Mirror/MarginaliaSearch
1
0
Fork 0
mirror of https://github.com/MarginaliaSearch/MarginaliaSearch.git synced 2025-02-24 05:18:58 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

(query-parsing) Drop search term elements that aren't indexed by the search engine

Browse Source
This commit is contained in:
Viktor Lofgren 2024-09-03 11:21:01 +02:00
parent bb5d946c26
commit 8290c19e24
3 changed files with 19 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
code/functions/search-query/java/nu/marginalia/functions/searchquery/query_parser/QueryParser.java
View File

@ -61,10 +61,20 @@ public class QueryParser {
if (str.isBlank()) if (str.isBlank())
return; return;
if (str.endsWith(":") || str.endsWith(".")) { // Remove trailing punctuation
int lastChar = str.charAt(str.length() - 1);
if (":.,!?$".indexOf(lastChar) >= 0)
entity.replace(new QueryToken.LiteralTerm(str.substring(0, str.length() - 1), lt.displayStr())); entity.replace(new QueryToken.LiteralTerm(str.substring(0, str.length() - 1), lt.displayStr()));
}
// Remove term elements that aren't indexed by the search engine
if (str.endsWith("()"))
entity.replace(new QueryToken.LiteralTerm(str.substring(0, str.length() - 2), lt.displayStr()));
if (str.startsWith("$"))
entity.replace(new QueryToken.LiteralTerm(str.substring(1), lt.displayStr()));
if (entity.isBlank()) {
entity.remove();
}
} }
private static void createNegatedTerms(TransformList<QueryToken>.Entity first, TransformList<QueryToken>.Entity second) { private static void createNegatedTerms(TransformList<QueryToken>.Entity first, TransformList<QueryToken>.Entity second) {

8
code/functions/search-query/java/nu/marginalia/util/transform_list/TransformList.java
View File

@ -1,5 +1,7 @@
package nu.marginalia.util.transform_list; package nu.marginalia.util.transform_list;
import nu.marginalia.functions.searchquery.query_parser.token.QueryToken;
import java.util.List; import java.util.List;
import java.util.function.BiConsumer; import java.util.function.BiConsumer;
import java.util.function.Consumer; import java.util.function.Consumer;
@ -30,7 +32,7 @@ import java.util.function.Predicate;
* </pre> * </pre>
* </code> * </code>
*/ */
public class TransformList<T> { public class TransformList<T extends QueryToken> {
private final List<T> backingList; private final List<T> backingList;
public TransformList(List<T> backingList) { public TransformList(List<T> backingList) {
@ -138,6 +140,10 @@ public class TransformList<T> {
value = newValue; value = newValue;
} }
public boolean isBlank() {
return value == null || value.str().isBlank();
}
public void remove() { public void remove() {
action = Action.REMOVE; action = Action.REMOVE;
} }

5
code/libraries/language-processing/java/nu/marginalia/language/WordPatterns.java
View File

@ -1,11 +1,6 @@
package nu.marginalia.language; package nu.marginalia.language;
/** Logic for deciding which words are eligible to be keywords. /** Logic for deciding which words are eligible to be keywords.
* <p/>
* This is in dire need of oversight. Here be towering dragons with names,
* a skull next to their HP bar, and their own Mick Gordon soundtrack just
* for the battle.
*
*/ */
public class WordPatterns { public class WordPatterns {
public static final int MIN_WORD_LENGTH = 1; public static final int MIN_WORD_LENGTH = 1;