X-Received: by 10.42.254.135 with SMTP id ne7mr41353608icb.30.1418836377579; Wed, 17 Dec 2014 09:12:57 -0800 (PST) X-Received: by 10.182.219.13 with SMTP id pk13mr9235obc.39.1418836377497; Wed, 17 Dec 2014 09:12:57 -0800 (PST) Path: csiph.com!v102.xanadu-bbs.net!xanadu-bbs.net!usenet.blueworldhosting.com!feeder01.blueworldhosting.com!peer01.iad.highwinds-media.com!news.highwinds-media.com!feed-me.highwinds-media.com!h15no24198421igd.0!news-out.google.com!h6ni559igv.0!nntp.google.com!h15no24198413igd.0!postnews.google.com!glegroupsg2000goo.googlegroups.com!not-for-mail Newsgroups: comp.lang.python Date: Wed, 17 Dec 2014 09:12:57 -0800 (PST) Complaints-To: groups-abuse@google.com Injection-Info: glegroupsg2000goo.googlegroups.com; posting-host=24.196.71.171; posting-account=ftUzqAkAAAB0ziTgdWrXKVPsILGert2L NNTP-Posting-Host: 24.196.71.171 User-Agent: G2/1.0 MIME-Version: 1.0 Message-ID: <79e004d7-67f7-4fbf-a5aa-4d422fd95b30@googlegroups.com> Subject: signal.signal handler arity check - PEP? From: Mike Pomraning Injection-Date: Wed, 17 Dec 2014 17:12:57 +0000 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable X-Received-Bytes: 2029 X-Received-Body-CRC: 2378651065 Xref: csiph.com comp.lang.python:82601 It strikes me as oddly dangerous that signal.signal() accepts callable hand= lers of the wrong arity: def timeout_cleanup(): ... signal.signal(signal.SIGALRM, timeout_cleanup) # I desire a TypeE= rror here signal.alarm(PROCESS_TIMEOUT) ... time passes ... TypeError: timeout_cleanup() takes 0 positional arguments but 2 were gi= ven (Similar TypeError with different text under python 2.) The downside to current behavior is that the coding error is revealed only = under certain conditions at runtime. In the example above, you don't know = that your seatbelt is defective until you need it. Does anyone disagree with modifying signal.signal's behavior to raise a Typ= eError when given a user-defined handler of the wrong arity? Any code rely= ing on this behavior is, IMO, contrived beyond plausibility. However, this= change might break code not yet known to be incorrect. If no disagreements, patch or PEP? -Mike