New module 'unicase/u8-ct-casefold'.

diff --git a/ChangeLog b/ChangeLog
index c361ce9..0b3885a 100644 (file)
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,11 @@
 2009-06-29  Bruno Haible  <bruno@clisp.org>
 
+       New module 'unicase/u8-ct-casefold'.
+       * lib/unicase/u8-ct-casefold.c: New file.
+       * lib/unicase/u-ct-casefold.h: New file, derived from
+       lib/unicase/u-casefold.h.
+       * modules/unicase/u8-ct-casefold: New file.
+
        New module 'unicase/u32-ct-totitle'.
        * lib/unicase/u32-ct-totitle.c: New file.
        * modules/unicase/u32-ct-totitle: New file.

diff --git a/lib/unicase/u-ct-casefold.h b/lib/unicase/u-ct-casefold.h
new file mode 100644 (file)
index 0000000..d97be26
--- /dev/null
+++ b/lib/unicase/u-ct-casefold.h
@@ -0,0 +1,107 @@
+/* Casefolding mapping for Unicode substrings (locale dependent).
+   Copyright (C) 2009 Free Software Foundation, Inc.
+   Written by Bruno Haible <bruno@clisp.org>, 2009.
+
+   This program is free software: you can redistribute it and/or modify it
+   under the terms of the GNU Lesser General Public License as published
+   by the Free Software Foundation; either version 3 of the License, 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
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
+
+UNIT *
+FUNC (const UNIT *s, size_t n,
+      casing_prefix_context_t prefix_context,
+      casing_suffix_context_t suffix_context,
+      const char *iso639_language,
+      uninorm_t nf,
+      UNIT *resultbuf, size_t *lengthp)
+{
+  /* Implement the three definitions of caseless matching, as described in
+     Unicode 5.0, section "Default caseless matching":
+       - If no normalization is requested, simply apply the casefolding.
+           X -> toCasefold(X).
+       - If canonical normalization is requested, apply it, and apply an NFD
+         before.
+           X -> NFD(toCasefold(NFD(X))).
+       - If compatibility normalization is requested, apply it twice, apply
+         the normalization after each, and apply an NFD before:
+           X -> NFKD(toCasefold(NFKD(toCasefold(NFD(X))))).  */
+  if (nf == NULL)
+    /* X -> toCasefold(X) */
+    return U_CASEMAP (s, n, prefix_context, suffix_context, iso639_language,
+                     uc_tocasefold, offsetof (struct special_casing_rule, casefold[0]),
+                     NULL,
+                     resultbuf, lengthp);
+  else
+    {
+      uninorm_t nfd = uninorm_decomposing_form (nf);
+      /* X -> nf(toCasefold(NFD(X))) or
+        X -> nf(toCasefold(nfd(toCasefold(NFD(X)))))  */
+      int repeat = (uninorm_is_compat_decomposing (nf) ? 2 : 1);
+      UNIT tmpbuf1[2048 / sizeof (UNIT)];
+      UNIT tmpbuf2[2048 / sizeof (UNIT)];
+      UNIT *tmp1;
+      size_t tmp1_length;
+      UNIT *tmp2;
+      size_t tmp2_length;
+
+      tmp1_length = sizeof (tmpbuf1) / sizeof (UNIT);
+      tmp1 = U_NORMALIZE (UNINORM_NFD, s, n, tmpbuf1, &tmp1_length);
+      if (tmp1 == NULL)
+       /* errno is set here.  */
+       return NULL;
+
+      do
+       {
+         tmp2_length = sizeof (tmpbuf2) / sizeof (UNIT);
+         tmp2 = U_CASEMAP (tmp1, tmp1_length,
+                           prefix_context, suffix_context, iso639_language,
+                           uc_tocasefold, offsetof (struct special_casing_rule, casefold[0]),
+                           NULL,
+                           tmpbuf2, &tmp2_length);
+         if (tmp2 == NULL)
+           {
+             int saved_errno = errno;
+             if (tmp1 != tmpbuf1)
+               free (tmp1);
+             errno = saved_errno;
+             return NULL;
+           }
+
+         if (tmp1 != tmpbuf1)
+           free (tmp1);
+
+         if (repeat > 1)
+           {
+             tmp1_length = sizeof (tmpbuf1) / sizeof (UNIT);
+             tmp1 = U_NORMALIZE (nfd, tmp2, tmp2_length,
+                                 tmpbuf1, &tmp1_length);
+           }
+         else
+           /* Last run through this loop.  */
+           tmp1 = U_NORMALIZE (nf, tmp2, tmp2_length,
+                               resultbuf, lengthp);
+         if (tmp1 == NULL)
+           {
+             int saved_errno = errno;
+             if (tmp2 != tmpbuf2)
+               free (tmp2);
+             errno = saved_errno;
+             return NULL;
+           }
+
+         if (tmp2 != tmpbuf2)
+           free (tmp2);
+       }
+      while (--repeat > 0);
+
+      return tmp1;
+    }
+}

diff --git a/lib/unicase/u8-ct-casefold.c b/lib/unicase/u8-ct-casefold.c
new file mode 100644 (file)
index 0000000..4cc149c
--- /dev/null
+++ b/lib/unicase/u8-ct-casefold.c
@@ -0,0 +1,35 @@
+/* Casefolding mapping for UTF-8 substrings (locale dependent).
+   Copyright (C) 2009 Free Software Foundation, Inc.
+   Written by Bruno Haible <bruno@clisp.org>, 2009.
+
+   This program is free software: you can redistribute it and/or modify it
+   under the terms of the GNU Lesser General Public License as published
+   by the Free Software Foundation; either version 3 of the License, 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
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public License
+   along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
+
+#include <config.h>
+
+/* Specification.  */
+#include "unicase.h"
+
+#include <errno.h>
+#include <stddef.h>
+#include <stdlib.h>
+
+#include "unicasemap.h"
+#include "special-casing.h"
+#include "casefold.h"
+
+#define FUNC u8_ct_casefold
+#define UNIT uint8_t
+#define U_CASEMAP u8_casemap
+#define U_NORMALIZE u8_normalize
+#include "u-ct-casefold.h"

diff --git a/modules/unicase/u8-ct-casefold b/modules/unicase/u8-ct-casefold
new file mode 100644 (file)
index 0000000..d82086c
--- /dev/null
+++ b/modules/unicase/u8-ct-casefold
@@ -0,0 +1,30 @@
+Description:
+Casefolding mapping for UTF-8 substrings (locale dependent).
+
+Files:
+lib/unicase/u8-ct-casefold.c
+lib/unicase/u-ct-casefold.h
+
+Depends-on:
+unicase/base
+unicase/u8-casemap
+unicase/special-casing
+unicase/tocasefold
+uninorm/decomposing-form
+uninorm/u8-normalize
+uninorm/nfd
+
+configure.ac:
+
+Makefile.am:
+lib_SOURCES += unicase/u8-ct-casefold.c
+
+Include:
+"unicase.h"
+
+License:
+LGPL
+
+Maintainer:
+Bruno Haible
+