crawler.transformer_basic.xml Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of fess-crawler-lasta Show documentation
This is LastaFlute support.
There is a newer version: 14.18.0
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE components PUBLIC "-//DBFLUTE//DTD LastaDi 1.0//EN"
	"http://dbflute.org/meta/lastadi10.dtd">
<components namespace="fessCrawler">
	<include path="crawler/container.xml" />

	<component name="binaryTransformer"
		class="org.codelibs.fess.crawler.transformer.impl.BinaryTransformer"
		instance="singleton">
		<property name="name">"binaryTransformer"</property>
	</component>

	<component name="defaultFeatureMap" class="java.util.LinkedHashMap">
		<postConstruct name="put">
			<arg>"http://xml.org/sax/features/namespaces"</arg>
			<arg>"false"</arg>
		</postConstruct>
	</component>

	<component name="defaultPropertyMap" class="java.util.LinkedHashMap">
	</component>

	<component name="allChildUrlRuleMap" class="java.util.LinkedHashMap">
		<postConstruct name="put">
			<arg>"//A"</arg>
			<arg>"href"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//AREA"</arg>
			<arg>"href"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//FRAME"</arg>
			<arg>"src"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//IFRAME"</arg>
			<arg>"src"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//IMG"</arg>
			<arg>"src"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//LINK"</arg>
			<arg>"href"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//SCRIPT"</arg>
			<arg>"src"</arg>
		</postConstruct>
	</component>

	<component name="htmlUrlRuleMap" class="java.util.LinkedHashMap">
		<postConstruct name="put">
			<arg>"//A"</arg>
			<arg>"href"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//AREA"</arg>
			<arg>"href"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//FRAME"</arg>
			<arg>"src"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//IFRAME"</arg>
			<arg>"src"</arg>
		</postConstruct>
		<postConstruct name="put">
			<arg>"//LINK"</arg>
			<arg>"href"</arg>
		</postConstruct>
	</component>

	<component name="fileTransformer"
		class="org.codelibs.fess.crawler.transformer.impl.FileTransformer" instance="singleton">
		<property name="name">"fileTransformer"</property>
		<property name="featureMap">defaultFeatureMap</property>
		<property name="propertyMap">defaultPropertyMap</property>
		<property name="childUrlRuleMap">allChildUrlRuleMap</property>
	</component>

	<component name="htmlTransformer"
		class="org.codelibs.fess.crawler.transformer.impl.HtmlTransformer" instance="singleton">
		<property name="name">"htmlTransformer"</property>
		<property name="featureMap">defaultFeatureMap</property>
		<property name="propertyMap">defaultPropertyMap</property>
		<property name="childUrlRuleMap">allChildUrlRuleMap</property>
	</component>

	<component name="htmlOnlyTransformer"
		class="org.codelibs.fess.crawler.transformer.impl.HtmlTransformer" instance="singleton">
		<property name="name">"htmlTransformer"</property>
		<property name="featureMap">defaultFeatureMap</property>
		<property name="propertyMap">defaultPropertyMap</property>
		<property name="childUrlRuleMap">htmlUrlRuleMap</property>
	</component>

	<component name="textTransformer"
		class="org.codelibs.fess.crawler.transformer.impl.TextTransformer" instance="singleton">
		<property name="name">"textTransformer"</property>
	</component>
</components>