X-Git-Url: http://sjero.net/git/?p=wget;a=blobdiff_plain;f=src%2Firi.c;h=d6067497c86a4755e929547edf27d087a60eb0e6;hp=5fb06d0992bde05c5dda9768e68e2b9b8d5ad454;hb=a9a2b34b052cfa903462124f59fbfeed7eaf374b;hpb=d3007f1b3a5d033babe40bc4c56a899eb3b10bfa

diff --git a/src/iri.c b/src/iri.c
index 5fb06d09..d6067497 100644
--- a/src/iri.c
+++ b/src/iri.c
@@ -1,6 +1,5 @@
 /* IRI related functions.
-   Copyright (C) 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007,
-   2008 Free Software Foundation, Inc.
+   Copyright (C) 2008 Free Software Foundation, Inc.
 
 This file is part of GNU Wget.
 
@@ -40,13 +39,12 @@ as that of the covered work.  */
 #include <errno.h>
 
 #include "utils.h"
-#include "iri.h"
 
+/* RFC3987 section 3.1 mandates STD3 ASCII RULES */
+#define IDNA_FLAGS  IDNA_USE_STD3_ASCII_RULES
 
-static iconv_t locale2utf8;
+/* Note: locale encoding is kept in options struct (opt.locale) */
 
-
-static bool open_locale_to_utf8 (void);
 static bool do_conversion (iconv_t cd, char *in, size_t inlen, char **out);
 
 
@@ -81,7 +79,7 @@ parse_charset (char *str)
       return NULL;
     }
 
-  logprintf (LOG_VERBOSE, "parse_charset: %s\n", quote (charset));
+  /*logprintf (LOG_VERBOSE, "parse_charset: %s\n", quote (charset));*/
 
   return charset;
 }
@@ -90,7 +88,6 @@ parse_charset (char *str)
 char *
 find_locale (void)
 {
-  /* sXXXav, made our own function or use libidn one ?! */
   return (char *) stringprep_locale_charset ();
 }
 
@@ -102,9 +99,9 @@ check_encoding_name (char *encoding)
 
   while (*s)
     {
-      if (!c_isascii(*s) || c_isspace(*s))
+      if (!c_isascii (*s) || c_isspace (*s))
         {
-          logprintf (LOG_VERBOSE, "Encoding %s isn't valid\n", quote(encoding));
+          logprintf (LOG_VERBOSE, _("Encoding %s isn't valid\n"), quote (encoding));
           return false;
         }
 
@@ -118,48 +115,44 @@ check_encoding_name (char *encoding)
 static bool
 open_locale_to_utf8 (void)
 {
-  if (locale2utf8)
-    return true;
-
-  /* sXXXav : That shouldn't happen, just in case */
-  if (!opt.locale)
-    {
-      logprintf (LOG_VERBOSE, "open_locale_to_utf8: locale is unset\n");
-      opt.locale = find_locale ();
-    }
-
-  if (!opt.locale)
-    return false;
 
-  locale2utf8 = iconv_open ("UTF-8", opt.locale);
-  if (locale2utf8 != (iconv_t)(-1))
-    return true;
-
-  logprintf (LOG_VERBOSE, "Conversion from %s to %s isn't supported\n",
-             quote (opt.locale), quote("UTF-8"));
-  locale2utf8 = NULL;
-  return false;
 }
 
-/* Return a new string */
+/* Try converting string str from locale to UTF-8. Return a new string
+   on success, or str on error or if conversion isn't needed. */
 const char *
 locale_to_utf8 (const char *str)
 {
+  iconv_t l2u;
   char *new;
 
-  if (!strcasecmp (opt.locale, "utf-8"))
-    return str;
+  /* That shouldn't happen, just in case */
+  if (!opt.locale)
+    {
+      logprintf (LOG_VERBOSE, _("locale_to_utf8: locale is unset\n"));
+      opt.locale = find_locale ();
+    }
 
-  if (!open_locale_to_utf8 ())
+  if (!opt.locale || !strcasecmp (opt.locale, "utf-8"))
     return str;
 
-  if (do_conversion (locale2utf8, (char *) str, strlen ((char *) str), &new))
+  l2u = iconv_open ("UTF-8", opt.locale);
+  if (l2u != (iconv_t)(-1))
+    { 
+      logprintf (LOG_VERBOSE, _("Conversion from %s to %s isn't supported\n"),
+                 quote (opt.locale), quote ("UTF-8"));
+      return str;
+    }
+
+  if (do_conversion (l2u, (char *) str, strlen ((char *) str), &new))
     return (const char *) new;
 
   return str;
 }
 
-/* */
+/* Do the conversion according to the passed conversion descriptor cd. *out
+   will contain the transcoded string on success. *out content is
+   unspecified otherwise. */
 static bool
 do_conversion (iconv_t cd, char *in, size_t inlen, char **out)
 {
@@ -173,7 +166,6 @@ do_conversion (iconv_t cd, char *in, size_t inlen, char **out)
   len = outlen;
   done = 0;
 
-  /* sXXXav : put a maximum looping factor ??? */
   for (;;)
     {
       if (iconv (cd, &in, &inlen, out, &outlen) != (size_t)(-1))
@@ -186,6 +178,10 @@ do_conversion (iconv_t cd, char *in, size_t inlen, char **out)
       /* Incomplete or invalid multibyte sequence */
       if (errno == EINVAL || errno == EILSEQ)
         {
+          if (!invalid)
+            logprintf (LOG_VERBOSE,
+                      _("Incomplete or invalid multibyte sequence encountered\n"));
+
           invalid++;
           **out = *in;
           in++;
@@ -193,7 +189,7 @@ do_conversion (iconv_t cd, char *in, size_t inlen, char **out)
           (*out)++;
           outlen--;
         }
-      else if (errno == E2BIG) /* Output buffer full */ 
+      else if (errno == E2BIG) /* Output buffer full */
         {
           char *new;
 
@@ -209,7 +205,7 @@ do_conversion (iconv_t cd, char *in, size_t inlen, char **out)
         }
       else /* Weird, we got an unspecified error */
         {
-          logprintf (LOG_VERBOSE, "Unhandled errno %d\n", errno);
+          logprintf (LOG_VERBOSE, _("Unhandled errno %d\n"), errno);
           break;
         }
     }
@@ -217,18 +213,47 @@ do_conversion (iconv_t cd, char *in, size_t inlen, char **out)
     return false;
 }
 
-/* Try to encode UTF-8 host to ASCII. Return the new domain on success or NULL
+/* Try to "ASCII encode" UTF-8 host. Return the new domain on success or NULL
    on error. */
-char *idn_encode (char *host)
+char *
+idn_encode (struct iri *i, char *host)
 {
   char *new;
   int ret;
 
+  /* Encode to UTF-8 if not done */
+  if (!i->utf8_encode)
+    {
+      if (!remote_to_utf8 (i, (const char *) host, (const char **) &new))
+          return NULL;  /* Nothing to encode or an error occured */
+      host = new;
+    }
+
   /* toASCII UTF-8 NULL terminated string */
-  ret = idna_to_ascii_8z (host, &new, 0);
+  ret = idna_to_ascii_8z (host, &new, IDNA_FLAGS);
+  if (ret != IDNA_SUCCESS)
+    {
+      /* sXXXav : free new when needed ! */
+      logprintf (LOG_VERBOSE, _("idn_encode failed (%d): %s\n"), ret,
+                 quote (idna_strerror (ret)));
+      return NULL;
+    }
+
+  return new;
+}
+
+/* Try to decode an "ASCII encoded" host. Return the new domain in the locale
+   on success or NULL on error. */
+char *
+idn_decode (char *host)
+{
+  char *new;
+  int ret;
+
+  ret = idna_to_unicode_8zlz (host, &new, IDNA_FLAGS);
   if (ret != IDNA_SUCCESS)
     {
-      logprintf (LOG_VERBOSE, "idn_encode failed (%d): %s\n", ret,
+      logprintf (LOG_VERBOSE, _("idn_decode failed (%d): %s\n"), ret,
                  quote (idna_strerror (ret)));
       return NULL;
     }
@@ -236,3 +261,101 @@ char *idn_encode (char *host)
   return new;
 }
 
+/* Try to transcode string str from remote encoding to UTF-8. On success, *new
+   contains the transcoded string. *new content is unspecified otherwise. */
+bool
+remote_to_utf8 (struct iri *i, const char *str, const char **new)
+{
+  iconv_t cd;
+  bool ret = false;
+
+  if (!i->uri_encoding)
+    return false;
+
+  cd = iconv_open ("UTF-8", i->uri_encoding);
+  if (cd == (iconv_t)(-1))
+    return false;
+
+  if (do_conversion (cd, (char *) str, strlen ((char *) str), (char **) new))
+    ret = true;
+
+  iconv_close (cd);
+
+  /* Test if something was converted */
+  if (!strcmp (str, *new))
+    {
+      xfree ((char *) *new);
+      return false;
+    }
+
+  return ret;
+}
+
+/* Allocate a new iri structure and return a pointer to it. */
+struct iri *
+iri_new (void)
+{
+  struct iri *i = xmalloc (sizeof *i);
+  i->uri_encoding = opt.encoding_remote ? xstrdup (opt.encoding_remote) : NULL;
+  i->content_encoding = NULL;
+  i->orig_url = NULL;
+  i->utf8_encode = opt.enable_iri;
+  return i;
+}
+
+struct iri *iri_dup (const struct iri *src)
+{
+  struct iri *i = xmalloc (sizeof *i);
+  i->uri_encoding = src->uri_encoding ? xstrdup (src->uri_encoding) : NULL;
+  i->content_encoding = (src->content_encoding ?
+                         xstrdup (src->content_encoding) : NULL);
+  i->orig_url = src->orig_url ? xstrdup (src->orig_url) : NULL;
+  i->utf8_encode = src->utf8_encode;
+  return i;
+}
+
+/* Completely free an iri structure. */
+void
+iri_free (struct iri *i)
+{
+  xfree_null (i->uri_encoding);
+  xfree_null (i->content_encoding);
+  xfree_null (i->orig_url);
+  xfree (i);
+}
+
+/* Set uri_encoding of struct iri i. If a remote encoding was specified, use
+   it unless force is true. */
+void
+set_uri_encoding (struct iri *i, char *charset, bool force)
+{
+  DEBUGP (("URI encoding = %s\n", charset ? quote (charset) : "None"));
+  if (!force && opt.encoding_remote)
+    return;
+  if (i->uri_encoding)
+    {
+      if (charset && !strcasecmp (i->uri_encoding, charset))
+        return;
+      xfree (i->uri_encoding);
+    }
+
+  i->uri_encoding = charset ? xstrdup (charset) : NULL;
+}
+
+/* Set content_encoding of struct iri i. */
+void
+set_content_encoding (struct iri *i, char *charset)
+{
+  DEBUGP (("URI content encoding = %s\n", charset ? quote (charset) : "None"));
+  if (opt.encoding_remote)
+    return;
+  if (i->content_encoding)
+    {
+      if (charset && !strcasecmp (i->content_encoding, charset))
+        return;
+      xfree (i->content_encoding);
+    }
+
+  i->content_encoding = charset ? xstrdup (charset) : NULL;
+}
+