Partial progress to fix frwiktionary parsing.

[DictionaryPC.git] / src / com / hughes / android / dictionary / parser / wiktionary / EnForeignParser.java

diff --git a/src/com/hughes/android/dictionary/parser/wiktionary/EnForeignParser.java b/src/com/hughes/android/dictionary/parser/wiktionary/EnForeignParser.java
index c4748e6e685d6be0f56a82fa2cde56c743a616ec..7dd933e38815ee1d595fb50777653f617120c705 100644 (file)
--- a/src/com/hughes/android/dictionary/parser/wiktionary/EnForeignParser.java
+++ b/src/com/hughes/android/dictionary/parser/wiktionary/EnForeignParser.java
@@ -54,6 +54,13 @@ public final class EnForeignParser extends EnParser {
             incrementCount("WARNING: Translations not in English section");
           } else if (headingName.equals("Pronunciation")) {
             //doPronunciation(wikiLineReader);
+          } else if (headingName.startsWith(" {{S|")) {
+            // HACK to support parsing frwiktionary
+            String[] parts = headingName.split("\\|");
+            if (parts.length > 2 && langCodePattern.matcher(parts[2]).find() &&
+                (parts.length < 4 || !parts[3].startsWith("flexion"))) {
+                doForeignPartOfSpeech(lang, headingName, wikiTokenizer.headingDepth(), wikiTokenizer);
+            }
           } else if (partOfSpeechHeader.matcher(headingName).matches()) {
             doForeignPartOfSpeech(lang, headingName, wikiTokenizer.headingDepth(), wikiTokenizer);
           }
@@ -212,7 +219,7 @@ public final class EnForeignParser extends EnParser {
       final String prefix = listSection.firstPrefix;
       if (prefix.length() > 1) {
         // Could just get looser and say that any prefix longer than first is a sublist.
-        LOG.warning("Prefix too long: " + listSection);
+        LOG.warning("Prefix '" + prefix + "' too long: " + listSection);
         incrementCount("WARNING: Prefix too long");
         return;
       }
@@ -241,7 +248,7 @@ public final class EnForeignParser extends EnParser {
       String lastForeign = null;
       for (int i = 0; i < listSection.nextPrefixes.size(); ++i) {
         final String nextPrefix = listSection.nextPrefixes.get(i);
-        final String nextLine = listSection.nextLines.get(i);
+        String nextLine = listSection.nextLines.get(i);
 
         // TODO: This splitting is not sensitive to wiki code.
         int dash = nextLine.indexOf("&mdash;");
@@ -263,14 +270,23 @@ public final class EnForeignParser extends EnParser {
             pairEntry.pairs.add(pair);
           }
           lastForeign = null;
-        } else if (nextPrefix.equals("#:") || nextPrefix.equals("##:")){
+        // TODO: make #* and #*: work
+        } else if (nextPrefix.equals("#:") || nextPrefix.equals("##:")/* || nextPrefix.equals("#*")*/){
           final Pair pair = new Pair("--", formatAndIndexExampleString(nextLine, null, indexedEntry), swap);
           lastForeign = nextLine;
           if (pair.lang1 != "--" && pair.lang1 != "--") {
             pairEntry.pairs.add(pair);
           }
-        } else if (nextPrefix.equals("#::") || nextPrefix.equals("#**")) {
+        } else if (nextPrefix.equals("#::") || nextPrefix.equals("#**")/* || nextPrefix.equals("#*:")*/) {
           if (lastForeign != null && pairEntry.pairs.size() > 0) {
+            if (i + 1 < listSection.nextPrefixes.size()) {
+              // Chinese has sometimes multiple foreign lines
+              final String nextNextPrefix = listSection.nextPrefixes.get(i + 1);
+              if (nextNextPrefix.equals("#::") || nextNextPrefix.equals("#**")) {
+                ++i;
+                nextLine += "\n" + listSection.nextLines.get(i);
+              }
+            }
             pairEntry.pairs.remove(pairEntry.pairs.size() - 1);
             final Pair pair = new Pair(formatAndIndexExampleString(nextLine, enIndexBuilder, indexedEntry), formatAndIndexExampleString(lastForeign, foreignIndexBuilder, indexedEntry), swap);
             if (pair.lang1 != "--" || pair.lang2 != "--") {