diff options
Diffstat (limited to 'dev-python/pyzor/files/read-stdin-as-binary-in-get_input_msg.patch')
| -rw-r--r-- | dev-python/pyzor/files/read-stdin-as-binary-in-get_input_msg.patch | 94 |
1 files changed, 67 insertions, 27 deletions
diff --git a/dev-python/pyzor/files/read-stdin-as-binary-in-get_input_msg.patch b/dev-python/pyzor/files/read-stdin-as-binary-in-get_input_msg.patch index 81668e369377..03031a976690 100644 --- a/dev-python/pyzor/files/read-stdin-as-binary-in-get_input_msg.patch +++ b/dev-python/pyzor/files/read-stdin-as-binary-in-get_input_msg.patch @@ -1,45 +1,85 @@ -From 6332a429ed415187599ecce7d8a169ee19f0bbe5 Mon Sep 17 00:00:00 2001 +From 66225b32d2774cf37fa7f702f7eb26cd94094482 Mon Sep 17 00:00:00 2001 From: Michael Orlitzky <michael@orlitzky.com> -Date: Sun, 4 Mar 2018 17:34:33 -0500 -Subject: [PATCH 1/1] scripts/pyzor: read stdin as binary in _get_input_msg(). +Date: Sun, 4 Mar 2018 17:27:01 -0500 +Subject: [PATCH 1/1] scripts/pyzor: replace the client with the git (+ issue + 64 fix) version. -Reading stdin in python-3.x is done as text, with a best-guess -encoding. But this can go awry: for example, if an iso-8859-1 message -is passed in and if python guesses the "utf-8" encoding, then read() -will fail with a UnicodeDecodeError on non-ASCII characters. For -example, the "copyright" symbol is a single byte 0xa9 in iso-8859-1, -and the utf-8 decoder can't handle it: - - UnicodeDecodeError: 'utf-8' codec can't decode byte 0xa9... invalid - start byte - -Instead -- and as was done in python-2.x -- we can read stdin as -binary using the new get_binary_stdin() function. Afterwards, we use -email.message_from_bytes() instead of the email.message_from_file() -constructor to parse the byte data. The resulting function is able to -correctly parse these messages. - -Closes: https://github.com/SpamExperts/pyzor/issues/64 --- - scripts/pyzor | 5 ++++- - 1 file changed, 4 insertions(+), 1 deletion(-) + scripts/pyzor | 33 +++++++++++++++++++++++++++------ + 1 file changed, 27 insertions(+), 6 deletions(-) diff --git a/scripts/pyzor b/scripts/pyzor -index 567a7f9..1ba632f 100755 +index 19b1d21..86c6f7d 100755 --- a/scripts/pyzor +++ b/scripts/pyzor -@@ -171,7 +171,10 @@ def _get_input_digests(dummy): +@@ -17,9 +17,9 @@ import tempfile + import threading + + try: +- import ConfigParser +-except ImportError: + import configparser as ConfigParser ++except ImportError: ++ import ConfigParser + + import pyzor.digest + import pyzor.client +@@ -110,7 +110,7 @@ def load_configuration(): + config = ConfigParser.ConfigParser() + # Set the defaults. + config.add_section("client") +- for key, value in defaults.iteritems(): ++ for key, value in defaults.items(): + config.set("client", key, value) + # Override with the configuration. + config.read(os.path.join(options.homedir, "config")) +@@ -171,14 +171,35 @@ def _get_input_digests(dummy): def _get_input_msg(digester): - msg = email.message_from_file(sys.stdin) -+ # Read and process stdin as bytes because we don't know its -+ # encoding. Python-3.x will try to guess -- and can sometimes -+ # guess wrong -- leading to decoding errors in read(). + msg = email.message_from_bytes(get_binary_stdin().read()) digested = digester(msg).value yield digested + ++def _is_binary_reader(stream, default=False): ++ try: ++ return isinstance(stream.read(0), bytes) ++ except Exception: ++ return default ++ ++ ++def get_binary_stdin(): ++ # sys.stdin might or might not be binary in some extra cases. By ++ # default it's obviously non binary which is the core of the ++ # problem but the docs recommend changing it to binary for such ++ # cases so we need to deal with it. ++ is_binary = _is_binary_reader(sys.stdin, False) ++ if is_binary: ++ return sys.stdin ++ buf = getattr(sys.stdin, 'buffer', None) ++ if buf is not None and _is_binary_reader(buf, True): ++ return buf ++ raise RuntimeError('Did not manage to get binary stdin') ++ ++ + def _get_input_mbox(digester): + tfile = tempfile.NamedTemporaryFile() +- tfile.write(sys.stdin.read().encode("utf8")) ++ tfile.write(get_binary_stdin().read()) + tfile.seek(0) + mbox = mailbox.mbox(tfile.name) + for msg in mbox: +@@ -372,7 +393,7 @@ def genkey(client, servers, config, hash_func=hashlib.sha1): + return False + # pylint: disable-msg=W0612 + salt = "".join([chr(random.randint(0, 255)) +- for unused in xrange(hash_func(b"").digest_size)]) ++ for unused in range(hash_func(b"").digest_size)]) + if sys.version_info >= (3, 0): + salt = salt.encode("utf8") + salt_digest = hash_func(salt) -- 2.13.6 |