(detect_eol, decode_eol): Handle text with DOS-style EOLs that also has

stray ^M characters.

diff --git a/src/ChangeLog b/src/ChangeLog
index 91d39e7b5807eed00dcbe29e9aadb60c90f42d18..f86e3c5afd4f637f92334aca8567f26a3bb8bfcc 100644 (file)
--- a/src/ChangeLog
+++ b/src/ChangeLog
@@ -1,3 +1,8 @@
+2009-01-30  Eli Zaretskii  <eliz@gnu.org>
+
+       * coding.c (detect_eol, decode_eol): Handle text with DOS-style
+       EOLs that also has stray ^M characters.
+
 2009-01-30  Juanma Barranquero  <lekktu@gmail.com>
 
        * atimer.c (run_timers, alarm_signal_handler):

diff --git a/src/coding.c b/src/coding.c
index 8c7ddf34db29141faa09240baeb2feb1d0b67cad..38cb3605464becdda981f7442dbe82a46d7c441f 100644 (file)
--- a/src/coding.c
+++ b/src/coding.c
@@ -5819,16 +5819,26 @@ detect_eol (source, src_bytes, category)
                       || src[lsb + 2] != '\n')
                this_eol = EOL_SEEN_CR;
              else
-               this_eol = EOL_SEEN_CRLF;
+               {
+                 this_eol = EOL_SEEN_CRLF;
+                 src += 2;
+               }
 
              if (eol_seen == EOL_SEEN_NONE)
                /* This is the first end-of-line.  */
                eol_seen = this_eol;
              else if (eol_seen != this_eol)
                {
-                 /* The found type is different from what found before.  */
-                 eol_seen = EOL_SEEN_LF;
-                 break;
+                 /* The found type is different from what found before.
+                    Allow for stray ^M characters in DOS EOL files.  */
+                 if (eol_seen == EOL_SEEN_CR && this_eol == EOL_SEEN_CRLF
+                     || eol_seen == EOL_SEEN_CRLF && this_eol == EOL_SEEN_CR)
+                   eol_seen = EOL_SEEN_CRLF;
+                 else
+                   {
+                     eol_seen = EOL_SEEN_LF;
+                     break;
+                   }
                }
              if (++total == MAX_EOL_CHECK_COUNT)
                break;
@@ -5857,9 +5867,16 @@ detect_eol (source, src_bytes, category)
                eol_seen = this_eol;
              else if (eol_seen != this_eol)
                {
-                 /* The found type is different from what found before.  */
-                 eol_seen = EOL_SEEN_LF;
-                 break;
+                 /* The found type is different from what found before.
+                    Allow for stray ^M characters in DOS EOL files.  */
+                 if (eol_seen == EOL_SEEN_CR && this_eol == EOL_SEEN_CRLF
+                     || eol_seen == EOL_SEEN_CRLF && this_eol == EOL_SEEN_CR)
+                   eol_seen = EOL_SEEN_CRLF;
+                 else
+                   {
+                     eol_seen = EOL_SEEN_LF;
+                     break;
+                   }
                }
              if (++total == MAX_EOL_CHECK_COUNT)
                break;
@@ -6114,7 +6131,12 @@ decode_eol (coding)
                eol_seen |= EOL_SEEN_CR;
            }
        }
-      if (eol_seen != EOL_SEEN_NONE
+      /* Handle DOS-style EOLs in a file with stray ^M characters.  */
+      if ((eol_seen & EOL_SEEN_CRLF) != 0
+         && (eol_seen & EOL_SEEN_CR) != 0
+         && (eol_seen & EOL_SEEN_LF) == 0)
+       eol_seen = EOL_SEEN_CRLF;
+      else if (eol_seen != EOL_SEEN_NONE
          && eol_seen != EOL_SEEN_LF
          && eol_seen != EOL_SEEN_CRLF
          && eol_seen != EOL_SEEN_CR)