From df050e44817819ec890145d7e9f647b58bd6310c Mon Sep 17 00:00:00 2001
From: Gabriel Wicke <gwicke@wikimedia.org>
Date: Fri, 13 Apr 2012 11:00:51 +0200
Subject: [PATCH] Convert external link syntax stops to stack

Eat unbalanced external link parts within template parameters. This does not
produce the same output as the PHP parser
(try echo '{{YouTube}}' | node parse.js), but preserves a level of sanity.
Need to check how common this is for external links. If it is rare enough,
moving the ']' after the parser function manually would fix the rendering for
the YouTube case.

Change-Id: I597d808efff36baa22191e7946a0061cc31120e8
---
 modules/parser/mediawiki.tokenizer.peg.js |  4 ++--
 modules/parser/pegTokenizer.pegjs.txt     | 14 ++++++++------
 2 files changed, 10 insertions(+), 8 deletions(-)

diff --git a/modules/parser/mediawiki.tokenizer.peg.js b/modules/parser/mediawiki.tokenizer.peg.js
index 42fb27742c..40fc194681 100644
--- a/modules/parser/mediawiki.tokenizer.peg.js
+++ b/modules/parser/mediawiki.tokenizer.peg.js
@@ -139,7 +139,7 @@ PegTokenizer.prototype.inline_breaks = function (input, pos, stops ) {
 			return counters.template && input[pos + 1] === "}" || null;
 		case ":":
 			return counters.colon &&
-				! counters.extlink &&
+				! stops.onStack( 'extlink' ) &&
 				! counters.linkdesc || null;
 		case "\r":
 			return stops.onStack( 'table' ) &&
@@ -151,7 +151,7 @@ PegTokenizer.prototype.inline_breaks = function (input, pos, stops ) {
 				input[pos + 1] === '|' ||
 				null;
 		case "]":
-			return counters.extlink ||
+			return stops.onStack( 'extlink' ) ||
 				( counters.linkdesc && input[pos + 1] === ']' ) ||
 				null;
 		case "<":
diff --git a/modules/parser/pegTokenizer.pegjs.txt b/modules/parser/pegTokenizer.pegjs.txt
index ad3d727a31..6e5404229d 100644
--- a/modules/parser/pegTokenizer.pegjs.txt
+++ b/modules/parser/pegTokenizer.pegjs.txt
@@ -667,7 +667,7 @@ behavior_switch
  **************************************************************/
 
 autolink
-  = ! { return stops.onCount('extlink') }
+  = ! { return stops.onStack('extlink') }
     (urllink / autoref / isbn)
 
 urllink
@@ -676,16 +676,16 @@ urllink
   }
 
 extlink
-  = ! { return stops.onCount('extlink') } // extlink cannot be nested
+  = ! { return stops.onStack('extlink') } // extlink cannot be nested
     (
         "[" 
-        & { return stops.inc('extlink'); }
+        & { return stops.push('extlink', true); }
         //target:urllink
         target:extlink_preprocessor_text
         text:(( space / [\u00A0\u1680\u180E\u2000-\u200A\u202F\u205F\u3000] )* 
               t:inlineline { return t } )?
         "]" {
-            stops.dec('extlink');
+            stops.pop('extlink');
             if ( text === '' ) {
                 // XXX: Link numbering should be implemented in post-processor.
                 text = [ "[" + linkCount + "]" ];
@@ -699,7 +699,7 @@ extlink
                 ] )
             ];
         }
-      / "[" & { stops.dec('extlink'); return false; }
+      / "[" & { return stops.pop('extlink'); }
     )
 
 autoref
@@ -939,10 +939,12 @@ template_param_text
           input.substr( pos +1, 9) ); */
         // re-enable tables within template parameters
         stops.push('table', false ); 
+        stops.push('extlink', false);
         return stops.inc('template') 
     }
     il:nested_block+ {
         stops.pop('table');
+        stops.pop('extlink');
         stops.dec('template');
         //console.warn( 'tpt match: ' + pp (il) + " stops: " + pp(stops));
         var r = flatten( il ),
@@ -962,7 +964,7 @@ template_param_text
 
         return r;
     }
-  / & { stops.pop('table'); return stops.dec('template'); }
+  / & { stops.pop('table'); stops.pop('extlink'); return stops.dec('template'); }
 
 
 // TODO: handle link prefixes as in al[[Razi]]