commit bc58a28bf07a646d3e1ced41891a03e31d023a7d Author: Damian Johnson atagar@torproject.org Date: Thu Apr 12 12:17:01 2018 -0700

Provide a User-Agent when downloading descriptors

By default urllib provides a user agent that simply says that we're using python. On reflection lets specify that the download is coming from Stem. This will give our dirauths a bit more visibility into DirPort requests they're servicing, and Stem versions that are actively being used to download descriptors. --- docs/change_log.rst | 1 + stem/descriptor/remote.py | 9 ++++++--- 2 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/docs/change_log.rst b/docs/change_log.rst index 9be7912d..ac827bdf 100644 --- a/docs/change_log.rst +++ b/docs/change_log.rst @@ -58,6 +58,7 @@ The following are only available within Stem's `git repository * Added zstd and lzma compression support (:spec:`1cb56af`) * Added server descriptor's new is_hidden_service_dir attribute * Don't retry downloading descriptors when we've timed out + * Supplying a User-Agent when downloading descriptors. * Reduced maximum descriptors fetched by the remote module to match tor's new limit (:trac:`24743`) * Consensus **shared_randomness_*_reveal_count** attributes undocumented, and unavailable if retrieved before their corresponding shared_randomness_*_value attribute (:trac:`25046`) * Allow 'proto' line to have blank values (:spec:`a8455f4`) diff --git a/stem/descriptor/remote.py b/stem/descriptor/remote.py index 3faace4c..30ae388b 100644 --- a/stem/descriptor/remote.py +++ b/stem/descriptor/remote.py @@ -106,11 +106,11 @@ import threading import time import zlib

+import stem import stem.descriptor import stem.prereq import stem.util.enum

-from stem import Flag from stem.util import _hash_attr, connection, log, str_tools, tor_tools

try: @@ -529,7 +529,10 @@ class Query(object): response = urllib.urlopen( urllib.Request( self.download_url, - headers = {'Accept-Encoding': ', '.join(self.compression)}, + headers = { + 'Accept-Encoding': ', '.join(self.compression), + 'User-Agent': 'Stem/%s' % stem.__version__, + } ), timeout = timeout, ) @@ -618,7 +621,7 @@ class DescriptorDownloader(object): consensus = list(self.get_consensus(document_handler = stem.descriptor.DocumentHandler.DOCUMENT).run())[0]

for desc in consensus.routers.values(): - if Flag.V2DIR in desc.flags: + if stem.Flag.V2DIR in desc.flags: new_endpoints.add((desc.address, desc.dir_port))

# we need our endpoints to be a list rather than set for random.choice()