Changeset 647 for trunk/lib

Timestamp:

Oct 25, 2018 12:35:07 AM (6 years ago)

Author:

anonymous

Message:

Update hyperlinkTxt() URL look-behind

File:

• trunk/lib/Utilities.inc.php (modified) (2 diffs)

trunk/lib/Utilities.inc.php

@@ -252 +252 @@
 * @return   string          Same input text, but URLs hyperlinked.
 * @author   Quinn Comendant <quinn@strangecode.com>
-* @version  2.0
+* @version  2.2
 * @since    22 Mar 2015 23:29:04
 */
@@ -261 +261 @@
 
     // Capture the full URL into the first match and only the first X characters into the second match.
-    // This will match URLs not preceeded by " ' or = (URLs inside an attribute) or ` (Markdown quoted) or double-scheme (http://http://www.asdf.com)
+    // This will match URLs not preceded by " ' or = (URLs inside an attribute) or ` (Markdown quoted) or double-scheme (http://http://www.asdf.com)
     // Valid URL characters: ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789-._~:/?#[]@!$&'()*+,;=
     $regex = '@
-        \b                              # Start with a word-boundary.
-        (?<!"|\'|=|>|`|[\w-]{2}://)     # Negative look-behind to exclude URLs already in <a> tag, Markdown quoted, or double SCHEME://
-        (                               # Begin match 1
-            (                           # Begin match 2
-                (?:%s)                  # URL starts with known scheme or www. if strict = false
-                [^\s/$.?#]+             # Any domain-valid characters
-                [^\s"`<>]{1,%s}         # Match 2 is limited to a maximum of LENGTH valid URL characters
+        \b                                 # Start with a word-boundary.
+        (?<!"|\'|=|>|`|\]\(|\[\d\] |[:/]/) # Negative look-behind to exclude URLs already in <a> tag, <tags>beween</tags>, `Markdown quoted`, [Markdown](link), [1] www.markdown.footnotes, and avoid broken:/ and doubled://schemes://
+        (                                  # Begin match 1
+            (                              # Begin match 2
+                (?:%s)                     # URL starts with known scheme or www. if strict = false
+                [^\s/$.?#]+                # Any domain-valid characters
+                [^\s"`<>]{1,%s}            # Match 2 is limited to a maximum of LENGTH valid URL characters
             )
-            [^\s"`<>]*                  # Match 1 continues with any further valid URL characters
-            ([^\P{Any}\s
<>«»"—–%s])    # Final character not a space or common end-of-sentence punctuation (.,:;?!, etc). Using double negation set, see http://stackoverflow.com/a/4786560/277303
+            [^\s"`<>]*                     # Match 1 continues with any further valid URL characters
+            ([^\P{Any}\s
<>«»"—–%s])       # Final character not a space or common end-of-sentence punctuation (.,:;?!, etc). Using double negation set, see http://stackoverflow.com/a/4786560/277303
         )
         @Suxi