Optimize several regexes from quadratic time to linear time

Part of the discussion in Python-Markdown#798.

Signed-off-by: Anders Kaseorg <[email protected]>

Author: andersk

markdown/inlinepatterns.py

@@ -147,13 +147,13 @@ def build_inlinepatterns(md, **kwargs):
 NOT_STRONG_RE = r'((^|\s)(\*|_)(\s|$))'
 
 # <http://www.123.com>
-AUTOLINK_RE = r'<((?:[Ff]|[Hh][Tt])[Tt][Pp][Ss]?://[^>]*)>'
+AUTOLINK_RE = r'<((?:[Ff]|[Hh][Tt])[Tt][Pp][Ss]?://[^<>]*)>'
 
 # <[email protected]>
-AUTOMAIL_RE = r'<([^> \!]*@[^> ]*)>'
+AUTOMAIL_RE = r'<([^<> !]*@[^@<> ]*)>'
 
 # <...>
-HTML_RE = r'(\<([a-zA-Z/][^\>]*?|\!--.*?--)\>)'
+HTML_RE = r'(<([a-zA-Z/][^<>]*|!--[^<>]*--)>)'
 
 # "&#38;" (decimal) or "&#x26;" (hex) or "&amp;" (named)
 ENTITY_RE = r'(&(?:\#[0-9]+|\#x[0-9a-fA-F]+|[a-zA-Z0-9]+);)'
@@ -433,7 +433,7 @@ def get_stash(m):
 
 class LinkInlineProcessor(InlineProcessor):
     """ Return a link element from the given match. """
-    RE_LINK = re.compile(r'''\(\s*(?:(<.*?>)\s*(?:(['"])(.*?)\2\s*)?\))?''', re.DOTALL | re.UNICODE)
+    RE_LINK = re.compile(r'''\(\s*(?:(<[^<>]*>)\s*(?:('[^']*'|"[^"]*")\s*)?\))?''', re.DOTALL | re.UNICODE)
     RE_TITLE_CLEAN = re.compile(r'\s')
 
     def handleMatch(self, m, data):
@@ -467,8 +467,8 @@ def getLink(self, data, index):
         if m and m.group(1):
             # Matches [Text](<link> "title")
             href = m.group(1)[1:-1].strip()
-            if m.group(3):
-                title = m.group(3)
+            if m.group(2):
+                title = m.group(2)[1:-1]
             index = m.end(0)
             handled = True
         elif m: