Changeset 33882 for other-projects/maori-lang-detection/src/org/greenstone/atea/MongoDBAccess.java

Timestamp:

2020-01-30T22:54:39+13:00 (4 years ago)

Author:

ak19

Message:

Code now writes both a listing of all non-autotranslated websites and a listing of overseas autotranslated sites.

File:

• other-projects/maori-lang-detection/src/org/greenstone/atea/MongoDBAccess.java (modified) (4 diffs)

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

@@ -96 +96 @@
     public static final int CONTAINS_MRI = 1;
 
+    /** Some reused fieldnames in the Websites collection */
+    private static final String FILTER_NUMPAGES_IN_MRI = "numPagesInMRI";
+    private static final String FILTER_NUMPAGES_CONTAINING_MRI = "numPagesContainingMRI";
+
     // configuration details, some with fallback values
     private String HOST = "localhost";
@@ -402 +406 @@
     ]);
     */
-    public void aggregateContainsMRIForNZ(Writer writer) throws IOException {
+    public void aggregateContainsMRIForNZ(Writer writer, int filterType) throws IOException {
     // working with the WebSites collection, not WebPages collection!
     MongoCollection<Document> collection = this.database.getCollection(WEBSITES_COLLECTION);
 
-    
-    //String isMRI_filter = 
+    String mriFilterString = (filterType == CONTAINS_MRI) ? "{numPagesContainingMRI: {$gt: 0}}" : "{numPagesInMRI: {$gt: 0}}";
     
     Bson orQuery = or(
@@ -414 +417 @@
               );
     Bson andQuery = and(
-        BasicDBObject.parse("{numPagesContainingMRI: {$gt: 0}}"),
+        BasicDBObject.parse(mriFilterString),
         orQuery);
     
@@ -456 +459 @@
     ]);
     */
-    public void aggregateContainsMRIForOverseas(Writer writer) throws UncheckedIOException {
+    public void aggregateContainsMRIForOverseas(Writer writer, int filterType,
+                        boolean isMiInURLPath) throws UncheckedIOException
+    {
     // working with the WebSites collection, not WebPages collection!
     MongoCollection<Document> collection = this.database.getCollection(WEBSITES_COLLECTION);
-
+    
+    String mriFilterString = (filterType == CONTAINS_MRI) ? "{numPagesContainingMRI: {$gt: 0}}" : "{numPagesInMRI: {$gt: 0}}";
     
     Bson orQuery = or(
               BasicDBObject.parse("{geoLocationCountryCode: \"AU\"}"),
-              BasicDBObject.parse("{urlContainsLangCodeInPath: false}")
+              BasicDBObject.parse("{urlContainsLangCodeInPath: "+ isMiInURLPath +"}")
+              // e.g. "{urlContainsLangCodeInPath: false}"
               );
     Bson andQuery = and(
         BasicDBObject.parse("{geoLocationCountryCode: {$ne: \"NZ\"}}"),
         BasicDBObject.parse("{domain: {$not: /\\.nz/}}"),
-        BasicDBObject.parse("{numPagesContainingMRI: {$gt: 0}}"),
+        BasicDBObject.parse(mriFilterString),
         orQuery);