backup/warcprox
1
0
Fork 0
mirror of https://github.com/internetarchive/warcprox.git synced 2025-01-18 13:22:09 +01:00
Code Issues Projects Releases Activity

catch exception handling special request (currently that means PUTMETA)

Browse Source
This commit is contained in:
Noah Levitt 2015-07-23 00:37:02 +00:00
parent f00602b764
commit eb7de9d3f9
2 changed files with 27 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
warcprox/mitmproxy.py
View File

@ -117,10 +117,10 @@ class MitmProxyHandler(http_server.BaseHTTPRequestHandler):
def do_COMMAND(self):
if not self.is_connect:
if self.command == 'PUTMETA':
self._prepare_custom_record(method=self.command, type_='metadata')
self._special_request(method=self.command, type_='metadata')
return
# if self.command == 'PUTRES':
# self._handle_custom_record(type_='resource')
# self._special_request(type_='resource')
# return
try:
@ -137,7 +137,7 @@ class MitmProxyHandler(http_server.BaseHTTPRequestHandler):
self._proxy_request()
def _handle_custom_record(self, method, type_):
def _special_request(self, method, type_):
raise Exception('Not supported')
def _proxy_request(self):

43
warcprox/warcprox.py
View File

@ -216,30 +216,35 @@ class WarcProxyHandler(warcprox.mitmproxy.MitmProxyHandler):
return recorded_url
def _handle_custom_record(self, method, type_):
self.url = self.path
def _special_request(self, method, type_):
try:
self.url = self.path
if 'Content-Length' in self.headers and 'Content-Type' in self.headers:
request_data = self.rfile.read(int(self.headers['Content-Length']))
if (method == 'PUTMETA' and 'Content-Length' in self.headers
and 'Content-Type' in self.headers):
request_data = self.rfile.read(int(self.headers['Content-Length']))
warcprox_meta = self.headers.get('Warcprox-Meta')
warcprox_meta = self.headers.get('Warcprox-Meta')
rec_custom = RecordedUrl(url=self.url,
request_data=request_data,
response_recorder=None,
remote_ip=b'',
warcprox_meta=warcprox_meta,
content_type=self.headers['Content-Type'].encode('latin1'),
custom_type=type_,
method=method,
status=204, size=len(request_data))
rec_custom = RecordedUrl(url=self.url,
request_data=request_data,
response_recorder=None,
remote_ip=b'',
warcprox_meta=warcprox_meta,
content_type=self.headers['Content-Type'].encode('latin1'),
custom_type=type_,
method=method,
status=204, size=len(request_data))
self.server.recorded_url_q.put(rec_custom)
self.send_response(204, 'OK')
else:
self.send_error(400, 'Bad request')
self.server.recorded_url_q.put(rec_custom)
self.send_response(204, 'OK')
else:
self.send_error(400, 'Bad request')
self.end_headers()
self.end_headers()
except:
self.logger.error("uncaught except in _special_request", exc_info=True)
raise
def log_error(self, fmt, *args):
# logging better handled elsewhere?