add basic wsgi app for parsing archivalurls, fallback on a referrer based redirect

pywb/__init__.py

@@ -0,0 +1,2 @@
+#Allow importing
+

pywb/aurl.py

@@ -3,26 +3,31 @@
 import re
 import rfc3987
 
+import wbexceptions
+
 # aurl : ArchivalUrl representation for WB
 
 class aurl:
     """
     # Replay Urls
     # ======================
-    >>> print_test(aurl('/20131010000506/example.com'))
-    ('replay', '20131010000506', None, 'example.com')
+    >>> repr(aurl('/20131010000506/example.com'))
+    "('replay', '20131010000506', '', 'example.com', '/20131010000506/example.com')"
 
-    >>> print_test(aurl('/20130102im_/example.com'))
-    ('replay', '20130102', 'im_', 'example.com')
+    >>> repr(aurl('/20130102im_/example.com'))
+    "('replay', '20130102', 'im_', 'example.com', '/20130102im_/example.com')"
 
-    >>> print_test(aurl('/https://example.com/xyz'))
-    ('latest_replay', None, None, 'https://example.com/xyz')
+    >>> repr(aurl('/cs_/example.com'))
+    "('latest_replay', '', 'cs_', 'example.com', '/cs_/example.com')"
+
+    >>> repr(aurl('/https://example.com/xyz'))
+    "('latest_replay', '', '', 'https://example.com/xyz', '/https://example.com/xyz')"
 
 
     # Query Urls
     # ======================
-    >>> print_test(aurl('/*/http://example.com/abc?def=a'))
-    ('query', None, None, 'http://example.com/abc?def=a')
+    >>> repr(aurl('/*/http://example.com/abc?def=a'))
+    "('query', '', '', 'http://example.com/abc?def=a', '/*/http://example.com/abc?def=a')"
 
 
     # Error Urls
@@ -33,34 +38,34 @@ class aurl:
 
     >>> x = aurl('/#$%#/')
     Traceback (most recent call last):
-    RequestParseException: Bad Request Url: #$%#/
+    BadUrlException: Bad Request Url: #$%#/
 
     >>> x = aurl('/http://example.com:abc/')
     Traceback (most recent call last):
-    RequestParseException: Bad Request Url: http://example.com:abc/
+    BadUrlException: Bad Request Url: http://example.com:abc/
     """
 
     # Regexs
     # ======================
     QUERY_REGEX = re.compile('^/(\d{1,14})?\*/(.*)$')
-    REPLAY_REGEX = re.compile('^(/(\d{1,14})([a-z]{2}_)?)?/(.*)$')
+    REPLAY_REGEX = re.compile('^/(\d{1,14})?([a-z]{2}_)?/?(.*)$')
     # ======================
 
 
     def __init__(self, url):
         self.original_url = url
         self.type = None
-        self.url = None
-        self.timestamp = None
-        self.mod = None
+        self.url = ''
+        self.timestamp = ''
+        self.mod = ''
 
         if not any (f(self, url) for f in [aurl._init_query, aurl._init_replay]):
-            raise RequestParseException('Invalid WB Request Url: ' + url)
+            raise wbexceptions.RequestParseException('Invalid WB Request Url: ' + url)
 
         matcher = rfc3987.match(self.url, 'IRI_reference')
 
         if not matcher:
-            raise RequestParseException('Bad Request Url: ' + self.url)
+            raise wbexceptions.BadUrlException('Bad Request Url: ' + self.url)
 
     # Match query regex
     # ======================
@@ -69,8 +74,10 @@ class aurl:
         if not query:
             return None
 
-        self.timestamp = query.group(1)
-        self.url = query.group(2)
+        res = query.groups('')
+
+        self.timestamp = res[0]
+        self.url = res[1]
         self.type = 'query'
         return True
 
@@ -81,9 +88,11 @@ class aurl:
         if not replay:
             return None
 
-        self.timestamp = replay.group(2)
-        self.mod = replay.group(3)
-        self.url = replay.group(4)
+        res = replay.groups('')
+
+        self.timestamp = res[0]
+        self.mod = res[1]
+        self.url = res[2]
         if self.timestamp:
             self.type = 'replay'
         else:
@@ -91,16 +100,25 @@ class aurl:
 
         return True
 
+    # Str Representation
+    # ====================
+    def __str__(self):
+        if self.type == 'query':
+            return "/*/" + self.url
+        else:
+            tsmod = self.timestamp + self.mod
+            if len(tsmod) > 0:
+                return "/" + tsmod + "/" + self.url
+            else:
+                return "/" + self.url
 
-class RequestParseException(Exception):
-    pass
-
-
+    def __repr__(self):
+        return str((self.type, self.timestamp, self.mod, self.url, str(self)))
 
 if __name__ == "__main__":
     import doctest
 
-    def print_test(self):
-        return self.type, self.timestamp, self.mod, self.url
+    #def print_test(self):
+    #    return self.type, self.timestamp, self.mod, self.url, str(self)
 
     doctest.testmod()

pywb/refer_redirect.py

@@ -0,0 +1,75 @@
+import aurl
+import urlparse
+from wbrequestresponse import WbRequest, WbResponse
+
+# Redirect urls that have 'fallen through' based on the referrer
+# settings
+class ReferRedirect:
+
+    """
+    >>> ReferRedirect('http://localhost:8080/').matchPrefixs
+    ['http://localhost:8080/']
+
+    >>> ReferRedirect(['http://example:9090/']).matchPrefixs
+    ['http://example:9090/']
+
+    >>> test_redir('http://localhost:8080/', '/other.html', 'http://localhost:8080/coll/20131010/http://example.com/path/page.html')
+    'http://localhost:8080/coll/20131010/http://example.com/path/other.html'
+
+    >>> test_redir('http://localhost:8080/', '/../other.html', 'http://localhost:8080/coll/20131010/http://example.com/path/page.html')
+    'http://localhost:8080/coll/20131010/http://example.com/other.html'
+
+    >>> test_redir('http://localhost:8080/', '/../../other.html', 'http://localhost:8080/coll/20131010/http://example.com/index.html')
+    'http://localhost:8080/coll/20131010/http://example.com/other.html'
+
+    >>> test_redir('http://example:8080/', '/other.html', 'http://localhost:8080/coll/20131010/http://example.com/path/page.html')
+    False
+    """
+
+    def __init__(self, matchPrefixs):
+        if isinstance(matchPrefixs, list):
+            self.matchPrefixs = matchPrefixs
+        else:
+            self.matchPrefixs = [matchPrefixs]
+
+    def run(self, wbrequest):
+        if wbrequest.referrer is None:
+            return None
+
+        if not any (wbrequest.referrer.startswith(i) for i in self.matchPrefixs):
+            return None
+
+        try:
+            ref_split = urlparse.urlsplit(wbrequest.referrer)
+            ref_path = ref_split.path[1:].split('/', 1)
+
+            ref_wb_url = aurl.aurl('/' + ref_path[1])
+
+            ref_wb_url.url = urlparse.urljoin(ref_wb_url.url, wbrequest.request_uri[1:])
+            ref_wb_url.url = ref_wb_url.url.replace('../', '')
+
+            final_url = urlparse.urlunsplit((ref_split.scheme, ref_split.netloc, ref_path[0] + str(ref_wb_url), '', ''))
+
+        except Exception as e:
+            return None
+
+        return WbResponse.redir_response(final_url)
+
+if __name__ == "__main__":
+    import doctest
+
+    def test_redir(matchHost, request_uri, referrer):
+        env = {'REQUEST_URI': request_uri, 'HTTP_REFERER': referrer}
+
+        redir = ReferRedirect(matchHost)
+        req = WbRequest(env)
+        rep = redir.run(req)
+        if not rep:
+            return False
+
+        return rep.get_header('Location')
+
+
+    doctest.testmod()
+
+

pywb/run.sh

@@ -0,0 +1,8 @@
+#!/bin/sh
+
+app=$1
+if [ -z "$app" ]; then
+  app=wbapp.py
+fi
+
+uwsgi --http :9090 --wsgi-file $app

pywb/wbapp.py

@@ -0,0 +1,59 @@
+from wbrequestresponse import WbRequest, WbResponse
+from refer_redirect import ReferRedirect
+import aurl
+
+class WBHandler:
+    def run(self, wbrequest):
+        wburl = aurl.aurl(wbrequest.wb_url)
+        return WbResponse.text_response(repr(wburl))
+
+class ArchivalParser:
+    def __init__(self, mappings, hostpaths=None):
+        self.mappings = mappings
+        self.fallback = ReferRedirect(hostpaths)
+
+    def find_handler(self, env):
+        request_uri = env['REQUEST_URI']
+
+        for key, value in self.mappings.iteritems():
+            if request_uri.startswith(key):
+                env['WB_URL'] = request_uri[len(key)-1:]
+                env['WB_COLL'] = key[1:-1]
+                #print "Found: " + str(value) + " for " + key
+                return value
+
+        return self.fallback
+
+    def handle_request(self, env):
+        handler = self.find_handler(env)
+        return handler.run(WbRequest(env))
+
+    def handle_exception(self, env, exc):
+        return WbResponse.text_response('Error: ' + str(exc), status = '400 Bad Request')
+
+    def handle_not_found(self, env):
+        return WbResponse.text_response('Not Found: ' + env['REQUEST_URI'], status = '404 Not Found')
+
+
+
+## ===========
+parser = ArchivalParser({'/web/': WBHandler()}, hostpaths = ['http://localhost:9090/'])
+## ===========
+
+
+def application(env, start_response):
+    response = None
+
+    try:
+        response = parser.handle_request(env)
+
+    except Exception as e:
+        last_exc = e
+        import traceback
+        traceback.print_exc()
+        response = parser.handle_exception(env, e)
+
+    if not response:
+        response = parser.handle_not_found(env)
+
+    return response(env, start_response)

pywb/wbexceptions.py

@@ -0,0 +1,8 @@
+
+class RequestParseException(Exception):
+    pass
+
+class BadUrlException(Exception):
+    pass
+
+

pywb/wbrequestresponse.py

@@ -0,0 +1,48 @@
+#from werkzeug.wrappers import BaseRequest, BaseResponse, CommonRequestDescriptorsMixin, CommonResponseDescriptorsMixin
+
+# WbRequest
+class WbRequest:
+    def __init__(self, env):
+        self.env = env
+        self.wb_url = env.get('WB_URL')
+        self.coll = env.get('WB_COLL')
+
+        setattr(self, 'request_uri', env.get('REQUEST_URI'))
+        setattr(self, 'referrer', env.get('HTTP_REFERER'))
+
+    def __repr__(self):
+        return self.coll + " " + self.wb_url
+
+
+class WbResponse:
+    def __init__(self, status, value = [], headersList = []):
+        self.status = status
+        self.body = value
+        self.headersList = headersList
+
+    @staticmethod
+    def text_response(text, status = '200 OK'):
+        return WbResponse(status, value = [text], headersList = [('Content-Type', 'text/plain')])
+
+    @staticmethod
+    def redir_response(location):
+        return WbResponse('302 Redirect', headersList = [('Location', location)])
+
+    def get_header(self, name):
+        name_upp = name.upper()
+        for value in self.headersList:
+            if (value[0].upper() == name_upp):
+                return value[1]
+
+    def __call__(self, env, start_response):
+        #headersList = []
+        #for key, value in self.headers.iteritems():
+        #    headersList.append((key, value))
+
+        start_response(self.status, self.headersList)
+        return self.body
+
+
+
+
+