Thanks Jaume!

I think the logic of the (new) test case is correct but let's see if
Daniel and/or Marcin can confirm this fix is right, then we can change
the rule you have.
This fix will allow much greater flexibility and will definitely help
me with Ukrainian rules.

Regards,
Andriy

2015-04-29 17:50 GMT-04:00 Jaume OrtolĂ  i Font <jaumeort...@gmail.com>:
> 2015-04-29 19:38 GMT+02:00 Andriy Rysin <ary...@gmail.com>:
>>
>> I just found out that if I have multiple tokens with min="0" my
>> patterns don't match. Looking at the code it seems like if min="0" we
>> only check for next pattern to match but that next may also have 0
>> mins.
>> I wrote little patch with tests that make my rules work but
>> unfortunately it breaks 1 rule in Catalan (where two consequitive min
>> 0s are used). All other languages tests pass.
>>
>> I would appreciate if somebody can take a look at the patch in general
>> and also why it breaks ca rule.
>
>
> This is the rule:
>
>             <rule>
>                 <pattern>
>                     <token inflected="yes">hi</token>
>                     <token postag="V.[MSI].*" postag_regexp="yes"/>
>                     <token min="0" postag="V.[NP].*" postag_regexp="yes"/>
>                     <token min="0" inflected="yes">de</token>
>                     <token postag="V.[NG].*" postag_regexp="yes"/>
>                     <token regexp="yes">[-']hi</token>
>                 </pattern>
>                 <message>Pronom feble duplicat.</message>
>                 <example correction=""><marker>Hi vam haver
> d'anar-hi</marker>.</example>
>                 <example correction=""><marker>Hi hem hagut
> d'anar-hi</marker>.</example>
>                 <example>Vam haver d'anar-hi.</example>
>             </rule>
>
> The problem, I think, is that there is an ambiguity because a word with
> postag="VMN000" ("haver") can match token number 3 and token number 5 (if
> both tokens with min="0" are skipped). In the first example it should match
> token 3, but it matches token 5 and the whole pattern doesn't match.
>
> If the bug can't be fixed, I can write the rule in some other way easily.
> That's not a problem
>
> Regards,
> Jaume
>
>
>
> ------------------------------------------------------------------------------
> One dashboard for servers and applications across Physical-Virtual-Cloud
> Widest out-of-the-box monitoring support with 50+ applications
> Performance metrics, stats and reports that give you Actionable Insights
> Deep dive visibility with transaction tracing using APM Insight.
> http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
> _______________________________________________
> Languagetool-devel mailing list
> Languagetool-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/languagetool-devel
>

------------------------------------------------------------------------------
One dashboard for servers and applications across Physical-Virtual-Cloud 
Widest out-of-the-box monitoring support with 50+ applications
Performance metrics, stats and reports that give you Actionable Insights
Deep dive visibility with transaction tracing using APM Insight.
http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
_______________________________________________
Languagetool-devel mailing list
Languagetool-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/languagetool-devel

Reply via email to