rewrite: fix WbUrl parsing for urls that start with a digit, eg. 1234.example.com

split latest replay url from timestamped replay regex
add additional rewrite tests

.coveragerc

@@ -5,6 +5,7 @@ omit =
     *.html
     *.js
     *.css
+    pywb/__init__.py
 
 [report]
 exclude_lines =

pywb/rewrite/test/test_wburl.py

@@ -20,6 +20,18 @@ ur"""
 >>> repr(WbUrl('cs_/example.com'))
 "('latest_replay', '', 'cs_', 'http://example.com', 'cs_/http://example.com')"
 
+>>> repr(WbUrl('im_/20130102.org'))
+"('latest_replay', '', 'im_', 'http://20130102.org', 'im_/http://20130102.org')"
+
+>>> repr(WbUrl('20130102.example.com'))
+"('latest_replay', '', '', 'http://20130102.example.com', 'http://20130102.example.com')"
+
+>>> repr(WbUrl('20130102.org/1'))
+"('latest_replay', '', '', 'http://20130102.org/1', 'http://20130102.org/1')"
+
+>>> repr(WbUrl('20130102/1.com'))
+"('replay', '20130102', '', 'http://1.com', '20130102/http://1.com')"
+
 >>> repr(WbUrl('https://example.com/xyz'))
 "('latest_replay', '', '', 'https://example.com/xyz', 'https://example.com/xyz')"
 

pywb/rewrite/wburl.py

@@ -86,7 +86,8 @@ class WbUrl(BaseWbUrl):
     # Regexs
     # ======================
     QUERY_REGEX = re.compile('^(?:([\w\-:]+)/)?(\d*)(?:-(\d+))?\*/?(.+)$')
-    REPLAY_REGEX = re.compile('^(\d*)([a-z]+_)?/{0,3}(.+)$')
+    REPLAY_REGEX = re.compile('^(\d*)([a-z]+_)?/{1,3}(.+)$')
+    #LATEST_REPLAY_REGEX = re.compile('^\w_)')
 
     DEFAULT_SCHEME = 'http://'
 
@@ -221,8 +222,15 @@ class WbUrl(BaseWbUrl):
     def _init_replay(self, url):
         replay = self.REPLAY_REGEX.match(url)
         if not replay:
+            if not url:
                 return None
 
+            self.timestamp = ''
+            self.mod = ''
+            self.url = url
+            self.type = self.LATEST_REPLAY
+            return True
+
         res = replay.groups('')
 
         self.timestamp = res[0]