From 3d9a22b6c7855f60dea5f2f772432d548a6e6fd4 Mon Sep 17 00:00:00 2001 From: Vangelis Banos Date: Sun, 29 Oct 2017 18:48:08 +0000 Subject: [PATCH] Return capture timestamp When client request has HTTP header ``Warcprox-Meta": {"return-capture-timestamp": 1}``, add to the response the WARC record timestamp in the following HTTP header: ``Warcprox-Meta: {"capture-timestamp": '%Y-%m-%d %H:%M:%S"}``. Add unit test. --- tests/test_warcprox.py | 15 +++++++++++++++ warcprox/mitmproxy.py | 14 +++++++++++--- warcprox/warcproxy.py | 7 ++++++- 3 files changed, 32 insertions(+), 4 deletions(-) diff --git a/tests/test_warcprox.py b/tests/test_warcprox.py index b24a5c8..22d4597 100755 --- a/tests/test_warcprox.py +++ b/tests/test_warcprox.py @@ -555,6 +555,21 @@ def test_limits(http_daemon, warcprox_, archiving_proxies): assert response.headers["content-type"] == "text/plain;charset=utf-8" assert response.raw.data == b"request rejected by warcprox: reached limit test_limits_bucket/total/urls=10\n" +def test_return_capture_timestamp(http_daemon, warcprox_, archiving_proxies): + url = 'http://localhost:{}/i/j'.format(http_daemon.server_port) + request_meta = {"return-capture-timestamp": 1} + headers = {"Warcprox-Meta": json.dumps(request_meta)} + response = requests.get(url, proxies=archiving_proxies, headers=headers, stream=True) + assert response.status_code == 200 + assert response.headers['Warcprox-Meta'] + data = json.loads(response.headers['Warcprox-Meta']) + assert data['capture-timestamp'] + try: + dt = datetime.datetime.strptime(data['capture-timestamp'], '%Y-%m-%d %H:%M:%S') + assert dt + except ValueError: + pytest.fail('Invalid capture-timestamp format %s', data['capture-timestamp']) + def test_dedup_buckets(https_daemon, http_daemon, warcprox_, archiving_proxies, playback_proxies): url1 = 'http://localhost:{}/k/l'.format(http_daemon.server_port) url2 = 'https://localhost:{}/k/l'.format(https_daemon.server_port) diff --git a/warcprox/mitmproxy.py b/warcprox/mitmproxy.py index 914fb52..e60c07e 100644 --- a/warcprox/mitmproxy.py +++ b/warcprox/mitmproxy.py @@ -45,6 +45,7 @@ try: http_client._MAXLINE = 4194304 # 4 MiB except ImportError: import httplib as http_client +import json import socket import logging import ssl @@ -163,13 +164,17 @@ class ProxyingRecordingHTTPResponse(http_client.HTTPResponse): self.fp, proxy_client, digest_algorithm, url=url) self.fp = self.recorder - def begin(self): + def begin(self, timestamp=None): http_client.HTTPResponse.begin(self) # reads status line, headers status_and_headers = 'HTTP/1.1 {} {}\r\n'.format( self.status, self.reason) self.msg['Via'] = via_header_value( self.msg.get('Via'), '%0.1f' % (self.version / 10.0)) + if timestamp: + rmeta = {"capture-timestamp": timestamp.strftime('%Y-%m-%d %H:%M:%S')} + self.msg['Warcprox-Meta'] = json.dumps(rmeta, separators=',:') + for k,v in self.msg.items(): if k.lower() not in ( 'connection', 'proxy-connection', 'keep-alive', @@ -361,12 +366,15 @@ class MitmProxyHandler(http_server.BaseHTTPRequestHandler): self.logger.error("exception proxying request", exc_info=True) raise - def _proxy_request(self): + def _proxy_request(self, timestamp=None): ''' Sends the request to the remote server, then uses a ProxyingRecorder to read the response and send it to the proxy client, while recording the bytes in transit. Returns a tuple (request, response) where request is the raw request bytes, and response is a ProxyingRecorder. + + :param timestamp: generated on warcprox._proxy_request. It is the + timestamp written in the WARC record for this request. ''' # Build request req_str = '{} {} {}\r\n'.format( @@ -407,7 +415,7 @@ class MitmProxyHandler(http_server.BaseHTTPRequestHandler): self._remote_server_sock, proxy_client=self.connection, digest_algorithm=self.server.digest_algorithm, url=self.url, method=self.command) - prox_rec_res.begin() + prox_rec_res.begin(timestamp=timestamp) buf = prox_rec_res.read(8192) while buf != b'': diff --git a/warcprox/warcproxy.py b/warcprox/warcproxy.py index 06983ed..48dc5cd 100644 --- a/warcprox/warcproxy.py +++ b/warcprox/warcproxy.py @@ -180,8 +180,13 @@ class WarcProxyHandler(warcprox.mitmproxy.MitmProxyHandler): remote_ip = self._remote_server_sock.getpeername()[0] timestamp = datetime.datetime.utcnow() + if warcprox_meta and 'return-capture-timestamp' in warcprox_meta: + return_timestamp = timestamp + else: + return_timestamp = None + req, prox_rec_res = warcprox.mitmproxy.MitmProxyHandler._proxy_request( - self) + self, timestamp=return_timestamp) content_type = None try: