commit e12e29ca7f758f688ebb432ccbea944e67326127
author Jean Chalard <jchalard@google.com> Wed Sep 28 16:21:23 2011 +0900
committer Jean Chalard <jchalard@google.com> Wed Sep 28 20:58:59 2011 +0900
tree fb2db24d84c873fe5cbc2e75996929ddb9e33a2c
parent a59f65ad0cadfd2e71ebbe7ca8b14285a4db6ae2

Fix a bug where ampersands in XML would not work

Change-Id: Icc1bce28b78c4da2d529c308f251e298ea94539f

tools/makedict/src/com/android/inputmethod/latin/XmlDictInputOutput.java

diff --git a/tools/makedict/src/com/android/inputmethod/latin/XmlDictInputOutput.java b/tools/makedict/src/com/android/inputmethod/latin/XmlDictInputOutput.java
index 096bfd1..35a7b51 100644
--- a/tools/makedict/src/com/android/inputmethod/latin/XmlDictInputOutput.java
+++ b/tools/makedict/src/com/android/inputmethod/latin/XmlDictInputOutput.java
@@ -60,6 +60,7 @@
         final FusionDictionary mDictionary;
         int mState; // the state of the parser
         int mFreq; // the currently read freq
+        String mWord; // the current word
         final HashMap<String, ArrayList<WeightedString>> mBigramsMap;
 
         /**
@@ -72,6 +73,7 @@
                 HashMap<String, ArrayList<WeightedString>> bigrams) {
             mDictionary = dict;
             mBigramsMap = bigrams;
+            mWord = "";
             mState = START;
             mFreq = 0;
         }
@@ -80,6 +82,7 @@
         public void startElement(String uri, String localName, String qName, Attributes attrs) {
             if (WORD_TAG.equals(localName)) {
                 mState = WORD;
+                mWord = "";
                 for (int attrIndex = 0; attrIndex < attrs.getLength(); ++attrIndex) {
                     final String attrName = attrs.getLocalName(attrIndex);
                     if (FREQUENCY_ATTR.equals(attrName)) {
@@ -94,14 +97,19 @@
         @Override
         public void characters(char[] ch, int start, int length) {
             if (WORD == mState) {
-                final String word = String.copyValueOf(ch, start, length);
-                mDictionary.add(word, mFreq, mBigramsMap.get(word));
+                // The XML parser is free to return text in arbitrary chunks one after the
+                // other. In particular, this happens in some implementations when it finds
+                // an escape code like "&amp;".
+                mWord += String.copyValueOf(ch, start, length);
             }
         }
 
         @Override
         public void endElement(String uri, String localName, String qName) {
-            if (WORD == mState) mState = START;
+            if (WORD == mState) {
+                mDictionary.add(mWord, mFreq, mBigramsMap.get(mWord));
+                mState = START;
+            }
         }
     }