error messaging: app path not found use default error.html template

- add AppPageNotFound() exception to differntiate app-level not found path from replay content not found
- add custom error messages for collectino not found and static file not found
tests: add tests for collection not found and static file not found errors

pywb/apps/frontendapp.py

@@ -251,7 +251,7 @@ class FrontEndApp(object):
                 response.add_access_control_headers(env=environ)
             return response
         except:
-            self.raise_not_found(environ, 'Static File Not Found: {0}'.format(filepath))
+            self.raise_not_found(environ, 'static_file_not_found', filepath)
 
     def get_metadata(self, coll):
         """Retrieve the metadata associated with a collection
@@ -282,7 +282,7 @@ class FrontEndApp(object):
         :rtype: WbResponse
         """
         if not self.is_valid_coll(coll):
-            self.raise_not_found(environ, 'No handler for "/{0}"'.format(coll))
+            self.raise_not_found(environ, 'coll_not_found', coll)
 
         self.setup_paths(environ, coll)
 
@@ -358,7 +358,7 @@ class FrontEndApp(object):
         :rtype: WbResponse
         """
         if not self.is_valid_coll(coll):
-            self.raise_not_found(environ, 'No handler for "/{0}"'.format(coll))
+            self.raise_not_found(environ, 'coll_not_found', coll)
 
         self.setup_paths(environ, coll, record)
 
@@ -440,14 +440,8 @@ class FrontEndApp(object):
         return (coll in self.warcserver.list_fixed_routes() or
                 coll in self.warcserver.list_dynamic_routes())
 
-    def raise_not_found(self, environ, msg):
-        """Utility function for raising a werkzeug.exceptions.NotFound execption with the supplied WSGI environment
-        and message.
-
-        :param dict environ: The WSGI environment dictionary for the request
-        :param str msg: The error message
-        """
-        raise NotFoundException(msg)
+    def raise_not_found(self, environ, err_type, url):
+        raise AppPageNotFound(err_type, url)
 
     def _check_refer_redirect(self, environ):
         """Returns a WbResponse for a HTTP 307 redirection if the HTTP referer header is the same as the HTTP host header
@@ -642,6 +636,13 @@ class FrontEndApp(object):
         return response
 
 
+# ============================================================================
+class AppPageNotFound(WbException):
+    @property
+    def status_code(self):
+        return 404
+
+
 # ============================================================================
 class MetadataCache(object):
     """This class holds the collection medata template string and

pywb/apps/rewriterapp.py

@@ -15,6 +15,9 @@ from pywb.rewrite.rewriteinputreq import RewriteInputRequest
 from pywb.rewrite.templateview import BaseInsertView, HeadInsertView, JinjaEnv, TopFrameView
 from pywb.rewrite.url_rewriter import IdentityUrlRewriter, UrlRewriter
 from pywb.rewrite.wburl import WbUrl
+from pywb.rewrite.url_rewriter import UrlRewriter, IdentityUrlRewriter
+
+from pywb.utils.wbexception import WbException, NotFoundException
 from pywb.rewrite.cookies import CookieTracker
 from pywb.utils.canonicalize import canonicalize
 from pywb.utils.io import BUFF_SIZE, OffsetLimitReader, no_except_close
@@ -566,7 +569,7 @@ class RewriterApp(object):
         return top_url
 
     def handle_error(self, environ, wbe):
-        if wbe.status_code == 404:
+        if isinstance(wbe, NotFoundException):
             return self._not_found_response(environ, wbe.url)
         else:
             return self._error_response(environ, wbe)

tests/test_integration.py

@@ -24,11 +24,13 @@ class TestWbIntegration(BaseConfigTest):
         assert resp.content_type == 'text/html'
         assert resp.status_int == 200
 
-    def test_pywb_invalid_path(self):
-        resp = self.testapp.head('/blah/', status=404)
+    def test_pywb_invalid_collection(self):
+        resp = self.testapp.get('/blah/http://example.com/', status=404)
         assert resp.content_type == 'text/html'
         assert resp.status_int == 404
 
+        assert 'Collection not found: <b>blah</b>' in resp.text
+
     def test_calendar_query(self):
         resp = self.testapp.get('/pywb/*/iana.org')
         self._assert_basic_html(resp)
@@ -429,6 +431,8 @@ class TestWbIntegration(BaseConfigTest):
         resp = self.testapp.get('/static/notfound.css', status = 404)
         assert resp.status_int == 404
 
+        assert 'Static file not found: <b>notfound.css</b>' in resp.text
+
     def test_cdx_server_filters(self):
         resp = self.testapp.get('/pywb/cdx?url=http://www.iana.org/_css/2013.1/screen.css&filter=mime:warc/revisit&filter=filename:dupes.warc.gz')
         assert resp.content_type == 'text/x-cdxj'
@@ -451,7 +455,6 @@ class TestWbIntegration(BaseConfigTest):
         origfilenames = list(map(lambda cdx: cdx['orig.filename'], cdxs))
         assert origfilenames == ['iana.warc.gz', 'iana.warc.gz', '-']
 
-
     # surt() no longer errors on this in 0.3b
     #def test_error(self):
     #    resp = self.testapp.get('/pywb/?abc', status = 400)