Changeset 33808 for other-projects/maori-lang-detection/src/org/greenstone/atea/morphia/WebsiteInfo.java

Timestamp:

2019-12-17T19:31:28+13:00 (4 years ago)

Author:

ak19

Message:

Storing not just whether /mi(/) suffix is in path, but also whether http(s):mi. is in path, as storing these can help reducing number of auto-translated sites too in a similar way.

File:

• other-projects/maori-lang-detection/src/org/greenstone/atea/morphia/WebsiteInfo.java (modified) (3 diffs)

other-projects/maori-lang-detection/src/org/greenstone/atea/morphia/WebsiteInfo.java

@@ -21 +21 @@
     
     public final String geoLocationCountryCode;
-    public final boolean urlContainsLangCodeInPath;
+    public final boolean urlContainsLangCodeInPathSuffix;
+    public final boolean urlContainsLangCodeInPathPrefix;
     
     public WebsiteInfo(/*int siteCount,*/ String siteFolderName, String domainOfSite,
@@ -27 +28 @@
                int numPagesInMRI, int numPagesContainingMRI,
                long siteCrawledTimestamp, boolean siteCrawlUnfinished, boolean redoCrawl,
-               String geoLocationCountryCode, boolean urlContainsLangCodeInPath)
+               String geoLocationCountryCode, boolean urlContainsLangCodeInPathSuffix, boolean urlContainsLangCodeInPathPrefix)
     {
     //this.id = siteCount;
@@ -44 +45 @@
     
     this.geoLocationCountryCode = geoLocationCountryCode;
-    this.urlContainsLangCodeInPath = urlContainsLangCodeInPath;
+    this.urlContainsLangCodeInPathSuffix = urlContainsLangCodeInPathSuffix;
+    this.urlContainsLangCodeInPathPrefix = urlContainsLangCodeInPathPrefix;
     }
 }