From 3c6b323300114b63d68cb83a70f96c40f86ba0e3 Mon Sep 17 00:00:00 2001
From: Dominique Fournier <dominique@fournier38.fr>
Date: Thu, 1 Jun 2017 14:38:49 +0000
Subject: [PATCH] markdown: allow to escape a part of string with backslash to
 not apply the markdown translation

git-svn-id: https://svn.fournier38.fr/svn/ProgSVN/trunk@3743 bf3deb0d-5f1a-0410-827f-c0cc1f45334c
---
 markdown.php | 90 ++++++++++++++++++++++++++++++++--------------------
 1 file changed, 56 insertions(+), 34 deletions(-)

diff --git a/markdown.php b/markdown.php
index daf80b2..b0bceca 100644
--- a/markdown.php
+++ b/markdown.php
@@ -57,44 +57,54 @@ class markdown
   {
     if ($this->debug)
       echo "CALL searchReplace ($line)\n";
+    // REMEMBER : THE $line is already in HTML ENTITIES !
+    // Quotes : &quot;
     $search = array ();
     $replace = array ();
     // Titles short
     // == TITRE1
-    $search[] = "/^==+ (.+)( ==+)?$/Um";
-    $replace[] = "</p>\n<h1>\\1</h1>\n<p>";
+    $search[] = '~^([^\\\\]|^)(==+ (.+)( ==+)?)$~Um';
+    $replace[] = '</p>'."\n".'<h1>\3</h1>'."\n".'<p>';
     // -- TITRE2
-    $search[] = "/^--+ (.+)( --+)?$/Um";
-    $replace[] = "</p>\n<h2>\\1</h2>\n<p>";
+    $search[] = '~^([^\\\\]|^)(--+ (.+)( --+)?)$~Um';
+    $replace[] = '</p>\n<h2>\3</h2>\n<p>';
+
     // EMPHASIS : _ or * for normal, __ or ** for strong
-    $search[] = "/__(.+)__/U";     $replace[] = "<strong>\\1</strong>";
-    $search[] = "/_(.+)_/U";       $replace[] = "<em>\\1</em>";
-    $search[] = "/\*\*(.+)\*\*/U"; $replace[] = "<strong>\\1</strong>";
-    $search[] = "/\*(.+)\*/U";     $replace[] = "<em>\\1</em>";
+    $search[] = '~([^\\\\]|^)(__(.+)__)~U';
+    $replace[] = '\1<strong>\3</strong>';
+    $search[] = '~([^\\\\_]|^)(_(.+)_)~U';
+    $replace[] = '\1<em>\3</em>';
+
+    $search[] = '~([^\\\\]|^)(\*\*(.+)\*\*)~U';
+    $replace[] = '\1<strong>\3</strong>';
+    $search[] = '~([^\\\\*]|^)(\*(.+)\*)~U';
+    $replace[] = '\1<em>\3</em>';
 
     // CODE : `code` -> <code>
-    $search[] = "/\\n?\`((\\n|.)+)\`/Um";
-    $replace[] = "<code>\\1</code>";
+    $search[] = "~\\n?([^\\\\]|^)(\`((\\n|.)+)\`)~Um";
+    $replace[] = '\1<code>\3</code>';
 
     // LINKS (can be relative)
     // images
-    $search[] = "(!\[(.+)\]\((.+)\))";
-    $replace[] = "<img src='\\2' alt='\\1'/>";
+    $search[] = '~([^\\\\]|^)(!\[(.+)\]\((.+)\))~';
+    $replace[] = '<img src=\'\4\' alt=\'\3\'/>';
     // [Google Site](http://google.fr/ "With help bubble")
-    $search[] = "(\[(.+)\]\((.+) \"(.+)\"\))";
-    $replace[] = "<a href='\\2' title='\\3'>\\1</a>";
+    $search[] = '~([^\\\\!]|^)(\[(.+)\]\((.+) &quot;(.+)&quot;\))~';
+    $replace[] = '\1<a href=\'\4\' title=\'\5\'>\3</a>';
     // [Google Site](http://google.fr/)
-    $search[] = "(\[(.+)\]\((.+)\))";  $replace[] = "<a href='\\2'>\\1</a>";
+    $search[] = '~([^\\\\!]|^)(\[(.+)\]\((.+)\))~U';
+    $replace[] = '\1<a href=\'\4\'>\3</a>';
+
     // Automatics links :
     // <http://dominique.fournier38.fr>
     // <dominique@fournier38.fr>
-    $search[] = "#&lt;(https?://.+)&gt;#U";
-    $replace[] = "<a href='\\1'>\\1</a>";
-    $search[] = "#&lt;(.+@.+)&gt;#U";
-    $replace[] = "<a href='mailto:\\1'>\\1</a>";
+    $search[] = '~([^\\\\]|^)(&lt;(https?://.+)&gt;)~U';
+    $replace[] = '\1<a href=\'\3\'>\3</a>';
+    $search[] = '~([^\\\\]|^)(&lt;(.+@.+)&gt;)~U';
+    $replace[] = '\1<a href=\'mailto:\3\'>\3</a>';
     // The links must not allow the <em> : redo the conversion
-    $search[] = "#(<a href='.*)<em>(.*)</em>(.*'>.*)<em>(.*)</em>(.*</a>)#";
-    $replace[] = "\\1_\\2_\\3_\\4_\\5";
+    $search[] = '~(<a href=\'.*)<em>(.*)</em>(.*\'>.*)<em>(.*)</em>(.*</a>)~';
+    $replace[] = '\1_\2_\3_\4_\5';
     // TODO : Links by reference :
     // Voici un petit texte écrit par [Michel Fortin][mf].
     // [mf]: http://michelf.ca/ "Mon site web"
@@ -102,24 +112,36 @@ class markdown
     // TITLES
     // Titles ATX (Optionnal sharp at the end)
     // ###### Title6
-    $search[] = "/^###### (.+)( +#+)?$/Um";
-    $replace[] = "</p><h6>\\1</h6><p>";
+    $search[] = '~^([^\\\\]|^)?(###### (.+)( +#+)?)$~Um';
+    $replace[] = '</p><h6>\3</h6><p>';
     // ##### Title5
-    $search[] = "/^##### (.+)( +#+)?$/Um";
-    $replace[] = "</p><h5>\\1</h5><p>";
+    $search[] = '~^([^\\\\]|^)?(##### (.+)( +#+)?)$~Um';
+    $replace[] = '</p><h5>\3</h5><p>';
     // #### Title4
-    $search[] = "/^#### (.+)( +#+)?$/Um";
-    $replace[] = "</p><h4>\\1</h4><p>";
+    $search[] = '~^([^\\\\]|^)?(#### (.+)( +#+)?)$~Um';
+    $replace[] = '</p><h4>\3</h4><p>';
     // ### Title3
-    $search[] = "/^### (.+)( +#+)?$/Um";
-    $replace[] = "</p><h3>\\1</h3><p>";
+    $search[] = '~^([^\\\\]|^)?(### (.+)( +#+)?)$~Um';
+    $replace[] = '</p><h3>\3</h3><p>';
     // ## Title2
-    $search[] = "/^## (.+)( +#+)?$/Um";
-    $replace[] = "</p><h2>\\1</h2><p>";
+    $search[] = '~^([^\\\\]|^)?(## (.+)( +#+)?)$~Um';
+    $replace[] = '</p><h2>\3</h2><p>';
     // # Title1
-    $search[] = "/^# (.+)( +#+)?$/Um";
-    $replace[] = "</p><h1>\\1</h1><p>";
-    return preg_replace ($search, $replace, $line);
+    $search[] = '~^([^\\\\]|^)?(# (.+)( +#+)?)$~Um';
+    $replace[] = '</p><h1>\3</h1><p>';
+    // Remove the backslashes on the existing regex
+    foreach ($search as $s)
+    {
+      $s = str_replace ('([^\\\\]|^)?', '([\\\\])', $s);
+      $s = str_replace ('([^\\\\]|^)', '([\\\\])', $s);
+      $s = str_replace ('([^\\\\!]|^)', '([\\\\])', $s);
+      $s = str_replace ('([^\\\\*]|^)', '([\\\\])', $s);
+      $s = str_replace ('([^\\\\_]|^)', '([\\\\])', $s);
+      $search[] = $s;
+      $replace[] = '\2';
+    }
+    $res =  preg_replace ($search, $replace, $line);
+    return $res;
   }
 
   /** Return HTML code corresponding to the code block