Make striconveh work better with stateful encodings.

diff --git a/ChangeLog b/ChangeLog
index 56059f2a6b26e989ac9312b8a3ced299abc2772a..cc9e8a54bb673629cbdff7aec75eb676c0008916 100644 (file)
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+2008-09-07  Bruno Haible  <bruno@clisp.org>
+
+       Make striconveh work better with stateful encodings.
+       * lib/striconveh.c (iconv_carefully, iconv_carefully_1): Don't assume
+       that iconv does not increment the inptr when returning -1/EINVAL.
+
 2008-09-07  Bruno Haible  <bruno@clisp.org>
 
        * build-aux/config.rpath: Update according to libtool-2.2.6.

diff --git a/lib/striconveh.c b/lib/striconveh.c
index 685117f81608fc93e7a2a6c17f19b21faffcbdfe..10cb06f9d65af1f4c065c49588c7d887cc9ac526 100644 (file)
--- a/lib/striconveh.c
+++ b/lib/striconveh.c
@@ -79,9 +79,13 @@ iconv_carefully (iconv_t cd,
                       &outptr, &outsize);
          if (!(res == (size_t)(-1) && errno == EINVAL))
            break;
-         /* We expect that no input bytes have been consumed so far.  */
-         if (inptr != inptr_before)
-           abort ();
+         /* iconv can eat up a shift sequence but give EINVAL while attempting
+            to convert the first character.  E.g. libiconv does this.  */
+         if (inptr > inptr_before)
+           {
+             res = 0;
+             break;
+           }
        }
 
       if (res == 0)
@@ -117,31 +121,36 @@ iconv_carefully (iconv_t cd,
 # endif
 
 /* iconv_carefully_1 is like iconv_carefully, except that it stops after
-   converting one character.  */
+   converting one character or one shift sequence.  */
 static size_t
 iconv_carefully_1 (iconv_t cd,
                   const char **inbuf, size_t *inbytesleft,
                   char **outbuf, size_t *outbytesleft,
                   bool *incremented)
 {
-  const char *inptr = *inbuf;
-  const char *inptr_end = inptr + *inbytesleft;
+  const char *inptr_before = *inbuf;
+  const char *inptr = inptr_before;
+  const char *inptr_end = inptr_before + *inbytesleft;
   char *outptr = *outbuf;
   size_t outsize = *outbytesleft;
-  const char *inptr_before = inptr;
   size_t res = (size_t)(-1);
   size_t insize;
 
-  for (insize = 1; inptr + insize <= inptr_end; insize++)
+  for (insize = 1; inptr_before + insize <= inptr_end; insize++)
     {
+      inptr = inptr_before;
       res = iconv (cd,
                   (ICONV_CONST char **) &inptr, &insize,
                   &outptr, &outsize);
       if (!(res == (size_t)(-1) && errno == EINVAL))
        break;
-      /* We expect that no input bytes have been consumed so far.  */
-      if (inptr != inptr_before)
-       abort ();
+      /* iconv can eat up a shift sequence but give EINVAL while attempting
+        to convert the first character.  E.g. libiconv does this.  */
+      if (inptr > inptr_before)
+       {
+         res = 0;
+         break;
+       }
     }
 
   *inbuf = inptr;