X-Git-Url: http://sjero.net/git/?p=wget;a=blobdiff_plain;f=src%2Firi.c;h=5fb06d0992bde05c5dda9768e68e2b9b8d5ad454;hp=fea7b150e7a5e3163ec642d68c0738b9bad9fbca;hb=5bb11da009c2f3bc4381bc8009c57007fd86534e;hpb=e6376b47433be6a0df64b0cd87b2d5c2c53a66f1

diff --git a/src/iri.c b/src/iri.c
index fea7b150..5fb06d09 100644
--- a/src/iri.c
+++ b/src/iri.c
@@ -34,13 +34,22 @@ as that of the covered work.  */
 #include <stdlib.h>
 #include <assert.h>
 #include <string.h>
-
+#include <iconv.h>
 #include <stringprep.h>
+#include <idna.h>
+#include <errno.h>
 
 #include "utils.h"
 #include "iri.h"
 
 
+static iconv_t locale2utf8;
+
+
+static bool open_locale_to_utf8 (void);
+static bool do_conversion (iconv_t cd, char *in, size_t inlen, char **out);
+
+
 /* Given a string containing "charset=XXX", return the encoding if found,
    or NULL otherwise */
 char *
@@ -77,7 +86,6 @@ parse_charset (char *str)
   return charset;
 }
 
-
 /* Find the locale used, or fall back on a default value */
 char *
 find_locale (void)
@@ -86,7 +94,6 @@ find_locale (void)
   return (char *) stringprep_locale_charset ();
 }
 
-
 /* Basic check of an encoding name. */
 bool
 check_encoding_name (char *encoding)
@@ -107,4 +114,125 @@ check_encoding_name (char *encoding)
   return true;
 }
 
+/* Try opening an iconv_t descriptor for conversion from locale to UTF-8 */
+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 */
+const char *
+locale_to_utf8 (const char *str)
+{
+  char *new;
+
+  if (!strcasecmp (opt.locale, "utf-8"))
+    return str;
+
+  if (!open_locale_to_utf8 ())
+    return str;
+
+  if (do_conversion (locale2utf8, (char *) str, strlen ((char *) str), &new))
+    return (const char *) new;
+
+  return str;
+}
+
+/* */
+static bool
+do_conversion (iconv_t cd, char *in, size_t inlen, char **out)
+{
+  /* sXXXav : hummm hard to guess... */
+  size_t len, done, outlen = inlen * 2;
+  int invalid = 0, tooshort = 0;
+  char *s;
+
+  s = xmalloc (outlen + 1);
+  *out = s;
+  len = outlen;
+  done = 0;
+
+  /* sXXXav : put a maximum looping factor ??? */
+  for (;;)
+    {
+      if (iconv (cd, &in, &inlen, out, &outlen) != (size_t)(-1))
+        {
+          *out = s;
+          *(s + len - outlen - done) = '\0';
+          return true;
+        }
+
+      /* Incomplete or invalid multibyte sequence */
+      if (errno == EINVAL || errno == EILSEQ)
+        {
+          invalid++;
+          **out = *in;
+          in++;
+          inlen--;
+          (*out)++;
+          outlen--;
+        }
+      else if (errno == E2BIG) /* Output buffer full */ 
+        {
+          char *new;
+
+          tooshort++;
+          done = len;
+          outlen = done + inlen * 2;
+          new = xmalloc (outlen + 1);
+          memcpy (new, s, done);
+          xfree (s);
+          s = new;
+          len = outlen;
+          *out = s + done;
+        }
+      else /* Weird, we got an unspecified error */
+        {
+          logprintf (LOG_VERBOSE, "Unhandled errno %d\n", errno);
+          break;
+        }
+    }
+
+    return false;
+}
+
+/* Try to encode UTF-8 host to ASCII. Return the new domain on success or NULL
+   on error. */
+char *idn_encode (char *host)
+{
+  char *new;
+  int ret;
+
+  /* toASCII UTF-8 NULL terminated string */
+  ret = idna_to_ascii_8z (host, &new, 0);
+  if (ret != IDNA_SUCCESS)
+    {
+      logprintf (LOG_VERBOSE, "idn_encode failed (%d): %s\n", ret,
+                 quote (idna_strerror (ret)));
+      return NULL;
+    }
+
+  return new;
+}