Make match and context size configurable (address #128)

Change-Id: Ieef96dd68adf4e3ce00f59fc21face545c2ce897

Changes

@@ -1,3 +1,7 @@
+0.62.4 2024-05-17
+    - [feature] Make match and context size configurable (address #128, 
+      diewald & margaretha)
+
 0.62.3 2024-04-16
     - [cleanup] Added getDocBitsSupplier to VirtualCorpusFilter (margaretha)
     - [feature] Make VC cache location customizable (margaretha)

pom.xml

@@ -35,7 +35,7 @@
 
   <groupId>de.ids-mannheim.korap.krill</groupId>
   <artifactId>Krill</artifactId>
-  <version>0.62.3</version>
+  <version>0.62.4</version>
   <packaging>jar</packaging>
 
   <name>Krill</name>

src/main/java/de/ids_mannheim/korap/Krill.java

@@ -60,6 +60,7 @@ public class Krill extends Response {
     private SpanQuery spanQuery;
     private JsonNode request;
 
+    private int maxTokenMatchSize;
     private final ObjectMapper mapper = new ObjectMapper();
 
     /**
@@ -347,4 +348,14 @@ public JsonNode retrieveFieldValues (String corpusQuery, KrillIndex index,
         VirtualCorpusResponse r = new VirtualCorpusResponse();
         return r.createKoralQueryForField(fieldName, fieldValues);
     }
+
+
+    public int getMaxTokenMatchSize () {
+        return maxTokenMatchSize;
+    }
+
+
+    public void setMaxTokenMatchSize (int maxMatchTokens) {
+        this.maxTokenMatchSize = maxMatchTokens;
+    }
 };

src/main/java/de/ids_mannheim/korap/KrillIndex.java

@@ -44,8 +44,8 @@
 import org.apache.lucene.search.spans.SpanQuery;
 import org.apache.lucene.search.spans.Spans;
 import org.apache.lucene.store.Directory;
-import org.apache.lucene.store.RAMDirectory;
 import org.apache.lucene.store.MMapDirectory;
+import org.apache.lucene.store.RAMDirectory;
 import org.apache.lucene.util.Bits;
 import org.apache.lucene.util.BytesRef;
 import org.apache.lucene.util.FixedBitSet;
@@ -175,22 +175,27 @@ public final class KrillIndex implements IndexInfo {
     private HashMap termContexts;
     private ObjectMapper mapper = new ObjectMapper();
 
+    private int maxTokenMatchSize;
+
     // private ByteBuffer bbTerm;
 
     // Some initializations ...
     {
         Properties prop = KrillProperties.loadDefaultProperties();
         Properties info = KrillProperties.loadInfo();
+
         if (info != null) {
             this.version = info.getProperty("krill.version");
             this.name = info.getProperty("krill.name");
         };
 
         // Check for auto commit value
         String autoCommitStr = null;
-        if (prop != null)
+        if (prop != null) {
             autoCommitStr = prop.getProperty("krill.index.commit.auto");
-
+            this.maxTokenMatchSize = KrillProperties.maxTokenMatchSize;
+        }
+
         if (autoCommitStr != null) {
             try {
                 this.autoCommit = Integer.parseInt(autoCommitStr);
@@ -429,8 +434,15 @@ public int getAutoCommit () {
     public void setAutoCommit (int value) {
         this.autoCommit = value;
     };
-
-
+
+    public int getMaxTokenMatchSize () {
+        return maxTokenMatchSize;
+    }
+
+    public void setMaxTokenMatchSize (int maxMatchTokens) {
+        this.maxTokenMatchSize = maxMatchTokens;
+    }
+
     /**
      * Update a document in the index as a {@link FieldDocument}
      * if it already exists (based on the textSigle), otherwise
@@ -972,11 +984,20 @@ public Match getMatchInfo (String idString, String field, boolean info,
                                boolean includeSnippets, boolean includeTokens,
                                boolean includeHighlights, boolean extendToSentence)
             throws QueryException {
-
+        return getMatchInfo(idString, field, info, foundry, layer, includeSpans,
+                includeSnippets, includeTokens, includeHighlights,
+                extendToSentence, maxTokenMatchSize);
+    };
+
+    public Match getMatchInfo (String idString, String field, boolean info,
+            List<String> foundry, List<String> layer, boolean includeSpans,
+            boolean includeSnippets, boolean includeTokens,
+            boolean includeHighlights, boolean extendToSentence,
+            int maxMatchTokens) throws QueryException {
         if (DEBUG)
             log.trace("Get info on {}", idString);
-
-        Match match = new Match(idString, includeHighlights);
+        
+        Match match = new Match(maxMatchTokens, idString, includeHighlights);
 
         if (this.getVersion() != null)
             match.setVersion(this.getVersion());
@@ -1202,8 +1223,8 @@ else if (includeSpans) {
                             && spanContext[0] < spanContext[1]) {
 
                         // Match needs to be cutted!
-                        if ((spanContext[1] - spanContext[0]) > match.getMaxMatchTokens()) {
-                            int contextLength = match.getMaxMatchTokens() - match.getLength();
+                        if ((spanContext[1] - spanContext[0]) > maxMatchTokens) {
+                            int contextLength = maxMatchTokens - match.getLength();
                             int halfContext = contextLength / 2;
 
                             // This is the extended context calculated
@@ -1216,8 +1237,8 @@ else if (includeSpans) {
                             }
                         }
 
-                        match.setStartPos(spanContext[0]);
-                        match.setEndPos(spanContext[1]);
+                        match.setStartPos(maxMatchTokens,spanContext[0]);
+                        match.setEndPos(maxMatchTokens,spanContext[1]);
 						match.potentialStartPosChar = spanContext[2];
 						match.potentialEndPosChar = spanContext[3];
                         match.startMore = false;
@@ -1569,9 +1590,14 @@ public Result search (Krill ks) {
                     final Document doc = (fields != null)
                             ? lreader.document(localDocID, fieldsSet)
                             : lreader.document(localDocID);
-
+
+                    int maxMatchSize = maxTokenMatchSize;
+                    if (ks.getMaxTokenMatchSize() > 0) {
+                        maxMatchSize = ks.getMaxTokenMatchSize();
+                    };
+
                     // Create new Match
-                    final Match match = new Match(pto, localDocID,
+                    final Match match = new Match(maxMatchSize, pto, localDocID,
                             spans.start(), spans.end());
 
                     // Add snippet if existing

src/main/java/de/ids_mannheim/korap/index/Indexer.java

@@ -180,7 +180,7 @@ public static void main (String[] argv) {
         Options options = new Options();
         options.addOption(Option.builder("c").longOpt("config")
                 .desc("configuration file (defaults to "
-                        + KrillProperties.defaultPropertiesLocation
+                        + KrillProperties.DEFAULT_PROPERTIES_LOCATION
                         + ").")
                 .hasArg().argName("properties file").required().build());
         options.addOption(Option.builder("i").longOpt("inputDir")

src/main/java/de/ids_mannheim/korap/response/Match.java

@@ -1,14 +1,23 @@
 package de.ids_mannheim.korap.response;
 
+import static de.ids_mannheim.korap.util.KrillByte.unsignedByte;
+import static de.ids_mannheim.korap.util.KrillString.codePointSubstring;
+
 import java.io.IOException;
 import java.nio.ByteBuffer;
-import java.util.*;
+import java.util.ArrayList;
+import java.util.Collections;
+import java.util.Comparator;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.LinkedList;
+import java.util.List;
 
 import org.apache.lucene.index.LeafReaderContext;
 import org.apache.lucene.index.Term;
 import org.apache.lucene.index.TermContext;
-import org.apache.lucene.search.spans.Spans;
 import org.apache.lucene.search.spans.SpanTermQuery;
+import org.apache.lucene.search.spans.Spans;
 import org.apache.lucene.util.Bits;
 import org.apache.lucene.util.FixedBitSet;
 import org.slf4j.Logger;
@@ -18,13 +27,11 @@
 import com.fasterxml.jackson.annotation.JsonInclude;
 import com.fasterxml.jackson.annotation.JsonInclude.Include;
 import com.fasterxml.jackson.annotation.JsonProperty;
-import com.fasterxml.jackson.databind.ObjectMapper;
 import com.fasterxml.jackson.databind.JsonNode;
+import com.fasterxml.jackson.databind.ObjectMapper;
 import com.fasterxml.jackson.databind.node.ArrayNode;
 import com.fasterxml.jackson.databind.node.ObjectNode;
 
-import static de.ids_mannheim.korap.util.KrillByte.*;
-import static de.ids_mannheim.korap.util.KrillString.codePointSubstring;
 import de.ids_mannheim.korap.index.AbstractDocument;
 import de.ids_mannheim.korap.index.PositionsToOffset;
 import de.ids_mannheim.korap.query.SpanElementQuery;
@@ -33,6 +40,7 @@
 import de.ids_mannheim.korap.response.match.MatchIdentifier;
 import de.ids_mannheim.korap.response.match.PosIdentifier;
 import de.ids_mannheim.korap.response.match.Relation;
+import de.ids_mannheim.korap.util.KrillProperties;
 
 /*
  * The snippet building algorithm is quite complicated for now
@@ -84,8 +92,6 @@ public class Match extends AbstractDocument {
 
     // Logger
     private final static Logger log = LoggerFactory.getLogger(Match.class);
-
-	private static final int MAX_MATCH_TOKENS = 50;
 
 	// end marker of highlights that are pagebreaks
 	private static final int PB_MARKER = -99999;
@@ -151,18 +157,19 @@ public class Match extends AbstractDocument {
     @JsonIgnore
     public boolean startMore = true, endMore = true;
 
-    private Collection<byte[]> payload;
+//    private Collection<byte[]> payload;
     private ArrayList<Highlight> highlight;
     private LinkedList<int[]> span;
 
     private PositionsToOffset positionsToOffset;
     private boolean processed = false;
-
-
+
     /**
      * Constructs a new Match object.
      * Todo: Maybe that's not necessary!
      * 
+     * @param maxTokenMatchSize
+     *            The maximum number of tokens a match may have
      * @param pto
      *            The PositionsToOffset object, containing relevant
      *            positional information for highlighting
@@ -177,12 +184,12 @@ public class Match extends AbstractDocument {
      * @see #snippetBrackets()
      * @see PositionsToOffset
      */
-    public Match (PositionsToOffset pto, int localDocID, int startPos,
-                  int endPos) {
+    public Match (int maxTokenMatchSize, PositionsToOffset pto,
+                  int localDocID, int startPos, int endPos) {
         this.positionsToOffset = pto;
         this.localDocID = localDocID;
-        this.setStartPos(startPos);
-        this.setEndPos(endPos);
+        this.setStartPos(maxTokenMatchSize, startPos);
+        this.setEndPos(maxTokenMatchSize, endPos);
     };
 
 
@@ -201,7 +208,7 @@ public Match (PositionsToOffset pto, int localDocID, int startPos,
      *            Boolean value indicating if possible provided
      *            highlight information should be ignored or not.
      */
-    public Match (String idString, boolean includeHighlights) {
+    public Match (int maxTokenMatchSize, String idString, boolean includeHighlights) {
         MatchIdentifier id = new MatchIdentifier(idString);
 
         if (id.getStartPos() > -1) {
@@ -215,8 +222,8 @@ public Match (String idString, boolean includeHighlights) {
             this.addString("ID", id.getDocID());
             // </legacy>
 
-            this.setStartPos(id.getStartPos());
-            this.setEndPos(id.getEndPos());
+            this.setStartPos(maxTokenMatchSize, id.getStartPos());
+            this.setEndPos(maxTokenMatchSize, id.getEndPos());
 
             if (includeHighlights)
                 for (int[] pos : id.getPos()) {
@@ -506,11 +513,6 @@ public void addPagebreak (int start, int pagenumber) {
 		this.addHighlight(new Highlight(start, pagenumber));
 	};
 
-    @JsonIgnore
-    public int getMaxMatchTokens () {
-        return MAX_MATCH_TOKENS;
-    }
-
     /**
      * Get document id.
      */
@@ -575,10 +577,10 @@ public int getStartPos (int number) {
      *            The positional offset.
      */
     @JsonIgnore
-    public void setStartPos (int pos) {
+    public void setStartPos (int maxTokenMatchSize, int pos) {
         this.startPos = pos;
-		if (this.endPos != -1 && (this.endPos - pos) > MAX_MATCH_TOKENS) {
-			this.endPos = pos + MAX_MATCH_TOKENS;
+		if (this.endPos != -1 && (this.endPos - pos) > maxTokenMatchSize) {
+			this.endPos = pos + maxTokenMatchSize;
 			this.endCutted = true;
 		};
     };
@@ -623,10 +625,15 @@ public int getEndPos (int number) {
      *            The positional offset.
      */
     @JsonIgnore
-    public void setEndPos (int pos) {
-		if (this.startPos != -1 && (pos - this.startPos) > MAX_MATCH_TOKENS) {
-			pos = this.startPos + MAX_MATCH_TOKENS;
-			this.endCutted = true;
+    public void setEndPos (int maxTokenMatchSize, int pos) {
+        if (maxTokenMatchSize > KrillProperties.maxTokenMatchSize) {
+            maxTokenMatchSize = KrillProperties.maxTokenMatchSize;
+            this.endCutted = true;
+        }
+
+        if (this.startPos != -1 && (pos - this.startPos) > maxTokenMatchSize) {
+			pos = this.startPos + maxTokenMatchSize;
+			    this.endCutted = true;
 		};
         this.endPos = pos;
     };