Revision: 6023
http://languagetool.svn.sourceforge.net/languagetool/?rev=6023&view=rev
Author: dominikoeo
Date: 2011-12-10 20:35:07 +0000 (Sat, 10 Dec 2011)
Log Message:
-----------
[br] - better disambiguation of the word "o" which
allows detecting more incorrect mutations such as:
*o patatez* ?\226?\134?\146 "o fatatez".
- fixed a couple of false positives.
Modified Paths:
--------------
trunk/JLanguageTool/src/resource/br/disambiguation.xml
trunk/JLanguageTool/src/rules/br/grammar.xml
Modified: trunk/JLanguageTool/src/resource/br/disambiguation.xml
===================================================================
--- trunk/JLanguageTool/src/resource/br/disambiguation.xml 2011-12-10
18:55:43 UTC (rev 6022)
+++ trunk/JLanguageTool/src/resource/br/disambiguation.xml 2011-12-10
20:35:07 UTC (rev 6023)
@@ -39,6 +39,13 @@
<disambig postag="P"/>
</rule>
</rulegroup>
+ <rule id="O_R" name="O pronoun">
+ <pattern mark="0">
+ <token>o</token>
+ <token postag="N .*" postag_regexp="yes"><exception postag="V
inf.*"/></token>
+ </pattern>
+ <disambig postag="R e p 3 obj"/>
+ </rule>
<rulegroup id="VERB_PARTICLE" name="verb particule">
<rule>
<pattern mark="0">
Modified: trunk/JLanguageTool/src/rules/br/grammar.xml
===================================================================
--- trunk/JLanguageTool/src/rules/br/grammar.xml 2011-12-10 18:55:43 UTC
(rev 6022)
+++ trunk/JLanguageTool/src/rules/br/grammar.xml 2011-12-10 20:35:07 UTC
(rev 6023)
@@ -699,6 +699,9 @@
<example type="incorrect">Re ziwezhat out va
<marker>baotr</marker>!</example>
<example type="correct">Re ziwezhat out va faotr!</example>
<example type="correct">Va beg digor</example>
+ <example type="incorrect">Bras eo o <marker>patatez</marker>.</example>
+ <example type="incorrect">Bras eo o <marker>batatez</marker>.</example>
+ <example type="correct">Bras eo o fatatez.</example>
</rule>
<rule>
<pattern mark_from="1">
@@ -888,7 +891,7 @@
<token regexp="yes">[au]n</token>
<and>
<token postag="N m s.*|N f p.*" postag_regexp="yes"
regexp="yes">[dz].*
- <exception regexp="yes">dra(-(mañ|se|hont))?</exception>
+ <exception regexp="yes">dra(ig)?(-(mañ|se|hont))?</exception>
<exception postag="N m s.*|N f p.*" postag_regexp="yes"
regexp="yes" inflected="yes">[dz].*</exception>
</token>
<token postag="N m s.*|N f p.*" postag_regexp="yes" regexp="yes"
inflected="yes">t.*
@@ -1563,7 +1566,7 @@
<rule id="AL" name="gerioù al, ul, el, hol">
<pattern mark_to="-1">
<token regexp="yes">[aeu][nr]|hor<exception postag="X EN_UR"/></token>
- <token regexp="yes">l..*</token>
+ <token regexp="yes">l..*<exception regexp="yes"
postag="UNKNOWN">L\p{Lu}*</exception></token>
</pattern>
<message>Ar ger «\1» a zlefe bezañ <suggestion><match no="1"
regexp_match="(.*)." regexp_replace="$1l"/></suggestion> dirak ur ger a grog
gant ar gensonenn «l».</message>
<example type="incorrect"><marker>Ur</marker> levr a zo ganin.</example>
@@ -1576,6 +1579,7 @@
<example type="correct">Tev eo hol levr.</example>
<example type="correct">Kouezhet on en ur lammat.</example>
<example type="correct">An l glebiet</example>
+ <example type="correct">An LTB</example>
</rule>
<rule id="AN" name="gerioù an, un, en, hon">
<pattern mark_to="-1">
@@ -1609,7 +1613,7 @@
<token regexp="yes">[aeu]n<exception scope="previous"
regexp="yes">e?z|ned</exception></token>
<token regexp="yes">[bcfgjkmprsvwyz].*
<exception postag="K .* o.*" postag_regexp="yes"></exception>
- <exception regexp="yes"
postag="UNKNOWN">[FLMNRS]\p{Lu}*</exception>
+ <exception regexp="yes"
postag="UNKNOWN">[FMNRS](\p{Lu}|-)*</exception>
<exception regexp="yes">[FLMNRS]\d*</exception>
</token>
</pattern>
@@ -1621,6 +1625,7 @@
<example type="correct">Ar ouiziegezh</example>
<example type="incorrect"><marker>AN</marker> FILM A BLIJ DIN</example>
<example type="correct">Gant an FSE</example>
+ <example type="correct">An SNJ-CGT</example>
<example type="correct">Un S.O.S.</example>
<example type="correct">Gustav ez an war-sav.</example>
</rule>
@@ -2141,13 +2146,14 @@
<rule id="E_KREIZ" name="e kreiz → e-kreiz (hag all)">
<pattern>
<token>e</token>
- <token
regexp="yes">barzh|douare|giz(-(all|mañ|se|hon))?|harz|hunan|keit(-se)?|kerz|kichen|krec’h|kreiz|lec’h|leizh|maez|pad|sell|ser|tailh|tal|touez|tro|unan</token>
+ <token
regexp="yes">barzh|douare|giz(-(all|mañ|se|hon))?|harz|hunan|keit(-se)?|kerz|kichen|krec’h|kreiz|leizh|maez|pad|sell|ser|tailh|tal|touez|tro|unan</token>
</pattern>
<message>Implijit <suggestion>\1-\2</suggestion>.</message>
<example type="incorrect"><marker>E kichen</marker> Gwitalmeze emañ o
chom.</example>
<example type="correct">E-kichen Gwitalmeze emañ o chom.</example>
<example type="incorrect">Rwanda zo ur vro <marker>e kreiz</marker>
Afrika.</example>
<example type="correct">Rwanda zo ur vro e-kreiz Afrika.</example>
+ <example type="correct">E breizh hag <marker>e lec’h</marker>
all.</example>
</rule>
<rule id="DIWAR_BENN" name="diwar benn → diwar-benn (hag all)">
@@ -2377,12 +2383,13 @@
</rulegroup>
<rule id="DIGAREZIT_DIN" name="Digarezit din → Digarezit ac’hanon">
<pattern mark_from="1">
- <token regexp="yes">digarez(it)?</token>
+ <token regexp="yes">digarez(it)?<exception
scope="previous">un</exception></token>
<token inflected="yes">da</token>
</pattern>
<message>Ne vez ket implijet an araogenn «da» gat ar verb «\1».</message>
<example type="incorrect">Digarezit <marker>dimp</marker></example>
<example type="correct">Digarezit ac’hanomp</example>
+ <example type="correct">Un digarez da ziskouez Breizh.</example>
</rule>
<rule id="A_RAOK_DA" name="Digarezit din → Digarezit ac’hanon">
<pattern mark_from="1" mark_to="-1">
This was sent by the SourceForge.net collaborative development platform, the
world's largest Open Source development site.
------------------------------------------------------------------------------
Learn Windows Azure Live! Tuesday, Dec 13, 2011
Microsoft is holding a special Learn Windows Azure training event for
developers. It will provide a great way to learn Windows Azure and what it
provides. You can attend the event by watching it streamed LIVE online.
Learn more at http://p.sf.net/sfu/ms-windowsazure
_______________________________________________
Languagetool-cvs mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/languagetool-cvs
