package com.hughes.android.dictionary.parser;
import java.io.BufferedReader;
+import java.io.BufferedInputStream;
+import java.io.DataInputStream;
import java.io.File;
import java.io.FileInputStream;
+import java.io.FileWriter;
import java.io.IOException;
import java.io.InputStreamReader;
import java.nio.charset.Charset;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
+import com.hughes.android.dictionary.engine.AbstractEntry;
import com.hughes.android.dictionary.engine.DictionaryBuilder;
import com.hughes.android.dictionary.engine.EntrySource;
import com.hughes.android.dictionary.engine.EntryTypeName;
+import com.hughes.android.dictionary.engine.HtmlEntry;
import com.hughes.android.dictionary.engine.IndexBuilder;
+import com.hughes.android.dictionary.engine.IndexBuilder.TokenData;
import com.hughes.android.dictionary.engine.IndexedEntry;
import com.hughes.android.dictionary.engine.Language;
import com.hughes.android.dictionary.engine.PairEntry;
final String fieldSplit;
final String subfieldSplit;
+ final boolean singleLang;
+
final DictionaryBuilder dictBuilder;
EntrySource entrySource;
public DictFileParser(final Charset charset, boolean flipCols,
final String fieldSplit, final String subfieldSplit,
+ final boolean singleLang,
final DictionaryBuilder dictBuilder) {
this.charset = charset;
this.flipCols = flipCols;
this.fieldSplit = fieldSplit;
this.subfieldSplit = subfieldSplit;
+ this.singleLang = singleLang;
this.dictBuilder = dictBuilder;
}
@Override
public void parse(final File file, final EntrySource entrySouce, final int pageLimit) throws IOException {
this.entrySource = entrySouce;
- final BufferedReader reader = new BufferedReader(new InputStreamReader(new FileInputStream(file), charset));
+ final BufferedReader reader = new BufferedReader(new InputStreamReader(new FileInputStream(file)));
String line;
int count = 0;
while ((line = reader.readLine()) != null) {
subfields[1] = new String[] { fields[1] };
}
- final PairEntry pairEntry = new PairEntry(entrySource);
for (int i = 0; i < subfields[0].length; ++i) {
subfields[0][i] = subfields[0][i].trim();
subfields[1][i] = subfields[1][i].trim();
if (subfields[1][i].isEmpty()) {
subfields[1][i] = "__";
}
- pairEntry.pairs.add(new PairEntry.Pair(subfields[0][i], subfields[1][i]));
}
- final IndexedEntry entryData = new IndexedEntry(pairEntry);
- entryData.isValid = true;
-
- for (int l = 0; l < 2; ++l) {
- // alreadyDone.clear();
-
- final IndexBuilder indexBuilder = dictBuilder.indexBuilders.get(l);
- for (int j = 0; j < subfields[l].length; ++j) {
- String subfield = subfields[l][j];
- if (indexBuilder.index.sortLanguage == Language.de) {
- subfield = parseField_DE(indexBuilder, subfield, entryData, j);
- } else if (indexBuilder.index.sortLanguage == Language.en) {
- subfield = parseField_EN(indexBuilder, subfield, entryData, j);
+
+ if (singleLang) {
+ HtmlEntry htmlEntry = new HtmlEntry(entrySource, fields[0]);
+ htmlEntry.html = StringUtil.escapeUnicodeToPureHtml(fields[1]);
+
+ final IndexBuilder titleIndexBuilder = dictBuilder.indexBuilders.get(0);
+ htmlEntry.addToDictionary(titleIndexBuilder.index.dict);
+
+ TokenData tokenData = titleIndexBuilder.getOrCreateTokenData(fields[0]);
+ tokenData.hasMainEntry = true;
+ tokenData.htmlEntries.add(0, htmlEntry);
+
+ final String[] tokens = NON_CHAR.split(fields[0]);
+ if (tokens.length > 1) {
+ for (final String token : tokens) {
+ assert token.length() >= 1;
+ assert token.indexOf("-") == -1;
+ if (/*!alreadyDone.contains(token) && */!token.isEmpty()) {
+ tokenData = titleIndexBuilder.getOrCreateTokenData(token);
+ tokenData.htmlEntries.add(htmlEntry);
+ }
+ }
+ }
+ final IndexedEntry entryData = new IndexedEntry(htmlEntry);
+ entryData.isValid = true;
+
+ } else {
+ PairEntry pairEntry = new PairEntry(entrySource);
+ for (int i = 0; i < subfields[0].length; ++i) {
+ pairEntry.pairs.add(new PairEntry.Pair(subfields[0][i], subfields[1][i]));
+ }
+ final IndexedEntry entryData = new IndexedEntry(pairEntry);
+ entryData.isValid = true;
+ for (int l = 0; l < 2; ++l) {
+ // alreadyDone.clear();
+
+ final IndexBuilder indexBuilder = dictBuilder.indexBuilders.get(l);
+ for (int j = 0; j < subfields[l].length; ++j) {
+ String subfield = subfields[l][j];
+ if (indexBuilder.index.sortLanguage == Language.de) {
+ subfield = parseField_DE(indexBuilder, subfield, entryData, j);
+ } else if (indexBuilder.index.sortLanguage == Language.en) {
+ subfield = parseField_EN(indexBuilder, subfield, entryData, j);
+ }
+ parseFieldGeneric(indexBuilder, subfield, entryData, j, subfields[l].length);
}
- parseFieldGeneric(indexBuilder, subfield, entryData, j, subfields[l].length);
}
}
+
}
private StringBuilder extractParenthesized(StringBuilder in, String startChar, String endChar) {