[EP-tech] Re: {Disarmed} Re: Re: IRStats2 Beta

Sebastien Francois sf2 at ecs.soton.ac.uk
Thu Jun 20 15:02:41 BST 2013


Hi JM,

That's a bug in EPrints::Database::prepare_select (I recall that bug 
nearly drove me mad!).

See https://github.com/eprints/irstats2/issues/1 and look at my comments 
for a way to fix that issue.

Seb.


On 19/06/13 08:51, Jean-Marie Le Bechec wrote:
> Hi Seb,
>
> the process is ok (create table, etc) :
>
> $ /appli/eprints/archives/oatao/bin/stats/process_stats oatao --setup 
> --verbose
> Generating set tables.
> Removed table 'irstats2_sets_authors'
> Removed table 'irstats2_sets_divisions'
> Removed table 'irstats2_sets_subjects'
> Removed table 'irstats2_sets_type'
> Removed table 'irstats2_groupings_authors'
> Removed table 'irstats2_groupings_divisions'
> Removed table 'irstats2_groupings_subjects'
> Removed table 'irstats2_groupings_type'
> Removed table 'irstats2_cache_set_values'
> Creating internal tables
> About to process dataset 'history'
> Creating table 'irstats2_history'
> history: starting from record 0
> history: 1083209 records to process.
> About to process dataset 'access'
> Creating table 'irstats2_referrer'
> Creating table 'irstats2_browsers'
> Creating table 'irstats2_search_terms'
> Creating table 'irstats2_downloads'
> Creating table 'irstats2_views'
> Access: accessid to process: from 0 to 1727808
> Access: it took 1 secs to parse 1 records ( average = 1.00 records/sec )
> Access: 1 records kept out of 1 ( ratio = 100.00% )
> About to process dataset 'eprint'
> Creating table 'irstats2_doc_access'
> Creating table 'irstats2_deposits'
> Creating table 'irstats2_doc_format'
> eprint: starting from record 0
> eprint: 2652 records to process.
> Leaving now
>
> but the results are not (or not complete). I see some informations 
> about eprints database but nothing about downloads! See attached images.
>
> Another thing : Looking where this problem could come I changed the 
> perl path in this file fp_stats.
>
> Thanks
>
> jean-marie
>
>
> Le 17/06/2013 17:01, Jean-Marie Le Bechec a écrit :
>> Sol. 1 !
>>
>> jean-Marie
>>
>> Le 17/06/2013 16:47, Sebastien Francois a écrit :
>>> Hello Jean-Marie!
>>>
>>> On 17/06/13 15:09, Jean-Marie Le Bechec wrote:
>>>> Hi Seb,
>>>>
>>>> When i try to execute the script : ./process_stats oatao --setup 
>>>> --verbose
>>>>
>>>> the prompt give me this :
>>>>
>>>> "Generating set tables.
>>>> Can't call method "clone" on an undefined value at 
>>>> /appli/eprints/perl_lib/EPrints/Database.pm line 931, <DATA> line 959."
>>>
>>> Yep, that rings a bell. The problem is that in EPrints 3.3.9, some 
>>> of the core DB API has changed (namely "create_tables()") which is 
>>> very naughty. My code abstracts this and tries to detect the changes 
>>> but in your instance it seems to have failed to detect your EPrints 
>>> version.
>>>
>>> You can either:
>>>
>>> (1) patch EPrints/Plugin/Stats/Handler.pm::_create_table (~line 
>>> 1198) to:
>>>
>>> if( (EPrints->VERSION()||'v3.2.0') gt *'v3.3.8' *)
>>>
>>> ^^ quotes v3.3.8 basically
>>>
>>> and test. if that doesn't work, just force the old API call:
>>>
>>> (2) same file, same lines:
>>>
>>> sub _create_table
>>> {
>>>         my( $self, $tablename, $setkey, @fields ) = @_;
>>>
>>> *#*        if( (EPrints->VERSION()||'v3.2.0') gt v3.3.8 )
>>> *#*       {
>>> *#*                return $self->{dbh}->create_table( $tablename, 
>>> $setkey, @fields );
>>> *#*        }
>>>
>>>         # pre-3.3.9 API
>>>         return $self->{dbh}->create_table( $tablename, undef, 
>>> $setkey, @fields );
>>> }
>>>
>>> and this should work.
>>>
>>>
>>>>
>>>> My eprints is : EPrints 3.2.4 (Victoria Sponge)
>>>>
>>>> I also have installed "Date::Calc" and added 
>>>> EP_TRIGGER_DYNAMIC_TEMPLATE => 10, (line 123)
>>>> EP_TRIGGER_DYNAMIC_TEMPLATE (line 153)
>>>>
>>> OK for Date::Calc.
>>>
>>> I don't think that adding the trigger's constants will help here 
>>> because you may have the constants defined nicely, but you'll be 
>>> missing the code which is actually calling the trigger.
>>>
>>> And since you're using v3.2.4, I think you'd need to back-port a bit 
>>> too much code.
>>>
>>> Anyway the point of that trigger is to insert some elements in the 
>>> <head> sections of your template, that's something you can easily do 
>>> by hand:
>>>
>>> - edit your template files (default.xml, perhaps secure.xml)
>>>
>>> - add the following into your <head>:
>>>
>>> <!-- IRStats2 -->
>>> <script type="text/javascript" src=*MailScanner has detected a 
>>> possible fraud attempt from "www.google.com" claiming to be* 
>>> "http://www.google.com/jsapi">// <!-- No script --></script>
>>> <script type="text/javascript">google.load("visualization", "1", 
>>> {packages:["corechart", "geochart"]});</script>
>>> <!-- end IRStats2 -->
>>>
>>> If you're editing secure.xml, you'll be loading the above page from 
>>> an HTTPS so technically you should use:
>>>
>>> <!-- IRStats2 -->
>>> <script type="text/javascript" 
>>> src="*https*://www.google.com/jsapi">// <!-- No script --></script>
>>> <script type="text/javascript">google.load("visualization", "1", 
>>> {packages:["corechart", "geochart"]});</script>
>>> <!-- end IRStats2 -->
>>>
>>> - reload apache
>>>
>>>
>>> Merci for your feedback JM - if you get to patch 
>>> Handler::_create_table, could you let me know which method worked (1 
>>> or 2)? This way I can patch the IRStats2.
>>>
>>> Seb.
>>>
>>>
>>>> in Const.pm
>>>>
>>>> Any idea ?
>>>>
>>>> Thanks !
>>>>
>>>> Jean-Marie
>>>>
>>>>
>>>> Le 05/06/2013 15:09, Sebastien Francois a écrit :
>>>>> Hi Pierre,
>>>>>
>>>>> I haven't tested on 3.1, I doubt it will work as most of the API 
>>>>> is using some "concepts" introduced in 3.2.
>>>>>
>>>>> Seb.
>>>>>
>>>>> On 05/06/13 13:47, Nault, Pierre wrote:
>>>>>>
>>>>>> Hi Sebastien,
>>>>>>
>>>>>> What is the compatibility with eprints 3.1 ?
>>>>>>
>>>>>> Pierre Nault
>>>>>>
>>>>>> *De :*eprints-tech-bounces at ecs.soton.ac.uk 
>>>>>> [mailto:eprints-tech-bounces at ecs.soton.ac.uk] *De la part de* 
>>>>>> Sebastien Francois
>>>>>> *Envoyé :* 5 juin 2013 05:56
>>>>>> *À :* eprints-tech at ecs.soton.ac.uk
>>>>>> *Objet :* [EP-tech] Re: IRStats2 Beta
>>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> It's a rewrite so I'm tempted to say that everything is different ;-)
>>>>>>
>>>>>> The main points are:
>>>>>>
>>>>>> - integrated to the EPrints API so configuring, extending etc 
>>>>>> should be easy if you know EPrints already. It's also rather easy 
>>>>>> to include graphs etc on pages
>>>>>> - report-based, which can be configured/managed in the local conf
>>>>>> - it can generate data from any EPrints' dataset, not just the 
>>>>>> download stats (for example: graph of deposits over time, size of 
>>>>>> the review over time etc...)
>>>>>> - data can be exported to XML, JSON, CSV (should be easy to add 
>>>>>> other formats if needed)
>>>>>> - uses Google Charts
>>>>>>
>>>>>> The main thing for me is that it's easy to customise/extend e.g. 
>>>>>> you could have reports on citations if you have this enabled on 
>>>>>> your repository.
>>>>>>
>>>>>> You can see a demo there: http://files.eprints.org/cgi/stats/report
>>>>>>
>>>>>> Seb.
>>>>>>
>>>>>>
>>>>>> On 04/06/13 18:34, Centro de Documentación wrote:
>>>>>>
>>>>>>     Hi Sebastian,
>>>>>>
>>>>>>     What are the main differences between version 1 and 2?
>>>>>>
>>>>>>     Of course, Google Charts and  ...
>>>>>>
>>>>>>     Regards,
>>>>>>
>>>>>>     On Thu, May 30, 2013 at 12:17 PM, Sebastien Francois
>>>>>>     <sf2 at ecs.soton.ac.uk <mailto:sf2 at ecs.soton.ac.uk>> wrote:
>>>>>>
>>>>>>     Hi all,
>>>>>>
>>>>>>     I have added IRStats2 to the eprints GitHub so you may now
>>>>>>     download and
>>>>>>     use/test it.
>>>>>>
>>>>>>     EPrints 3.3 users may use the bazaar package available from
>>>>>>     there:
>>>>>>     https://github.com/eprints/irstats2/blob/master/epm/irstats2-0.0.2.epm
>>>>>>     (note that it's not yet live on the Bazaar-store). Even so
>>>>>>     you can
>>>>>>     install the package easily this way, you still need to run a
>>>>>>     script to
>>>>>>     process the stats. That is not yet automated.
>>>>>>
>>>>>>     EPrints 3.2 users will have to copy files by hand, it's pretty
>>>>>>     straight-forward and the INSTALL file contains some instructions.
>>>>>>
>>>>>>     It's not quite the final/stable release yet but it has been
>>>>>>     thoroughly
>>>>>>     tested on a few "real" repositories and it's unlikely to break a
>>>>>>     repository anyway.
>>>>>>
>>>>>>     I hope you'll enjoy, please contribute (ideas / comments /
>>>>>>     bugs etc)
>>>>>>     directly to me or to GitHub.
>>>>>>
>>>>>>     Seb.
>>>>>>
>>>>>>     PS: yes, it can run in parallel to IRStats1.
>>>>>>     *** Options:
>>>>>>     http://mailman.ecs.soton.ac.uk/mailman/listinfo/eprints-tech
>>>>>>     *** Archive: http://www.eprints.org/tech.php/
>>>>>>     *** EPrints community wiki: http://wiki.eprints.org/
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>     *** Options:http://mailman.ecs.soton.ac.uk/mailman/listinfo/eprints-tech
>>>>>>
>>>>>>     *** Archive:http://www.eprints.org/tech.php/
>>>>>>
>>>>>>     *** EPrints community wiki:http://wiki.eprints.org/
>>>>>>
>>>>>>
>>>>>>
>>>>>> *** Options:http://mailman.ecs.soton.ac.uk/mailman/listinfo/eprints-tech
>>>>>> *** Archive:http://www.eprints.org/tech.php/
>>>>>> *** EPrints community wiki:http://wiki.eprints.org/
>>>>>
>>>>>
>>>>>
>>>>> *** Options:http://mailman.ecs.soton.ac.uk/mailman/listinfo/eprints-tech
>>>>> *** Archive:http://www.eprints.org/tech.php/
>>>>> *** EPrints community wiki:http://wiki.eprints.org/
>>>>
>>>> -- 
>>>>
>>>> ***********************************************
>>>> Jean Marie Le Bechec
>>>> Service Commun de la Documentation
>>>> Responsable ingenierie documentaire
>>>> &
>>>> Direction du Systeme d'Information
>>>> Referent Etudes
>>>>
>>>> Institut National Polytechnique de Toulouse
>>>> 6 allee Emile Monso - bp 34038 -
>>>> 31029 Toulouse cedex 4
>>>> Tel : 05 34 32 31 16
>>>> Mail :lebechec at inp-toulouse.fr
>>>> ***********************************************
>>>
>>
>> -- 
>>
>> ***********************************************
>> Jean Marie Le Bechec
>> Service Commun de la Documentation
>> Responsable ingenierie documentaire
>> &
>> Direction du Systeme d'Information
>> Referent Etudes
>>
>> Institut National Polytechnique de Toulouse
>> 6 allee Emile Monso - bp 34038 -
>> 31029 Toulouse cedex 4
>> Tel : 05 34 32 31 16
>> Mail :lebechec at inp-toulouse.fr
>> ***********************************************
>>
>>
>> *** Options:http://mailman.ecs.soton.ac.uk/mailman/listinfo/eprints-tech
>> *** Archive:http://www.eprints.org/tech.php/
>> *** EPrints community wiki:http://wiki.eprints.org/
>
> -- 
>
> ***********************************************
> Jean Marie Le Bechec
> Service Commun de la Documentation
> Responsable ingenierie documentaire
> &
> Direction du Systeme d'Information
> Referent Etudes
>
> Institut National Polytechnique de Toulouse
> 6 allee Emile Monso - bp 34038 -
> 31029 Toulouse cedex 4
> Tel : 05 34 32 31 16
> Mail :lebechec at inp-toulouse.fr
> ***********************************************

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ecs.soton.ac.uk/pipermail/eprints-tech/attachments/20130620/25b5b73b/attachment-0001.html 


More information about the Eprints-tech mailing list