org.apache.commons.codec.language.bm.Rule Maven / Gradle / Ivy
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.commons.codec.language.bm;
import java.io.InputStream;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collections;
import java.util.Comparator;
import java.util.EnumMap;
import java.util.HashMap;
import java.util.HashSet;
import java.util.List;
import java.util.Map;
import java.util.Scanner;
import java.util.Set;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
/**
*
* A phoneme rule.
*
*
* Rules have a pattern, left context, right context, output phoneme, set of languages for which they apply and a logical flag indicating if
* all lanugages must be in play. A rule matches if:
*
* - the pattern matches at the current position
* - the string up until the beginning of the pattern matches the left context
* - the string from the end of the pattern matches the right context
* - logical is ALL and all languages are in scope; or
* - logical is any other value and at least one language is in scope
*
*
*
* Rules are typically generated by parsing rules resources. In normal use, there will be no need for the user to explicitly construct their
* own.
*
*
* Rules are immutable and thread-safe.
*
Rules resources
*
* Rules are typically loaded from resource files. These are UTF-8 encoded text files. They are systematically named following the pattern:
*
org/apache/commons/codec/language/bm/${NameType#getName}_${RuleType#getName}_${language}.txt
*
*
* The format of these resources is the following:
*
* - Rules: whitespace separated, double-quoted strings. There should be 4 columns to each row, and these will be interpreted as:
*
* - pattern
* - left context
* - right context
* - phoneme
*
*
* - End-of-line comments: Any occurance of '//' will cause all text following on that line to be discarded as a comment.
* - Multi-line comments: Any line starting with '/*' will start multi-line commenting mode. This will skip all content until a
* line ending in '*' and '/' is found.
* - Blank lines: All blank lines will be skipped.
*
*
*
* @author Apache Software Foundation
* @since 1.6
*/
public class Rule {
public static final class Phoneme implements PhonemeExpr {
public static final Comparator COMPARATOR = new Comparator() {
public int compare(Phoneme o1, Phoneme o2) {
for (int i = 0; i < o1.phonemeText.length(); i++) {
if (i >= o2.phonemeText.length()) {
return +1;
}
int c = o1.phonemeText.charAt(i) - o2.phonemeText.charAt(i);
if (c != 0) {
return c;
}
}
if (o1.phonemeText.length() < o2.phonemeText.length()) {
return -1;
}
return 0;
}
};
private final CharSequence phonemeText;
private final Languages.LanguageSet languages;
public Phoneme(CharSequence phonemeText, Languages.LanguageSet languages) {
this.phonemeText = phonemeText;
this.languages = languages;
}
public Phoneme append(CharSequence str) {
return new Phoneme(this.phonemeText.toString() + str.toString(), this.languages);
}
public Languages.LanguageSet getLanguages() {
return this.languages;
}
public Iterable getPhonemes() {
return Collections.singleton(this);
}
public CharSequence getPhonemeText() {
return this.phonemeText;
}
public Phoneme join(Phoneme right) {
return new Phoneme(this.phonemeText.toString() + right.phonemeText.toString(), this.languages.restrictTo(right.languages));
}
}
public interface PhonemeExpr {
Iterable getPhonemes();
}
public static final class PhonemeList implements PhonemeExpr {
private final List phonemes;
public PhonemeList(List phonemes) {
this.phonemes = phonemes;
}
public List getPhonemes() {
return this.phonemes;
}
}
/**
* A minimal wrapper around the functionality of Pattern that we use, to allow for alternate implementations.
*/
public static interface RPattern {
boolean isMatch(CharSequence input);
}
public static final RPattern ALL_STRINGS_RMATCHER = new RPattern() {
public boolean isMatch(CharSequence input) {
return true;
}
};
public static final String ALL = "ALL";
private static final String DOUBLE_QUOTE = "\"";
private static final String HASH_INCLUDE = "#include";
private static final Map>>> RULES = new EnumMap>>>(
NameType.class);
static {
for (NameType s : NameType.values()) {
Map>> rts = new EnumMap>>(RuleType.class);
for (RuleType rt : RuleType.values()) {
Map> rs = new HashMap>();
Languages ls = Languages.getInstance(s);
for (String l : ls.getLanguages()) {
try {
rs.put(l, parseRules(createScanner(s, rt, l), createResourceName(s, rt, l)));
} catch (IllegalStateException e) {
throw new IllegalStateException("Problem processing " + createResourceName(s, rt, l), e);
}
}
if (!rt.equals(RuleType.RULES)) {
rs.put("common", parseRules(createScanner(s, rt, "common"), createResourceName(s, rt, "common")));
}
rts.put(rt, Collections.unmodifiableMap(rs));
}
RULES.put(s, Collections.unmodifiableMap(rts));
}
}
private static boolean contains(CharSequence chars, char input) {
for (int i = 0; i < chars.length(); i++) {
if (chars.charAt(i) == input) {
return true;
}
}
return false;
}
private static String createResourceName(NameType nameType, RuleType rt, String lang) {
return String.format("org/apache/commons/codec/language/bm/%s_%s_%s.txt", nameType.getName(), rt.getName(), lang);
}
private static Scanner createScanner(NameType nameType, RuleType rt, String lang) {
String resName = createResourceName(nameType, rt, lang);
InputStream rulesIS = Languages.class.getClassLoader().getResourceAsStream(resName);
if (rulesIS == null) {
throw new IllegalArgumentException("Unable to load resource: " + resName);
}
return new Scanner(rulesIS, ResourceConstants.ENCODING);
}
private static Scanner createScanner(String lang) {
String resName = String.format("org/apache/commons/codec/language/bm/%s.txt", lang);
InputStream rulesIS = Languages.class.getClassLoader().getResourceAsStream(resName);
if (rulesIS == null) {
throw new IllegalArgumentException("Unable to load resource: " + resName);
}
return new Scanner(rulesIS, ResourceConstants.ENCODING);
}
private static boolean endsWith(CharSequence input, CharSequence suffix) {
if (suffix.length() > input.length()) {
return false;
}
for (int i = input.length() - 1, j = suffix.length() - 1; j >= 0; i--, j--) {
if (input.charAt(i) != suffix.charAt(j)) {
return false;
}
}
return true;
}
/**
* Gets rules for a combination of name type, rule type and languages.
*
* @param nameType
* the NameType to consider
* @param rt
* the RuleType to consider
* @param langs
* the set of languages to consider
* @return a list of Rules that apply
*/
public static List getInstance(NameType nameType, RuleType rt, Languages.LanguageSet langs) {
return langs.isSingleton() ? getInstance(nameType, rt, langs.getAny()) : getInstance(nameType, rt, Languages.ANY);
}
/**
* Gets rules for a combination of name type, rule type and a single language.
*
* @param nameType
* the NameType to consider
* @param rt
* the RuleType to consider
* @param lang
* the language to consider
* @return a list rules for a combination of name type, rule type and a single language.
*/
public static List getInstance(NameType nameType, RuleType rt, String lang) {
List rules = RULES.get(nameType).get(rt).get(lang);
if (rules == null) {
throw new IllegalArgumentException(String.format("No rules found for %s, %s, %s.", nameType.getName(), rt.getName(), lang));
}
return rules;
}
private static Phoneme parsePhoneme(String ph) {
int open = ph.indexOf("[");
if (open >= 0) {
if (!ph.endsWith("]")) {
throw new IllegalArgumentException("Phoneme expression contains a '[' but does not end in ']'");
}
String before = ph.substring(0, open);
String in = ph.substring(open + 1, ph.length() - 1);
Set langs = new HashSet(Arrays.asList(in.split("[+]")));
return new Phoneme(before, Languages.LanguageSet.from(langs));
} else {
return new Phoneme(ph, Languages.ANY_LANGUAGE);
}
}
private static PhonemeExpr parsePhonemeExpr(String ph) {
if (ph.startsWith("(")) { // we have a bracketed list of options
if (!ph.endsWith(")")) {
throw new IllegalArgumentException("Phoneme starts with '(' so must end with ')'");
}
List phs = new ArrayList();
String body = ph.substring(1, ph.length() - 1);
for (String part : body.split("[|]")) {
phs.add(parsePhoneme(part));
}
if (body.startsWith("|") || body.endsWith("|")) {
phs.add(new Phoneme("", Languages.ANY_LANGUAGE));
}
return new PhonemeList(phs);
} else {
return parsePhoneme(ph);
}
}
private static List parseRules(final Scanner scanner, final String location) {
List lines = new ArrayList();
int currentLine = 0;
boolean inMultilineComment = false;
while (scanner.hasNextLine()) {
currentLine++;
String rawLine = scanner.nextLine();
String line = rawLine;
if (inMultilineComment) {
if (line.endsWith(ResourceConstants.EXT_CMT_END)) {
inMultilineComment = false;
} else {
// skip
}
} else {
if (line.startsWith(ResourceConstants.EXT_CMT_START)) {
inMultilineComment = true;
} else {
// discard comments
int cmtI = line.indexOf(ResourceConstants.CMT);
if (cmtI >= 0) {
line = line.substring(0, cmtI);
}
// trim leading-trailing whitespace
line = line.trim();
if (line.length() == 0) {
continue; // empty lines can be safely skipped
}
if (line.startsWith(HASH_INCLUDE)) {
// include statement
String incl = line.substring(HASH_INCLUDE.length()).trim();
if (incl.contains(" ")) {
System.err.println("Warining: malformed import statement: " + rawLine);
} else {
lines.addAll(parseRules(createScanner(incl), location + "->" + incl));
}
} else {
// rule
String[] parts = line.split("\\s+");
if (parts.length != 4) {
System.err.println("Warning: malformed rule statement split into " + parts.length + " parts: " + rawLine);
} else {
try {
String pat = stripQuotes(parts[0]);
String lCon = stripQuotes(parts[1]);
String rCon = stripQuotes(parts[2]);
PhonemeExpr ph = parsePhonemeExpr(stripQuotes(parts[3]));
final int cLine = currentLine;
Rule r = new Rule(pat, lCon, rCon, ph) {
private final int myLine = cLine;
private final String loc = location;
@Override
public String toString() {
final StringBuilder sb = new StringBuilder();
sb.append("Rule");
sb.append("{line=").append(myLine);
sb.append(", loc='").append(loc).append('\'');
sb.append('}');
return sb.toString();
}
};
lines.add(r);
} catch (IllegalArgumentException e) {
throw new IllegalStateException("Problem parsing line " + currentLine, e);
}
}
}
}
}
}
return lines;
}
/**
* Attempts to compile the regex into direct string ops, falling back to Pattern and Matcher in the worst case.
*
* @param regex
* the regular expression to compile
* @return an RPattern that will match this regex
*/
private static RPattern pattern(final String regex) {
boolean startsWith = regex.startsWith("^");
boolean endsWith = regex.endsWith("$");
final String content = regex.substring(startsWith ? 1 : 0, endsWith ? regex.length() - 1 : regex.length());
boolean boxes = content.contains("[");
if (!boxes) {
if (startsWith && endsWith) {
// exact match
if (content.length() == 0) {
// empty
return new RPattern() {
public boolean isMatch(CharSequence input) {
return input.length() == 0;
}
};
} else {
return new RPattern() {
public boolean isMatch(CharSequence input) {
return input.equals(content);
}
};
}
} else if ((startsWith || endsWith) && content.length() == 0) {
// matches every string
return ALL_STRINGS_RMATCHER;
} else if (startsWith) {
// matches from start
return new RPattern() {
public boolean isMatch(CharSequence input) {
return startsWith(input, content);
}
};
} else if (endsWith) {
// matches from start
return new RPattern() {
public boolean isMatch(CharSequence input) {
return endsWith(input, content);
}
};
}
} else {
boolean startsWithBox = content.startsWith("[");
boolean endsWithBox = content.endsWith("]");
if (startsWithBox && endsWithBox) {
String boxContent = content.substring(1, content.length() - 1);
if (!boxContent.contains("[")) {
// box containing alternatives
boolean negate = boxContent.startsWith("^");
if (negate) {
boxContent = boxContent.substring(1);
}
final String bContent = boxContent;
final boolean shouldMatch = !negate;
if (startsWith && endsWith) {
// exact match
return new RPattern() {
public boolean isMatch(CharSequence input) {
return input.length() == 1 && (contains(bContent, input.charAt(0)) == shouldMatch);
}
};
} else if (startsWith) {
// first char
return new RPattern() {
public boolean isMatch(CharSequence input) {
return input.length() > 0 && (contains(bContent, input.charAt(0)) == shouldMatch);
}
};
} else if (endsWith) {
// last char
return new RPattern() {
public boolean isMatch(CharSequence input) {
return input.length() > 0 && (contains(bContent, input.charAt(input.length() - 1)) == shouldMatch);
}
};
}
}
}
}
// System.out.println("Couldn't optimize regex: " + regex);
return new RPattern() {
Pattern pattern = Pattern.compile(regex);
public boolean isMatch(CharSequence input) {
Matcher matcher = pattern.matcher(input);
return matcher.find();
}
};
}
private static boolean startsWith(CharSequence input, CharSequence prefix) {
if (prefix.length() > input.length()) {
return false;
}
for (int i = 0; i < prefix.length(); i++) {
if (input.charAt(i) != prefix.charAt(i)) {
return false;
}
}
return true;
}
private static String stripQuotes(String str) {
if (str.startsWith(DOUBLE_QUOTE)) {
str = str.substring(1);
}
if (str.endsWith(DOUBLE_QUOTE)) {
str = str.substring(0, str.length() - 1);
}
return str;
}
private final RPattern lContext;
private final String pattern;
private final PhonemeExpr phoneme;
private final RPattern rContext;
/**
* Creates a new rule.
*
* @param pattern
* the pattern
* @param lContext
* the left context
* @param rContext
* the right context
* @param phoneme
* the resulting phoneme
*/
public Rule(String pattern, String lContext, String rContext, PhonemeExpr phoneme) {
this.pattern = pattern;
this.lContext = pattern(lContext + "$");
this.rContext = pattern("^" + rContext);
this.phoneme = phoneme;
}
/**
* Gets the left context. This is a regular expression that must match to the left of the pattern.
*
* @return the left context Pattern
*/
public RPattern getLContext() {
return this.lContext;
}
/**
* Gets the pattern. This is a string-literal that must exactly match.
*
* @return the pattern
*/
public String getPattern() {
return this.pattern;
}
/**
* Gets the phoneme. If the rule matches, this is the phoneme associated with the pattern match.
*
* @return the phoneme
*/
public PhonemeExpr getPhoneme() {
return this.phoneme;
}
/**
* Gets the right context. This is a regular expression that must match to the right of the pattern.
*
* @return the right context Pattern
*/
public RPattern getRContext() {
return this.rContext;
}
/**
* Decides if the pattern and context match the input starting at a position. It is a match if the
* lContext matches input up to i, pattern matches at i and
* rContext matches from the end of the match of pattern to the end of input.
*
* @param input
* the input String
* @param i
* the int position within the input
* @return true if the pattern and left/right context match, false otherwise
*/
public boolean patternAndContextMatches(CharSequence input, int i) {
if (i < 0) {
throw new IndexOutOfBoundsException("Can not match pattern at negative indexes");
}
int patternLength = this.pattern.length();
int ipl = i + patternLength;
if (ipl > input.length()) {
// not enough room for the pattern to match
return false;
}
// fixme: this is a readability/speed trade-off - these 3 expressions should be inlined for speed to avoid
// evaluating latter ones if earlier ones have already failed, but that would make the code a lot harder to
// read
boolean patternMatches = input.subSequence(i, ipl).equals(this.pattern);
boolean rContextMatches = this.rContext.isMatch(input.subSequence(ipl, input.length()));
boolean lContextMatches = this.lContext.isMatch(input.subSequence(0, i));
return patternMatches && rContextMatches && lContextMatches;
}
}