Fix an issue where the "%%%" parser could match too many lines in unterminated blocks

Summary: Fixes T13530. The block parser could match too many lines in an unterminated "%%%" literal block. Adjust the logic to stop doing this (and hopefully be a little easier to read).

Test Plan: Added a failing test, made it pass.

Maniphest Tasks: T13530

Differential Revision: https://secure.phabricator.com/D21208

src/infrastructure/markup/blockrule/PhutilRemarkupLiteralBlockRule.php

@@ -22,51 +22,69 @@ final class PhutilRemarkupLiteralBlockRule extends PhutilRemarkupBlockRule {
     // output text and assists automated escaping of blocks coming into the
     // system.
 
-    $num_lines = 0;
+    $start_pattern = '(^\s*%%%)';
-    while (preg_match('/^\s*%%%/', $lines[$cursor])) {
+    $end_pattern = '(%%%\s*$)';
-      $num_lines++;
+    $trivial_pattern = '(^\s*%%%\s*$)';
 
-      // If the line has ONLY "%%%", the block opener doesn't get to double
+    if (!preg_match($start_pattern, $lines[$cursor])) {
-      // up as a block terminator.
+      return 0;
-      if (preg_match('/^\s*%%%\s*\z/', $lines[$cursor])) {
+    }
-        $num_lines++;
+
-        $cursor++;
+    $start_cursor = $cursor;
+
+    $found_empty = false;
+    $block_start = null;
+    while (true) {
+      if (!isset($lines[$cursor])) {
+        break;
       }
 
-      while (isset($lines[$cursor])) {
+      $line = $lines[$cursor];
-        if (!preg_match('/%%%\s*$/', $lines[$cursor])) {
+
-          $num_lines++;
+      if ($block_start === null) {
+        $is_start = preg_match($start_pattern, $line);
+
+        // If we've matched a block and then consumed one or more empty lines
+        // after it, stop merging more blocks into the match.
+        if ($found_empty) {
+          break;
+        }
+
+        if ($is_start) {
+          $block_start = $cursor;
+        }
+      }
+
+      if ($block_start !== null) {
+        $is_end = preg_match($end_pattern, $line);
+
+        // If a line contains only "%%%", it will match both the start and
+        // end patterns, but it only counts as a block start.
+        if ($is_end && ($cursor === $block_start)) {
+          $is_trivial = preg_match($trivial_pattern, $line);
+          if ($is_trivial) {
+            $is_end = false;
+          }
+        }
+
+        if ($is_end) {
+          $block_start = null;
           $cursor++;
           continue;
         }
-        break;
+      }
+
+      if ($block_start === null) {
+        if (strlen(trim($line))) {
+          break;
+        }
+        $found_empty = true;
       }
 
       $cursor++;
-
-      $found_empty = false;
-      while (isset($lines[$cursor])) {
-        if (!strlen(trim($lines[$cursor]))) {
-          $num_lines++;
-          $cursor++;
-          $found_empty = true;
-          continue;
-        }
-        break;
-      }
-
-      if ($found_empty) {
-        // If there's an empty line after the block, stop merging blocks.
-        break;
-      }
-
-      if (!isset($lines[$cursor])) {
-        // If we're at the end of the input, stop looking for more lines.
-        break;
-      }
     }
 
-    return $num_lines;
+    return ($cursor - $start_cursor);
   }
 
   public function markupText($text, $children) {

src/infrastructure/markup/remarkup/PhutilRemarkupEngine.php

@@ -204,6 +204,7 @@ final class PhutilRemarkupEngine extends PhutilMarkupEngine {
           }
 
           $cursor += $num_lines;
+
           break;
         }
       }

src/infrastructure/markup/remarkup/__tests__/remarkup/percent-block-unterminated.txt

@@ -0,0 +1,5 @@
+%%%xyz
+~~~~~~~~~~
+<p class="remarkup-literal">xyz</p>
+~~~~~~~~~~
+xyz