Mercurial > hg4j

annotate src/org/tmatesoft/hg/repo/HgIgnore.java @ 343:58016b1b8554

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
HgIgnore: glob patterns not to match substring when not meant to
author Artem Tikhomirov <tikhomirov.artem@gmail.com>
date Sat, 19 Nov 2011 01:07:27 +0100
parents 516b817415ba
children e732521a9eb4
rev   line source
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
1 /*
74
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
2 * Copyright (c) 2010-2011 TMate Software Ltd
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
3 *
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
4 * This program is free software; you can redistribute it and/or modify
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
5 * it under the terms of the GNU General Public License as published by
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
6 * the Free Software Foundation; version 2 of the License.
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
7 *
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
8 * This program is distributed in the hope that it will be useful,
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
9 * but WITHOUT ANY WARRANTY; without even the implied warranty of
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
10 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
11 * GNU General Public License for more details.
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
12 *
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
13 * For information on how to redistribute this software under
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
14 * the terms of a license other than GNU General Public License
102
a3a2e5deb320 Updated contact address to support@hg4j.com
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 91
diff changeset
15 * contact TMate Software at support@hg4j.com
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
16 */
74
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
17 package org.tmatesoft.hg.repo;
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
18
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
19 import java.io.BufferedReader;
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
20 import java.io.File;
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
21 import java.io.FileReader;
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
22 import java.io.IOException;
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
23 import java.util.ArrayList;
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
24 import java.util.Collections;
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
25 import java.util.List;
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
26 import java.util.regex.Pattern;
335
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
27 import java.util.regex.PatternSyntaxException;
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
28
141
8248aae33f7d Adopt FileIterator moving towards WCStatusCollector parameterizing. Improved path handling, move 'em around
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 114
diff changeset
29 import org.tmatesoft.hg.util.Path;
8248aae33f7d Adopt FileIterator moving towards WCStatusCollector parameterizing. Improved path handling, move 'em around
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 114
diff changeset
30
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
31 /**
141
8248aae33f7d Adopt FileIterator moving towards WCStatusCollector parameterizing. Improved path handling, move 'em around
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 114
diff changeset
32 * Handling of ignored paths according to .hgignore configuration
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
33 *
74
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
34 * @author Artem Tikhomirov
6f1b88693d48 Complete refactoring to org.tmatesoft
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 15
diff changeset
35 * @author TMate Software Ltd.
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
36 */
289
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
37 public class HgIgnore implements Path.Matcher {
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
38
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
39 private List<Pattern> entries;
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
40
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
41 HgIgnore() {
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
42 entries = Collections.emptyList();
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
43 }
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
44
335
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
45 /* package-local */List<String> read(File hgignoreFile) throws IOException {
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
46 if (!hgignoreFile.exists()) {
335
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
47 return null;
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
48 }
269
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
49 BufferedReader fr = new BufferedReader(new FileReader(hgignoreFile));
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
50 try {
335
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
51 return read(fr);
269
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
52 } finally {
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
53 fr.close();
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
54 }
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
55 }
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
56
335
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
57 /* package-local */List<String> read(BufferedReader content) throws IOException {
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
58 ArrayList<String> errors = new ArrayList<String>();
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
59 ArrayList<Pattern> result = new ArrayList<Pattern>(entries); // start with existing
269
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
60 String syntax = "regexp"; // or "glob"
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
61 String line;
269
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
62 while ((line = content.readLine()) != null) {
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
63 line = line.trim();
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
64 if (line.startsWith("syntax:")) {
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
65 syntax = line.substring("syntax:".length()).trim();
269
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
66 if (!"regexp".equals(syntax) && !"glob".equals(syntax)) {
335
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
67 errors.add(line);
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
68 continue;
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
69 //throw new IllegalStateException(line);
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
70 }
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
71 } else if (line.length() > 0) {
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
72 // shall I account for local paths in the file (i.e.
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
73 // back-slashed on windows)?
269
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
74 int x, s = 0;
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
75 while ((x = line.indexOf('#', s)) >= 0) {
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
76 if (x > 0 && line.charAt(x-1) == '\\') {
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
77 // remove escape char
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
78 line = line.substring(0, x-1).concat(line.substring(x));
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
79 s = x; // with exclusion of char at [x], s now points to what used to be at [x+1]
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
80 } else {
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
81 line = line.substring(0, x).trim();
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
82 }
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
83 }
269
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
84 if (line.length() == 0) {
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
85 continue;
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
86 }
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
87 if ("glob".equals(syntax)) {
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
88 // hgignore(5)
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
89 // (http://www.selenic.com/mercurial/hgignore.5.html) says slashes '\' are escape characters,
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
90 // hence no special treatment of Windows path
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
91 // however, own attempts make me think '\' on Windows are not treated as escapes
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
92 line = glob2regex(line);
342
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
93 } else {
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
94 assert "regexp".equals(syntax);
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
95 // regular expression patterns need not match start of the line unless demanded explicitly
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
96 line = line.charAt(0) == '^' ? line : ".*" + line;
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
97 }
335
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
98 try {
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
99 result.add(Pattern.compile(line)); // case-sensitive
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
100 } catch (PatternSyntaxException ex) {
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
101 errors.add(line + "@" + ex.getMessage());
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
102 }
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
103 }
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
104 }
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
105 result.trimToSize();
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
106 entries = result;
335
3d41dc148d14 Do not fail with exception on syntax errors in .hgignore
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 289
diff changeset
107 return errors.isEmpty() ? null : errors;
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
108 }
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
109
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
110 // note, #isIgnored(), even if queried for directories and returned positive reply, may still get
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
111 // a file from that ignored folder to get examined. Thus, patterns like "bin" shall match not only a folder,
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
112 // but any file under that folder as well
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
113 // Alternatively, file walker may memorize folder is ignored and uses this information for all nested files. However,
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
114 // this approach would require walker (a) return directories (b) provide nesting information. This may become
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
115 // troublesome when one walks not over io.File, but Eclipse's IResource or any other custom VFS.
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
116 //
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
117 //
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
118 // might be interesting, although looks like of no direct use in my case
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
119 // @see http://stackoverflow.com/questions/1247772/is-there-an-equivalent-of-java-util-regex-for-glob-type-patterns
289
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
120 //
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
121 // TODO consider refactoring to reuse in PathGlobMatcher#glob2regexp
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
122 private static String glob2regex(String line) {
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
123 assert line.length() > 0;
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
124 StringBuilder sb = new StringBuilder(line.length() + 10);
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
125 int start = 0, end = line.length() - 1;
342
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
126 sb.append("(?:|.*/)"); // glob patterns shall match file in any directory
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
127
269
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
128 int inCurly = 0;
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
129 for (int i = start; i <= end; i++) {
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
130 char ch = line.charAt(i);
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
131 if (ch == '.' || ch == '\\') {
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
132 sb.append('\\');
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
133 } else if (ch == '?') {
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
134 // simple '.' substitution might work out, however, more formally
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
135 // a char class seems more appropriate to avoid accidentally
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
136 // matching a subdirectory with ? char (i.e. /a/b?d against /a/bad, /a/bed and /a/b/d)
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
137 // @see http://pubs.opengroup.org/onlinepubs/009695399/utilities/xcu_chap02.html#tag_02_13_03
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
138 // quote: "The slash character in a pathname shall be explicitly matched by using one or more slashes in the pattern;
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
139 // it shall neither be matched by the asterisk or question-mark special characters nor by a bracket expression"
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
140 sb.append("[^/]");
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
141 continue;
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
142 } else if (ch == '*') {
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
143 sb.append("[^/]*?");
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
144 continue;
269
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
145 } else if (ch == '{') {
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
146 // XXX in fact, need to respect if last char was escaping ('\\'), then don't need to treat this as special
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
147 // see link at javadoc above for reasonable example
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
148 inCurly++;
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
149 sb.append('(');
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
150 continue;
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
151 } else if (ch == '}') {
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
152 if (inCurly > 0) {
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
153 inCurly--;
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
154 sb.append(')');
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
155 continue;
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
156 }
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
157 } else if (ch == ',' && inCurly > 0) {
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
158 sb.append('|');
7af843ecc378 Respect glob pattern with alternatives {a,b}
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 225
diff changeset
159 continue;
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
160 }
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
161 sb.append(ch);
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
162 }
343
58016b1b8554 HgIgnore: glob patterns not to match substring when not meant to
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 342
diff changeset
163 sb.append("(?:/|$)");
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
164 return sb.toString();
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
165 }
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
166
289
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
167 /**
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
168 * @param path file or directory name in question
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
169 * @return <code>true</code> if matches repository configuration of ignored files.
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
170 */
141
8248aae33f7d Adopt FileIterator moving towards WCStatusCollector parameterizing. Improved path handling, move 'em around
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 114
diff changeset
171 public boolean isIgnored(Path path) {
342
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
172 String ps = path.toString();
91
c2ce1cfaeb9e ignore file with regex and 'honest' glob support
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 74
diff changeset
173 for (Pattern p : entries) {
342
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
174 int x = ps.indexOf('/'); // reset for each pattern
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
175 if (p.matcher(ps).find()) {
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
176 return true;
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
177 }
342
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
178 while (x != -1 && x+1 != ps.length() /*skip very last segment not to check complete string twice*/) {
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
179 String fragment = ps.substring(0, x);
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
180 if (p.matcher(fragment).matches()) {
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
181 return true;
339
863356c2847e Issue 16: respect glob patterns in HgIgnore for sub-directories
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 335
diff changeset
182 }
342
516b817415ba HgIgnore: regex patterns to match part of the filename do not work
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 339
diff changeset
183 x = ps.indexOf('/', x+1);
339
863356c2847e Issue 16: respect glob patterns in HgIgnore for sub-directories
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 335
diff changeset
184 }
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
185 }
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
186 return false;
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
187 }
289
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
188
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
189 /**
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
190 * A handy wrap of {@link #isIgnored(Path)} into {@link Path.Matcher}. Yields same result as {@link #isIgnored(Path)}.
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
191 * @return <code>true</code> if file is deemed ignored.
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
192 */
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
193 public boolean accept(Path path) {
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
194 return isIgnored(path);
086a326f181f Provide public access to ignored files configuration to use in alternative file walkers
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents: 271
diff changeset
195 }
15
865bf07f381f Basic hgignore handling
Artem Tikhomirov <tikhomirov.artem@gmail.com>
parents:
diff changeset
196 }