about summary refs log tree commit diff
path: root/locale/programs
diff options
context:
space:
mode:
authorUlrich Drepper <drepper@redhat.com>2000-04-07 02:38:44 +0000
committerUlrich Drepper <drepper@redhat.com>2000-04-07 02:38:44 +0000
commita0dc52061fd8d47c37bbd363533ae0a4307e507c (patch)
tree7b44e3c280a8c39bc7f4c13b426eee8a40dda66d /locale/programs
parent372f94ba46f2b64aa2d99085112816789db3fe41 (diff)
downloadglibc-a0dc52061fd8d47c37bbd363533ae0a4307e507c.tar.gz
glibc-a0dc52061fd8d47c37bbd363533ae0a4307e507c.tar.xz
glibc-a0dc52061fd8d47c37bbd363533ae0a4307e507c.zip
Update.
2000-04-06  Ulrich Drepper  <drepper@redhat.com>

	* locale/programs/charmap.c (charmap_new_char): Add parameter step.
	Support ..(2).. ellipsis.
	(parse_charmap): Recognize ..(2).. etc and pass step down.
	Correctly generate names for UCS4 characters.

	* locale/programs/ld-ctype.c (struct translit_ignore_t): Add step.
	(ctype_finish): We know the wide character value for <SP>,
	don't search.
	(charclass_symbolic_ellipsis): Handle ..(2).. ellipsis.
	(charclass_ucs4_ellipsis): Likewise.
	(read_translit_ignore_entry): Store ellipsis step.
	(ctype_read): Recognize ..(2).. etc and pass step down.

	* locale/programs/linereader.c (lr_token): When seeing comment
	character ignore only rest of line in sources but stop at escaped
	newline.
	Recognize ..(2).. and ....(2).....

	* locale/programs/locfile-token.h (enum token_t): Add tok_ellipsis2_2
	and tok_ellipsis4_2.
Diffstat (limited to 'locale/programs')
-rw-r--r--locale/programs/charmap.c32
-rw-r--r--locale/programs/ld-ctype.c69
-rw-r--r--locale/programs/linereader.c26
-rw-r--r--locale/programs/locfile-token.h4
4 files changed, 98 insertions, 33 deletions
diff --git a/locale/programs/charmap.c b/locale/programs/charmap.c
index 37047a0407..3ccebcac48 100644
--- a/locale/programs/charmap.c
+++ b/locale/programs/charmap.c
@@ -53,7 +53,7 @@ static void new_width (struct linereader *cmfile, struct charmap_t *result,
 		       unsigned long int width);
 static void charmap_new_char (struct linereader *lr, struct charmap_t *cm,
 			      int nbytes, char *bytes, const char *from,
-			      const char *to, int decimal_ellipsis);
+			      const char *to, int decimal_ellipsis, int step);
 
 
 struct charmap_t *
@@ -225,6 +225,7 @@ parse_charmap (struct linereader *cmfile)
   char *from_name = NULL;
   char *to_name = NULL;
   enum token_t ellipsis = 0;
+  int step = 1;
 
   /* We don't want symbolic names in string to be translated.  */
   cmfile->translate_strings = 0;
@@ -461,7 +462,7 @@ character sets with locking states are not supported"));
 						now->val.str.lenmb);
 	  else
 	    {
-	      obstack_printf (&result->mem_pool, "<%08X>",
+	      obstack_printf (&result->mem_pool, "U%08X",
 			      cmfile->token.val.ucs4);
 	      obstack_1grow (&result->mem_pool, '\0');
 	      from_name = (char *) obstack_finish (&result->mem_pool);
@@ -475,9 +476,20 @@ character sets with locking states are not supported"));
 	  /* We have two possibilities: We can see an ellipsis or an
 	     encoding value.  */
 	  if (nowtok == tok_ellipsis3 || nowtok == tok_ellipsis4
-	      || nowtok == tok_ellipsis2)
+	      || nowtok == tok_ellipsis2 || nowtok == tok_ellipsis4_2
+	      || nowtok == tok_ellipsis2_2)
 	    {
 	      ellipsis = nowtok;
+	      if (nowtok == tok_ellipsis4_2)
+		{
+		  step = 2;
+		  nowtok = tok_ellipsis4;
+		}
+	      else if (nowtok == tok_ellipsis2_2)
+		{
+		  step = 2;
+		  nowtok = tok_ellipsis2;
+		}
 	      state = 4;
 	      continue;
 	    }
@@ -502,13 +514,15 @@ character sets with locking states are not supported"));
 	  else
 	    charmap_new_char (cmfile, result, now->val.charcode.nbytes,
 			      now->val.charcode.bytes, from_name, to_name,
-			      ellipsis != tok_ellipsis2);
+			      ellipsis != tok_ellipsis2, step);
 
 	  /* Ignore trailing comment silently.  */
 	  lr_ignore_rest (cmfile, 0);
 
 	  from_name = NULL;
 	  to_name = NULL;
+	  ellipsis = tok_none;
+	  step = 1;
 
 	  state = 2;
 	  continue;
@@ -531,7 +545,7 @@ character sets with locking states are not supported"));
 					      cmfile->token.val.str.lenmb);
 	  else
 	    {
-	      obstack_printf (&result->mem_pool, "<%08X>",
+	      obstack_printf (&result->mem_pool, "U%08X",
 			      cmfile->token.val.ucs4);
 	      obstack_1grow (&result->mem_pool, '\0');
 	      to_name = (char *) obstack_finish (&result->mem_pool);
@@ -814,7 +828,7 @@ charmap_find_value (const struct charmap_t *cm, const char *name, size_t len)
 static void
 charmap_new_char (struct linereader *lr, struct charmap_t *cm,
 		  int nbytes, char *bytes, const char *from, const char *to,
-		  int decimal_ellipsis)
+		  int decimal_ellipsis, int step)
 {
   hash_table *ht = &cm->char_table;
   hash_table *bt = &cm->byte_table;
@@ -833,7 +847,7 @@ charmap_new_char (struct linereader *lr, struct charmap_t *cm,
       newp = (struct charseq *) obstack_alloc (ob, sizeof (*newp) + nbytes);
       newp->nbytes = nbytes;
       memcpy (newp->bytes, bytes, nbytes);
-      newp->name = obstack_copy (ob, from, len1 + 1);
+      newp->name = from;
 
       newp->ucs4 = UNINITIALIZED_CHAR_VALUE;
       if ((from[0] == 'U' || from[0] == 'P') && (len1 == 5 || len1 == 9))
@@ -852,7 +866,7 @@ charmap_new_char (struct linereader *lr, struct charmap_t *cm,
 	  char *endp;
 
 	  errno = 0;
-	  newp->ucs4 = strtoul (from, &endp, 16);
+	  newp->ucs4 = strtoul (from + 1, &endp, 16);
 	  if (endp - from != len1
 	      || (newp->ucs4 == ULONG_MAX && errno == ERANGE)
 	      || newp->ucs4 >= 0x80000000)
@@ -916,7 +930,7 @@ hexadecimal range format should use only capital characters"));
       return;
     }
 
-  for (cnt = from_nr; cnt <= to_nr; ++cnt)
+  for (cnt = from_nr; cnt <= to_nr; cnt += step)
     {
       char *name_end;
       obstack_printf (ob, decimal_ellipsis ? "%.*s%0*d" : "%.*s%0*X",
diff --git a/locale/programs/ld-ctype.c b/locale/programs/ld-ctype.c
index 23ca2368c8..9cf4d2c5f5 100644
--- a/locale/programs/ld-ctype.c
+++ b/locale/programs/ld-ctype.c
@@ -97,6 +97,7 @@ struct translit_ignore_t
 {
   uint32_t from;
   uint32_t to;
+  uint32_t step;
 
   const char *fname;
   size_t lineno;
@@ -504,18 +505,13 @@ character '%s' in class `%s' must not be in class `%s'"),
     }
 
   /* ... and now test <SP> as a special case.  */
-  space_value = repertoire_find_value (ctype->repertoire, "SP", 2);
-  if (space_value == ILLEGAL_CHAR_VALUE)
-    {
-      if (!be_quiet)
-	error (0, 0, _("character <SP> not defined in character map"));
-    }
-  else if (((cnt = BITPOS (tok_space),
-	     (ELEM (ctype, class_collection, , space_value)
-	      & BITw (tok_space)) == 0)
-	    || (cnt = BITPOS (tok_blank),
-		(ELEM (ctype, class_collection, , space_value)
-		 & BITw (tok_blank)) == 0)))
+  space_value = 32;
+  if (((cnt = BITPOS (tok_space),
+	(ELEM (ctype, class_collection, , space_value)
+	 & BITw (tok_space)) == 0)
+       || (cnt = BITPOS (tok_blank),
+	   (ELEM (ctype, class_collection, , space_value)
+	    & BITw (tok_blank)) == 0)))
     {
       if (!be_quiet)
 	error (0, 0, _("<SP> character not in class `%s'"),
@@ -1236,7 +1232,8 @@ get_character (struct token *now, struct charmap_t *charmap,
 }
 
 
-/* Ellipsis like in `<foo123>..<foo12a>' or `<j1234>....<j1245>'.  */
+/* Ellipsis like in `<foo123>..<foo12a>' or `<j1234>....<j1245>' and
+   the .(2). counterparts.  */
 static void
 charclass_symbolic_ellipsis (struct linereader *ldfile,
 			     struct locale_ctype_t *ctype,
@@ -1246,7 +1243,7 @@ charclass_symbolic_ellipsis (struct linereader *ldfile,
 			     const char *last_str,
 			     unsigned long int class256_bit,
 			     unsigned long int class_bit, int base,
-			     int ignore_content, int handle_digits)
+			     int ignore_content, int handle_digits, int step)
 {
   const char *nowstr = now->val.str.startmb;
   char tmp[now->val.str.lenmb + 1];
@@ -1288,7 +1285,7 @@ charclass_symbolic_ellipsis (struct linereader *ldfile,
   if (!ignore_content)
     {
       now->val.str.startmb = tmp;
-      while (++from <= to)
+      while ((from += step) <= to)
 	{
 	  struct charseq *seq;
 	  uint32_t wch;
@@ -1346,7 +1343,7 @@ charclass_symbolic_ellipsis (struct linereader *ldfile,
 }
 
 
-/* Ellipsis like in `<U1234>..<U2345>'.  */
+/* Ellipsis like in `<U1234>..<U2345>' or `<U1234>..(2)..<U2345>'.  */
 static void
 charclass_ucs4_ellipsis (struct linereader *ldfile,
 			 struct locale_ctype_t *ctype,
@@ -1355,7 +1352,7 @@ charclass_ucs4_ellipsis (struct linereader *ldfile,
 			 struct token *now, uint32_t last_wch,
 			 unsigned long int class256_bit,
 			 unsigned long int class_bit, int ignore_content,
-			 int handle_digits)
+			 int handle_digits, int step)
 {
   if (last_wch > now->val.ucs4)
     {
@@ -1367,7 +1364,7 @@ to-value <U%0*X> of range is smaller than from-value <U%0*X>"),
     }
 
   if (!ignore_content)
-    while (++last_wch <= now->val.ucs4)
+    while ((last_wch += step) <= now->val.ucs4)
       {
 	/* We have to find out whether there is a byte sequence corresponding
 	   to this UCS4 value.  */
@@ -1376,6 +1373,11 @@ to-value <U%0*X> of range is smaller than from-value <U%0*X>"),
 
 	snprintf (utmp, sizeof (utmp), "U%08X", last_wch);
 	seq = charmap_find_value (charmap, utmp, 9);
+	if (seq == NULL)
+	  {
+	    snprintf (utmp, sizeof (utmp), "U%04X", last_wch);
+	    seq = charmap_find_value (charmap, utmp, 5);
+	  }
 
 	if (seq == NULL)
 	  /* Try looking in the repertoire map.  */
@@ -1779,6 +1781,7 @@ read_translit_ignore_entry (struct linereader *ldfile,
 	    obstack_alloc (&ctype->mempool, sizeof (struct translit_ignore_t));
 	  newp->from = from;
 	  newp->to = from;
+	  newp->step = 1;
 
 	  newp->next = ctype->translit_ignore;
 	  ctype->translit_ignore = newp;
@@ -1788,11 +1791,12 @@ read_translit_ignore_entry (struct linereader *ldfile,
 	 line.  */
       now = lr_token (ldfile, charmap, repertoire);
 
-      if (now->tok == tok_ellipsis2)
+      if (now->tok == tok_ellipsis2 || now->tok == tok_ellipsis2_2)
 	{
 	  /* XXX Should we bother implementing `....'?  `...' certainly
 	     will not be implemented.  */
 	  uint32_t to;
+	  int step = now->tok == tok_ellipsis2_2 ? 2 : 1;
 
 	  now = lr_token (ldfile, charmap, repertoire);
 
@@ -1823,7 +1827,10 @@ read_translit_ignore_entry (struct linereader *ldfile,
 	    {
 	      /* Make sure the `to'-value is larger.  */
 	      if (to >= from)
-		newp->to = to;
+		{
+		  newp->to = to;
+		  newp->step = step;
+		}
 	      else
 		lr_error (ldfile, _("\
 to-value <U%0*X> of range is smaller than from-value <U%0*X>"),
@@ -1866,6 +1873,7 @@ ctype_read (struct linereader *ldfile, struct localedef_t *result,
   uint32_t last_wch = 0;
   enum token_t last_token;
   enum token_t ellipsis_token;
+  int step;
   char last_charcode[16];
   size_t last_charcode_len = 0;
   const char *last_str = NULL;
@@ -2040,6 +2048,7 @@ ctype_read (struct linereader *ldfile, struct localedef_t *result,
 	  ctype->class_done |= class_bit;
 	  last_token = tok_none;
 	  ellipsis_token = tok_none;
+	  step = 1;
 	  now = lr_token (ldfile, charmap, NULL);
 	  while (now->tok != tok_eol && now->tok != tok_eof)
 	    {
@@ -2140,7 +2149,7 @@ the absolute ellipsis `...' must not be used"));
 						    == tok_ellipsis4
 						    ? 10 : 16),
 						   ignore_content,
-						   handle_digits);
+						   handle_digits, step);
 		    }
 		  else if (last_token == tok_ucs4)
 		    {
@@ -2151,7 +2160,8 @@ with UCS range values one must use the hexadecimal symbolic ellipsis `..'"));
 		      charclass_ucs4_ellipsis (ldfile, ctype, charmap,
 					       repertoire, now, last_wch,
 					       class256_bit, class_bit,
-					       ignore_content, handle_digits);
+					       ignore_content, handle_digits,
+					       step);
 		    }
 		  else
 		    {
@@ -2180,9 +2190,21 @@ with character code range values one must use the absolute ellipsis `...'"));
 		break;
 
 	      if (last_token != tok_none
-		  && now->tok >= tok_ellipsis2 && now->tok <= tok_ellipsis4)
+		  && now->tok >= tok_ellipsis2 && now->tok <= tok_ellipsis4_2)
 		{
+		  if (now->tok == tok_ellipsis2_2)
+		    {
+		      now->tok = tok_ellipsis2;
+		      step = 2;
+		    }
+		  else if (now->tok == tok_ellipsis4_2)
+		    {
+		      now->tok = tok_ellipsis4;
+		      step = 2;
+		    }
+
 		  ellipsis_token = now->tok;
+
 		  now = lr_token (ldfile, charmap, NULL);
 		  continue;
 		}
@@ -2194,6 +2216,7 @@ with character code range values one must use the absolute ellipsis `...'"));
 	      now = lr_token (ldfile, charmap, NULL);
 
 	      ellipsis_token = tok_none;
+	      step = 1;
 	    }
 	  break;
 
diff --git a/locale/programs/linereader.c b/locale/programs/linereader.c
index f6532a4792..36dd0cd2d0 100644
--- a/locale/programs/linereader.c
+++ b/locale/programs/linereader.c
@@ -185,6 +185,16 @@ lr_token (struct linereader *lr, const struct charmap_t *charmap,
       if (ch != lr->comment_char)
 	break;
 
+      /* Is there an newline at the end of the buffer?  */
+      if (lr->buf[lr->bufact - 1] != '\n')
+	{
+	  /* No.  Some people want this to mean that only the line in
+	     the file not the logical, concatenated line is ignored.
+	     Let's try this.  */
+	  lr->idx = lr->bufact;
+	  continue;
+	}
+
       /* Ignore rest of line.  */
       lr_ignore_rest (lr, 0);
       lr->token.tok = tok_eol;
@@ -198,6 +208,14 @@ lr_token (struct linereader *lr, const struct charmap_t *charmap,
   /* Match ellipsis.  */
   if (ch == '.')
     {
+      if (strncmp (&lr->buf[lr->idx], "...(2)....", 10) == 0)
+	{
+	  int cnt;
+	  for (cnt = 0; cnt < 10; ++cnt)
+	    lr_getc (lr);
+	  lr->token.tok = tok_ellipsis4_2;
+	  return &lr->token;
+	}
       if (strncmp (&lr->buf[lr->idx], "...", 3) == 0)
 	{
 	  lr_getc (lr);
@@ -213,6 +231,14 @@ lr_token (struct linereader *lr, const struct charmap_t *charmap,
 	  lr->token.tok = tok_ellipsis3;
 	  return &lr->token;
 	}
+      if (strncmp (&lr->buf[lr->idx], ".(2)..", 6) == 0)
+	{
+	  int cnt;
+	  for (cnt = 0; cnt < 6; ++cnt)
+	    lr_getc (lr);
+	  lr->token.tok = tok_ellipsis2_2;
+	  return &lr->token;
+	}
       if (lr->buf[lr->idx] == '.')
 	{
 	  lr_getc (lr);
diff --git a/locale/programs/locfile-token.h b/locale/programs/locfile-token.h
index e1cd5f7a74..6eecc5618a 100644
--- a/locale/programs/locfile-token.h
+++ b/locale/programs/locfile-token.h
@@ -1,4 +1,4 @@
-/* Copyright (C) 1996, 1997, 1998, 1999 Free Software Foundation, Inc.
+/* Copyright (C) 1996, 1997, 1998, 1999, 2000 Free Software Foundation, Inc.
    This file is part of the GNU C Library.
    Contributed by Ulrich Drepper <drepper@gnu.org>, 1996.
 
@@ -31,6 +31,8 @@ enum token_t
   tok_ellipsis2,
   tok_ellipsis3,
   tok_ellipsis4,
+  tok_ellipsis2_2,
+  tok_ellipsis4_2,
   tok_semicolon,
   tok_comma,
   tok_open_brace,