backup/pywb
1
0
Fork 0
mirror of https://github.com/webrecorder/pywb.git synced 2025-03-15 00:03:28 +01:00
Code Issues Projects Releases Activity

Improve handling of exceptions in wsgi_wrappers, fixes #219 (#220)

Browse Source 
* Improve handling of exceptions in wsgi_wrappers, fixes #219

* Update Common Crawl public data set location
This commit is contained in:
Sebastian Nagel 2017-06-17 12:41:52 +02:00 committed by Ilya Kreymer
parent 29da503321
commit 3e8e590c1b
2 changed files with 16 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
pywb/framework/wsgi_wrappers.py
View File

@ -86,7 +86,19 @@ class WSGIApp(object):
error_view = self.wb_router.error_view
if hasattr(exc, 'status'):
status = exc.status()
if callable(exc.status):
status = exc.status()
else:
status = exc.status
# wsgi requires status
# - to have at least 4 characters and
# - to start with a number / integer
if type(status) == int:
status = '{} Exception {}'.format(status, type(exc).__name__)
elif type(status) == str and status[0].isdigit():
pass
else:
status = '500 Internal Server Error'
else:
status = '500 Internal Server Error'
@ -96,7 +108,7 @@ class WSGIApp(object):
err_url = None
if len(exc.args):
err_msg = exc.args[0]
err_msg = str(exc.args[0])
if print_trace:
import traceback
@ -125,7 +137,7 @@ class WSGIApp(object):
#msg = msg.encode('utf-8', 'ignore')
return WbResponse.text_response(msg,
status=status)
status=status)
#=================================================================
DEFAULT_CONFIG_FILE = 'config.yaml'

2
pywb/utils/test/test_loaders.py
View File

@ -166,7 +166,7 @@ def seek_read_full(seekable_reader, offset):
def test_s3_read_1():
pytest.importorskip('boto')
res = BlockLoader().load('s3://aws-publicdatasets/common-crawl/crawl-data/CC-MAIN-2015-11/segments/1424936462700.28/warc/CC-MAIN-20150226074102-00159-ip-10-28-5-156.ec2.internal.warc.gz',
res = BlockLoader().load('s3://commoncrawl/crawl-data/CC-MAIN-2015-11/segments/1424936462700.28/warc/CC-MAIN-20150226074102-00159-ip-10-28-5-156.ec2.internal.warc.gz',
offset=53235662,
length=2526)