Ensure toLowerCase is independent of locale.

[Dictionary.git] / src / com / hughes / android / dictionary / engine / Index.java

diff --git a/src/com/hughes/android/dictionary/engine/Index.java b/src/com/hughes/android/dictionary/engine/Index.java
index 5b7b43d5aadd20a39a3de5e047b3ca348cdd7b83..b9300ba0462c9846398229ad96c561f45343fe36 100644 (file)
--- a/src/com/hughes/android/dictionary/engine/Index.java
+++ b/src/com/hughes/android/dictionary/engine/Index.java
@@ -14,8 +14,11 @@
 
 package com.hughes.android.dictionary.engine;
 
+import java.io.BufferedOutputStream;
 import java.io.DataInput;
 import java.io.DataOutput;
+import java.io.DataOutputStream;
+import java.io.FileOutputStream;
 import java.io.IOException;
 import java.io.PrintStream;
 import java.io.RandomAccessFile;
@@ -28,6 +31,7 @@ import java.util.EnumMap;
 import java.util.HashMap;
 import java.util.HashSet;
 import java.util.List;
+import java.util.Locale;
 import java.util.Map;
 import java.util.Set;
 import java.util.concurrent.atomic.AtomicBoolean;
@@ -136,10 +140,11 @@ public final class Index {
                 stoplist.add(raf.readUTF());
             }
         } else if (dict.dictFileVersion >= 4) {
-            raf.readInt(); // length
-            raf.skipBytes(52);
             stoplist = new HashSet<>();
-            byte b;
+            raf.readInt(); // length
+            raf.skipBytes(18);
+            byte b = raf.readByte();
+            raf.skipBytes(b == 'L' ? 71 : 33);
             while ((b = raf.readByte()) == 0x74) {
                 stoplist.add(raf.readUTF());
             }
@@ -165,7 +170,9 @@ public final class Index {
         for (String i : stoplist) {
             raf.writeUTF(i);
         }
-        UniformRAFList.write(raf, rows, new RowBase.Serializer(this), 3 /* bytes per entry */);
+        DataOutputStream outb = new DataOutputStream(new BufferedOutputStream(new FileOutputStream(raf.getFD())));
+        UniformRAFList.write(outb, rows, new RowBase.Serializer(this), 3 /* bytes per entry */);
+        outb.flush();
     }
 
     public void print(final PrintStream out) {
@@ -194,14 +201,14 @@ public final class Index {
         public List<HtmlEntry> htmlEntries;
 
         public IndexEntry(final Index index, final String token, final String normalizedToken,
-                          final int startRow, final int numRows) {
+                          final int startRow, final int numRows, final List<HtmlEntry> htmlEntries) {
             assert token.equals(token.trim());
             assert token.length() > 0;
             this.token = token;
             this.normalizedToken = normalizedToken;
             this.startRow = startRow;
             this.numRows = numRows;
-            this.htmlEntries = new ArrayList<>();
+            this.htmlEntries = htmlEntries;
         }
 
         IndexEntry(final Index index, final DataInput raf) throws IOException {
@@ -311,6 +318,7 @@ public final class Index {
     }
 
     private int findInsertionPointIndex(String token, final AtomicBoolean interrupted) {
+        String orig_token = token;
         token = normalizeToken(token);
 
         int start = 0;
@@ -328,7 +336,8 @@ public final class Index {
             if (comp == 0)
                 comp = sortCollator.compare(token, midEntry.normalizedToken());
             if (comp == 0) {
-                return windBackCase(token, mid, interrupted);
+                start = end = mid;
+                break;
             } else if (comp < 0) {
                 // System.out.println("Upper bound: " + midEntry + ", norm=" +
                 // midEntry.normalizedToken() + ", mid=" + mid);
@@ -365,6 +374,23 @@ public final class Index {
                 start--;
         }
 
+        // If the search term was normalized, try to find an exact match first
+        if (!orig_token.equalsIgnoreCase(token)) {
+            int matchLen = findMatchLen(sortCollator, token, sortedIndexEntries.get(start).normalizedToken());
+            int scan = start;
+            while (scan >= 0 && scan < sortedIndexEntries.size()) {
+                IndexEntry e = sortedIndexEntries.get(scan);
+                if (e.token.equalsIgnoreCase(orig_token))
+                {
+                    return scan;
+                }
+                if (matchLen > findMatchLen(sortCollator, token, e.normalizedToken()))
+                    break;
+                if (interrupted.get()) return start;
+                scan++;
+            }
+        }
+
         // If we search for a substring of a string that's in there, return
         // that.
         int result = Math.min(start, sortedIndexEntries.size() - 1);
@@ -530,7 +556,7 @@ public final class Index {
             swapPairEntries);
         for (final Collection<RowBase> rows : matches.values()) {
             final List<RowBase> ordered = new ArrayList<>(rows);
-            ordered.sort(lengthComparator);
+            Collections.sort(ordered, lengthComparator);
             result.addAll(ordered);
         }
 
@@ -544,7 +570,7 @@ public final class Index {
             return normalizer.transliterate(searchToken);
         } else {
             // Do our best since the Transliterators aren't up yet.
-            return searchToken.toLowerCase();
+            return searchToken.toLowerCase(Locale.US);
         }
     }