rewriting: support setting cookie_scope at collection level

js rewriting: add custom url rewrite option to per-url rewrite rules

pywb/framework/archivalrouter.py

@@ -61,7 +61,8 @@ class ArchivalRouter(object):
                               coll=coll,
                               use_abs_prefix=use_abs_prefix,
                               wburl_class=route.handler.get_wburl_type(),
-                              urlrewriter_class=UrlRewriter)
+                              urlrewriter_class=UrlRewriter,
+                              cookie_scope=route.cookie_scope)
 
         # Allow for applying of additional filters
         route.apply_filters(wbrequest, matcher)
@@ -99,6 +100,7 @@ class Route(object):
         self.request_class = request_class
         # collection id from regex group (default 0)
         self.coll_group = coll_group
+        self.cookie_scope = config.get('cookie_scope')
         self._custom_init(config)
 
     def is_handling(self, request_uri):

pywb/framework/wbrequestresponse.py

@@ -37,7 +37,8 @@ class WbRequest(object):
                  use_abs_prefix=False,
                  wburl_class=None,
                  urlrewriter_class=None,
-                 is_proxy=False):
+                 is_proxy=False,
+                 cookie_scope=None):
 
         self.env = env
 
@@ -69,10 +70,12 @@ class WbRequest(object):
         # wb_url present and not root page
         if wb_url_str != '/' and wburl_class:
             self.wb_url = wburl_class(wb_url_str)
-            self.urlrewriter = urlrewriter_class(self.wb_url, self.wb_prefix,
+            self.urlrewriter = urlrewriter_class(self.wb_url,
+                                                 self.wb_prefix,
                                                  host_prefix + rel_prefix,
                                                  rel_prefix,
-                                                 env.get('SCRIPT_NAME', '/'))
+                                                 env.get('SCRIPT_NAME', '/'),
+                                                 cookie_scope)
         else:
         # no wb_url, just store blank wb_url
             self.wb_url = None

pywb/rewrite/cookie_rewriter.py

@@ -77,8 +77,8 @@ class RootScopeCookieRewriter(WbUrlBaseCookieRewriter):
 
 
 #=================================================================
-def get_cookie_rewriter(rule):
-    if rule and rule.cookie_scope == 'root':
+def get_cookie_rewriter(cookie_scope):
+    if cookie_scope == 'root':
         return RootScopeCookieRewriter
     else:
         return MinimalScopeCookieRewriter

pywb/rewrite/regex_rewriters.py

@@ -90,13 +90,19 @@ class RegexRewriter(object):
 
     @staticmethod
     def parse_rules_from_config(config):
-        def parse_rule(obj):
-            match = obj.get('match')
-            replace = RegexRewriter.format(obj.get('replace', '{0}'))
-            group = obj.get('group', 0)
-            result = (match, replace, group)
-            return result
-        return map(parse_rule, config)
+        def run_parse_rules(rewriter):
+            def parse_rule(obj):
+                match = obj.get('match')
+                if 'rewrite' in obj:
+                    replace = RegexRewriter.archival_rewrite(rewriter)
+                else:
+                    replace = RegexRewriter.format(obj.get('replace', '{0}'))
+                group = obj.get('group', 0)
+                result = (match, replace, group)
+                return result
+            
+            return map(parse_rule, config)
+        return run_parse_rules
 
 
 #=================================================================

pywb/rewrite/rewriterules.py

@@ -52,11 +52,11 @@ class RewriteRules(BaseRule):
 
         rewriter_cls = self.rewriters[field]
 
-        rule_def_tuples = RegexRewriter.parse_rules_from_config(regexs)
+        #rule_def_tuples = RegexRewriter.parse_rules_from_config(regexs)
+        parse_rules_func = RegexRewriter.parse_rules_from_config(regexs)
 
         def extend_rewriter_with_regex(urlrewriter):
-            #import sys
-            #sys.stderr.write('\n\nEXTEND: ' + str(rule_def_tuples))
+            rule_def_tuples = parse_rules_func(urlrewriter)
             return rewriter_cls(urlrewriter, rule_def_tuples)
 
         self.rewriters[field] = extend_rewriter_with_regex

pywb/rewrite/url_rewriter.py

@@ -18,12 +18,14 @@ class UrlRewriter(object):
 
     PROTOCOLS = ['http:', 'https:', 'ftp:', 'mms:', 'rtsp:', 'wais:']
 
-    def __init__(self, wburl, prefix, full_prefix=None, rel_prefix=None, root_path=None):
+    def __init__(self, wburl, prefix, full_prefix=None, rel_prefix=None, root_path=None,
+                 cookie_scope=None):
         self.wburl = wburl if isinstance(wburl, WbUrl) else WbUrl(wburl)
         self.prefix = prefix
         self.full_prefix = full_prefix
         self.rel_prefix = rel_prefix if rel_prefix else prefix
         self.root_path = root_path if root_path else '/'
+        self.cookie_scope = cookie_scope
 
     def rewrite(self, url, mod=None):
         # if special protocol, no rewriting at all
@@ -84,8 +86,12 @@ class UrlRewriter(object):
         new_wburl = WbUrl(new_url)
         return UrlRewriter(new_wburl, self.prefix)
 
-    def get_cookie_rewriter(self, rule=None):
-        cls = get_cookie_rewriter(rule)
+    def get_cookie_rewriter(self, scope=None):
+        # collection scope overrides rule scope?
+        if self.cookie_scope:
+            scope = self.cookie_scope
+
+        cls = get_cookie_rewriter(scope)
         return cls(self)
 
     def __repr__(self):
@@ -151,5 +157,5 @@ class HttpsUrlRewriter(UrlRewriter):
     def rebase_rewriter(self, new_url):
         return self
 
-    def get_cookie_rewriter(self, rule=None):
+    def get_cookie_rewriter(self, scope=None):
         return None

pywb/rules.yaml

@@ -45,14 +45,20 @@ rules:
 
         parse_comments: true
 
-
-    # instagram rules
-    #=================================================================
-    - url_prefix: 'com,instagram'
+    - url_prefix: 'com,facebook'
       rewrite:
           cookie_scope: root
 
 
+    # instagram rules
+    #=================================================================
+    - url_prefix: 'net,cloudfront,'
+      rewrite:
+          js_regexs:
+              - match: '\burl\((//[^)]+)\)'
+                rewrite: true
+                group: 1
+
      # flickr rules
     #=================================================================
     - url_prefix: ['com,yimg,l)/g/combo', 'com,yimg,s)/pw/combo', 'com,yahooapis,yui)/combo']

pywb/webapp/pywb_init.py

@@ -179,7 +179,7 @@ def create_wb_router(passed_config={}):
     for name, value in collections.iteritems():
         if isinstance(value, BaseHandler):
             handler_dict[name] = value
-            routes.append(Route(name, value))
+            routes.append(Route(name, value, config=route_config))
             continue
 
         route_config = init_route_config(value, config)
@@ -187,7 +187,7 @@ def create_wb_router(passed_config={}):
         if route_config.get('index_paths') == '$liveweb':
             live = create_live_handler(route_config)
             handler_dict[name] = live
-            routes.append(Route(name, live))
+            routes.append(Route(name, live, config=route_config))
             continue
 
         query_handler = init_collection(route_config)