public List<RuleMatch> startMatch(
RuleMatch ruleMatch,
RuleApply ruleApply,
ComposedRuleElementMatch containerMatch,
RuleElement entryPoint,
RutaStream stream,
InferenceCrowd crowd) {
List<RuleMatch> result = new ArrayList<RuleMatch>();
ComposedRuleElementMatch composedMatch = createComposedMatch(ruleMatch, containerMatch, stream);
boolean allMatched = true;
for (RuleElement each : elements) {
List<RuleMatch> startMatch =
each.startMatch(ruleMatch, null, composedMatch, each, stream, crowd);
boolean oneMatched = false;
;
for (RuleMatch eachRuleMatch : startMatch) {
boolean matched = eachRuleMatch.matched();
if (matched) {
oneMatched = true;
break;
}
}
allMatched &= oneMatched;
result.addAll(startMatch);
}
for (RuleMatch each : result) {
if (!each.isApplied()) {
ruleApply.add(each);
if (each.matched() && allMatched) {
each.getRule().getRoot().applyRuleElements(each, stream, crowd);
}
each.setApplied(true);
}
}
return result;
}
@Override
public final RuleMatch[] match(final AnalyzedSentence text) {
final List<RuleMatch> ruleMatches = new ArrayList<RuleMatch>();
final AnalyzedTokenReadings[] tokens = text.getTokens();
String prevToken = "";
String prevPrevToken = "";
boolean prevWhite = false;
int pos = 0;
int prevLen = 0;
for (int i = 0; i < tokens.length; i++) {
final String token = tokens[i].getToken();
final boolean isWhite = tokens[i].isWhitespace() || tokens[i].isFieldCode();
pos += token.length();
String msg = null;
int fixLen = 0;
String suggestionText = null;
if (isWhite && isLeftBracket(prevToken)) {
msg = messages.getString("no_space_after");
suggestionText = prevToken;
fixLen = 1;
} else if (!isWhite
&& prevToken.equals(",")
&& isNotQuoteOrHyphen(token)
&& containsNoNumber(prevPrevToken)
&& containsNoNumber(token)
&& !",".equals(prevPrevToken)) {
msg = messages.getString("missing_space_after_comma");
suggestionText = ", ";
} else if (prevWhite) {
if (isRightBracket(token)) {
msg = messages.getString("no_space_before");
suggestionText = token;
fixLen = 1;
} else if (token.equals(",")) {
msg = messages.getString("space_after_comma");
suggestionText = ",";
fixLen = 1;
// exception for duplicated comma (we already have another rule for that)
if (i + 1 < tokens.length && ",".equals(tokens[i + 1].getToken())) {
msg = null;
}
} else if (token.equals(".")) {
msg = messages.getString("no_space_before_dot");
suggestionText = ".";
fixLen = 1;
// exception case for figures such as ".5" and ellipsis
if (i + 1 < tokens.length && isNumberOrDot(tokens[i + 1].getToken())) {
msg = null;
}
}
}
if (msg != null) {
final int fromPos = tokens[i - 1].getStartPos();
final int toPos = tokens[i - 1].getStartPos() + fixLen + prevLen;
// TODO: add some good short comment here
final RuleMatch ruleMatch = new RuleMatch(this, fromPos, toPos, msg);
ruleMatch.setSuggestedReplacement(suggestionText);
ruleMatches.add(ruleMatch);
}
prevPrevToken = prevToken;
prevToken = token;
prevWhite = isWhite && !tokens[i].isFieldCode(); // OOo code before comma/dot
prevLen = tokens[i].getToken().length();
}
return toRuleMatchArray(ruleMatches);
}
