access system work for ukwa/ukwa-pywb#7

- 'acl_paths' config can accept a list of files or directories, a file or a directory string
- tests_acl: test collection with acl list, single file, dir

pywb/utils/merge.py

@@ -35,8 +35,10 @@ else:  #pragma: no cover
         the input streams is already sorted (smallest to largest).
         >>> list(merge([1,3,5,7], [0,2,4,8], [5,10,15,20], [], [25]))
         [0, 1, 2, 3, 4, 5, 5, 7, 8, 10, 15, 20, 25]
+
         If *key* is not None, applies a key function to each element to determine
         its sort order.
+
         >>> list(merge(['dog', 'horse'], ['cat', 'fish', 'kangaroo'], key=len))
         ['dog', 'cat', 'fish', 'horse', 'kangaroo']
         '''

pywb/warcserver/access_checker.py

@@ -1,5 +1,6 @@
 from pywb.warcserver.index.indexsource import FileIndexSource
 from pywb.warcserver.index.aggregator import DirectoryIndexSource, CacheDirectoryMixin
+from pywb.warcserver.index.aggregator import SimpleAggregator
 
 from pywb.utils.binsearch import search
 from pywb.utils.merge import merge
@@ -18,13 +19,21 @@ class FileAccessIndexSource(FileIndexSource):
 
 
 # ============================================================================
-class DirectoryAccessSource(DirectoryIndexSource):
-    INDEX_SOURCES = [('.aclj', FileAccessIndexSource)]
-
+class ReverseMergeMixin(object):
     def _merge(self, iter_list):
         return merge(*(iter_list), reverse=True)
 
 
+# ============================================================================
+class AccessRulesAggregator(ReverseMergeMixin, SimpleAggregator):
+    pass
+
+
+# ============================================================================
+class DirectoryAccessSource(ReverseMergeMixin, DirectoryIndexSource):
+    INDEX_SOURCES = [('.aclj', FileAccessIndexSource)]
+
+
 # ============================================================================
 class CacheDirectoryAccessSource(CacheDirectoryMixin, DirectoryAccessSource):
     pass
@@ -32,15 +41,25 @@ class CacheDirectoryAccessSource(CacheDirectoryMixin, DirectoryAccessSource):
 
 # ============================================================================
 class AccessChecker(object):
-    def __init__(self, access_source_file, default_access='allow'):
-        if isinstance(access_source_file, str):
-            self.aggregator = self.create_access_aggregator(access_source_file)
+    def __init__(self, access_source, default_access='allow'):
+        if isinstance(access_source, str):
+            self.aggregator = self.create_access_aggregator([access_source])
+        elif isinstance(access_source, list):
+            self.aggregator = self.create_access_aggregator(access_source)
         else:
-            self.aggregator = access_source_file
+            self.aggregator = access_source
 
         self.default_rule = {'urlkey': '', 'access': default_access}
 
-    def create_access_aggregator(self, filename):
+    def create_access_aggregator(self, source_files):
+        sources = {}
+        for filename in source_files:
+            sources[filename] = self.create_access_source(filename)
+
+        aggregator = AccessRulesAggregator(sources)
+        return aggregator
+
+    def create_access_source(self, filename):
         if os.path.isdir(filename):
             return CacheDirectoryAccessSource(filename)
 
@@ -52,18 +71,18 @@ class AccessChecker(object):
 
     def find_access_rule(self, url, ts=None, urlkey=None):
         params = {'url': url, 'urlkey': urlkey}
-        cdx_iter, errs = self.aggregator(params)
+        acl_iter, errs = self.aggregator(params)
         if errs:
             print(errs)
 
         key = params['key'].decode('utf-8')
 
-        for cdx in cdx_iter:
-            if 'urlkey' not in cdx:
+        for acl in acl_iter:
+            if 'urlkey' not in acl:
                 continue
 
-            if key.startswith(cdx['urlkey']):
-                return cdx
+            if key.startswith(acl['urlkey']):
+                return acl
 
         return self.default_rule
 

pywb/warcserver/warcserver.py

@@ -178,9 +178,9 @@ class WarcServer(BaseWarcServer):
         else:
             raise Exception('collection config must be string or dict')
 
+        # INDEX CONFIG
         if index:
             agg = init_index_agg({name: index})
-
         else:
             if not isinstance(coll_config, dict):
                 raise Exception('collection config missing')
@@ -196,9 +196,11 @@ class WarcServer(BaseWarcServer):
             timeout = int(coll_config.get('timeout', 0))
             agg = init_index_agg(index_group, True, timeout)
 
+        # ARCHIVE CONFIG
         if not archive_paths:
             archive_paths = self.config.get('archive_paths')
 
+        # ACCESS CONFIG
         access_checker = None
         if acl_paths:
             access_checker = AccessChecker(acl_paths, default_access)

sample_archive/access/list2.aclj

@@ -1,2 +1,3 @@
+org,httpbin)/ - {"access": "allow"}
 com,example)/ - {"access": "allow"}
 bo,example)/ - {"access": "exclude"}

tests/config_test_access.yaml

@@ -4,6 +4,24 @@ collections:
     pywb:
         index_paths: ./sample_archive/cdx/
         archive_paths: ./sample_archive/warcs/
-        acl_paths: ./sample_archive/access/
+        acl_paths: ./sample_archive/access/pywb.aclj
+
         default_access: block
 
+    pywb-acl-list:
+        index_paths: ./sample_archive/cdx/
+        archive_paths: ./sample_archive/warcs/
+        acl_paths:
+            - ./sample_archive/access/pywb.aclj
+            - ./sample_archive/access/list2.aclj
+
+        default_access: block
+
+    pywb-acl-dir:
+        index_paths: ./sample_archive/cdx/
+        archive_paths: ./sample_archive/warcs/
+        acl_paths: ./sample_archive/access/
+
+        default_access: block
+
+

tests/test_acl.py

@@ -12,9 +12,10 @@ class TestACLApp(BaseConfigTest):
     def setup_class(cls):
         super(TestACLApp, cls).setup_class('config_test_access.yaml')
 
-    def query(self, url, is_error=False, **params):
+    def query(self, url, coll='pywb'):
+        params = {}
         params['url'] = url
-        return self.testapp.get('/pywb/cdx?' + urlencode(params, doseq=1), expect_errors=is_error)
+        return self.testapp.get('/{coll}/cdx?'.format(coll=coll) + urlencode(params, doseq=1))
 
     def test_excluded_url(self):
         resp = self.query('http://www.iana.org/')
@@ -52,4 +53,23 @@ class TestACLApp(BaseConfigTest):
 
         assert 'Access Blocked' in resp.text
 
+    def test_allowed_different_coll_acl_list(self):
+        resp = self.query('http://httpbin.org/anything/resource.json', coll='pywb-acl-list')
+
+        assert len(resp.text.splitlines()) > 0
+
+        resp = self.testapp.get('/pywb-acl-list/mp_/http://httpbin.org/anything/resource.json')
+
+        assert '"http://httpbin.org/anything/resource.json"' in resp.text
+
+    def test_allowed_different_coll_acl_dir(self):
+        resp = self.query('http://httpbin.org/anything/resource.json', coll='pywb-acl-dir')
+
+        assert len(resp.text.splitlines()) > 0
+
+        resp = self.testapp.get('/pywb-acl-dir/mp_/http://httpbin.org/anything/resource.json')
+
+        assert '"http://httpbin.org/anything/resource.json"' in resp.text
+
+