Retrospective commit from 2009-1219.

Fix reordering of Arabic text in etc/HELLO.
Extend .gdbinit commands to support bidirectional display.

 buffer.c (Fbuffer_swap_text): Swap the values of
 bidi_display_reordering and bidi_paragraph_direction.

 bidi.c (bidi_resolve_weak): Fix nesting of conditions for Wn
 processing.  Move W3 after W1 and W2.  Simplify W4 because it is
 now always after W1.

 .gdbinit (pbiditype): New command.
 (pgx): Use it to display bidi level and type of the glyph.

diff --git a/src/.gdbinit b/src/.gdbinit
index e8a64f5dfe419e5be586830448ced2f384b95fe4..61e3da6c68c4b25ab2f89c5ba6b5c342bf7b238a 100644 (file)
--- a/src/.gdbinit
+++ b/src/.gdbinit
@@ -447,6 +447,33 @@ document pwin
 Pretty print window structure w.
 end
 
+define pbiditype
+  if ($arg0 == 1)
+    printf "L"
+  end
+  if ($arg0 == 2)
+    printf "R"
+  end
+  if ($arg0 == 3)
+    printf "EN"
+  end
+  if ($arg0 == 4)
+    printf "AN"
+  end
+  if ($arg0 == 5)
+    printf "BN"
+  end
+  if ($arg0 == 6)
+    printf "B"
+  end
+  if ($arg0 < 1 || $arg0 > 6)
+    printf "%d??", $arg0
+  end
+end
+document pbiditype
+Print textual description of bidi type given as first argument.
+end
+
 define pgx
   set $g = $arg0
   # CHAR_GLYPH
@@ -475,6 +502,11 @@ define pgx
   else
     printf " pos=%d", $g->charpos
   end
+  # For characters, print their resolved level and bidi type
+  if ($g->type == 0)
+    printf " blev=%d,btyp=", $g->resolved_level
+    pbiditype $g->bidi_type
+  end
   printf " w=%d a+d=%d+%d", $g->pixel_width, $g->ascent, $g->descent
   # If not DEFAULT_FACE_ID
   if ($g->face_id != 0)

diff --git a/src/ChangeLog.bidi b/src/ChangeLog.bidi
index ef911354df823bc149a9b1c31d8fc2d76bf9d9bf..b9fb7457cfd9a6245b1b321e82032e95ca31c6a6 100644 (file)
--- a/src/ChangeLog.bidi
+++ b/src/ChangeLog.bidi
@@ -1,3 +1,15 @@
+2009-12-19  Eli Zaretskii  <eliz@gnu.org>
+
+       * buffer.c (Fbuffer_swap_text): Swap the values of
+       bidi_display_reordering and bidi_paragraph_direction.
+
+       * bidi.c (bidi_resolve_weak): Fix nesting of conditions for Wn
+       processing.  Move W3 after W1 and W2.  Simplify W4 because it is
+       now always after W1.
+
+       * .gdbinit (pbiditype): New command.
+       (pgx): Use it to display bidi level and type of the glyph.
+
 2009-12-12  Eli Zaretskii  <eliz@gnu.org>
 
        * dispextern.h (struct it): New members prev_stop and

diff --git a/src/bidi.c b/src/bidi.c
index 102326b41e8bd2fcf4f4e3ec1c831b0dd47e98ea..d18629279d6f2f639d79e33aa32e77bc2befdc03 100644 (file)
--- a/src/bidi.c
+++ b/src/bidi.c
@@ -1342,123 +1342,121 @@ bidi_resolve_weak (struct bidi_it *bidi_it)
     type = STRONG_R;
   else if (override == L2R)
     type = STRONG_L;
-  else if (type == STRONG_AL)
-    type = STRONG_R;           /* W3 */
-  else if (type == WEAK_NSM)   /* W1 */
-    {
-      /* Note that we don't need to consider the case where the prev
-        character has its type overridden by an RLO or LRO: such
-        characters are outside the current level run, and thus not
-        relevant to this NSM.  Thus, NSM gets the orig_type of the
-        previous character.  */
-      if (bidi_it->prev.type != UNKNOWN_BT)
-       type = bidi_it->prev.orig_type;
-      else if (bidi_it->sor == R2L)
-       type = STRONG_R;
-      else if (bidi_it->sor == L2R)
-       type = STRONG_L;
-      else /* shouldn't happen! */
-       abort ();
-      if (type == WEAK_EN      /* W2 after W1 */
-         && bidi_it->last_strong.type_after_w1 == STRONG_AL)
-       type = WEAK_AN;
-    }
-  else if (type == WEAK_EN     /* W2 */
-          && bidi_it->last_strong.type_after_w1 == STRONG_AL)
-    type = WEAK_AN;
-  else if ((type == WEAK_ES
-           && (bidi_it->prev.type_after_w1 == WEAK_EN           /* W4 */
-               && (bidi_it->prev.orig_type == WEAK_EN
-                   || bidi_it->prev.orig_type == WEAK_NSM)))    /* aft W1 */
-          || (type == WEAK_CS
-              && ((bidi_it->prev.type_after_w1 == WEAK_EN
-                   && (bidi_it->prev.orig_type == WEAK_EN       /* W4 */
-                       || bidi_it->prev.orig_type == WEAK_NSM)) /* a/W1 */
-                  || bidi_it->prev.type_after_w1 == WEAK_AN)))  /* W4 */
+  else
     {
-      next_char =
-       bidi_it->bytepos + bidi_it->ch_len >= ZV_BYTE
-       ? BIDI_EOB : FETCH_CHAR (bidi_it->bytepos + bidi_it->ch_len);
-      type_of_next = bidi_get_type (next_char, override);
-
-      if (type_of_next == WEAK_BN
-         || bidi_explicit_dir_char (next_char))
+      if (type == WEAK_NSM)    /* W1 */
        {
-         bidi_copy_it (&saved_it, bidi_it);
-         while (bidi_resolve_explicit (bidi_it) == new_level
-                && bidi_it->type == WEAK_BN)
-           ;
-         type_of_next = bidi_it->type;
-         bidi_copy_it (bidi_it, &saved_it);
-       }
-
-      /* If the next character is EN, but the last strong-type
-        character is AL, that next EN will be changed to AN when we
-        process it in W2 above.  So in that case, this ES should not
-        be changed into EN.  */
-      if (type == WEAK_ES
-         && type_of_next == WEAK_EN
-         && bidi_it->last_strong.type_after_w1 != STRONG_AL)
-       type = WEAK_EN;
-      else if (type == WEAK_CS)
-       {
-         if (bidi_it->prev.type_after_w1 == WEAK_AN
-             && (type_of_next == WEAK_AN
-                 /* If the next character is EN, but the last
-                    strong-type character is AL, EN will be later
-                    changed to AN when we process it in W2 above.  So
-                    in that case, this ES should not be changed into
-                    EN.  */
-                 || (type_of_next == WEAK_EN
-                     && bidi_it->last_strong.type_after_w1 == STRONG_AL)))
-           type = WEAK_AN;
-         else if (bidi_it->prev.type_after_w1 == WEAK_EN
-                  && type_of_next == WEAK_EN
-                  && bidi_it->last_strong.type_after_w1 != STRONG_AL)
-           type = WEAK_EN;
+         /* Note that we don't need to consider the case where the
+            prev character has its type overridden by an RLO or LRO:
+            such characters are outside the current level run, and
+            thus not relevant to this NSM.  Thus, NSM gets the
+            orig_type of the previous character.  */
+         if (bidi_it->prev.type != UNKNOWN_BT)
+           type = bidi_it->prev.orig_type;
+         else if (bidi_it->sor == R2L)
+           type = STRONG_R;
+         else if (bidi_it->sor == L2R)
+           type = STRONG_L;
+         else /* shouldn't happen! */
+           abort ();
        }
-    }
-  else if (type == WEAK_ET     /* W5: ET with EN before or after it */
-          || type == WEAK_BN)  /* W5/Retaining */
-    {
-      if (bidi_it->prev.type_after_w1 == WEAK_EN /* ET/BN with EN before it */
-         || bidi_it->next_en_pos > bidi_it->charpos)
-       type = WEAK_EN;
-      /* W5: ET with EN after it.  */
-      else
+      if (type == WEAK_EN      /* W2 */
+         && bidi_it->last_strong.type_after_w1 == STRONG_AL)
+       type = WEAK_AN;
+      else if (type == STRONG_AL) /* W3 */
+       type = STRONG_R;
+      else if ((type == WEAK_ES        /* W4 */
+               && bidi_it->prev.type_after_w1 == WEAK_EN
+               && bidi_it->prev.orig_type == WEAK_EN)
+              || (type == WEAK_CS
+                  && ((bidi_it->prev.type_after_w1 == WEAK_EN
+                       && bidi_it->prev.orig_type == WEAK_EN)
+                      || bidi_it->prev.type_after_w1 == WEAK_AN)))
        {
-         EMACS_INT en_pos = bidi_it->charpos + 1;
-
          next_char =
            bidi_it->bytepos + bidi_it->ch_len >= ZV_BYTE
            ? BIDI_EOB : FETCH_CHAR (bidi_it->bytepos + bidi_it->ch_len);
          type_of_next = bidi_get_type (next_char, override);
 
-         if (type_of_next == WEAK_ET
-             || type_of_next == WEAK_BN
+         if (type_of_next == WEAK_BN
              || bidi_explicit_dir_char (next_char))
            {
              bidi_copy_it (&saved_it, bidi_it);
              while (bidi_resolve_explicit (bidi_it) == new_level
-                    && (bidi_it->type == WEAK_BN || bidi_it->type == WEAK_ET))
+                    && bidi_it->type == WEAK_BN)
                ;
              type_of_next = bidi_it->type;
-             en_pos = bidi_it->charpos;
              bidi_copy_it (bidi_it, &saved_it);
            }
-         if (type_of_next == WEAK_EN)
+
+         /* If the next character is EN, but the last strong-type
+            character is AL, that next EN will be changed to AN when
+            we process it in W2 above.  So in that case, this ES
+            should not be changed into EN.  */
+         if (type == WEAK_ES
+             && type_of_next == WEAK_EN
+             && bidi_it->last_strong.type_after_w1 != STRONG_AL)
+           type = WEAK_EN;
+         else if (type == WEAK_CS)
            {
-             /* If the last strong character is AL, the EN we've
-                found will become AN when we get to it (W2). */
-             if (bidi_it->last_strong.type_after_w1 != STRONG_AL)
+             if (bidi_it->prev.type_after_w1 == WEAK_AN
+                 && (type_of_next == WEAK_AN
+                     /* If the next character is EN, but the last
+                        strong-type character is AL, EN will be later
+                        changed to AN when we process it in W2 above.
+                        So in that case, this ES should not be
+                        changed into EN.  */
+                     || (type_of_next == WEAK_EN
+                         && bidi_it->last_strong.type_after_w1 == STRONG_AL)))
+               type = WEAK_AN;
+             else if (bidi_it->prev.type_after_w1 == WEAK_EN
+                      && type_of_next == WEAK_EN
+                      && bidi_it->last_strong.type_after_w1 != STRONG_AL)
+               type = WEAK_EN;
+           }
+       }
+      else if (type == WEAK_ET /* W5: ET with EN before or after it */
+              || type == WEAK_BN)      /* W5/Retaining */
+       {
+         if (bidi_it->prev.type_after_w1 == WEAK_EN /* ET/BN w/EN before it */
+             || bidi_it->next_en_pos > bidi_it->charpos)
+           type = WEAK_EN;
+         else                  /* W5: ET/BN with EN after it.  */
+           {
+             EMACS_INT en_pos = bidi_it->charpos + 1;
+
+             next_char =
+               bidi_it->bytepos + bidi_it->ch_len >= ZV_BYTE
+               ? BIDI_EOB : FETCH_CHAR (bidi_it->bytepos + bidi_it->ch_len);
+             type_of_next = bidi_get_type (next_char, override);
+
+             if (type_of_next == WEAK_ET
+                 || type_of_next == WEAK_BN
+                 || bidi_explicit_dir_char (next_char))
+               {
+                 bidi_copy_it (&saved_it, bidi_it);
+                 while (bidi_resolve_explicit (bidi_it) == new_level
+                        && (bidi_it->type == WEAK_BN
+                            || bidi_it->type == WEAK_ET))
+                   ;
+                 type_of_next = bidi_it->type;
+                 en_pos = bidi_it->charpos;
+                 bidi_copy_it (bidi_it, &saved_it);
+               }
+             if (type_of_next == WEAK_EN)
                {
-                 type = WEAK_EN;
-                 /* Remember this EN position, to speed up processing
-                    of the next ETs.  */
-                 bidi_it->next_en_pos = en_pos;
+                 /* If the last strong character is AL, the EN we've
+                    found will become AN when we get to it (W2). */
+                 if (bidi_it->last_strong.type_after_w1 != STRONG_AL)
+                   {
+                     type = WEAK_EN;
+                     /* Remember this EN position, to speed up processing
+                        of the next ETs.  */
+                     bidi_it->next_en_pos = en_pos;
+                   }
+                 else if (type == WEAK_BN)
+                   type = NEUTRAL_ON; /* W6/Retaining */
                }
-             else if (type == WEAK_BN)
-               type = NEUTRAL_ON; /* W6/Retaining */
            }
        }
     }

diff --git a/src/buffer.c b/src/buffer.c
index 521fe9b50198f1971a85973a89381a77d44eef93..ce9dc1c274d8e054b5e5af0497d0d14bf3d577f5 100644 (file)
--- a/src/buffer.c
+++ b/src/buffer.c
@@ -2261,6 +2261,8 @@ DEFUN ("buffer-swap-text", Fbuffer_swap_text, Sbuffer_swap_text,
   swapfield (undo_list, Lisp_Object);
   swapfield (mark, Lisp_Object);
   swapfield (enable_multibyte_characters, Lisp_Object);
+  swapfield (bidi_display_reordering, Lisp_Object);
+  swapfield (bidi_paragraph_direction, Lisp_Object);
   /* FIXME: Not sure what we should do with these *_marker fields.
      Hopefully they're just nil anyway.  */
   swapfield (pt_marker, Lisp_Object);