import java.util.Arrays;
import java.util.Collections;
import java.util.LinkedHashSet;
-import java.util.List;
import java.util.Set;
import java.util.logging.Logger;
import java.util.regex.Matcher;
import com.hughes.android.dictionary.engine.DictionaryBuilder;
import com.hughes.android.dictionary.engine.EntrySource;
-import com.hughes.android.dictionary.engine.IndexedEntry;
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.Language;
import com.hughes.android.dictionary.engine.PairEntry;
-import com.hughes.android.dictionary.engine.PairEntry.Pair;
+import com.hughes.util.StringUtil;
public class DictFileParser implements Parser {
-
- static final Logger logger = Logger.getLogger(DictFileParser.class.getName());
-
- // Dictcc
- public static final Pattern TAB = Pattern.compile("\\t");
-
- // Chemnitz
- public static final Pattern DOUBLE_COLON = Pattern.compile(" :: ");
- public static final Pattern PIPE = Pattern.compile("\\|");
-
- static final Pattern SPACES = Pattern.compile("\\s+");
-// static final Pattern DE_NOUN = Pattern.compile("([^ ]+) *\\{(m|f|n|pl)\\}");
-// static final Pattern EN_VERB = Pattern.compile("^to ([^ ]+)");
-
- static final Pattern BRACKETED = Pattern.compile("\\[([^]]+)\\]");
- static final Pattern PARENTHESIZED = Pattern.compile("\\(([^)]+)\\)");
- static final Pattern CURLY_BRACED = Pattern.compile("\\{([^}]+)\\}");
-
- static final Pattern NON_CHAR_DASH = Pattern.compile("[^-'\\p{L}0-9]+");
- public static final Pattern NON_CHAR = Pattern.compile("[^\\p{L}0-9]+");
-
- static final Pattern TRIM_PUNC = Pattern.compile("^[^\\p{L}0-9]+|[^\\p{L}0-9]+$");
-
- final Charset charset;
- final boolean flipCols;
-
- final Pattern fieldSplit;
- final Pattern subfieldSplit;
-
- final DictionaryBuilder dictBuilder;
- final IndexBuilder[] langIndexBuilders;
- final IndexBuilder bothIndexBuilder;
-
- EntrySource entrySource;
-
- // final Set<String> alreadyDone = new HashSet<String>();
-
- public DictFileParser(final Charset charset, boolean flipCols,
- final Pattern fieldSplit, final Pattern subfieldSplit,
- final DictionaryBuilder dictBuilder, final IndexBuilder[] langIndexBuilders,
- final IndexBuilder bothIndexBuilder) {
- this.charset = charset;
- this.flipCols = flipCols;
- this.fieldSplit = fieldSplit;
- this.subfieldSplit = subfieldSplit;
- this.dictBuilder = dictBuilder;
- this.langIndexBuilders = langIndexBuilders;
- this.bothIndexBuilder = bothIndexBuilder;
- }
-
- @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));
- String line;
- int count = 0;
- while ((line = reader.readLine()) != null) {
- if (pageLimit >= 0 && count >= pageLimit) {
- return;
- }
- if (count % 10000 == 0) {
- logger.info("count=" + count + ", line=" + line);
- }
- parseLine(line);
- ++count;
- }
- }
-
- private void parseLine(final String line) {
- if (line.startsWith("#") || line.length() == 0) {
- logger.info("Skipping comment line: " + line);
- return;
- }
- final String[] fields = fieldSplit.split(line);
- if (fields.length != 2) {
- logger.warning("Malformed line: " + line);
- return;
- }
-
- fields[0] = SPACES.matcher(fields[0]).replaceAll(" ").trim();
- fields[1] = SPACES.matcher(fields[1]).replaceAll(" ").trim();
- if (flipCols) {
- final String temp = fields[0];
- fields[0] = fields[1];
- fields[1] = temp;
- }
- final String[][] subfields = new String[2][];
- if (subfieldSplit != null) {
- subfields[0] = subfieldSplit.split(fields[0]);
- subfields[1] = subfieldSplit.split(fields[1]);
- if (subfields[0].length != subfields[1].length) {
- logger.warning("Number of subfields doesn't match: " + line);
- return;
- }
- } else {
- subfields[0] = new String[] { fields[0] };
- 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[0][i].length() == 0 && subfields[1][i].length() == 0) {
- logger.warning("Empty pair: " + line);
- continue;
- }
- if (subfields[0][i].length() == 0) {
- subfields[0][i] = "__";
- }
- if (subfields[1][i].length() == 0) {
- subfields[1][i] = "__";
- }
- pairEntry.pairs.add(new Pair(subfields[0][i], subfields[1][i]));
+ static final Logger logger = Logger.getLogger(DictFileParser.class.getName());
+
+ // Dictcc
+ public static final String TAB = "\t";
+
+ // Chemnitz
+ public static final String DOUBLE_COLON = " :: ";
+ public static final String PIPE = "|";
+
+ static final Pattern SPACES = Pattern.compile("\\s+");
+
+ static final Pattern CURLY_BRACED = Pattern.compile("\\{([^}]+)\\}");
+
+ // http://www.regular-expressions.info/unicode.html
+ static final Pattern NON_CHAR_DASH = Pattern.compile("[^-'\\p{L}\\p{M}\\p{N}]+");
+ public static final Pattern NON_CHAR = Pattern.compile("[^\\p{L}\\p{M}\\p{N}]+");
+
+ static final Pattern TRIM_PUNC = Pattern.compile("^[^\\p{L}\\p{M}\\p{N}]+|[^\\p{L}\\p{M}\\p{N}]+$");
+
+ final Charset charset;
+ final boolean flipCols;
+
+ final String fieldSplit;
+ final String subfieldSplit;
+
+ final DictionaryBuilder dictBuilder;
+
+ EntrySource entrySource;
+
+ // final Set<String> alreadyDone = new HashSet<String>();
+
+ public DictFileParser(final Charset charset, boolean flipCols,
+ final String fieldSplit, final String subfieldSplit,
+ final DictionaryBuilder dictBuilder) {
+ this.charset = charset;
+ this.flipCols = flipCols;
+ this.fieldSplit = fieldSplit;
+ this.subfieldSplit = subfieldSplit;
+ this.dictBuilder = dictBuilder;
}
- final IndexedEntry entryData = new IndexedEntry(pairEntry);
-
- for (int l = 0; l < 2; ++l) {
- // alreadyDone.clear();
-
- for (int j = 0; j < subfields[l].length; ++j) {
- String subfield = subfields[l][j];
- final IndexBuilder indexBuilder = langIndexBuilders[l];
- 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);
+
+ @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));
+ String line;
+ int count = 0;
+ while ((line = reader.readLine()) != null) {
+ if (pageLimit >= 0 && count >= pageLimit) {
+ return;
+ }
+ if (count % 10000 == 0) {
+ logger.info("count=" + count + ", line=" + line);
+ }
+ parseLine(line);
+ ++count;
}
- parseFieldGeneric(indexBuilder, subfield, entryData, j, subfields[l].length);
- }
}
- }
-
- private void parseFieldGeneric(final IndexBuilder indexBuilder, String field,
- final IndexedEntry entryData, final int subfieldIdx, final int numSubFields) {
- // remove bracketed and parenthesized stuff.
- final StringBuilder bracketed = new StringBuilder();
- final StringBuilder parenthesized = new StringBuilder();
-
- Matcher matcher;
- while ((matcher = BRACKETED.matcher(field)).find()) {
- bracketed.append(matcher.group(1)).append(" ");
- field = matcher.replaceFirst(" ");
+
+ private void parseLine(final String line) {
+ if (line.startsWith("#") || line.isEmpty()) {
+ logger.info("Skipping comment line: " + line);
+ return;
+ }
+ final String[] fields = StringUtil.split(line, fieldSplit);
+ if (fields.length < 2 || fields.length > 4) {
+ logger.warning("Malformed line, expected 3 or 4 fields, got " + fields.length + ": " + line);
+ return;
+ }
+
+ fields[0] = SPACES.matcher(fields[0]).replaceAll(" ").trim();
+ fields[1] = SPACES.matcher(fields[1]).replaceAll(" ").trim();
+ if (flipCols) {
+ final String temp = fields[0];
+ fields[0] = fields[1];
+ fields[1] = temp;
+ }
+
+ final String[][] subfields = new String[2][];
+ if (subfieldSplit != null) {
+ subfields[0] = StringUtil.split(fields[0], subfieldSplit);
+ subfields[1] = StringUtil.split(fields[1], subfieldSplit);
+ if (subfields[0].length != subfields[1].length) {
+ logger.warning("Number of subfields doesn't match: " + line);
+ return;
+ }
+ } else {
+ subfields[0] = new String[] { fields[0] };
+ 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[0][i].isEmpty() && subfields[1][i].isEmpty()) {
+ logger.warning("Empty pair: " + line);
+ continue;
+ }
+ if (subfields[0][i].isEmpty()) {
+ subfields[0][i] = "__";
+ }
+ 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);
+ }
+ parseFieldGeneric(indexBuilder, subfield, entryData, j, subfields[l].length);
+ }
+ }
}
- while ((matcher = PARENTHESIZED.matcher(field)).find()) {
- parenthesized.append(matcher.group(1)).append(" ");
- field = matcher.replaceFirst(" ");
+ private StringBuilder extractParenthesized(StringBuilder in, String startChar, String endChar) {
+ StringBuilder res = new StringBuilder();
+ int pos = 0;
+ while ((pos = in.indexOf(startChar, pos)) != -1) {
+ int end = in.indexOf(endChar, pos + 1);
+ if (end == -1) break;
+ res.append(in, pos + 1, end).append(" ");
+ in.replace(pos, end + 1, " ");
+ pos++; // skip the just appended space
+ }
+ return res;
}
-
- field = SPACES.matcher(field).replaceAll(" ").trim();
-
- // split words on non -A-z0-9, do them.
- final String[] tokens = NON_CHAR_DASH.split(field);
-
- final EntryTypeName entryTypeName;
- if (numSubFields == 1) {
- assert subfieldIdx == 0;
- if (tokens.length == 1) {
- entryTypeName = EntryTypeName.ONE_WORD;
- } else if (tokens.length == 2) {
- entryTypeName = EntryTypeName.TWO_WORDS;
- } else if (tokens.length == 3) {
- entryTypeName = EntryTypeName.THREE_WORDS;
- } else if (tokens.length == 4) {
- entryTypeName = EntryTypeName.FOUR_WORDS;
- } else {
- entryTypeName = EntryTypeName.FIVE_OR_MORE_WORDS;
- }
- } else {
- assert numSubFields > 1;
- if (subfieldIdx == 0) {
- if (tokens.length == 1) {
- entryTypeName = EntryTypeName.MULTIROW_HEAD_ONE_WORD;
+
+ private void parseFieldGeneric(final IndexBuilder indexBuilder, String field,
+ final IndexedEntry entryData, final int subfieldIdx, final int numSubFields) {
+ final StringBuilder fieldsb = new StringBuilder(field);
+ // remove bracketed and parenthesized stuff.
+ final StringBuilder bracketed = extractParenthesized(fieldsb, "[", "]");
+ final StringBuilder parenthesized = extractParenthesized(fieldsb, "(", ")");
+
+ field = fieldsb.toString().trim();
+
+ // split words on non -A-z0-9, do them.
+ final String[] tokens = NON_CHAR_DASH.split(field);
+
+ final EntryTypeName entryTypeName;
+ if (numSubFields == 1) {
+ assert subfieldIdx == 0;
+ if (tokens.length == 1) {
+ entryTypeName = EntryTypeName.ONE_WORD;
+ } else if (tokens.length == 2) {
+ entryTypeName = EntryTypeName.TWO_WORDS;
+ } else if (tokens.length == 3) {
+ entryTypeName = EntryTypeName.THREE_WORDS;
+ } else if (tokens.length == 4) {
+ entryTypeName = EntryTypeName.FOUR_WORDS;
+ } else {
+ entryTypeName = EntryTypeName.FIVE_OR_MORE_WORDS;
+ }
} else {
- entryTypeName = EntryTypeName.MULTIROW_HEAD_MANY_WORDS;
+ assert numSubFields > 1;
+ if (subfieldIdx == 0) {
+ if (tokens.length == 1) {
+ entryTypeName = EntryTypeName.MULTIROW_HEAD_ONE_WORD;
+ } else {
+ entryTypeName = EntryTypeName.MULTIROW_HEAD_MANY_WORDS;
+ }
+ } else {
+ assert subfieldIdx > 0;
+ if (tokens.length == 1) {
+ entryTypeName = EntryTypeName.MULTIROW_TAIL_ONE_WORD;
+ } else {
+ entryTypeName = EntryTypeName.MULTIROW_TAIL_MANY_WORDS;
+ }
+ }
}
- } else {
- assert subfieldIdx > 0;
- if (tokens.length == 1) {
- entryTypeName = EntryTypeName.MULTIROW_TAIL_ONE_WORD;
- } else {
- entryTypeName = EntryTypeName.MULTIROW_TAIL_MANY_WORDS;
+
+ for (String token : tokens) {
+ token = TRIM_PUNC.matcher(token).replaceAll("");
+ if (/*!alreadyDone.contains(token) && */!token.isEmpty()) {
+ indexBuilder.addEntryWithTokens(entryData, Collections.singleton(token), entryTypeName);
+ // alreadyDone.add(token);
+
+ // also split words on dashes, do them, too.
+ if (token.indexOf('-') != -1) {
+ final String[] dashed = StringUtil.split(token, "-");
+ for (final String dashedToken : dashed) {
+ if (/*!alreadyDone.contains(dashedToken) && */!dashedToken.isEmpty()) {
+ indexBuilder.addEntryWithTokens(entryData, Collections.singleton(dashedToken), EntryTypeName.PART_OF_HYPHENATED);
+ }
+ }
+ }
+
+ } // if (!alreadyDone.contains(token)) {
+ } // for (final String token : tokens) {
+
+ // process bracketed stuff (split on spaces and dashes always)
+ if (bracketed.length() > 0) {
+ final String[] bracketedTokens = NON_CHAR.split(bracketed.toString());
+ for (final String token : bracketedTokens) {
+ assert token.indexOf("-") == -1;
+ if (/*!alreadyDone.contains(token) && */!token.isEmpty()) {
+ indexBuilder.addEntryWithTokens(entryData, Collections.singleton(token), EntryTypeName.BRACKETED);
+ }
+ }
}
- }
- }
- for (String token : tokens) {
- token = TRIM_PUNC.matcher(token).replaceAll("");
- if (/*!alreadyDone.contains(token) && */token.length() > 0) {
- indexBuilder.addEntryWithTokens(entryData, Collections.singleton(token), entryTypeName);
- // alreadyDone.add(token);
-
- // also split words on dashes, do them, too.
- if (token.contains("-")) {
- final String[] dashed = token.split("-");
- for (final String dashedToken : dashed) {
- if (/*!alreadyDone.contains(dashedToken) && */dashedToken.length() > 0) {
- indexBuilder.addEntryWithTokens(entryData, Collections.singleton(dashedToken), EntryTypeName.PART_OF_HYPHENATED);
+ // process paren stuff
+ if (parenthesized.length() > 0) {
+ final String[] parenTokens = NON_CHAR.split(parenthesized.toString());
+ for (final String token : parenTokens) {
+ assert token.indexOf("-") == -1;
+ if (/*!alreadyDone.contains(token) && */!token.isEmpty()) {
+ indexBuilder.addEntryWithTokens(entryData, Collections.singleton(token), EntryTypeName.PARENTHESIZED);
+ }
}
- }
}
- } // if (!alreadyDone.contains(token)) {
- } // for (final String token : tokens) {
-
- // process bracketed stuff (split on spaces and dashes always)
- final String[] bracketedTokens = NON_CHAR.split(bracketed.toString());
- for (final String token : bracketedTokens) {
- assert !token.contains("-");
- if (/*!alreadyDone.contains(token) && */token.length() > 0) {
- indexBuilder.addEntryWithTokens(entryData, Collections.singleton(token), EntryTypeName.BRACKETED);
- }
- }
-
- // process paren stuff
- final String[] parenTokens = NON_CHAR.split(parenthesized.toString());
- for (final String token : parenTokens) {
- assert !token.contains("-");
- if (/*!alreadyDone.contains(token) && */token.length() > 0) {
- indexBuilder.addEntryWithTokens(entryData, Collections.singleton(token), EntryTypeName.PARENTHESIZED);
- }
}
-
- }
- private String parseField_DE(final IndexBuilder indexBuilder, String field,
- final IndexedEntry entryData, final int subfieldIdx) {
-
+ private String parseField_DE(final IndexBuilder indexBuilder, String field,
+ final IndexedEntry entryData, final int subfieldIdx) {
+
// final Matcher matcher = DE_NOUN.matcher(field);
// while (matcher.find()) {
// final String noun = matcher.group(1);
- //final String gender = matcher.group(2);
+ //final String gender = matcher.group(2);
// if (alreadyDone.add(noun)) {
// System.out.println("Found DE noun " + noun + ", " + gender);
// final List<EntryData> entries = indexBuilder.getOrCreateEntries(noun, EntryTypeName.NOUN);
// }
// }
- // In English, curly braces are used for different tenses.
- field = CURLY_BRACED.matcher(field).replaceAll(" ");
+ if (field.indexOf('{') == -1) return field;
+
+ // In English, curly braces are used for different tenses.
+ field = CURLY_BRACED.matcher(field).replaceAll(" ");
+
+ return field;
+ }
+
+ private String parseField_EN(final IndexBuilder indexBuilder, String field,
+ final IndexedEntry entryData, final int subfieldIdx) {
+ if (field.startsWith("to ")) {
+ field = field.substring(3);
+ }
+ return field;
+ }
- return field;
- }
-
- private String parseField_EN(final IndexBuilder indexBuilder, String field,
- final IndexedEntry entryData, final int subfieldIdx) {
- if (field.startsWith("to ")) {
- field = field.substring(3);
+ public static Set<String> tokenize(final String text, final Pattern pattern) {
+ final String[] split = pattern.split(text);
+ final Set<String> result = new LinkedHashSet<>(Arrays.asList(split));
+ result.remove("");
+ return result;
}
- return field;
- }
-
- public static final Set<String> tokenize(final String text, final Pattern pattern) {
- final String[] split = pattern.split(text);
- final Set<String> result = new LinkedHashSet<String>(Arrays.asList(split));
- result.remove("");
- return result;
- }
}
projects / DictionaryPC.git / blobdiff