improved Mojo::DOM::HTML performance slightly

mojolicious · Jul 15, 2014 · 502e791 · 502e791
1 parent 7176ed7
commit 502e791
Show file tree

Hide file tree

Showing 3 changed files with 23 additions and 23 deletions.
diff --git a/Changes b/Changes
@@ -1,5 +1,6 @@
 
 5.15  2014-07-15
+  - Improved Mojo::DOM::HTML performance slightly.
 
 5.14  2014-07-14
   - Improved all_text performance in Mojo::DOM.

diff --git a/lib/Mojo/DOM/CSS.pm b/lib/Mojo/DOM/CSS.pm
@@ -27,10 +27,7 @@ my $TOKEN_RE        = qr/
   ((?:[^[\\:\s,]|$ESCAPE_RE\s?)+)?   # Element
   ($PSEUDO_CLASS_RE*)?               # Pseudoclass
   ((?:$ATTR_RE)*)?                   # Attributes
-  (?:
-    \s*
-    ([>+~])                          # Combinator
-  )?
+  (?:\s*([>+~]))?                    # Combinator
 /x;
 
 sub match {

diff --git a/lib/Mojo/DOM/HTML.pm b/lib/Mojo/DOM/HTML.pm
@@ -22,29 +22,31 @@ my $ATTR_RE = qr/
   \s*
 /x;
 my $TOKEN_RE = qr/
-  ([^<]+)?                                          # Text
+  ([^<]+)?                                            # Text
   (?:
-    <\?(.*?)\?>                                     # Processing Instruction
-  |
-    <!--(.*?)--\s*>                                 # Comment
-  |
-    <!\[CDATA\[(.*?)\]\]>                           # CDATA
-  |
-    <!DOCTYPE(
-      \s+\w+
-      (?:(?:\s+\w+)?(?:\s+(?:"[^"]*"|'[^']*'))+)?   # External ID
-      (?:\s+\[.+?\])?                               # Int Subset
-      \s*
-    )>
-  |
-    <(
-      \s*
-      [^<>\s]+                                      # Tag
+    <(?:
+      \?(.*?)\?                                       # Processing Instruction
+    |
+      !(?:
+        --(.*?)--\s*                                  # Comment
+      |
+        \[CDATA\[(.*?)\]\]                            # CDATA
+      |
+        DOCTYPE(
+        \s+\w+                                        # Doctype
+        (?:(?:\s+\w+)?(?:\s+(?:"[^"]*"|'[^']*'))+)?   # External ID
+        (?:\s+\[.+?\])?                               # Int Subset
+        \s*)
+      )
+    |
+      (\s*
+      [^<>\s]+                                        # Tag
       \s*
-      (?:(?:$ATTR_RE){0,32766})*+                   # Attributes
+      (?:(?:$ATTR_RE){0,32766})*+                     # Attributes
+      )
     )>
   |
-    (<)                                             # Runaway "<"
+    (<)                                               # Runaway "<"
   )??
 /xis;