import java.util.List;
import java.util.Map;
import java.util.Set;
+import java.util.logging.Level;
import java.util.logging.Logger;
import java.util.regex.Pattern;
+import com.hughes.android.dictionary.engine.EntrySource;
import com.hughes.android.dictionary.engine.EntryTypeName;
import com.hughes.android.dictionary.engine.IndexBuilder;
import com.hughes.android.dictionary.engine.IndexedEntry;
import com.hughes.android.dictionary.engine.PairEntry;
import com.hughes.android.dictionary.engine.PairEntry.Pair;
import com.hughes.android.dictionary.parser.WikiTokenizer;
-import com.hughes.util.ListUtil;
public class EnWiktionaryXmlParser {
- private static final String TRANSLITERATION_FORMAT = " (tr. %s)";
-
static final Logger LOG = Logger.getLogger(EnWiktionaryXmlParser.class.getName());
// TODO: process {{ttbc}} lines
"Preposition|Proper noun|Article|Prepositional phrase|Acronym|" +
"Abbreviation|Initialism|Contraction|Prefix|Suffix|Symbol|Letter|" +
"Ligature|Idiom|Phrase|\\{\\{acronym\\}\\}|\\{\\{initialism\\}\\}|" +
+ "\\{\\{abbreviation\\}\\}|" +
// These are @deprecated:
"Noun form|Verb form|Adjective form|Nominal phrase|Noun phrase|" +
"Verb phrase|Transitive verb|Intransitive verb|Reflexive verb|" +
"Particle|Interjection|Pronominal adverb" +
"Han character|Hanzi|Hanja|Kanji|Katakana character|Syllable");
+ EntrySource entrySource;
final IndexBuilder enIndexBuilder;
final IndexBuilder foreignIndexBuilder;
final Pattern langPattern;
}
- public void parse(final File file, final int pageLimit) throws IOException {
+ public void parse(final File file, final EntrySource entrySource, final int pageLimit) throws IOException {
+ this.entrySource = entrySource;
int pageCount = 0;
final DataInputStream dis = new DataInputStream(new BufferedInputStream(new FileInputStream(file)));
while (true) {
try {
title = dis.readUTF();
} catch (EOFException e) {
- LOG.warning("Error reading split!");
+ LOG.log(Level.INFO, "EOF reading split.");
dis.close();
return;
}
} else if (headerName.equals("Translations")) {
if (pos == null) {
- LOG.warning("Translations without POS: " + title);
+ LOG.info("Translations without POS (but using anyway): " + title);
}
doTranslations(wikiTokenizer, pos);
} else if (headerName.equals("Pronunciation")) {
}
} else if (wikiTokenizer.isFunction()) {
final String name = wikiTokenizer.functionName();
- if (name.equals("head")) {
+ if (name.equals("head") && pos == null) {
LOG.warning("{{head}} without POS: " + title);
}
}
private void doTranslationLine(final String line, final String lang, final String pos, final String sense, final String rest) {
state = State.TRANSLATION_LINE;
// Good chance we'll actually file this one...
- final PairEntry pairEntry = new PairEntry();
+ final PairEntry pairEntry = new PairEntry(entrySource);
final IndexedEntry indexedEntry = new IndexedEntry(pairEntry);
final StringBuilder foreignText = new StringBuilder();
if (!pairsAdded.add(pair.toString())) {
LOG.warning("Duplicate pair: " + pair.toString());
}
- if (pair.toString().equals("libero {m} :: free (adjective)")) {
- System.out.println();
- }
-
}
final Collection<String> wordForms = new ArrayList<String>();
boolean titleAppended = false;
-
private void doForeignPartOfSpeech(final String lang, String posHeading, final int posDepth, WikiTokenizer wikiTokenizer) {
if (++foreignCount % 1000 == 0) {
LOG.info("***" + lang + ", " + title + ", pos=" + posHeading + ", foreignCount=" + foreignCount);
foreign = String.format("(%s) %s", lang, foreign);
}
for (final ListSection listSection : listSections) {
- doForeignListItem(foreign, title, wordForms, listSection);
+ doForeignListSection(foreign, title, wordForms, listSection);
}
}
}
"sc",
"sort",
"cat",
- "xs"));
+ "xs",
+ "nodot"));
- private void doForeignListItem(final String foreignText, String title, final Collection<String> forms, final ListSection listSection) {
+ public boolean entryIsFormOfSomething = false;
+
+ private void doForeignListSection(final String foreignText, String title, final Collection<String> forms, final ListSection listSection) {
state = State.ENGLISH_DEF_OF_FOREIGN;
final String prefix = listSection.firstPrefix;
if (prefix.length() > 1) {
return;
}
- final PairEntry pairEntry = new PairEntry();
+ final PairEntry pairEntry = new PairEntry(entrySource);
final IndexedEntry indexedEntry = new IndexedEntry(pairEntry);
-
+
+ entryIsFormOfSomething = false;
final StringBuilder englishBuilder = new StringBuilder();
final String mainLine = listSection.firstLine;
-
appendAndIndexWikiCallback.reset(englishBuilder, indexedEntry);
appendAndIndexWikiCallback.dispatch(mainLine, enIndexBuilder, EntryTypeName.WIKTIONARY_ENGLISH_DEF);
if (english.length() > 0) {
final Pair pair = new Pair(english, trim(foreignText), this.swap);
pairEntry.pairs.add(pair);
- foreignIndexBuilder.addEntryWithString(indexedEntry, title, EntryTypeName.WIKTIONARY_TITLE_MULTI);
+ foreignIndexBuilder.addEntryWithString(indexedEntry, title, entryIsFormOfSomething ? EntryTypeName.WIKTIONARY_IS_FORM_OF_SOMETHING_ELSE : EntryTypeName.WIKTIONARY_TITLE_MULTI);
for (final String form : forms) {
foreignIndexBuilder.addEntryWithString(indexedEntry, form, EntryTypeName.WIKTIONARY_INFLECTED_FORM_MULTI);
}
for (int i = 0; i < listSection.nextPrefixes.size(); ++i) {
final String nextPrefix = listSection.nextPrefixes.get(i);
final String nextLine = listSection.nextLines.get(i);
+
+ // TODO: This splitting is not sensitive to wiki code.
int dash = nextLine.indexOf("—");
int mdashLen = 7;
if (dash == -1) {
}
private String formatAndIndexExampleString(final String example, final IndexBuilder indexBuilder, final IndexedEntry indexedEntry) {
- final WikiTokenizer wikiTokenizer = new WikiTokenizer(example, false);
+ // TODO:
+// if (wikiTokenizer.token().equals("'''")) {
+// insideTripleQuotes = !insideTripleQuotes;
+// }
final StringBuilder builder = new StringBuilder();
- boolean insideTripleQuotes = false;
- while (wikiTokenizer.nextToken() != null) {
- if (wikiTokenizer.isPlainText()) {
- builder.append(wikiTokenizer.token());
- if (indexBuilder != null) {
- indexBuilder.addEntryWithStringNoSingle(indexedEntry, wikiTokenizer.token(), EntryTypeName.WIKTIONARY_EXAMPLE);
- }
- } else if (wikiTokenizer.isWikiLink()) {
- final String text = wikiTokenizer.wikiLinkText().replaceAll("'", "");
- builder.append(text);
- if (indexBuilder != null) {
- indexBuilder.addEntryWithStringNoSingle(indexedEntry, text, EntryTypeName.WIKTIONARY_EXAMPLE);
- }
- } else if (wikiTokenizer.isFunction()) {
- builder.append(wikiTokenizer.token());
- } else if (wikiTokenizer.isMarkup()) {
- if (wikiTokenizer.token().equals("'''")) {
- insideTripleQuotes = !insideTripleQuotes;
- }
- } else if (wikiTokenizer.isComment() || wikiTokenizer.isNewline()) {
- // Do nothing.
- } else {
- LOG.warning("unexpected token: " + wikiTokenizer.token());
- }
+ appendAndIndexWikiCallback.reset(builder, indexedEntry);
+ appendAndIndexWikiCallback.entryTypeName = EntryTypeName.WIKTIONARY_EXAMPLE;
+ appendAndIndexWikiCallback.entryTypeNameSticks = true;
+ try {
+ // TODO: this is a hack needed because we don't safely split on the dash.
+ appendAndIndexWikiCallback.dispatch(example, indexBuilder, EntryTypeName.WIKTIONARY_EXAMPLE);
+ } catch (AssertionError e) {
+ return "--";
}
final String result = trim(builder.toString());
return result.length() > 0 ? result : "--";