From 24cb9ee5475b5b870143bb0063918c0ebf674095 Mon Sep 17 00:00:00 2001 From: Anton Leherbauer Date: Wed, 5 Nov 2014 10:39:26 +0100 Subject: [PATCH] Bug 450057 - Open Element fails to find function for certain wildcard patterns --- .../cdt/internal/ui/util/StringMatcher.java | 247 +++++++++--------- 1 file changed, 120 insertions(+), 127 deletions(-) diff --git a/core/org.eclipse.cdt.ui/src/org/eclipse/cdt/internal/ui/util/StringMatcher.java b/core/org.eclipse.cdt.ui/src/org/eclipse/cdt/internal/ui/util/StringMatcher.java index 7ccdd88120a..d24fc15fe39 100644 --- a/core/org.eclipse.cdt.ui/src/org/eclipse/cdt/internal/ui/util/StringMatcher.java +++ b/core/org.eclipse.cdt.ui/src/org/eclipse/cdt/internal/ui/util/StringMatcher.java @@ -1,5 +1,5 @@ /******************************************************************************* - * Copyright (c) 2000, 2008 IBM Corporation and others. + * Copyright (c) 2000, 2014 IBM Corporation and others. * All rights reserved. This program and the accompanying materials * are made available under the terms of the Eclipse Public License v1.0 * which accompanies this distribution, and is available at @@ -26,6 +26,7 @@ public class StringMatcher { /* boundary value beyond which we don't need to search in the text */ protected int fBound= 0; + protected static final char fSingleWildCard= '\u0000'; public static class Position { @@ -42,104 +43,33 @@ public class StringMatcher { return end; } } - /** - * Find the first occurrence of the pattern between startend(exclusive). - * @param text the String object to search in - * @param start the starting index of the search range, inclusive - * @param end the ending index of the search range, exclusive - * @return an StringMatcher.Position object that keeps the starting - * (inclusive) and ending positions (exclusive) of the first occurrence of the - * pattern in the specified range of the text; return null if not found or subtext - * is empty (start==end). A pair of zeros is returned if pattern is empty string - * Note that for pattern like "*abc*" with leading and trailing stars, position of "abc" - * is returned. For a pattern like"*??*" in text "abcdf", (1,3) is returned - */ - - public StringMatcher.Position find(String text, int start, int end) { - if (fPattern == null || text == null) - throw new IllegalArgumentException(); - - int tlen= text.length(); - if (start < 0) - start= 0; - if (end > tlen) - end= tlen; - if (end < 0 || start >= end) - return null; - if (fLength == 0) - return new Position(start, start); - if (fIgnoreWildCards) { - int x= posIn(text, start, end); - if (x < 0) - return null; - return new Position(x, x + fLength); - } - - int segCount= fSegments.length; - if (segCount == 0) //pattern contains only '*'(s) - return new Position(start, end); - - int curPos= start; - int matchStart= -1; - for (int i= 0; i < segCount && curPos < end; ++i) { - String current= fSegments[i]; - int nextMatch= regExpPosIn(text, curPos, end, current); - if (nextMatch < 0) - return null; - if (i == 0) - matchStart= nextMatch; - curPos= nextMatch + current.length(); - } - return new Position(matchStart, curPos); - } - /** - * StringMatcher constructor takes in a String object that is a simple - * pattern which may contain *  for 0 and many characters and - * ?  for exactly one character. Also takes as parameter a boolean object - * specifying if case should be ignored - * @deprecated Use StringMatcher(pattern, ignoreCase, ignoreWildCards). - */ - @Deprecated - public StringMatcher(String aPattern, boolean ignoreCase) { - this(aPattern, ignoreCase, false); - } - /** - * StringMatcher constructor takes in a String object that is a simple - * pattern which may contain *  for 0 and many characters and - * ?  for exactly one character. + * StringMatcher constructor takes in a String object that is a simple + * pattern. The pattern may contain '*' for 0 and many characters and + * '?' for exactly one character. * - * Literal '*' and '?' characters must be escaped in the pattern + * Literal '*' and '?' characters must be escaped in the pattern * e.g., "\*" means literal "*", etc. * - * Escaping any other character (including the escape character itself), + * Escaping any other character (including the escape character itself), * just results in that character in the pattern. * e.g., "\a" means "a" and "\\" means "\" * * If invoking the StringMatcher with string literals in Java, don't forget * escape characters are represented by "\\". * - * @param aPattern the pattern to match text against + * @param pattern the pattern to match text against * @param ignoreCase if true, case is ignored * @param ignoreWildCards if true, wild cards and their escape sequences are ignored * (everything is taken literally). */ - public StringMatcher(String aPattern, boolean ignoreCase, boolean ignoreWildCards) { + public StringMatcher(String pattern, boolean ignoreCase, boolean ignoreWildCards) { + if (pattern == null) + throw new IllegalArgumentException(); fIgnoreCase= ignoreCase; fIgnoreWildCards= ignoreWildCards; - fLength= aPattern.length(); - - /* convert case */ - if (fIgnoreCase) { - char[] chars= aPattern.toCharArray(); - for (int i = 0; i < chars.length; i++) { - chars[i]= Character.toUpperCase(chars[i]); - } - fPattern= new String(chars); - } else { - fPattern= aPattern; - } + fPattern= pattern; + fLength= pattern.length(); if (fIgnoreWildCards) { parseNoWildCards(); @@ -148,24 +78,85 @@ public class StringMatcher { } } /** - * Given the starting (inclusive) and the ending (exclusive) poisitions in the - * text, determine if the given substring matches with aPattern + * Find the first occurrence of the pattern between startend(exclusive). + * @param text the String object to search in + * @param start the starting index of the search range, inclusive + * @param end the ending index of the search range, exclusive + * @return an StringMatcher.Position object that keeps the starting + * (inclusive) and ending positions (exclusive) of the first occurrence of the + * pattern in the specified range of the text; return null if not found or subtext + * is empty (start==end). A pair of zeros is returned if pattern is empty string + * Note that for pattern like "*abc*" with leading and trailing stars, position of "abc" + * is returned. For a pattern like"*??*" in text "abcdf", (1,3) is returned + */ + public StringMatcher.Position find(String text, int start, int end) { + if (text == null) + throw new IllegalArgumentException(); + + int tlen= text.length(); + if (start < 0) + start= 0; + if (end > tlen) + end= tlen; + if (end < 0 ||start >= end ) + return null; + if (fLength == 0) + return new Position(start, start); + if (fIgnoreWildCards) { + int x= posIn(text, start, end); + if (x < 0) + return null; + return new Position(x, x+fLength); + } + + int segCount= fSegments.length; + if (segCount == 0)//pattern contains only '*'(s) + return new Position (start, end); + + int curPos= start; + int matchStart= -1; + int i; + for (i= 0; i < segCount && curPos < end; ++i) { + String current= fSegments[i]; + int nextMatch= regExpPosIn(text, curPos, end, current); + if (nextMatch < 0 ) + return null; + if(i == 0) + matchStart= nextMatch; + curPos= nextMatch + current.length(); + } + if (i < segCount) + return null; + return new Position(matchStart, curPos); + } + /** + * match the given text with the pattern + * @return true if matched eitherwise false + * @param text a String object + */ + public boolean match(String text) { + return match(text, 0, text.length()); + } + /** + * Given the starting (inclusive) and the ending (exclusive) positions in the + * text, determine if the given substring matches with aPattern * @return true if the specified portion of the text matches the pattern - * @param text a String object that contains the substring to match + * @param text a String object that contains the substring to match * @param start marks the starting position (inclusive) of the substring - * @param end marks the ending index (exclusive) of the substring + * @param end marks the ending index (exclusive) of the substring */ public boolean match(String text, int start, int end) { - if (null == fPattern || null == text) + if (null == text) throw new IllegalArgumentException(); if (start > end) return false; if (fIgnoreWildCards) - return fPattern.regionMatches(fIgnoreCase, 0, text, start, fLength); + return (end - start == fLength) && fPattern.regionMatches(fIgnoreCase, 0, text, start, fLength); int segCount= fSegments.length; - if (segCount == 0) //pattern contains only '*'(s) or empty pattern + if (segCount == 0 && (fHasLeadingStar || fHasTrailingStar)) // pattern contains only '*'(s) return true; if (start == end) return fLength == 0; @@ -180,23 +171,27 @@ public class StringMatcher { int tCurPos= start; int bound= end - fBound; - if (bound < 0) + if ( bound < 0) return false; - int i= 0; + int i=0; String current= fSegments[i]; int segLength= current.length(); /* process first segment */ - if (!fHasLeadingStar) { - if (!regExpRegionMatches(text, start, current, 0, segLength)) { + if (!fHasLeadingStar){ + if(!regExpRegionMatches(text, start, current, 0, segLength)) { return false; + } else { + ++i; + tCurPos= tCurPos + segLength; } - ++i; - tCurPos= tCurPos + segLength; } - + if ((fSegments.length == 1) && (!fHasLeadingStar) && (!fHasTrailingStar)) { + // only one segment to match, no wildcards specified + return tCurPos == end; + } /* process middle segments */ - for (; i < segCount && tCurPos <= bound; ++i) { + while (i < segCount) { current= fSegments[i]; int currentMatch; int k= current.indexOf(fSingleWildCard); @@ -210,6 +205,7 @@ public class StringMatcher { return false; } tCurPos= currentMatch + current.length(); + i++; } /* process final segment */ @@ -217,16 +213,9 @@ public class StringMatcher { int clen= current.length(); return regExpRegionMatches(text, end - clen, current, 0, clen); } - return i == segCount; - } - /** - * match the given text with the pattern - * @return true if matched eitherwise false - * @param text a String object - */ - public boolean match(String text) { - return match(text, 0, text.length()); + return i == segCount ; } + /** * This method parses the given pattern into segments seperated by wildcard '*' characters. * Since wildcards are not being used in this case, the pattern consists of a single segment. @@ -237,12 +226,12 @@ public class StringMatcher { fBound= fLength; } /** - * This method parses the given pattern into segments seperated by wildcard '*' characters. + * Parses the given pattern into segments seperated by wildcard '*' characters. */ private void parseWildCards() { - if (fPattern.startsWith("*")) //$NON-NLS-1$ + if(fPattern.startsWith("*"))//$NON-NLS-1$ fHasLeadingStar= true; - if (fPattern.endsWith("*")) { //$NON-NLS-1$ + if(fPattern.endsWith("*")) {//$NON-NLS-1$ /* make sure it's not an escaped wildcard */ if (fLength > 1 && fPattern.charAt(fLength - 2) != '\\') { fHasTrailingStar= true; @@ -256,7 +245,7 @@ public class StringMatcher { while (pos < fLength) { char c= fPattern.charAt(pos++); switch (c) { - case '\\' : + case '\\': if (pos >= fLength) { buf.append(c); } else { @@ -270,20 +259,20 @@ public class StringMatcher { buf.append(next); } } - break; - case '*' : + break; + case '*': if (buf.length() > 0) { /* new segment */ temp.addElement(buf.toString()); fBound += buf.length(); buf.setLength(0); } - break; - case '?' : + break; + case '?': /* append special character representing single match wildcard */ buf.append(fSingleWildCard); - break; - default : + break; + default: buf.append(c); } } @@ -297,13 +286,13 @@ public class StringMatcher { fSegments= new String[temp.size()]; temp.copyInto(fSegments); } - /** + /** * @param text a string which contains no wildcard * @param start the starting index in the text for search, inclusive * @param end the stopping point of search, exclusive - * @return the starting index in the text of the pattern , or -1 if not found + * @return the starting index in the text of the pattern , or -1 if not found */ - protected int posIn(String text, int start, int end) { //no wild card in pattern + protected int posIn(String text, int start, int end) {//no wild card in pattern int max= end - fLength; if (!fIgnoreCase) { @@ -320,12 +309,12 @@ public class StringMatcher { return -1; } - /** + /** * @param text a simple regular expression that may only contain '?'(s) * @param start the starting index in the text for search, inclusive * @param end the stopping point of search, exclusive * @param p a simple regular expression that may contains '?' - * @return the starting index in the text of the pattern , or -1 if not found + * @return the starting index in the text of the pattern , or -1 if not found */ protected int regExpPosIn(String text, int start, int end, String p) { int plen= p.length(); @@ -338,6 +327,7 @@ public class StringMatcher { return -1; } + protected boolean regExpRegionMatches(String text, int tStart, String p, int pStart, int plen) { while (plen-- > 0) { char tchar= text.charAt(tStart++); @@ -353,20 +343,23 @@ public class StringMatcher { if (pchar == tchar) continue; if (fIgnoreCase) { - char tc= Character.toUpperCase(tchar); - if (tc == pchar) + if (Character.toUpperCase(tchar) == Character.toUpperCase(pchar)) + continue; + // comparing after converting to upper case doesn't handle all cases; + // also compare after converting to lower case + if (Character.toLowerCase(tchar) == Character.toLowerCase(pchar)) continue; } return false; } return true; } - /** + /** * @param text the string to match * @param start the starting index in the text for search, inclusive * @param end the stopping point of search, exclusive * @param p a string that has no wildcard - * @return the starting index in the text of the pattern , or -1 if not found + * @return the starting index in the text of the pattern , or -1 if not found */ protected int textPosIn(String text, int start, int end, String p) { @@ -380,7 +373,7 @@ public class StringMatcher { return i; } - for (int i= 0; i <= max; ++i) { + for (int i= start; i <= max; ++i) { if (text.regionMatches(true, i, p, 0, plen)) return i; }