 18:42:29 +0000
committerRoland McGrath <roland@gnu.org>1995-09-28 18:42:29 +0000
commit24906b43b9bd9108aef17d18dfbb8764212085f7 (patch)
tree48d641dc176c9941b3b036e66c383b00a6974bad /intl/dcgettext.c
parent91f62ce6b5797f04006ac5aff08416ecd61bf972 (diff)
downloadglibc-24906b43b9bd9108aef17d18dfbb8764212085f7.tar.gz
glibc-24906b43b9bd9108aef17d18dfbb8764212085f7.tar.xz
glibc-24906b43b9bd9108aef17d18dfbb8764212085f7.zip
Thu Sep 28 13:05:54 1995 Roland McGrath <roland@churchy.gnu.ai.mit.edu>
	Merge new message handling code from GNU gettext, by Drepper.
	* intl: New directory.
	* Makefile (subdirs): Add intl.

	* sysdeps/generic/dl-sysdep.c (_dl_sysdep_start): Return
	USER_ENTRY instead of storing it on our stack.

	* elf/rtld.c (rtld_command): Variable removed.
	(_dl_skip_args): New variable.
	(dl_main): Increment _dl_skip_args instead of setting rtld_command.
	If the link_map for the executable itself is not first in the chain,
	make it so.
	* sysdeps/i386/dl-machine.h (RTLD_START): Use _dl_skip_args as
	count of args to skip.

Thu Sep 28 09:20:04 1995  Ulrich Drepper  <drepper@gnu.ai.mit.edu>

	* stdlib/strtod.c (STRTOF): Fix handling of numbers with lots of
	leading zeroes. 
	
Diffstat (limited to 'intl/dcgettext.c')
-rw-r--r--intl/dcgettext.c523
1 files changed, 523 insertions, 0 deletions
diff --git a/intl/dcgettext.c b/intl/dcgettext.c
new file mode 100644
index 0000000000..91025be89b
--- /dev/null
+++ b/intl/dcgettext.c
@@ -0,0 +1,523 @@
+/* dcgettext.c -- implemenatation of the dcgettext(3) function
+   Copyright (C) 1995 Software Foundation, Inc.
+
+This program is free software; you can redistribute it and/or modify
+it under the terms of the GNU General Public License as published by
+the Free Software Foundation; either version 2, or (at your option)
+any later version.
+
+This program is distributed in the hope that it will be useful,
+but WITHOUT ANY WARRANTY; without even the implied warranty of
+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+GNU General Public License for more details.
+
+You should have received a copy of the GNU General Public License
+along with this program; if not, write to the Free Software
+Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
+
+#ifdef HAVE_CONFIG_H
+# include <config.h>
+#endif
+
+#include <sys/types.h>
+
+#ifdef __GNUC__
+# define alloca __builtin_alloca
+#else
+# ifdef HAVE_ALLOCA_H || defined _LIBC
+#  include <alloca.h>
+# else
+#  ifdef _AIX
+ #pragma alloca
+#  else
+#   ifndef alloca
+char *alloca ();
+#   endif
+#  endif
+# endif
+#endif
+
+#include <errno.h>
+#ifndef errno
+extern int errno;
+#endif
+
+#if defined STDC_HEADERS || defined _LIBC
+# include <stdlib.h>
+#else
+char *getenv ();
+# ifdef HAVE_MALLOC_H
+#  include <malloc.h>
+# else
+void free ();
+# endif
+#endif
+
+#if defined HAVE_STRING_H || defined _LIBC
+# include <string.h>
+#else
+# include <strings.h>
+#endif
+#if !HAVE_STRCHR && !defined _LIBC
+# ifndef strchr
+#  define strchr index
+# endif
+#endif
+
+#if defined HAVE_UNISTD_H || defined _LIBC
+# include <unistd.h>
+#endif
+
+#include "gettext.h"
+#include "gettextP.h"
+#ifdef _LIBC
+# include <libintl.h>
+#else
+# include "libgettext.h"
+#endif
+#include "hash-string.h"
+
+/* @@ end of prolog @@ */
+
+#ifdef _LIBC
+/* Rename the non ANSI C functions.  This is required by the standard
+   because some ANSI C functions will require linking with this object
+   file and the name space must not be polluted.  */
+# define getcwd __getcwd
+# define stpcpy __stpcpy
+#endif
+
+#if !defined HAVE_GETCWD && !defined _LIBC
+char *getwd ();
+# define getcwd(buf, max) getwd (buf)
+#else
+char *getcwd ();
+#endif
+
+/* Amount to increase buffer size by in each try.  */
+#define PATH_INCR 32
+
+/* The following is from pathmax.h.  */
+/* Non-POSIX BSD systems might have gcc's limits.h, which doesn't define
+   PATH_MAX but might cause redefinition warnings when sys/param.h is
+   later included (as on MORE/BSD 4.3).  */
+#if defined(_POSIX_VERSION) || (defined(HAVE_LIMITS_H) && !defined(__GNUC__))
+# include <limits.h>
+#endif
+
+#ifndef _POSIX_PATH_MAX
+# define _POSIX_PATH_MAX 255
+#endif
+
+#if !defined(PATH_MAX) && defined(_PC_PATH_MAX)
+# define PATH_MAX (pathconf ("/", _PC_PATH_MAX) < 1 ? 1024 : pathconf ("/", _PC_PATH_MAX))
+#endif
+
+/* Don't include sys/param.h if it already has been.  */
+#if defined(HAVE_SYS_PARAM_H) && !defined(PATH_MAX) && !defined(MAXPATHLEN)
+# include <sys/param.h>
+#endif
+
+#if !defined(PATH_MAX) && defined(MAXPATHLEN)
+# define PATH_MAX MAXPATHLEN
+#endif
+
+#ifndef PATH_MAX
+# define PATH_MAX _POSIX_PATH_MAX
+#endif
+
+/* XPG3 defines the result of `setlocale (category, NULL)' as:
+   ``Directs `setlocale()' to query `category' and return the current
+     setting of `local'.''
+   However it does not specify the exact format.  And even worse: POSIX
+   defines this not at all.  So we can use this feature only on selected
+   system (e.g. those using GNU C Library).  */
+#ifdef _LIBC
+# define HAVE_LOCALE_NULL
+#endif
+
+/* Name of the default domain used for gettext(3) prior any call to
+   textdomain(3).  The default value for this is "messages".  */
+const char _nl_default_default_domain[] = "messages";
+
+/* Value used as the default domain for gettext(3).  */
+const char *_nl_current_default_domain = _nl_default_default_domain;
+
+/* Contains the default location of the message catalogs.  */
+const char _nl_default_dirname[] = GNULOCALEDIR;
+
+/* List with bindings of specific domains created by bindtextdomain()
+   calls.  */
+struct binding *_nl_domain_bindings;
+
+/* Prototypes for local functions.  */
+static char *find_msg __P ((struct loaded_domain *domain, const char *msgid));
+static const char *category_to_name __P((int category));
+static const char *guess_category_value __P((int category,
+					     const char *categoryname));
+
+
+/* Names for the libintl functions are a problem.  They must not clash
+   with existing names and they should follow ANSI C.  But this source
+   code is also used in GNU C Library where the names have a __
+   prefix.  So we have to make a difference here.  */
+#ifdef _LIBC
+# define DCGETTEXT __dcgettext
+#else
+# define DCGETTEXT dcgettext__
+#endif
+
+/* Look up MSGID in the DOMAINNAME message catalog for the current CATEGORY
+   locale.  */
+char *
+DCGETTEXT (domainname, msgid, category)
+     const char *domainname;
+     const char *msgid;
+     int category;
+{
+  struct loaded_domain *domain;
+  struct binding *binding;
+  const char *categoryname;
+  const char *categoryvalue;
+  char *dirname, *xdomainname;
+  char *single_locale;
+  char *retval;
+
+  /* If no real MSGID is given return NULL.  */
+  if (msgid == NULL)
+    return NULL;
+
+  /* If DOMAINNAME is NULL, we are interested in the default domain.  If
+     CATEGORY is not LC_MESSAGES this might not make much sense but the
+     defintion left this undefined.  */
+  if (domainname == NULL)
+    domainname = _nl_current_default_domain;
+
+  /* First find matching binding.  */
+  for (binding = _nl_domain_bindings; binding != NULL; binding = binding->next)
+    {
+      int compare = strcmp (domainname, binding->domainname);
+      if (compare == 0)
+	/* We found it!  */
+	break;
+      if (compare < 0)
+	{
+	  /* It is not in the list.  */
+	  binding = NULL;
+	  break;
+	}
+    }
+
+  if (binding == NULL)
+    dirname = (char *) _nl_default_dirname;
+  else if (binding->dirname[0] == '/')
+    dirname = binding->dirname;
+  else
+    {
+      /* We have a relative path.  Make it absolute now.  */
+      size_t dirname_len = strlen (binding->dirname) + 1;
+      size_t path_max;
+      char *ret;
+
+      path_max = (unsigned) PATH_MAX;
+      path_max += 2;		/* The getcwd docs say to do this.  */
+
+      dirname = (char *) alloca (path_max + dirname_len);
+
+      errno = 0;
+      while ((ret = getcwd (dirname, path_max)) == NULL && errno == ERANGE)
+	{
+	  path_max += PATH_INCR;
+	  dirname = (char *) alloca (path_max + dirname_len);
+	  errno = 0;
+	}
+
+      if (ret == NULL)
+	/* We cannot get the current working directory.  Don't signal an
+	   error but simply return the default string.  */
+	return (char *) msgid;
+
+      /* We don't want libintl.a to depend on any other library.  So
+	 we avoid the non-standard function stpcpy.  In GNU C Library
+	 this function is available, though.  Also allow the symbol
+	 HAVE_STPCPY to be defined.  */
+#if defined _LIBC || defined HAVE_STPCPY
+      stpcpy (stpcpy (strchr (dirname, '\0'), "/"), binding->dirname);
+#else
+      strcat (dirname, "/");
+      strcat (dirname, binding->dirname);
+#endif
+    }
+
+  /* Now determine the symbolic name of CATEGORY and its value.  */
+  categoryname = category_to_name (category);
+  categoryvalue = guess_category_value (category, categoryname);
+
+  xdomainname = (char *) alloca (strlen (categoryname)
+				 + strlen (domainname) + 5);
+  /* We don't want libintl.a to depend on any other library.  So we
+     avoid the non-standard function stpcpy.  In GNU C Library this
+     function is available, though.  Also allow the symbol HAVE_STPCPY
+     to be defined.  */
+#if defined _LIBC || defined HAVE_STPCPY
+  stpcpy (stpcpy (stpcpy (stpcpy (xdomainname, categoryname), "/"),
+		  domainname),
+	  ".mo");
+#else
+  strcpy (xdomainname, categoryname);
+  strcat (xdomainname, "/");
+  strcat (xdomainname, domainname);
+  strcat (xdomainname, ".mo");
+#endif
+
+  /* Creating working area.  */
+  single_locale = (char *) alloca (strlen (categoryvalue) + 1);
+
+
+  /* Search for the given string.  This is a loop because we perhaps
+     got an ordered list of languages to consider for th translation.  */
+  while (1)
+    {
+      /* Make CATEGORYVALUE point to the next element of the list.  */
+      while (categoryvalue[0] != '\0' && categoryvalue[0] == ':')
+	++categoryvalue;
+      if (categoryvalue[0] == '\0')
+	{
+	  /* The whole contents of CATEGORYVALUE has been searched but
+	     no valid entry has been found.  We solve this situation
+	     by implicitely appending a "C" entry, i.e. no translation
+	     will take place.  */
+	  single_locale[0] = 'C';
+	  single_locale[1] = '\0';
+	}
+      else
+	{
+	  char *cp = single_locale;
+	  while (categoryvalue[0] != '\0' && categoryvalue[0] != ':')
+	    *cp++ = *categoryvalue++;
+	  *cp = '\0';
+	}
+
+      /* If the current locale value is C (or POSIX) we don't load a
+	 domain.  Return the MSGID.  */
+      if (strcmp (single_locale, "C") == 0
+	  || strcmp (single_locale, "POSIX") == 0)
+	return (char *) msgid;
+
+
+      /* Find structure describing the message catalog matching the
+	 DOMAINNAME and CATEGORY.  */
+      domain = _nl_find_domain (dirname, single_locale, xdomainname);
+
+      if (domain != NULL)
+	{
+	  retval = find_msg (domain, msgid);
+
+	  if (retval == NULL)
+	    {
+	      int cnt;
+
+	      for (cnt = 6; cnt >= 0 && retval == NULL; --cnt)
+		if (domain->successor[cnt] != NULL)
+		  {
+		    retval = find_msg (domain->successor[cnt], msgid);
+
+ 		    if (domain->successor[cnt]->data == NULL)
+		      domain->successor[cnt] = NULL;
+		  }
+	    }
+
+	  if (retval != NULL)
+	    return retval;
+	}
+    }
+  /* NOTREACHED */
+}
+
+#ifdef _LIBC
+/* Alias for function name in GNU C Library.  */
+weak_alias (__dcgettext, dcgettext);
+#endif
+
+
+static char *
+find_msg (domain, msgid)
+     struct loaded_domain *domain;
+     const char *msgid;
+{
+  size_t top, act, bottom;
+
+  if (domain->decided == 0)
+    _nl_load_domain (domain);
+
+  if (domain->data == NULL)
+    return NULL;
+
+  /* Locate the MSGID and its translation.  */
+  if (domain->hash_size > 2 && domain->hash_tab != NULL)
+    {
+      /* Use the hashing table.  */
+      nls_uint32 len = strlen (msgid);
+      nls_uint32 hash_val = hash_string (msgid);
+      nls_uint32 idx = hash_val % domain->hash_size;
+      nls_uint32 incr = 1 + (hash_val % (domain->hash_size - 2));
+      nls_uint32 nstr = W (domain->must_swap, domain->hash_tab[idx]);
+
+      if (nstr == 0)
+	/* Hash table entry is empty.  */
+	return NULL;
+
+      if (W (domain->must_swap, domain->orig_tab[nstr - 1].length) == len
+	  && strcmp (msgid,
+		     domain->data + W (domain->must_swap,
+				       domain->orig_tab[nstr - 1].offset)) == 0)
+	return (char *) domain->data + W (domain->must_swap,
+					  domain->trans_tab[nstr - 1].offset);
+
+      while (1)
+	{
+	  if (idx >= W (domain->must_swap, domain->hash_size) - incr)
+	    idx -= W (domain->must_swap, domain->hash_size) - incr;
+	  else
+	    idx += incr;
+
+	  nstr = W (domain->must_swap, domain->hash_tab[idx]);
+	  if (nstr == 0)
+	    /* Hash table entry is empty.  */
+	    return NULL;
+
+	  if (W (domain->must_swap, domain->orig_tab[nstr - 1].length) == len
+	      && strcmp (msgid,
+			 domain->data + W (domain->must_swap,
+					   domain->orig_tab[nstr - 1].offset))
+	         == 0)
+	    return (char *) domain->data
+	      + W (domain->must_swap, domain->trans_tab[nstr - 1].offset);
+	}
+      /* NOTREACHED */
+    }
+
+  /* Now we try the default method:  binary search in the sorted
+     array of messages.  */
+  bottom = 0;
+  top = domain->nstrings;
+  while (bottom < top)
+    {
+      int cmp_val;
+
+      act = (bottom + top) / 2;
+      cmp_val = strcmp (msgid, domain->data
+			       + W (domain->must_swap,
+				    domain->orig_tab[act].offset));
+      if (cmp_val < 0)
+	top = act;
+      else if (cmp_val > 0)
+	bottom = act + 1;
+      else
+	break;
+    }
+
+  /* If an translation is found return this.  */
+  return bottom >= top ? NULL : (char *) domain->data
+                                + W (domain->must_swap,
+				     domain->trans_tab[act].offset);
+}
+
+
+/* Return string representation of locale CATEGORY.  */
+static const char *category_to_name (category)
+     int category;
+{
+  const char *retval;
+
+  switch (category)
+  {
+#ifdef LC_COLLATE
+  case LC_COLLATE:
+    retval = "LC_COLLATE";
+    break;
+#endif
+#ifdef LC_CTYPE
+  case LC_CTYPE:
+    retval = "LC_CTYPE";
+    break;
+#endif
+#ifdef LC_MONETARY
+  case LC_MONETARY:
+    retval = "LC_MONETARY";
+    break;
+#endif
+#ifdef LC_NUMERIC
+  case LC_NUMERIC:
+    retval = "LC_NUMERIC";
+    break;
+#endif
+#ifdef LC_TIME
+  case LC_TIME:
+    retval = "LC_TIME";
+    break;
+#endif
+#ifdef LC_MESSAGES
+  case LC_MESSAGES:
+    retval = "LC_MESSAGES";
+    break;
+#endif
+#ifdef LC_RESPONSE
+  case LC_RESPONSE:
+    retval = "LC_RESPONSE";
+    break;
+#endif
+#ifdef LC_ALL
+  case LC_ALL:
+    /* This might not make sense but is perhaps better than any other
+       value.  */
+    retval = "LC_ALL";
+    break;
+#endif
+  default:
+    /* If you have a better idea for a default value let me know.  */
+    retval = "LC_XXX";
+  }
+
+  return retval;
+}
+
+/* Guess value of current locale from value of the environment variables.  */
+static const char *guess_category_value (category, categoryname)
+     int category;
+     const char *categoryname;
+{
+  const char *retval;
+
+  /* The highest priority value is the `LANGUAGE' environment
+     variable.  This is a GNU extension.  */
+  retval = getenv ("LANGUAGE");
+  if (retval != NULL && retval[0] != '\0')
+    return retval;
+
+  /* `LANGUAGE' is not set.  So we have to proceed with the POSIX
+     methods of looking to `LC_ALL', `LC_xxx', and `LANG'.  On some
+     systems this can be done by the `setlocale' function itself.  */
+#if defined HAVE_SETLOCALE && defined HAVE_LC_MESSAGES && defined HAVE_LOCALE_NULL
+  return setlocale (category, NULL);
+#else
+  /* Setting of LC_ALL overwrites all other.  */
+  retval = getenv ("LC_ALL");
+  if (retval != NULL && retval[0] != '\0')
+    return retval;
+
+  /* Next comes the name of the desired category.  */
+  retval = getenv (categoryname);
+  if (retval != NULL && retval[0] != '\0')
+    return retval;
+
+  /* Last possibility is the LANG environment variable.  */
+  retval = getenv ("LANG");
+  if (retval != NULL && retval[0] != '\0')
+    return retval;
+
+  /* We use C as the default domain.  POSIX says this is implementation
+     defined.  */
+  return "C";
+#endif
+}