From 9b9b92be06db68ac24ea0a5b25753e84c5ac62ce Mon Sep 17 00:00:00 2001
From: Oliver Schmidt <ol.sc@web.de>
Date: Mon, 25 May 2015 16:46:23 +0200
Subject: [PATCH] Improved parsing of <SCRIPT> tag.

Complex script code tends to contain other tags inside strings. As we generally don't parse strings we erroneously interpreted those tags. The easiest workaround is to not interpret tags at all until the </SCRIPT> tag is found.
---
 apps/webbrowser/htmlparser.c | 14 +++++++++++---
 1 file changed, 11 insertions(+), 3 deletions(-)

diff --git a/apps/webbrowser/htmlparser.c b/apps/webbrowser/htmlparser.c
index b3c46574e..b66307965 100644
--- a/apps/webbrowser/htmlparser.c
+++ b/apps/webbrowser/htmlparser.c
@@ -137,7 +137,7 @@ G * (<br>, <p>, <h>), the <li> tag (but does not even try to
 #define MAJORSTATE_LINK       2
 #define MAJORSTATE_FORM       3
 #define MAJORSTATE_DISCARD    4
-
+#define MAJORSTATE_SCRIPT     5
 
 struct htmlparser_state {
 
@@ -303,7 +303,7 @@ do_word(void)
       if(s.word[s.wordlen - 1] != ISO_space) {
 	add_char(ISO_space);
       }
-    } else if(s.majorstate == MAJORSTATE_DISCARD) {
+    } else if(s.majorstate >= MAJORSTATE_DISCARD) {
       s.wordlen = 0;
     } else {
       s.word[s.wordlen] = '\0';
@@ -363,11 +363,17 @@ static void
 parse_tag(void)
 {
   static char *tagattrparam;
+  static unsigned char tag;
   static unsigned char size;
 
+  tag = find_tag(s.tag);
+  if(s.majorstate == MAJORSTATE_SCRIPT && tag != TAG_SLASHSCRIPT) {
+    return;
+  }
+
   PRINTF(("Parsing tag '%s' '%s' '%s'\n", s.tag, s.tagattr, s.tagattrparam));
 
-  switch(find_tag(s.tag)) {
+  switch(tag) {
   case TAG_P:
   case TAG_H1:
   case TAG_H2:
@@ -389,6 +395,8 @@ parse_tag(void)
     }
     break;
   case TAG_SCRIPT:
+    switch_majorstate(MAJORSTATE_SCRIPT);
+    break;
   case TAG_STYLE:
   case TAG_SELECT:
     switch_majorstate(MAJORSTATE_DISCARD);