X-Git-Url: http://sjero.net/git/?p=wget;a=blobdiff_plain;f=src%2Fhtml-parse.c;h=20791cd83450272210c17a129ecd6904aec86231;hp=fdf5b99bc791c8c420633be631903b8ab3f1be64;hb=HEAD;hpb=ec84142901fc685d7a08267fc0be8962e468968d

diff --git a/src/html-parse.c b/src/html-parse.c
index fdf5b99b..20791cd8 100644
--- a/src/html-parse.c
+++ b/src/html-parse.c
@@ -1,6 +1,6 @@
 /* HTML parser for Wget.
    Copyright (C) 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006,
-   2007, 2008 Free Software Foundation, Inc.
+   2007, 2008, 2009, 2010, 2011 Free Software Foundation, Inc.
 
 This file is part of GNU Wget.
 
@@ -272,6 +272,94 @@ struct pool {
    to "<foo", but "&lt,foo" to "<,foo".  */
 #define SKIP_SEMI(p, inc) (p += inc, p < end && *p == ';' ? ++p : p)
 
+struct tagstack_item {
+  const char *tagname_begin;
+  const char *tagname_end;
+  const char *contents_begin;
+  struct tagstack_item *prev;
+  struct tagstack_item *next;
+};
+
+static struct tagstack_item *
+tagstack_push (struct tagstack_item **head, struct tagstack_item **tail)
+{
+  struct tagstack_item *ts = xmalloc(sizeof(struct tagstack_item));
+  if (*head == NULL)
+    {
+      *head = *tail = ts;
+      ts->prev = ts->next = NULL;
+    }
+  else
+    {
+      (*tail)->next = ts;
+      ts->prev = *tail;
+      *tail = ts;
+      ts->next = NULL;
+    }
+
+  return ts;
+}
+
+/* remove ts and everything after it from the stack */
+static void
+tagstack_pop (struct tagstack_item **head, struct tagstack_item **tail,
+              struct tagstack_item *ts)
+{
+  if (*head == NULL)
+    return;
+
+  if (ts == *tail)
+    {
+      if (ts == *head)
+        {
+          xfree (ts);
+          *head = *tail = NULL;
+        }
+      else
+        {
+          ts->prev->next = NULL;
+          *tail = ts->prev;
+          xfree (ts);
+        }
+    }
+  else
+    {
+      if (ts == *head)
+        {
+          *head = NULL;
+        }
+      *tail = ts->prev;
+
+      if (ts->prev)
+        {
+          ts->prev->next = NULL;
+        }
+      while (ts)
+        {
+          struct tagstack_item *p = ts->next;
+          xfree (ts);
+          ts = p;
+        }
+    }
+}
+
+static struct tagstack_item *
+tagstack_find (struct tagstack_item *tail, const char *tagname_begin,
+               const char *tagname_end)
+{
+  int len = tagname_end - tagname_begin;
+  while (tail)
+    {
+      if (len == (tail->tagname_end - tail->tagname_begin))
+        {
+          if (0 == strncasecmp (tail->tagname_begin, tagname_begin, len))
+            return tail;
+        }
+      tail = tail->prev;
+    }
+  return NULL;
+}
+
 /* Decode the HTML character entity at *PTR, considering END to be end
    of buffer.  It is assumed that the "&" character that marks the
    beginning of the entity has been seen at *PTR-1.  If a recognized
@@ -440,13 +528,14 @@ convert_and_copy (struct pool *pool, const char *beg, const char *end, int flags
      * whitespace
      * 8-bit and control chars
      * characters that clearly cannot be part of name:
-       '=', '>', '/'.
+       '=', '<', '>', '/'.
 
    This only affects attribute and tag names; attribute values allow
    an even greater variety of characters.  */
 
 #define NAME_CHAR_P(x) ((x) > 32 && (x) < 127                           \
-                        && (x) != '=' && (x) != '>' && (x) != '/')
+                        && (x) != '=' && (x) != '<' && (x) != '>'       \
+                        && (x) != '/')
 
 #ifdef STANDALONE
 static int comment_backout_count;
@@ -531,6 +620,7 @@ advance_declaration (const char *beg, const char *end)
             case '\n':
               ch = *p++;
               break;
+            case '<':
             case '>':
               state = AC_S_DONE;
               break;
@@ -757,6 +847,9 @@ map_html_tags (const char *text, int size,
   bool attr_pair_resized = false;
   struct attr_pair *pairs = attr_pair_initial_storage;
 
+  struct tagstack_item *head = NULL;
+  struct tagstack_item *tail = NULL;
+
   if (!size)
     return;
 
@@ -823,7 +916,19 @@ map_html_tags (const char *text, int size,
       goto look_for_tag;
     tag_name_end = p;
     SKIP_WS (p);
-    if (end_tag && *p != '>')
+
+    if (!end_tag)
+      {
+        struct tagstack_item *ts = tagstack_push (&head, &tail);
+        if (ts)
+          {
+            ts->tagname_begin  = tag_name_begin;
+            ts->tagname_end    = tag_name_end;
+            ts->contents_begin = NULL;
+          }
+      }
+
+    if (end_tag && *p != '>' && *p != '<')
       goto backout_tag;
 
     if (!name_allowed (allowed_tags, tag_name_begin, tag_name_end))
@@ -855,12 +960,12 @@ map_html_tags (const char *text, int size,
             /*              ^  */
             ADVANCE (p);
             SKIP_WS (p);
-            if (*p != '>')
+            if (*p != '<' && *p != '>')
               goto backout_tag;
           }
 
         /* Check for end of tag definition. */
-        if (*p == '>')
+        if (*p == '<' || *p == '>')
           break;
 
         /* Establish bounds of attribute name. */
@@ -875,7 +980,8 @@ map_html_tags (const char *text, int size,
 
         /* Establish bounds of attribute value. */
         SKIP_WS (p);
-        if (NAME_CHAR_P (*p) || *p == '/' || *p == '>')
+
+        if (NAME_CHAR_P (*p) || *p == '/' || *p == '<' || *p == '>')
           {
             /* Minimized attribute syntax allows `=' to be omitted.
                For example, <UL COMPACT> is a valid shorthand for <UL
@@ -912,7 +1018,7 @@ map_html_tags (const char *text, int size,
                         newline_seen = true;
                         continue;
                       }
-                    else if (newline_seen && *p == '>')
+                    else if (newline_seen && (*p == '<' || *p == '>'))
                       break;
                     ADVANCE (p);
                   }
@@ -937,7 +1043,7 @@ map_html_tags (const char *text, int size,
                    violated by, for instance, `%' in `width=75%'.
                    We'll be liberal and allow just about anything as
                    an attribute value.  */
-                while (!c_isspace (*p) && *p != '>')
+                while (!c_isspace (*p) && *p != '<' && *p != '>')
                   ADVANCE (p);
                 attr_value_end = p; /* <foo bar=baz qux=quix> */
                                     /*             ^          */
@@ -984,6 +1090,11 @@ map_html_tags (const char *text, int size,
         ++nattrs;
       }
 
+    if (!end_tag && tail && (tail->tagname_begin == tag_name_begin))
+      {
+        tail->contents_begin = p+1;
+      }
+
     if (uninteresting_tag)
       {
         ADVANCE (p);
@@ -995,6 +1106,7 @@ map_html_tags (const char *text, int size,
     {
       int i;
       struct taginfo taginfo;
+      struct tagstack_item *ts = NULL;
 
       taginfo.name      = pool.contents;
       taginfo.end_tag_p = end_tag;
@@ -1011,8 +1123,26 @@ map_html_tags (const char *text, int size,
       taginfo.attrs = pairs;
       taginfo.start_position = tag_start_position;
       taginfo.end_position   = p + 1;
+      taginfo.contents_begin = NULL;
+      taginfo.contents_end = NULL;
+
+      if (end_tag)
+        {
+          ts = tagstack_find (tail, tag_name_begin, tag_name_end);
+          if (ts)
+            {
+              if (ts->contents_begin)
+                {
+                  taginfo.contents_begin = ts->contents_begin;
+                  taginfo.contents_end   = tag_start_position;
+                }
+              tagstack_pop (&head, &tail, ts);
+            }
+        }
+
       mapfun (&taginfo, maparg);
-      ADVANCE (p);
+      if (*p != '<')
+        ADVANCE (p);
     }
     goto look_for_tag;
 
@@ -1030,6 +1160,8 @@ map_html_tags (const char *text, int size,
   POOL_FREE (&pool);
   if (attr_pair_resized)
     xfree (pairs);
+  /* pop any tag stack that's left */
+  tagstack_pop (&head, &tail, head);
 }
 
 #undef ADVANCE