From noreply at sourceforge.net Sun Mar 4 02:07:13 2012 From: noreply at sourceforge.net (SourceForge.net) Date: Sun Mar 4 02:07:33 2012 Subject: [CWB] [ cwb-Bugs-3427282 ] CQP output redirect doesn't always check SIGPIPE Message-ID: <20120304010724.B7DC698C38@einstein.sslmit.unibo.it> Bugs item #3427282, was opened at 2011-10-22 13:05 Message generated for change (Comment added) made by nobody You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=722303&aid=3427282&group_id=131809 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: CQP interface Group: None Status: Open Resolution: None Priority: 4 Private: No Submitted By: Stefan Evert (schtepf) Assigned to: Stefan Evert (schtepf) Summary: CQP output redirect doesn't always check SIGPIPE Initial Comment: When redirecting "cat;" output to a pipe, CQP installs a signal handler to check for broken pipe signals, so the output loop can be aborted (important e.g. when displaying a large query result in the interactive pager). Other output commands with redirect ? count, tabulate, etc. ? fail to do so. The signal handler code should be merged into the open_stream() and close_stream() functions in . All CQP output commands that support redirects through this mechanism will then only have to check the global variable broken_pipe. ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2012-03-03 17:07 Message: vR0tit mtxcnbmwxred, [url=http://iddhlksuwgvq.com/]iddhlksuwgvq[/url], [link=http://uxrhdaiuevnn.com/]uxrhdaiuevnn[/link], http://vnbfnkffxzdj.com/ ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=722303&aid=3427282&group_id=131809 From yversley at gmail.com Tue Mar 13 15:34:00 2012 From: yversley at gmail.com (Yannick Versley) Date: Tue Mar 13 15:34:08 2012 Subject: [CWB] cl.h and C++ Message-ID: Hi, I was trying to use cl.h (respectively, the Cython code that uses it) together with another module that is written in C++ and there is a problem because C++ includes the functions from cl.h but mangles their names (thinking that they are C++ functions and not C ones). This can be easily fixed by putting a #ifdef __cplusplus extern "C" { #endif and a #ifdef __cplusplus } #endif around the function declarations in cl.h, which would allow C++ code to use the C-level API for CWB. I'll probably have to make do with a workaround for now (which is easy enough), but maybe it would be useful for the benefit of future generations ;-) Best, Yannick -------------- next part -------------- An HTML attachment was scrubbed... URL: http://liste.sslmit.unibo.it/pipermail/cwb/attachments/20120313/9003f25a/attachment.htm From noreply at sourceforge.net Tue Mar 13 16:55:36 2012 From: noreply at sourceforge.net (SourceForge.net) Date: Tue Mar 13 16:55:54 2012 Subject: [CWB] [ cwb-Feature Requests-2817253 ] CQPweb: subcorpus-as-corpus Message-ID: <20120313155546.5C43298DFC@einstein.sslmit.unibo.it> Feature Requests item #2817253, was opened at 2009-07-05 22:24 Message generated for change (Comment added) made by nobody You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=722306&aid=2817253&group_id=131809 Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: CQPweb Group: None Status: Open Priority: 3 Private: No Submitted By: Andrew Hardie (andrewhardie) Assigned to: Andrew Hardie (andrewhardie) Summary: CQPweb: subcorpus-as-corpus Initial Comment: Allow subcorpora to be assigned their own web-folders, settings file, user access permissions, etc and accessed as though they were independently-indexed corpora. So, for instance, you could have brown-family as a corpus with subcorpora borwn, flob, lob, frown... ... and then each of brown, frown, lob, flob could also be accessed independently (if set up for this by the admin user) and the average user would not be able to tell the difference between this and an independently indexed copy of Brown (or whatever). This will involve duplication of (some or all) mysql data, but not of CWB indexed files. (This is a rather major feature, so not something for the immediate future!) ---------------------------------------------------------------------- Comment By: Nobody/Anonymous (nobody) Date: 2012-03-13 08:55 Message: RBT5Gk waxpcbpicqiv, [url=http://kixgdrsnofhq.com/]kixgdrsnofhq[/url], [link=http://uquuvfeiwynh.com/]uquuvfeiwynh[/link], http://sdynowqucxer.com/ ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=722306&aid=2817253&group_id=131809 From sylvain.loiseau at wanadoo.fr Thu Mar 29 14:06:49 2012 From: sylvain.loiseau at wanadoo.fr (Sylvain LOISEAU) Date: Thu Mar 29 14:06:58 2012 Subject: [CWB] [CQPWeb] dealing with large corpora Message-ID: <233094870.100558.1333022810005.JavaMail.www@wwinf1j15> Dear all, I've installed a rather large corpus on CQPWeb (500 million+ tokens). However I've not been able to categorize the corpus neither to create frequency list: it seems that the corpus is to big and "time out" is reached before the processes have completed. Do you know how to cope with that? Does a list of the command-lines / SQL instructions performed by CQP web is available, so that I try to execute them one by one, in a terminal? Best regards, Sylvain Loiseau From sylvain.loiseau at wanadoo.fr Thu Mar 29 14:07:11 2012 From: sylvain.loiseau at wanadoo.fr (Sylvain LOISEAU) Date: Thu Mar 29 14:07:15 2012 Subject: [CWB] [CQPWeb] dealing with large corpora Message-ID: <1439776698.100579.1333022831977.JavaMail.www@wwinf1j15> Dear all, I've installed a rather large corpus on CQPWeb (500 million+ tokens). However I've not been able to categorize the corpus neither to create frequency list: it seems that the corpus is to big and "time out" is reached before the processes have completed. Do you know how to cope with that? Does a list of the command-lines / SQL instructions performed by CQP web is available, so that I try to execute them one by one, in a terminal? Best regards, Sylvain Loiseau From a.hardie at lancaster.ac.uk Fri Mar 30 19:05:52 2012 From: a.hardie at lancaster.ac.uk (Hardie, Andrew) Date: Fri Mar 30 19:06:01 2012 Subject: [CWB] [CQPWeb] dealing with large corpora In-Reply-To: <1439776698.100579.1333022831977.JavaMail.www@wwinf1j15> References: <1439776698.100579.1333022831977.JavaMail.www@wwinf1j15> Message-ID: <28078EC3FBF1B940A3EF3D0D19BE351D059376@EX-0-MB1.lancs.local> Hi Sylvain, You can circumvent the timeout (which is to limit the amount of time on the server that users can occupy) by running frequency list creation outside the web interface. To do this: - go on the commandline to the webdirectory of the corpus in question, e.g. /home/www/cqpweb/my_corpus or whatever - run this command: php ../bin/offline-freqlists.php That should do the trick. Let me know if it doesn't. Alternatively, if you don't need the timeout protection, you can disable it in your php.ini file. best Andrew. -----Original Message----- From: cwb-bounces@sslmit.unibo.it [mailto:cwb-bounces@sslmit.unibo.it] On Behalf Of Sylvain LOISEAU Sent: 29 March 2012 13:07 To: cwb@sslmit.unibo.it Subject: [CWB] [CQPWeb] dealing with large corpora Dear all, I've installed a rather large corpus on CQPWeb (500 million+ tokens). However I've not been able to categorize the corpus neither to create frequency list: it seems that the corpus is to big and "time out" is reached before the processes have completed. Do you know how to cope with that? Does a list of the command-lines / SQL instructions performed by CQP web is available, so that I try to execute them one by one, in a terminal? Best regards, Sylvain Loiseau _______________________________________________ CWB mailing list CWB@sslmit.unibo.it http://devel.sslmit.unibo.it/mailman/listinfo/cwb