Bug#470390: newsbeuter: dies when reading feeds with bad accents
Hi Gregor, * gregor herrmann [EMAIL PROTECTED] [2008-03-11 12:40]: On Mon, 10 Mar 2008 22:59:32 +0100, Riccardo Stagni wrote: I tried to reproduce this using the same locale ([EMAIL PROTECTED]), setting it as LC_ALL and invoking newsbeuter in urxvt. Even using exactly the same library versions as you I fail to reproduce this. No crash, everything gets displayed fine. I have the same problems as the submitter (i.e. the crash with the same error). I just had no idea yet what might be the cause :) please try setting the LANG variable: with LANG=C it works but displays ? instead the offending à. If instead I use the usual [EMAIL PROTECTED] (or also it_IT.UTF-8) it crashes. Similar here: [EMAIL PROTECTED]:~$ locale [...] Thanks, confirmed. This seems to be a problem when converting from multibyte to widechar from the database. Upstream is working on a patch. This was introduced with the new format string feature. Kind regards Nico -- Nico Golde - http://www.ngolde.de - [EMAIL PROTECTED] - GPG: 0x73647CFF For security reasons, all text in this mail is double-rot13 encrypted. pgpk2Ydi9Z7kZ.pgp Description: PGP signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
Hi gregor, * gregor herrmann [EMAIL PROTECTED] [2008-03-11 12:40]: On Mon, 10 Mar 2008 23:38:29 +0100, Nico Golde wrote: please try setting the LANG variable: with LANG=C it works but displays ? instead the offending à. If instead I use the usual [EMAIL PROTECTED] (or also it_IT.UTF-8) it crashes. No problem, still not crashing :/ This is indeed weird; you get no crashes, Riccardo gets them both with latin9 and utf8, I get them only with latin9 ... Could this be terminal-related in some way? FWIW: I used plain xterm (and uxterm). The reason seems to be that on my system all locale variables have the same values, on yours not. Kind regards Nico -- Nico Golde - http://www.ngolde.de - [EMAIL PROTECTED] - GPG: 0x73647CFF For security reasons, all text in this mail is double-rot13 encrypted. pgpOVpSi0rQ0E.pgp Description: PGP signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
Hi Riccardo, * Riccardo Stagni [EMAIL PROTECTED] [2008-03-11 12:42]: On Tue, Mar 11, 2008 at 12:17:43AM +0100, gregor herrmann wrote: [...] I hope there is a patch at the end of the day so I can upload a quick fix. P.S. Did you experience this also with older versions or are you a new user? Here 0.7-4 works fine (but I know that you know that I'm not a new user :)) I am, instead, a new user: I've read Nico's orphaning message for snownews and I tought giving a try to newsbeuter. Sorry to stop your experience with this! ;-P Cheers Nico -- Nico Golde - http://www.ngolde.de - [EMAIL PROTECTED] - GPG: 0x73647CFF For security reasons, all text in this mail is double-rot13 encrypted. pgp5wVqESk4HG.pgp Description: PGP signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
On Tue, 11 Mar 2008 13:17:12 +0100, Nico Golde wrote: [EMAIL PROTECTED]:~$ locale Thanks, confirmed. This seems to be a problem when converting from multibyte to widechar from the database. Upstream is working on a patch. This was introduced with the new format string feature. Great, congratulations on tracking down this issue so fast and thanks for all your efforts! Cheers, gregor -- .''`. http://info.comodo.priv.at/ | gpg key ID: 0x00F3CFE4 : :' : debian: the universal operating system - http://www.debian.org/ `. `' member of https://www.vibe.at/ | how to reply: http://got.to/quote/ `-Warp 7 -- It's a law we can live with. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#470449: Bug#470390: newsbeuter: dies when reading feeds with bad accents
Hi, can you try to reproduce this with: http://people.debian.org/~nion/newsbeuter/newsbeuter_0.8-1_i386.deb It contains a fix by the upstream author which may fix it. Kind regards Nico -- Nico Golde - http://www.ngolde.de - [EMAIL PROTECTED] - GPG: 0x73647CFF For security reasons, all text in this mail is double-rot13 encrypted. pgp794qo5MAyW.pgp Description: PGP signature
Bug#470449: Bug#470390: newsbeuter: dies when reading feeds with bad accents
On Tue, Mar 11, 2008 at 02:45:49PM +0100, Nico Golde wrote: can you try to reproduce this with: http://people.debian.org/~nion/newsbeuter/newsbeuter_0.8-1_i386.deb It contains a fix by the upstream author which may fix it. Same results (works with LANG=C, doesn't with [EMAIL PROTECTED],.UTF-8}). Also, there are new dependencies: the new package now asks for libruby1.8 (and libcurl3 instead of libcurl3-gnutls). ciao R, signature.asc Description: Digital signature
Bug#470390: Bug#470449: Bug#470390: newsbeuter: dies when reading feeds with bad accents
Hi Riccardo, * Riccardo Stagni [EMAIL PROTECTED] [2008-03-11 20:47]: On Tue, Mar 11, 2008 at 02:45:49PM +0100, Nico Golde wrote: can you try to reproduce this with: http://people.debian.org/~nion/newsbeuter/newsbeuter_0.8-1_i386.deb It contains a fix by the upstream author which may fix it. Same results (works with LANG=C, doesn't with [EMAIL PROTECTED],.UTF-8}). Also, there are new dependencies: the new package now asks for libruby1.8 (and libcurl3 instead of libcurl3-gnutls). http://people.debian.org/~nion/newsbeuter/newsbeuter_0.8-1_i386.deb Please give it another try, just added another patch. Kind regards Nico -- Nico Golde - http://www.ngolde.de - [EMAIL PROTECTED] - GPG: 0x73647CFF For security reasons, all text in this mail is double-rot13 encrypted. pgpbThub2g09l.pgp Description: PGP signature
Bug#470449: Bug#470390: newsbeuter: dies when reading feeds with bad accents
Hi Riccardo, * Riccardo Stagni [EMAIL PROTECTED] [2008-03-11 20:47]: On Tue, Mar 11, 2008 at 02:45:49PM +0100, Nico Golde wrote: can you try to reproduce this with: http://people.debian.org/~nion/newsbeuter/newsbeuter_0.8-1_i386.deb It contains a fix by the upstream author which may fix it. Same results (works with LANG=C, doesn't with [EMAIL PROTECTED],.UTF-8}). Ok thanks for testing, I'll check back with upstream. Also, there are new dependencies: the new package now asks for libruby1.8 (and libcurl3 instead of libcurl3-gnutls). Yeah might be. I wasn't able to build this package in a clean chroot because of libmrss0 was broken so I needed to build that on my normal installation. Kind regatrds Nico -- Nico Golde - http://www.ngolde.de - [EMAIL PROTECTED] - GPG: 0x73647CFF For security reasons, all text in this mail is double-rot13 encrypted. pgpFzMNxFzJX3.pgp Description: PGP signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
On Tue, 11 Mar 2008 14:45:49 +0100, Nico Golde wrote:
can you try to reproduce this with:
http://people.debian.org/~nion/newsbeuter/newsbeuter_0.8-1_i386.deb
It contains a fix by the upstream author which may fix it.
Sorry, but it fails in the same way in my default locale (see my mail
from yesterday).
It works with:
$ [EMAIL PROTECTED] [EMAIL PROTECTED] newsbeuter
$ LANG=de_AT.utf8 LC_ALL=de_AT.utf8 newsbeuter
$ LANG=de_AT.iso88591 LC_ALL=de_AT.iso88591 newsbeuter
$ LANG=C LC_ALL=C newsbeuter
$ LANG=de_AT.utf8 newsbeuter
$ [EMAIL PROTECTED] newsbeuter
$ LANG=de_AT.iso88591 newsbeuter
$ LANG=en_US.iso885915 newsbeuter
It looks like the mismatch of LANG and LC_CTYPE is no problem and
LC_ALL is not needed; just LANG={C,POSIX} breaks.
Interestingly
$ LANG=C [EMAIL PROTECTED] newsbeuter
works -- and that's the same as my default locale.
$ export LANG=C; export [EMAIL PROTECTED]
$ newsbeuter
fails;
$ LANG=C [EMAIL PROTECTED] newsbeuter
now fails too.
Hu?!
New xterm, old locale ...
$ LC_MESSAGES=en_US.iso885915 newsbeuter
$ [EMAIL PROTECTED] newsbeuter
both work
$ [EMAIL PROTECTED] newsbeuter
$ LC_CTYPE=en_US.iso885915 newsbeuter
both fail
Maybe that helps more:
$ newsbeuter -d debug.log -l 6
The end of debug.log is:
[2008-03-11 21:00:40] DEBUG: utils::convert_text: before: Chrisu JÀhnl
[2008-03-11 21:00:40] DEBUG: utils::convert_text: after: Chrisu Jähnl
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = i value = 4
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = i done
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = u value =
(1/34)
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = u done
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = n value = N
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = n done
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = t value =
Chrisu Jähnl
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = t done
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = l value =
http://www.google.com/calendar/[..]
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = l done
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = L value =
http://www.google.com/calendar/[..]
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = L done
[2008-03-11 21:00:40] DEBUG: utils::convert_text: before:
[2008-03-11 21:00:40] DEBUG: utils::convert_text: after:
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = d value =
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::register_fmt: char = d done
[2008-03-11 21:00:40] DEBUG: utils::str2wstr: current locale: C
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::do_format: fmt = `%4i %n %11u
%t' width = 110
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::do_format: fmtlen = 14
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::do_format: number = 4
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::do_format: swprintf result =4
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::do_format: number = 11
[2008-03-11 21:00:40] DEBUG: fmtstr_formatter::do_format: swprintf result =
(1/34)
[2008-03-11 21:00:40] DEBUG: utils::wstr2str: locale = C input = ` 4 N
(1/34) Chrisu Jähnl'
[2008-03-11 21:00:40] ERROR: utils::wstr2str: conversion of ` 4 N (1/34)
Chrisu Jähnl' failed.
(URLs truncated, this is a private feed for a Google calendar of a friend)
Cheers,
gregor
--
.''`. http://info.comodo.priv.at/ | gpg key ID: 0x00F3CFE4
: :' : debian: the universal operating system - http://www.debian.org/
`. `' member of https://www.vibe.at/ | how to reply: http://got.to/quote/
`-NP: Kurt Ostbahn Die Kombo: So gern so vü
signature.asc
Description: Digital signature
Bug#470449: Bug#470390: newsbeuter: dies when reading feeds with bad accents
On Tue, Mar 11, 2008 at 09:22:44PM +0100, Nico Golde wrote: http://people.debian.org/~nion/newsbeuter/newsbeuter_0.8-1_i386.deb Please give it another try, just added another patch. works for me! :) great work! R, signature.asc Description: Digital signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
Hi Gregor, * gregor herrmann [EMAIL PROTECTED] [2008-03-11 21:49]: On Tue, 11 Mar 2008 14:45:49 +0100, Nico Golde wrote: can you try to reproduce this with: http://people.debian.org/~nion/newsbeuter/newsbeuter_0.8-1_i386.deb It contains a fix by the upstream author which may fix it. Sorry, but it fails in the same way in my default locale (see my mail from yesterday). [...] Did you try the updated .deb? 605aa95410502f1a4e96841d50d83b91 newsbeuter_0.8-1_i386.deb Cheers Nico -- Nico Golde - http://www.ngolde.de - [EMAIL PROTECTED] - GPG: 0x73647CFF For security reasons, all text in this mail is double-rot13 encrypted. pgp36bHKKkPSa.pgp Description: PGP signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
On Tue, 11 Mar 2008 21:59:57 +0100, Nico Golde wrote: Sorry, but it fails in the same way in my default locale (see my mail from yesterday). [...] Did you try the updated .deb? 605aa95410502f1a4e96841d50d83b91 newsbeuter_0.8-1_i386.deb Now I did :) (Your new mail came when I was playing with the old version, and I just found it now). Anyway, here's the good news: works for me! Thanks again, gregor -- .''`. http://info.comodo.priv.at/ | gpg key ID: 0x00F3CFE4 : :' : debian: the universal operating system - http://www.debian.org/ `. `' member of https://www.vibe.at/ | how to reply: http://got.to/quote/ `-NP: Rod Stewart: Dancin Alone signature.asc Description: Digital signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
Package: newsbeuter Version: 0.8-1 Severity: important Seems that newsbeuter dislikes feeds with unescaped accents: eg. http://syndication.splinder.com/spettrodellabolognesita/rss2.xml contains bolognesità within the feed title. Newsbeuter dies with a terminate called after throwing an instance of 'char*' error. (severity important because, since the accented letter is in the feed title, it prevents newsbeuter starting, making it die when reading cache) ciao Riccardo -- System Information: Debian Release: lenny/sid APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.6.24-1-686 (SMP w/2 CPU cores) Locale: [EMAIL PROTECTED], [EMAIL PROTECTED] (charmap=ISO-8859-15) Shell: /bin/sh linked to /bin/bash Versions of packages newsbeuter depends on: ii libc6 2.7-9 GNU C Library: Shared libraries ii libcurl3-gnutls 7.18.0-1 Multi-protocol file transfer libra ii libgcc1 1:4.3.0-1 GCC support library ii libkrb53 1.6.dfsg.3~beta1-3 MIT Kerberos runtime libraries ii libmrss0 [libmrss-abi 0.19.0-2 C library for parsing, writing and ii libncursesw5 5.6+20080203-1 Shared libraries for terminal hand ii libnxml0 [libnxml-abi 0.18.1-4 C library for parsing, writing and ii libsqlite3-0 3.5.6-3SQLite 3 shared library ii libstdc++64.3.0-1The GNU Standard C++ Library v3 newsbeuter recommends no packages. -- no debconf information
Bug#470390: newsbeuter: dies when reading feeds with bad accents
tags 470390 + upstream moreinfo thanks Hi Riccardo, * Riccardo Stagni [EMAIL PROTECTED] [2008-03-10 22:22]: [...] Kernel: Linux 2.6.24-1-686 (SMP w/2 CPU cores) Locale: [EMAIL PROTECTED], [EMAIL PROTECTED] (charmap=ISO-8859-15) Shell: /bin/sh linked to /bin/bash Versions of packages newsbeuter depends on: ii libc6 2.7-9 GNU C Library: Shared libraries ii libcurl3-gnutls 7.18.0-1 Multi-protocol file transfer libra ii libgcc1 1:4.3.0-1 GCC support library ii libkrb53 1.6.dfsg.3~beta1-3 MIT Kerberos runtime libraries ii libmrss0 [libmrss-abi 0.19.0-2 C library for parsing, writing and ii libncursesw5 5.6+20080203-1 Shared libraries for terminal hand ii libnxml0 [libnxml-abi 0.18.1-4 C library for parsing, writing and ii libsqlite3-0 3.5.6-3SQLite 3 shared library ii libstdc++64.3.0-1The GNU Standard C++ Library v3 I tried to reproduce this using the same locale ([EMAIL PROTECTED]), setting it as LC_ALL and invoking newsbeuter in urxvt. Even using exactly the same library versions as you I fail to reproduce this. No crash, everything gets displayed fine. Can you provide any additional information on how to reproduce this? Kind regards Nico -- Nico Golde - http://www.ngolde.de - [EMAIL PROTECTED] - GPG: 0x73647CFF For security reasons, all text in this mail is double-rot13 encrypted. pgpggzB3H3AVq.pgp Description: PGP signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
On Mon, Mar 10, 2008 at 10:35:32PM +0100, Nico Golde wrote: I tried to reproduce this using the same locale ([EMAIL PROTECTED]), setting it as LC_ALL and invoking newsbeuter in urxvt. Even using exactly the same library versions as you I fail to reproduce this. No crash, everything gets displayed fine. Can you provide any additional information on how to reproduce this? please try setting the LANG variable: with LANG=C it works but displays ? instead the offending à. If instead I use the usual [EMAIL PROTECTED] (or also it_IT.UTF-8) it crashes. regards R, signature.asc Description: Digital signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
Hi Riccardo, * Riccardo Stagni [EMAIL PROTECTED] [2008-03-10 23:29]: On Mon, Mar 10, 2008 at 10:35:32PM +0100, Nico Golde wrote: I tried to reproduce this using the same locale ([EMAIL PROTECTED]), setting it as LC_ALL and invoking newsbeuter in urxvt. Even using exactly the same library versions as you I fail to reproduce this. No crash, everything gets displayed fine. Can you provide any additional information on how to reproduce this? please try setting the LANG variable: with LANG=C it works but displays ? instead the offending à. If instead I use the usual [EMAIL PROTECTED] (or also it_IT.UTF-8) it crashes. No problem, still not crashing :/ newsbeuter 0.8 - Articoli nel feed 'Il blog di Bologna Spettro della bolognesità' (10 non letti, 10 totale) So everything seems to be fine. I forwarded this to the upstream author, maybe he has an idea. Kind regards Nico P.S. Did you experience this also with older versions or are you a new user? -- Nico Golde - http://www.ngolde.de - [EMAIL PROTECTED] - GPG: 0x73647CFF For security reasons, all text in this mail is double-rot13 encrypted. pgpNh1yIEOHG8.pgp Description: PGP signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
On Mon, 10 Mar 2008 22:59:32 +0100, Riccardo Stagni wrote: I tried to reproduce this using the same locale ([EMAIL PROTECTED]), setting it as LC_ALL and invoking newsbeuter in urxvt. Even using exactly the same library versions as you I fail to reproduce this. No crash, everything gets displayed fine. I have the same problems as the submitter (i.e. the crash with the same error). I just had no idea yet what might be the cause :) please try setting the LANG variable: with LANG=C it works but displays ? instead the offending à. If instead I use the usual [EMAIL PROTECTED] (or also it_IT.UTF-8) it crashes. Similar here: [EMAIL PROTECTED]:~$ locale LANG=C [EMAIL PROTECTED] LC_NUMERIC=C LC_TIME=C LC_COLLATE=C LC_MONETARY=C LC_MESSAGES=C LC_PAPER=C LC_NAME=C LC_ADDRESS=C LC_TELEPHONE=C LC_MEASUREMENT=C LC_IDENTIFICATION=C LC_ALL= [EMAIL PROTECTED]:~$ newsbeuter [start messages, screen blank] terminate called after throwing an instance of 'char*' Aborted [EMAIL PROTECTED]:~$ LANG=C newsbeuter the same (well, LC_CTYPE is not overriden) [EMAIL PROTECTED]:~$ LANG=C LC_ALL=C newsbeuter Works. (Without non-ascii characters obviously.) [EMAIL PROTECTED]:~$ LANG=de_AT.utf8 LC_ALL=de_AT.utf8 uxterm newsbeuter Works, too. Interesting. Here's the backtrace from the invokation in my normal locale: terminate called after throwing an instance of 'char*' Program received signal SIGABRT, Aborted. [Switching to Thread 0xb789e6d0 (LWP 20840)] 0xb7bb9546 in raise () from /lib/libc.so.6 (gdb) bt full #0 0xb7bb9546 in raise () from /lib/libc.so.6 No symbol table info available. #1 0xb7bbad58 in abort () from /lib/libc.so.6 No symbol table info available. #2 0xb7dbc748 in __gnu_cxx::__verbose_terminate_handler () from /usr/lib/libstdc++.so.6 No symbol table info available. #3 0xb7dba605 in ?? () from /usr/lib/libstdc++.so.6 No symbol table info available. #4 0xb7dba642 in std::terminate () from /usr/lib/libstdc++.so.6 No symbol table info available. #5 0xb7dba77a in __cxa_throw () from /usr/lib/libstdc++.so.6 No symbol table info available. #6 0x080b0ef7 in ?? () No symbol table info available. #7 0x080b572d in ?? () No symbol table info available. #8 0x080873c1 in ?? () No symbol table info available. #9 0x08065620 in ?? () No symbol table info available. #10 0x0806eaf3 in ?? () No symbol table info available. #11 0x0808a194 in ?? () No symbol table info available. #12 0x08066875 in ?? () No symbol table info available. #13 0x08071a87 in ?? () No symbol table info available. #14 0x0804e005 in ?? () No symbol table info available. #15 0xb7ba5456 in __libc_start_main () from /lib/libc.so.6 No symbol table info available. #16 0x0804dcf1 in ?? () No symbol table info available. (gdb) Cheers, gregor -- .''`. http://info.comodo.priv.at/ | gpg key ID: 0x00F3CFE4 : :' : debian: the universal operating system - http://www.debian.org/ `. `' member of https://www.vibe.at/ | how to reply: http://got.to/quote/ `-NP: Dire Straits: Down To The Waterline signature.asc Description: Digital signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
On Mon, 10 Mar 2008 23:38:29 +0100, Nico Golde wrote: please try setting the LANG variable: with LANG=C it works but displays ? instead the offending à. If instead I use the usual [EMAIL PROTECTED] (or also it_IT.UTF-8) it crashes. No problem, still not crashing :/ This is indeed weird; you get no crashes, Riccardo gets them both with latin9 and utf8, I get them only with latin9 ... Could this be terminal-related in some way? FWIW: I used plain xterm (and uxterm). P.S. Did you experience this also with older versions or are you a new user? Here 0.7-4 works fine (but I know that you know that I'm not a new user :)) Cheers, gregor -- .''`. http://info.comodo.priv.at/ | gpg key ID: 0x00F3CFE4 : :' : debian: the universal operating system - http://www.debian.org/ `. `' member of https://www.vibe.at/ | how to reply: http://got.to/quote/ `-NP: Rolling Stones: Beast Of Burden signature.asc Description: Digital signature
Bug#470390: newsbeuter: dies when reading feeds with bad accents
On Tue, Mar 11, 2008 at 12:17:43AM +0100, gregor herrmann wrote: please try setting the LANG variable: with LANG=C it works but displays ? instead the offending à. If instead I use the usual [EMAIL PROTECTED] (or also it_IT.UTF-8) it crashes. No problem, still not crashing :/ This is indeed weird; you get no crashes, Riccardo gets them both with latin9 and utf8, I get them only with latin9 ... Could this be terminal-related in some way? FWIW: I used plain xterm (and uxterm). I've just tried and it crashes with xterm and with uxterm. P.S. Did you experience this also with older versions or are you a new user? Here 0.7-4 works fine (but I know that you know that I'm not a new user :)) I am, instead, a new user: I've read Nico's orphaning message for snownews and I tought giving a try to newsbeuter. R, signature.asc Description: Digital signature
