def AddTorrentSkipHashCheck(self, logger, torrentPath, downloadPath):
logger.info(
"Adding torrent '%s' without hash checking to rTorrent to '%s'." %
(torrentPath, downloadPath))
sourceDirectory, sourceFilename = os.path.split(torrentPath)
sourceFilename = "fast resume " + sourceFilename
destinationTorrentPath = os.path.join(sourceDirectory, sourceFilename)
if os.path.exists(destinationTorrentPath):
raise PtpUploaderException(
"Can't create fast resume torrent because path '%s' already exists."
% destinationTorrentPath)
shutil.copyfile(torrentPath, destinationTorrentPath)
metainfo = bencode.bread(destinationTorrentPath)
metafile.add_fast_resume(metainfo, downloadPath.encode('utf-8'))
bencode.bwrite(destinationTorrentPath, metainfo)
infoHash = ""
try:
infoHash = self.AddTorrent(logger, destinationTorrentPath,
downloadPath)
finally:
# We always remove the fast resume torrent regardless of result of adding the torrent to rTorrent.
# This ensures that even if adding to rTorent fails, then resuming the job will work.
os.remove(destinationTorrentPath)
return infoHash
def AddTorrentSkipHashCheck(self, logger, torrentPath, downloadPath): logger.info( "Adding torrent '%s' without hash checking to rTorrent to '%s'." % ( torrentPath, downloadPath ) ); sourceDirectory, sourceFilename = os.path.split( torrentPath ); sourceFilename = "fast resume " + sourceFilename; destinationTorrentPath = os.path.join( sourceDirectory, sourceFilename ); if os.path.exists( destinationTorrentPath ): raise PtpUploaderException( "Can't create fast resume torrent because path '%s' already exists." % destinationTorrentPath ) shutil.copyfile( torrentPath, destinationTorrentPath ); metainfo = bencode.bread( destinationTorrentPath ) metafile.add_fast_resume( metainfo, downloadPath.encode( 'utf-8' ) ) bencode.bwrite( destinationTorrentPath, metainfo ) infoHash = "" try: infoHash = self.AddTorrent( logger, destinationTorrentPath, downloadPath ) finally: # We always remove the fast resume torrent regardless of result of adding the torrent to rTorrent. # This ensures that even if adding to rTorent fails, then resuming the job will work. os.remove( destinationTorrentPath ) return infoHash
def mainloop(self):
""" The main loop.
"""
if len(self.args) == 1 and "=urn:btih:" in self.args[0]:
# Handle magnet link
self.make_magnet_meta(self.args[0])
return
if not self.args:
self.parser.print_help()
self.parser.exit()
elif len(self.args) < 2:
self.parser.error("Expected a path and at least one announce URL, got: %s" % (' '.join(self.args),))
# Create and configure metafile factory
datapath = self.args[0].rstrip(os.sep)
metapath = datapath
if self.options.output_filename:
metapath = self.options.output_filename
if os.path.isdir(metapath):
metapath = os.path.join(metapath, os.path.basename(datapath))
if not metapath.endswith(".torrent"):
metapath += ".torrent"
torrent = metafile.Metafile(metapath)
torrent.ignore.extend(self.options.exclude)
def callback(meta):
"Callback to set label and resume data."
if self.options.cross_seed:
meta["info"]["x_cross_seed_label"] = self.options.cross_seed
if self.options.no_cross_seed:
del meta["info"]["x_cross_seed"]
# Set specific keys?
metafile.assign_fields(meta, self.options.set)
# Create and write the metafile(s)
# TODO: make it work better with multiple trackers (hash only once), also create fast-resume file for each tracker
meta = torrent.create(datapath, self.args[1:],
progress=None if self.options.quiet else metafile.console_progress(),
root_name=self.options.root_name, private=self.options.private, no_date=self.options.no_date,
comment=self.options.comment, created_by="PyroScope %s" % self.version, callback=callback
)
# Create second metafile with fast-resume?
if self.options.hashed:
try:
metafile.add_fast_resume(meta, datapath)
except EnvironmentError, exc:
self.fatal("Error making fast-resume data (%s)" % (exc,))
raise
hashed_path = re.sub(r"\.torrent$", "", metapath) + "-resume.torrent"
self.LOG.info("Writing fast-resume metafile %r..." % (hashed_path,))
try:
bencode.bwrite(hashed_path, meta)
except EnvironmentError, exc:
self.fatal("Error writing fast-resume metafile %r (%s)" % (hashed_path, exc,))
raise
class MetafileChanger(ScriptBaseWithConfig):
""" Change attributes of a bittorrent metafile.
"""
# argument description for the usage information
ARGS_HELP = "<metafile>..."
# Keys of rTorrent session data
RT_RESUMT_KEYS = ('libtorrent_resume', 'log_callback', 'err_callback', 'rtorrent')
def add_options(self):
""" Add program options.
"""
super(MetafileChanger, self).add_options()
self.add_bool_option("-n", "--dry-run",
help="don't write changes to disk, just tell what would happen")
self.add_bool_option("-V", "--no-skip",
help="do not skip broken metafiles that fail the integrity check")
self.add_value_option("-o", "--output-directory", "PATH",
help="optional output directory for the modified metafile(s)")
self.add_bool_option("-p", "--make-private",
help="make torrent private (DHT/PEX disabled)")
self.add_bool_option("-P", "--make-public",
help="make torrent public (DHT/PEX enabled)")
self.add_value_option("-s", "--set", "KEY=VAL [-s ...]",
action="append", default=[],
help="set a specific key to the given value")
self.add_value_option("-r", "--regex", "KEYcREGEXcSUBSTc [-r ...]",
action="append", default=[],
help="replace pattern in a specific key by the given substitution")
self.add_bool_option("-C", "--clean",
help="remove all non-standard data from metafile outside the info dict")
self.add_bool_option("-A", "--clean-all",
help="remove all non-standard data from metafile including inside the info dict")
self.add_bool_option("-X", "--clean-xseed",
help="like --clean-all, but keep libtorrent resume information")
self.add_bool_option("-R", "--clean-rtorrent",
help="remove all rTorrent session data from metafile")
self.add_value_option("-H", "--hashed", "--fast-resume", "DATAPATH",
help="add libtorrent fast-resume information (use {} in place of the torrent's name in DATAPATH)")
# TODO: chtor --tracker
##self.add_value_option("-T", "--tracker", "DOMAIN",
## help="filter given torrents for a tracker domain")
self.add_value_option("-a", "--reannounce", "URL",
help="set a new announce URL, but only if the old announce URL matches the new one")
self.add_value_option("--reannounce-all", "URL",
help="set a new announce URL on ALL given metafiles")
self.add_bool_option("--no-ssl",
help="force announce URL to 'http'")
self.add_bool_option("--no-cross-seed",
help="when using --reannounce-all, do not add a non-standard field to the info dict ensuring unique info hashes")
self.add_value_option("--comment", "TEXT",
help="set a new comment (an empty value deletes it)")
self.add_bool_option("--bump-date",
help="set the creation date to right now")
self.add_bool_option("--no-date",
help="remove the 'creation date' field")
def mainloop(self):
""" The main loop.
"""
if not self.args:
self.parser.error("No metafiles given, nothing to do!")
if 1 < sum(bool(i) for i in (self.options.no_ssl, self.options.reannounce, self.options.reannounce_all)):
self.parser.error("Conflicting options --no-ssl, --reannounce and --reannounce-all!")
# Set filter criteria for metafiles
filter_url_prefix = None
if self.options.reannounce:
# <scheme>://<netloc>/<path>?<query>
filter_url_prefix = urlparse.urlsplit(self.options.reannounce, allow_fragments=False)
filter_url_prefix = urlparse.urlunsplit((
filter_url_prefix.scheme, filter_url_prefix.netloc, '/', '', '' # bogus pylint: disable=E1103
))
self.LOG.info("Filtering for metafiles with announce URL prefix %r..." % filter_url_prefix)
if self.options.reannounce_all:
self.options.reannounce = self.options.reannounce_all
else:
# When changing the announce URL w/o changing the domain, don't change the info hash!
self.options.no_cross_seed = True
# Resolve tracker alias, if URL doesn't look like an URL
if self.options.reannounce and not urlparse.urlparse(self.options.reannounce).scheme:
tracker_alias, idx = self.options.reannounce, "0"
if '.' in tracker_alias:
tracker_alias, idx = tracker_alias.split('.', 1)
try:
idx = int(idx, 10)
_, tracker_url = config.lookup_announce_alias(tracker_alias)
self.options.reannounce = tracker_url[idx]
except (KeyError, IndexError, TypeError, ValueError), exc:
raise error.UserError("Unknown tracker alias or bogus URL %r (%s)!" % (
self.options.reannounce, exc))
# go through given files
bad = 0
changed = 0
for filename in self.args:
try:
# Read and remember current content
metainfo = bencode.bread(filename)
old_metainfo = bencode.bencode(metainfo)
except (EnvironmentError, KeyError, bencode.BencodeError), exc:
self.LOG.warning("Skipping bad metafile %r (%s: %s)" % (filename, type(exc).__name__, exc))
bad += 1
else:
# Check metafile integrity
try:
metafile.check_meta(metainfo)
except ValueError, exc:
self.LOG.warn("Metafile %r failed integrity check: %s" % (filename, exc,))
if not self.options.no_skip:
continue
# Skip any metafiles that don't meet the pre-conditions
if filter_url_prefix and not metainfo['announce'].startswith(filter_url_prefix):
self.LOG.warn("Skipping metafile %r no tracked by %r!" % (filename, filter_url_prefix,))
continue
# Keep resume info safe
libtorrent_resume = {}
if "libtorrent_resume" in metainfo:
try:
libtorrent_resume["bitfield"] = metainfo["libtorrent_resume"]["bitfield"]
except KeyError:
pass # nothing to remember
libtorrent_resume["files"] = copy.deepcopy(metainfo["libtorrent_resume"]["files"])
# Change private flag?
if self.options.make_private and not metainfo["info"].get("private", 0):
self.LOG.info("Setting private flag...")
metainfo["info"]["private"] = 1
if self.options.make_public and metainfo["info"].get("private", 0):
self.LOG.info("Clearing private flag...")
del metainfo["info"]["private"]
# Remove non-standard keys?
if self.options.clean or self.options.clean_all or self.options.clean_xseed:
metafile.clean_meta(metainfo, including_info=not self.options.clean, logger=self.LOG.info)
# Restore resume info?
if self.options.clean_xseed:
if libtorrent_resume:
self.LOG.info("Restoring key 'libtorrent_resume'...")
metainfo.setdefault("libtorrent_resume", {})
metainfo["libtorrent_resume"].update(libtorrent_resume)
else:
self.LOG.warn("No resume information found!")
# Clean rTorrent data?
if self.options.clean_rtorrent:
for key in self.RT_RESUMT_KEYS:
if key in metainfo:
self.LOG.info("Removing key %r..." % (key,))
del metainfo[key]
# Change announce URL?
if self.options.reannounce:
metainfo['announce'] = self.options.reannounce
if "announce-list" in metainfo:
del metainfo["announce-list"]
if not self.options.no_cross_seed:
# Enforce unique hash per tracker
metainfo["info"]["x_cross_seed"] = hashlib.md5(self.options.reannounce).hexdigest()
if self.options.no_ssl:
# We're assuming here the same (default) port is used
metainfo['announce'] = (metainfo['announce']
.replace("https://", "http://").replace(":443/", ":80/"))
# Change comment or creation date?
if self.options.comment is not None:
if self.options.comment:
metainfo["comment"] = self.options.comment
elif "comment" in metainfo:
del metainfo["comment"]
if self.options.bump_date:
metainfo["creation date"] = long(time.time())
if self.options.no_date and "creation date" in metainfo:
del metainfo["creation date"]
# Add fast-resume data?
if self.options.hashed:
try:
metafile.add_fast_resume(metainfo, self.options.hashed.replace("{}", metainfo["info"]["name"]))
except EnvironmentError, exc:
self.fatal("Error making fast-resume data (%s)" % (exc,))
raise
if ratio >= args.ratio or seeding_time >= max_seconds:
D2Rlog.info("[ " + str(torrent_index + 1) + " ] " + " Selected <" +
details['name'] + "> Ratio: " + str(ratio) +
" Seed Time: " + str(seeding_time))
delugeTorrent = os.path.join(delugeBasePath, torrent + ".torrent")
# Read Torrent
metainfo = bencode.bread(delugeTorrent)
payloadPath = os.path.join(details["save_path"],
metainfo["info"]["name"])
# Add RTorrent Fast Resume
rTorrent = metafile.add_fast_resume(metainfo, payloadPath)
# Determine where to save the new torrent
destination = watchFolder
if args.label:
labeled = os.path.join(watchFolder, details["label"])
# Label Folder must exist, otherwise add torrent to base watch folder (no label)
if os.path.exists(labeled):
destination = labeled
torrentFile = os.path.join(destination, torrent + ".torrent")
# Write out rtorrent torrent to watch folder destination
try:
def mainloop(self):
""" The main loop.
"""
if not self.args:
self.parser.error("No metafiles given, nothing to do!")
if 1 < sum(bool(i) for i in (self.options.no_ssl, self.options.reannounce, self.options.reannounce_all)):
self.parser.error("Conflicting options --no-ssl, --reannounce and --reannounce-all!")
# Set filter criteria for metafiles
filter_url_prefix = None
if self.options.reannounce:
# <scheme>://<netloc>/<path>?<query>
filter_url_prefix = urlparse.urlsplit(self.options.reannounce, allow_fragments=False)
filter_url_prefix = urlparse.urlunsplit((
filter_url_prefix.scheme, filter_url_prefix.netloc, '/', '', '' # bogus pylint: disable=E1103
))
self.LOG.info("Filtering for metafiles with announce URL prefix %r..." % filter_url_prefix)
if self.options.reannounce_all:
self.options.reannounce = self.options.reannounce_all
else:
# When changing the announce URL w/o changing the domain, don't change the info hash!
self.options.no_cross_seed = True
# Resolve tracker alias, if URL doesn't look like an URL
if self.options.reannounce and not urlparse.urlparse(self.options.reannounce).scheme:
tracker_alias, idx = self.options.reannounce, "0"
if '.' in tracker_alias:
tracker_alias, idx = tracker_alias.split('.', 1)
try:
idx = int(idx, 10)
_, tracker_url = config.lookup_announce_alias(tracker_alias)
self.options.reannounce = tracker_url[idx]
except (KeyError, IndexError, TypeError, ValueError) as exc:
raise error.UserError("Unknown tracker alias or bogus URL %r (%s)!" % (
self.options.reannounce, exc))
# go through given files
bad = 0
changed = 0
for filename in self.args:
try:
# Read and remember current content
metainfo = bencode.bread(filename)
old_metainfo = bencode.bencode(metainfo)
except (EnvironmentError, KeyError, bencode.BencodeError) as exc:
self.LOG.warning("Skipping bad metafile %r (%s: %s)" % (filename, type(exc).__name__, exc))
bad += 1
else:
# Check metafile integrity
try:
metafile.check_meta(metainfo)
except ValueError as exc:
self.LOG.warn("Metafile %r failed integrity check: %s" % (filename, exc,))
if not self.options.no_skip:
continue
# Skip any metafiles that don't meet the pre-conditions
if filter_url_prefix and not metainfo['announce'].startswith(filter_url_prefix):
self.LOG.warn("Skipping metafile %r no tracked by %r!" % (filename, filter_url_prefix,))
continue
# Keep resume info safe
libtorrent_resume = {}
if "libtorrent_resume" in metainfo:
try:
libtorrent_resume["bitfield"] = metainfo["libtorrent_resume"]["bitfield"]
except KeyError:
pass # nothing to remember
libtorrent_resume["files"] = copy.deepcopy(metainfo["libtorrent_resume"]["files"])
# Change private flag?
if self.options.make_private and not metainfo["info"].get("private", 0):
self.LOG.info("Setting private flag...")
metainfo["info"]["private"] = 1
if self.options.make_public and metainfo["info"].get("private", 0):
self.LOG.info("Clearing private flag...")
del metainfo["info"]["private"]
# Remove non-standard keys?
if self.options.clean or self.options.clean_all or self.options.clean_xseed:
metafile.clean_meta(metainfo, including_info=not self.options.clean, logger=self.LOG.info)
# Restore resume info?
if self.options.clean_xseed:
if libtorrent_resume:
self.LOG.info("Restoring key 'libtorrent_resume'...")
metainfo.setdefault("libtorrent_resume", {})
metainfo["libtorrent_resume"].update(libtorrent_resume)
else:
self.LOG.warn("No resume information found!")
# Clean rTorrent data?
if self.options.clean_rtorrent:
for key in self.RT_RESUMT_KEYS:
if key in metainfo:
self.LOG.info("Removing key %r..." % (key,))
del metainfo[key]
# Change announce URL?
if self.options.reannounce:
metainfo['announce'] = self.options.reannounce
if "announce-list" in metainfo:
del metainfo["announce-list"]
if not self.options.no_cross_seed:
# Enforce unique hash per tracker
metainfo["info"]["x_cross_seed"] = hashlib.md5(self.options.reannounce).hexdigest()
if self.options.no_ssl:
# We're assuming here the same (default) port is used
metainfo['announce'] = (metainfo['announce']
.replace("https://", "http://").replace(":443/", ":80/"))
# Change comment or creation date?
if self.options.comment is not None:
if self.options.comment:
metainfo["comment"] = self.options.comment
elif "comment" in metainfo:
del metainfo["comment"]
if self.options.bump_date:
metainfo["creation date"] = int(time.time())
if self.options.no_date and "creation date" in metainfo:
del metainfo["creation date"]
# Add fast-resume data?
if self.options.hashed:
try:
metafile.add_fast_resume(metainfo, self.options.hashed.replace("{}", metainfo["info"]["name"]))
except EnvironmentError as exc:
self.fatal("Error making fast-resume data (%s)" % (exc,))
raise
# Set specific keys?
metafile.assign_fields(metainfo, self.options.set, self.options.debug)
replace_fields(metainfo, self.options.regex)
# Write new metafile, if changed
new_metainfo = bencode.bencode(metainfo)
if new_metainfo != old_metainfo:
if self.options.output_directory:
filename = os.path.join(self.options.output_directory, os.path.basename(filename))
self.LOG.info("Writing %r..." % filename)
if not self.options.dry_run:
bencode.bwrite(filename, metainfo)
if "libtorrent_resume" in metainfo:
# Also write clean version
filename = filename.replace(".torrent", "-no-resume.torrent")
del metainfo["libtorrent_resume"]
self.LOG.info("Writing %r..." % filename)
bencode.bwrite(filename, metainfo)
else:
self.LOG.info("Changing %r..." % filename)
if not self.options.dry_run:
# Write to temporary file
tempname = os.path.join(
os.path.dirname(filename),
'.' + os.path.basename(filename),
)
self.LOG.debug("Writing %r..." % tempname)
bencode.bwrite(tempname, metainfo)
# Replace existing file
if os.name != "posix":
# cannot rename to existing target on WIN32
os.remove(filename)
try:
os.rename(tempname, filename)
except EnvironmentError as exc:
# TODO: Try to write directly, keeping a backup!
raise error.LoggableError("Can't rename tempfile %r to %r (%s)" % (
tempname, filename, exc
))
changed += 1
# Print summary
if changed:
self.LOG.info("%s %d metafile(s)." % (
"Would've changed" if self.options.dry_run else "Changed", changed
))
if bad:
self.LOG.warn("Skipped %d bad metafile(s)!" % (bad))
def mainloop(self):
""" The main loop.
"""
if len(self.args) == 1 and "=urn:btih:" in self.args[0]:
# Handle magnet link
self.make_magnet_meta(self.args[0])
return
if not self.args:
self.parser.print_help()
self.parser.exit()
elif len(self.args) < 2:
self.parser.error(
"Expected a path and at least one announce URL, got: %s" %
(' '.join(self.args), ))
# Create and configure metafile factory
datapath = self.args[0].rstrip(os.sep)
metapath = datapath
if self.options.output_filename:
metapath = self.options.output_filename
if os.path.isdir(metapath):
metapath = os.path.join(metapath, os.path.basename(datapath))
if not metapath.endswith(".torrent"):
metapath += ".torrent"
torrent = metafile.Metafile(metapath)
torrent.ignore.extend(self.options.exclude)
def callback(meta):
"Callback to set label and resume data."
if self.options.cross_seed:
if self.options.cross_seed == "@entropy":
meta["info"]["entropy"] = format(
random.getrandbits(self.ENTROPY_BITS),
'x').zfill(self.ENTROPY_BITS // 4)
else:
meta["info"][
"x_cross_seed_label"] = self.options.cross_seed
if self.options.no_cross_seed:
del meta["info"]["x_cross_seed"]
# Set specific keys?
metafile.assign_fields(meta, self.options.set, self.options.debug)
# Create and write the metafile(s)
# TODO: make it work better with multiple trackers (hash only once), also create fast-resume file for each tracker
meta = torrent.create(datapath,
self.args[1:],
progress=None if self.options.quiet else
metafile.console_progress(),
root_name=self.options.root_name,
private=self.options.private,
no_date=self.options.no_date,
comment=self.options.comment,
created_by="PyroScope %s" % self.version,
callback=callback)
tied_file = metapath
# Create second metafile with fast-resume?
if self.options.hashed:
try:
metafile.add_fast_resume(meta, datapath)
except EnvironmentError as exc:
self.fatal("Error making fast-resume data (%s)" % (exc, ))
raise
hashed_path = re.sub(r"\.torrent$", "",
metapath) + "-resume.torrent"
self.LOG.info("Writing fast-resume metafile %r..." %
(hashed_path, ))
try:
bencode.bwrite(hashed_path, meta)
tied_file = hashed_path
except EnvironmentError as exc:
self.fatal("Error writing fast-resume metafile %r (%s)" % (
hashed_path,
exc,
))
raise
# Load into client on demand
if self.options.load or self.options.start:
proxy = config.engine.open()
info_hash = metafile.info_hash(meta)
try:
item_name = proxy.d.name(info_hash, fail_silently=True)
except xmlrpc.HashNotFound:
load_item = proxy.load.start_verbose if self.options.start else proxy.load.verbose
load_item(xmlrpc.NOHASH, os.path.abspath(tied_file))
time.sleep(.05) # let things settle
try:
item_name = proxy.d.name(info_hash, fail_silently=True)
self.LOG.info(
"OK: Item #%s %s client.", info_hash,
'started in' if self.options.start else 'loaded into')
except xmlrpc.HashNotFound as exc:
self.fatal("Error while loading item #%s into client: %s" %
(
info_hash,
exc,
))
else:
self.LOG.warning(
"Item #%s already exists in client, --load/--start is ignored!",
info_hash)
def mainloop(self):
""" The main loop.
"""
if len(self.args) == 1 and "=urn:btih:" in self.args[0]:
# Handle magnet link
self.make_magnet_meta(self.args[0])
return
if not self.args:
self.parser.print_help()
self.parser.exit()
elif len(self.args) < 2:
self.parser.error("Expected a path and at least one announce URL, got: %s" % (' '.join(self.args),))
# Create and configure metafile factory
datapath = self.args[0].rstrip(os.sep)
metapath = datapath
if self.options.output_filename:
metapath = self.options.output_filename
if os.path.isdir(metapath):
metapath = os.path.join(metapath, os.path.basename(datapath))
if not metapath.endswith(".torrent"):
metapath += ".torrent"
torrent = metafile.Metafile(metapath)
torrent.ignore.extend(self.options.exclude)
def callback(meta):
"Callback to set label and resume data."
if self.options.cross_seed:
if self.options.cross_seed == "@entropy":
meta["info"]["entropy"] = format(random.getrandbits(self.ENTROPY_BITS),
'x').zfill(self.ENTROPY_BITS//4)
else:
meta["info"]["x_cross_seed_label"] = self.options.cross_seed
if self.options.no_cross_seed:
del meta["info"]["x_cross_seed"]
# Set specific keys?
metafile.assign_fields(meta, self.options.set)
# Create and write the metafile(s)
# TODO: make it work better with multiple trackers (hash only once), also create fast-resume file for each tracker
meta = torrent.create(datapath, self.args[1:],
progress=None if self.options.quiet else metafile.console_progress(),
root_name=self.options.root_name, private=self.options.private, no_date=self.options.no_date,
comment=self.options.comment, created_by="PyroScope %s" % self.version, callback=callback
)
# Create second metafile with fast-resume?
if self.options.hashed:
try:
metafile.add_fast_resume(meta, datapath)
except EnvironmentError as exc:
self.fatal("Error making fast-resume data (%s)" % (exc,))
raise
hashed_path = re.sub(r"\.torrent$", "", metapath) + "-resume.torrent"
self.LOG.info("Writing fast-resume metafile %r..." % (hashed_path,))
try:
bencode.bwrite(hashed_path, meta)
except EnvironmentError as exc:
self.fatal("Error writing fast-resume metafile %r (%s)" % (hashed_path, exc,))
raise