0ct0pu5/fess
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

fix #2803 Include aside tag in HTML crawler's pruned tags to enhance content relevance.

Browse source
This commit is contained in:
Shinsuke Sugaya 2024-02-05 22:20:57 +09:00
parent cff8b1ccde
commit f6b2ef3b06
2 changed files with 4 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
src/main/java/org/codelibs/fess/mylasta/direction/FessConfig.java
View file

@ -412,7 +412,7 @@ public interface FessConfig extends FessEnv, org.codelibs.fess.mylasta.direction
/** The key of the configuration. e.g. //LINK[@rel='canonical'][1]/@href */
String CRAWLER_DOCUMENT_HTML_CANONICAL_XPATH = "crawler.document.html.canonical.xpath";
/** The key of the configuration. e.g. noscript,script,style,header,footer,nav,a[rel=nofollow] */
/** The key of the configuration. e.g. noscript,script,style,header,footer,aside,nav,a[rel=nofollow] */
String CRAWLER_DOCUMENT_HTML_PRUNED_TAGS = "crawler.document.html.pruned.tags";
/** The key of the configuration. e.g. 120 */
@ -2972,7 +2972,7 @@ public interface FessConfig extends FessEnv, org.codelibs.fess.mylasta.direction
/**
* Get the value for the key 'crawler.document.html.pruned.tags'. <br>
* The value is, e.g. noscript,script,style,header,footer,nav,a[rel=nofollow] <br>
* The value is, e.g. noscript,script,style,header,footer,aside,nav,a[rel=nofollow] <br>
* @return The value of found property. (NotNull: if not found, exception but basically no way)
*/
String getCrawlerDocumentHtmlPrunedTags();
@ -10916,7 +10916,7 @@ public interface FessConfig extends FessEnv, org.codelibs.fess.mylasta.direction
defaultMap.put(FessConfig.CRAWLER_DOCUMENT_HTML_LANG_XPATH, "//HTML/@lang");
defaultMap.put(FessConfig.CRAWLER_DOCUMENT_HTML_DIGEST_XPATH, "//META[@name='description']/@content");
defaultMap.put(FessConfig.CRAWLER_DOCUMENT_HTML_CANONICAL_XPATH, "//LINK[@rel='canonical'][1]/@href");
defaultMap.put(FessConfig.CRAWLER_DOCUMENT_HTML_PRUNED_TAGS, "noscript,script,style,header,footer,nav,a[rel=nofollow]");
defaultMap.put(FessConfig.CRAWLER_DOCUMENT_HTML_PRUNED_TAGS, "noscript,script,style,header,footer,aside,nav,a[rel=nofollow]");
defaultMap.put(FessConfig.CRAWLER_DOCUMENT_HTML_MAX_DIGEST_LENGTH, "120");
defaultMap.put(FessConfig.CRAWLER_DOCUMENT_HTML_DEFAULT_LANG, "");
defaultMap.put(FessConfig.CRAWLER_DOCUMENT_HTML_DEFAULT_INCLUDE_INDEX_PATTERNS, "");

2
src/main/resources/fess_config.properties
View file

@ -239,7 +239,7 @@ crawler.document.html.content.xpath=//BODY
crawler.document.html.lang.xpath=//HTML/@lang
crawler.document.html.digest.xpath=//META[@name='description']/@content
crawler.document.html.canonical.xpath=//LINK[@rel='canonical'][1]/@href
crawler.document.html.pruned.tags=noscript,script,style,header,footer,nav,a[rel=nofollow]
crawler.document.html.pruned.tags=noscript,script,style,header,footer,aside,nav,a[rel=nofollow]
crawler.document.html.max.digest.length=120
crawler.document.html.default.lang=
crawler.document.html.default.include.index.patterns=