You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@uima.apache.org by mb...@apache.org on 2007/09/25 12:58:12 UTC
svn commit: r579196 - in
/incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc:
RegExAnnotator.xml concepts.xml
Author: mbaessler
Date: Tue Sep 25 03:58:04 2007
New Revision: 579196
URL: http://svn.apache.org/viewvc?rev=579196&view=rev
Log:
UIMA-539
update regex rules file
https://issues.apache.org/jira/browse/UIMA-539
Modified:
incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/RegExAnnotator.xml
incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/concepts.xml
Modified: incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/RegExAnnotator.xml
URL: http://svn.apache.org/viewvc/incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/RegExAnnotator.xml?rev=579196&r1=579195&r2=579196&view=diff
==============================================================================
--- incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/RegExAnnotator.xml (original)
+++ incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/RegExAnnotator.xml Tue Sep 25 03:58:04 2007
@@ -52,38 +52,24 @@
<typeSystemDescription>
<types>
<typeDescription>
- <name>org.apache.uima.TestAnnot</name>
+ <name>org.apache.uima.EmailAddress</name>
<description/>
<supertypeName>uima.tcas.Annotation</supertypeName>
<features>
<featureDescription>
- <name>testFeature</name>
+ <name>localPart</name>
<description/>
<rangeTypeName>uima.cas.String</rangeTypeName>
</featureDescription>
<featureDescription>
- <name>testFeature1</name>
- <description/>
- <rangeTypeName>org.apache.uima.TestAnnot1</rangeTypeName>
- </featureDescription>
- <featureDescription>
- <name>confidenceValue</name>
- <description/>
- <rangeTypeName>uima.cas.Float</rangeTypeName>
- </featureDescription>
- <featureDescription>
- <name>ruleId</name>
+ <name>domainPart</name>
<description/>
<rangeTypeName>uima.cas.String</rangeTypeName>
</featureDescription>
</features>
</typeDescription>
- <typeDescription>
- <name>org.apache.uima.TestAnnot1</name>
- <description/>
- <supertypeName>uima.tcas.Annotation</supertypeName>
- </typeDescription>
- </types>
+
+ </types>
</typeSystemDescription>
<capabilities>
<capability>
Modified: incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/concepts.xml
URL: http://svn.apache.org/viewvc/incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/concepts.xml?rev=579196&r1=579195&r2=579196&view=diff
==============================================================================
--- incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/concepts.xml (original)
+++ incubator/uima/sandbox/trunk/RegularExpressionAnnotator/desc/concepts.xml Tue Sep 25 03:58:04 2007
@@ -19,59 +19,18 @@
-->
<conceptSet xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:noNamespaceSchemaLocation="concept.xsd">
- <concept name="carConcept">
+ <concept name="emailAddressDetection">
<rules>
- <rule ruleId="ID1" regEx="TestRegex" matchStrategy="matchAll" matchType="uima.tcas.DocumentAnnotation" confidence="1.0">
- <matchTypeFilter>
- <feature name="language">en</feature>
- </matchTypeFilter>
- <updateMatchTypeAnnotation>
- <setFeature name="language" type="String">$0</setFeature>
- </updateMatchTypeAnnotation>
- <ruleExceptions>
- <exception matchType="uima.tcas.DocumentAnnotation">Exception</exception>
- </ruleExceptions>
-
- </rule>
- <rule regEx="TestRegexSubrule1" matchStrategy="matchAll" matchType="uima.tcas.DocumentAnnotation" confidence="0.7">
- <matchTypeFilter>
- <feature name="language">en</feature>
- </matchTypeFilter>
- </rule>
- <rule regEx="TestRegexSubrule2" matchStrategy="matchAll" matchType="uima.tcas.DocumentAnnotation" confidence="0.3">
- <matchTypeFilter>
- <feature name="language">en</feature>
- </matchTypeFilter>
-
- </rule>
+ <rule regEx="([a-zA-Z0-9!#$%*+'/=?^_-`{|}~.\x26]+)@([a-zA-Z0-9._-]+[a-zA-Z]{2,4})" matchStrategy="matchAll" matchType="uima.tcas.DocumentAnnotation"/>
</rules>
<createAnnotations>
- <annotation id="testannot" type="org.apache.uima.TestAnnot">
- <begin group="0"/>
- <end group="0"/>
- <setFeature name="testFeature" type="String">$0</setFeature>
- <setFeature name="testFeature1" type="Reference">testannot1</setFeature>
- <setFeature name="confidenceValue" type="Confidence"/>
- <setFeature name="ruleId" type="RuleId"/>
- </annotation>
-
- <annotation id="testannot1" type="org.apache.uima.TestAnnot1">
- <begin group="0"/>
- <end group="0"/>
- </annotation>
- </createAnnotations>
- </concept>
-
- <concept name="simpleConcept">
- <rules>
- <rule regEx="RestRegex" matchStrategy="matchAll" matchType="uima.tcas.DocumentAnnotation"/>
- </rules>
- <createAnnotations>
- <annotation id="testannotation" type="org.apache.uima.TestAnnot1">
+ <annotation id="emailAnnot" type="org.apache.uima.EmailAddress">
<begin group="0"/>
<end group="0"/>
+ <setFeature name="localPart" type="String">$1</setFeature>
+ <setFeature name="domainPart" type="String">$2</setFeature>
</annotation>
</createAnnotations>
</concept>
-
+
</conceptSet>