net.sf.saxon.regex.JRegexIterator Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of Saxon-HE Show documentation
Show all versions of Saxon-HE Show documentation
The XSLT and XQuery Processor
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Copyright (c) 2018-2022 Saxonica Limited
// This Source Code Form is subject to the terms of the Mozilla Public License, v. 2.0.
// If a copy of the MPL was not distributed with this file, You can obtain one at http://mozilla.org/MPL/2.0/.
// This Source Code Form is "Incompatible With Secondary Licenses", as defined by the Mozilla Public License, v. 2.0.
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
package net.sf.saxon.regex;
import net.sf.saxon.expr.LastPositionFinder;
import net.sf.saxon.functions.Count;
import net.sf.saxon.str.EmptyUnicodeString;
import net.sf.saxon.str.StringView;
import net.sf.saxon.str.UnicodeString;
import net.sf.saxon.trans.XPathException;
import net.sf.saxon.value.StringValue;
import net.sf.saxon.str.UnicodeBuilder;
import net.sf.saxon.z.IntHashMap;
import net.sf.saxon.z.IntToIntHashMap;
import java.util.ArrayList;
import java.util.List;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
//import java.util.regex.Matcher;
//import java.util.regex.Pattern;
/**
* Class JRegexIterator - provides an iterator over matched and unmatched substrings.
* This implementation of RegexIterator uses the JDK regular expression engine.
*/
public class JRegexIterator implements RegexIterator, LastPositionFinder {
private final String theString; // the input string being matched
private final Pattern pattern; // the regex against which the string is matched
private final Matcher matcher; // the Matcher object that does the matching, and holds the state
/*@Nullable*/ private String current; // the string most recently returned by the iterator
private String nextSubstring; // if the last string was a matching string, null; otherwise the next substring
// matched by the regex
private int prevEnd = 0; // the position in the input string of the end of the last match or non-match
private IntToIntHashMap nestingTable = null;
// evaluated on demand: a table that indicates for each captured group,
// what its immediately-containing captured group is.
/**
* Construct a RegexIterator. Note that the underlying matcher.find() method is called once
* to obtain each matching substring. But the iterator also returns non-matching substrings
* if these appear between the matching substrings.
*
* @param str the string to be analysed
* @param pattern the regular expression
*/
public JRegexIterator(String str, Pattern pattern) {
theString = str;
this.pattern = pattern;
matcher = pattern.matcher(str);
nextSubstring = null;
}
@Override
public boolean supportsGetLength() {
return true;
}
@Override
public int getLength() {
JRegexIterator another = new JRegexIterator(theString, pattern);
return Count.steppingCount(another);
}
/**
* Get the next item in the sequence
*
* @return the next item in the sequence
*/
@Override
public StringValue next() {
if (nextSubstring == null && prevEnd >= 0) {
// we've returned a match (or we're at the start), so find the next match
if (matcher.find()) {
int start = matcher.start();
int end = matcher.end();
if (prevEnd == start) {
// there's no intervening non-matching string to return
nextSubstring = null;
current = theString.substring(start, end);
prevEnd = end;
} else {
// return the non-matching substring first
current = theString.substring(prevEnd, start);
nextSubstring = theString.substring(start, end);
}
} else {
// there are no more regex matches, we must return the final non-matching text if any
if (prevEnd < theString.length()) {
current = theString.substring(prevEnd);
nextSubstring = null;
} else {
// this really is the end...
current = null;
prevEnd = -1;
return null;
}
prevEnd = -1;
}
} else {
// we've returned a non-match, so now return the match that follows it, if there is one
if (prevEnd >= 0) {
current = nextSubstring;
nextSubstring = null;
prevEnd = matcher.end();
} else {
current = null;
return null;
}
}
return StringValue.makeStringValue(current);
}
/**
* Determine whether the current item is a matching item or a non-matching item
*
* @return true if the current item (the one most recently returned by next()) is
* an item that matches the regular expression, or false if it is an item that
* does not match
*/
@Override
public boolean isMatching() {
return nextSubstring == null && prevEnd >= 0;
}
/**
* Get a substring that matches a parenthesised group within the regular expression
*
* @param number the number of the group to be obtained
* @return the substring of the current item that matches the n'th parenthesized group
* within the regular expression
*/
@Override
public UnicodeString getRegexGroup(int number) {
if (!isMatching()) {
return null;
}
if (number > matcher.groupCount() || number < 0) {
return EmptyUnicodeString.getInstance();
}
String s = matcher.group(number);
if (s == null) {
return EmptyUnicodeString.getInstance();
}
return StringView.of(s).tidy();
}
/**
* Get the number of captured groups
*/
@Override
public int getNumberOfGroups() {
return matcher.groupCount();
}
/**
* Process a matching substring, performing specified actions at the start and end of each captured
* subgroup. This method will always be called when operating in "push" mode; it writes its
* result to context.getReceiver(). The matching substring text is all written to the receiver,
* interspersed with calls to the methods onGroupStart() and onGroupEnd().
*
* @param action defines the processing to be performed at the start and end of a group
*/
@Override
public void processMatchingSubstring(RegexMatchHandler action) throws XPathException {
int c = matcher.groupCount();
if (c == 0) {
action.characters(StringView.of(current));
} else {
// Create a map from positions in the string to lists of actions.
// The "actions" in each list are: +N: start group N; -N: end group N.
IntHashMap> actions = new IntHashMap>(c);
for (int i = 1; i <= c; i++) {
int start = matcher.start(i) - matcher.start();
if (start != -1) {
int end = matcher.end(i) - matcher.start();
if (start < end) {
// Add the start action after all other actions on the list for the same position
List s = actions.get(start);
if (s == null) {
s = new ArrayList(4);
actions.put(start, s);
}
s.add(i);
// Add the end action before all other actions on the list for the same position
List e = actions.get(end);
if (e == null) {
e = new ArrayList(4);
actions.put(end, e);
}
e.add(0, -i);
} else {
// zero-length group (start==end). The problem here is that the information available
// from Java isn't sufficient to determine the nesting of groups: match("a", "(a(b?))")
// and match("a", "(a)(b?)") will both give the same result for group 2 (start=1, end=1).
// So we need to go back to the original regex to determine the group nesting
if (nestingTable == null) {
computeNestingTable();
}
int parentGroup = nestingTable.get(i);
// insert the start and end events immediately before the end event for the parent group,
// if present; otherwise after all existing events for this position
List s = actions.get(start);
if (s == null) {
s = new ArrayList(4);
actions.put(start, s);
s.add(i);
s.add(-i);
} else {
int pos = s.size();
for (int e = 0; e < s.size(); e++) {
if (s.get(e) == -parentGroup) {
pos = e;
break;
}
}
s.add(pos, -i);
s.add(pos, i);
}
}
}
}
UnicodeBuilder buff = new UnicodeBuilder();
for (int i = 0; i < current.length() + 1; i++) {
List events = actions.get(i);
if (events != null) {
if (buff.length() > 0) {
action.characters(buff.toUnicodeString());
buff.clear();
}
for (int group : events) {
if (group > 0) {
action.onGroupStart(group);
} else {
action.onGroupEnd(-group);
}
}
}
if (i < current.length()) {
buff.append(current.charAt(i));
}
}
if (buff.length() > 0) {
action.characters(buff.toUnicodeString());
}
}
}
/**
* Compute a table showing for each captured group number (opening paren in the regex),
* the number of its parent group. This is done by reparsing the source of the regular
* expression. This is needed when the result of a match includes an empty group, to determine
* its position relative to other groups finishing at the same character position.
*/
private void computeNestingTable() {
nestingTable = new IntToIntHashMap(16);
String s = pattern.pattern();
int[] stack = new int[s.length()];
int tos = 0;
int group = 1;
int inBrackets = 0;
stack[tos++] = 0;
for (int i = 0; i < s.length(); i++) {
char ch = s.charAt(i);
if (ch == '\'') {
i++;
} else if (ch == '[') {
inBrackets++;
} else if (ch == ']') {
inBrackets--;
} else if (ch == '(' && s.charAt(i + 1) != '?' && inBrackets == 0) {
nestingTable.put(group, stack[tos - 1]);
stack[tos++] = group++;
} else if (ch == ')' && inBrackets == 0) {
tos--;
}
}
}
}