Directory: | cvmfs/ |
---|---|
File: | cvmfs/pathspec/pathspec.cc |
Date: | 2025-02-09 02:34:19 |
Exec | Total | Coverage | |
---|---|---|---|
Lines: | 201 | 210 | 95.7% |
Branches: | 120 | 174 | 69.0% |
Line | Branch | Exec | Source |
---|---|---|---|
1 | /** | ||
2 | * This file is part of the CernVM File System. | ||
3 | */ | ||
4 | |||
5 | #include "pathspec.h" | ||
6 | |||
7 | #include <cassert> | ||
8 | |||
9 | #include "util/logging.h" | ||
10 | #include "util/smalloc.h" | ||
11 | |||
12 | 165 | Pathspec::Pathspec(const std::string &spec) : | |
13 | 165 | regex_(NULL), | |
14 | 165 | relaxed_regex_(NULL), | |
15 | 165 | prefix_regex_(NULL), | |
16 | 165 | regex_compiled_(false), | |
17 | 165 | relaxed_regex_compiled_(false), | |
18 | 165 | prefix_regex_compiled_(false), | |
19 | 165 | glob_string_compiled_(false), | |
20 | 165 | glob_string_sequence_compiled_(false), | |
21 | 165 | valid_(true), | |
22 | 165 | absolute_(false) | |
23 | { | ||
24 |
1/2✓ Branch 1 taken 165 times.
✗ Branch 2 not taken.
|
165 | Parse(spec); |
25 |
2/2✓ Branch 1 taken 1 times.
✓ Branch 2 taken 164 times.
|
165 | if (patterns_.size() == 0) { |
26 | 1 | valid_ = false; | |
27 | } | ||
28 | |||
29 | 165 | ElementPatterns::const_iterator i = patterns_.begin(); | |
30 | 165 | const ElementPatterns::const_iterator iend = patterns_.end(); | |
31 |
2/2✓ Branch 2 taken 363 times.
✓ Branch 3 taken 165 times.
|
528 | for (; i != iend; ++i) { |
32 |
2/2✓ Branch 2 taken 2 times.
✓ Branch 3 taken 361 times.
|
363 | if (!i->IsValid()) { |
33 | 2 | valid_ = false; | |
34 | } | ||
35 | } | ||
36 | 165 | } | |
37 | |||
38 | // Compiled regex structure cannot be duplicated and needs to be re-compiled | ||
39 | // Note: the copy-constructed object will perform a lazy evaluation again | ||
40 | 181 | Pathspec::Pathspec(const Pathspec &other) : | |
41 | 181 | patterns_(other.patterns_), | |
42 | 181 | regex_(NULL), | |
43 | 181 | relaxed_regex_(NULL), | |
44 | 181 | prefix_regex_(NULL), | |
45 |
1/2✓ Branch 1 taken 181 times.
✗ Branch 2 not taken.
|
181 | glob_string_(other.glob_string_), |
46 |
1/2✓ Branch 1 taken 181 times.
✗ Branch 2 not taken.
|
181 | glob_string_sequence_(other.glob_string_sequence_), |
47 | 181 | regex_compiled_(false), | |
48 | 181 | relaxed_regex_compiled_(false), | |
49 | 181 | prefix_regex_compiled_(false), | |
50 | 181 | glob_string_compiled_(other.glob_string_compiled_), | |
51 | 181 | glob_string_sequence_compiled_(other.glob_string_sequence_compiled_), | |
52 | 181 | valid_(other.valid_), | |
53 | 181 | absolute_(other.absolute_) {} | |
54 | |||
55 | 346 | Pathspec::~Pathspec() { | |
56 | 346 | DestroyRegularExpressions(); | |
57 | 346 | } | |
58 | |||
59 | 3 | Pathspec& Pathspec::operator=(const Pathspec &other) { | |
60 |
1/2✓ Branch 0 taken 3 times.
✗ Branch 1 not taken.
|
3 | if (this != &other) { |
61 | 3 | DestroyRegularExpressions(); // see: copy c'tor for details | |
62 | 3 | patterns_ = other.patterns_; | |
63 | |||
64 | 3 | glob_string_compiled_ = other.glob_string_compiled_; | |
65 | 3 | glob_string_ = other.glob_string_; | |
66 | |||
67 | 3 | glob_string_sequence_compiled_ = other.glob_string_sequence_compiled_; | |
68 | 3 | glob_string_sequence_ = other.glob_string_sequence_; | |
69 | |||
70 | 3 | valid_ = other.valid_; | |
71 | 3 | absolute_ = other.absolute_; | |
72 | } | ||
73 | |||
74 | 3 | return *this; | |
75 | } | ||
76 | |||
77 | |||
78 | 165 | void Pathspec::Parse(const std::string &spec) { | |
79 | // parsing is done using std::string iterators to walk through the entire | ||
80 | // pathspec parameter. Thus, all parsing methods receive references to these | ||
81 | // iterators and increment itr as they pass along. | ||
82 | 165 | std::string::const_iterator itr = spec.begin(); | |
83 | 165 | const std::string::const_iterator end = spec.end(); | |
84 | |||
85 | 165 | absolute_ = (*itr == kSeparator); | |
86 |
2/2✓ Branch 1 taken 689 times.
✓ Branch 2 taken 165 times.
|
854 | while (itr != end) { |
87 |
2/2✓ Branch 1 taken 326 times.
✓ Branch 2 taken 363 times.
|
689 | if (*itr == kSeparator) { |
88 | 326 | ++itr; | |
89 | 326 | continue; | |
90 | } | ||
91 |
1/2✓ Branch 1 taken 363 times.
✗ Branch 2 not taken.
|
363 | ParsePathElement(end, &itr); |
92 | } | ||
93 | 165 | } | |
94 | |||
95 | 363 | void Pathspec::ParsePathElement( | |
96 | const std::string::const_iterator &end, | ||
97 | std::string::const_iterator *itr | ||
98 | ) { | ||
99 | // find the end of the current pattern element (next directory boundary) | ||
100 | 363 | const std::string::const_iterator begin_element = *itr; | |
101 |
6/6✓ Branch 1 taken 1742 times.
✓ Branch 2 taken 157 times.
✓ Branch 4 taken 1536 times.
✓ Branch 5 taken 206 times.
✓ Branch 6 taken 1536 times.
✓ Branch 7 taken 363 times.
|
1899 | while (*itr != end && **itr != kSeparator) { |
102 | 1536 | ++(*itr); | |
103 | } | ||
104 | 363 | const std::string::const_iterator end_element = *itr; | |
105 | |||
106 | // create a PathspecElementPattern out of this directory description | ||
107 |
2/4✓ Branch 1 taken 363 times.
✗ Branch 2 not taken.
✓ Branch 4 taken 363 times.
✗ Branch 5 not taken.
|
363 | patterns_.push_back(PathspecElementPattern(begin_element, end_element)); |
108 | 363 | } | |
109 | |||
110 | 407 | bool Pathspec::IsMatching(const std::string &query_path) const { | |
111 |
1/2✗ Branch 1 not taken.
✓ Branch 2 taken 407 times.
|
407 | assert(IsValid()); |
112 | |||
113 |
2/2✓ Branch 1 taken 17 times.
✓ Branch 2 taken 390 times.
|
407 | if (query_path.empty()) { |
114 | 17 | return false; | |
115 | } | ||
116 | |||
117 | 390 | const bool query_is_absolute = (query_path[0] == kSeparator); | |
118 |
4/6✓ Branch 0 taken 301 times.
✓ Branch 1 taken 89 times.
✓ Branch 3 taken 288 times.
✓ Branch 4 taken 13 times.
✗ Branch 5 not taken.
✗ Branch 6 not taken.
|
767 | return (!query_is_absolute || this->IsAbsolute()) && |
119 |
2/2✓ Branch 1 taken 149 times.
✓ Branch 2 taken 228 times.
|
767 | IsPathspecMatching(query_path); |
120 | } | ||
121 | |||
122 | 12 | bool Pathspec::IsPrefixMatching(const std::string &query_path) const { | |
123 |
1/2✗ Branch 1 not taken.
✓ Branch 2 taken 12 times.
|
12 | assert(IsValid()); |
124 |
1/2✗ Branch 1 not taken.
✓ Branch 2 taken 12 times.
|
12 | assert(IsAbsolute()); |
125 | |||
126 |
1/2✗ Branch 1 not taken.
✓ Branch 2 taken 12 times.
|
12 | if (query_path.empty()) { |
127 | ✗ | return false; | |
128 | } | ||
129 | |||
130 | 12 | const bool query_is_absolute = (query_path[0] == kSeparator); | |
131 |
4/4✓ Branch 0 taken 11 times.
✓ Branch 1 taken 1 times.
✓ Branch 3 taken 7 times.
✓ Branch 4 taken 4 times.
|
12 | return (query_is_absolute && IsPathspecPrefixMatching(query_path)); |
132 | } | ||
133 | |||
134 | 315 | bool Pathspec::IsMatchingRelaxed(const std::string &query_path) const { | |
135 |
1/2✗ Branch 1 not taken.
✓ Branch 2 taken 315 times.
|
315 | assert(IsValid()); |
136 | |||
137 |
2/2✓ Branch 1 taken 36 times.
✓ Branch 2 taken 279 times.
|
315 | if (query_path.empty()) { |
138 | 36 | return false; | |
139 | } | ||
140 | |||
141 | 279 | return IsPathspecMatchingRelaxed(query_path); | |
142 | } | ||
143 | |||
144 | 377 | bool Pathspec::IsPathspecMatching(const std::string &query_path) const { | |
145 | 377 | return ApplyRegularExpression(query_path, GetRegularExpression()); | |
146 | } | ||
147 | |||
148 | 11 | bool Pathspec::IsPathspecPrefixMatching(const std::string &query_path) const { | |
149 | 11 | return ApplyRegularExpression(query_path, GetPrefixRegularExpression()); | |
150 | } | ||
151 | |||
152 | 279 | bool Pathspec::IsPathspecMatchingRelaxed(const std::string &query_path) const { | |
153 | 279 | return ApplyRegularExpression(query_path, GetRelaxedRegularExpression()); | |
154 | } | ||
155 | |||
156 | 667 | bool Pathspec::ApplyRegularExpression(const std::string &query_path, | |
157 | regex_t *regex) const { | ||
158 | 667 | const char *path = query_path.c_str(); | |
159 | 667 | const int retval = regexec(regex, path, 0, NULL, 0); | |
160 | |||
161 |
3/4✓ Branch 0 taken 459 times.
✓ Branch 1 taken 208 times.
✗ Branch 2 not taken.
✓ Branch 3 taken 459 times.
|
667 | if (retval != 0 && retval != REG_NOMATCH) { |
162 | ✗ | PrintRegularExpressionError(retval); | |
163 | } | ||
164 | |||
165 | 667 | return (retval == 0); | |
166 | } | ||
167 | |||
168 | 377 | regex_t* Pathspec::GetRegularExpression() const { | |
169 |
2/2✓ Branch 0 taken 75 times.
✓ Branch 1 taken 302 times.
|
377 | if (!regex_compiled_) { |
170 | 75 | const bool is_relaxed = false; | |
171 |
1/2✓ Branch 1 taken 75 times.
✗ Branch 2 not taken.
|
75 | const std::string regex = GenerateRegularExpression(is_relaxed); |
172 |
1/2✓ Branch 2 taken 75 times.
✗ Branch 3 not taken.
|
75 | LogCvmfs(kLogPathspec, kLogDebug, "compiled regex: %s", regex.c_str()); |
173 | |||
174 |
1/2✓ Branch 1 taken 75 times.
✗ Branch 2 not taken.
|
75 | regex_ = CompileRegularExpression(regex); |
175 | 75 | regex_compiled_ = true; | |
176 | 75 | } | |
177 | |||
178 | 377 | return regex_; | |
179 | } | ||
180 | |||
181 | 11 | regex_t* Pathspec::GetPrefixRegularExpression() const { | |
182 |
2/2✓ Branch 0 taken 2 times.
✓ Branch 1 taken 9 times.
|
11 | if (!prefix_regex_compiled_) { |
183 | 2 | const bool is_relaxed = false; | |
184 | 2 | const bool is_prefix = true; | |
185 |
1/2✓ Branch 1 taken 2 times.
✗ Branch 2 not taken.
|
2 | const std::string regex = GenerateRegularExpression(is_relaxed, is_prefix); |
186 |
1/2✓ Branch 2 taken 2 times.
✗ Branch 3 not taken.
|
2 | LogCvmfs(kLogPathspec, kLogDebug, "compiled regex: %s", regex.c_str()); |
187 | |||
188 |
1/2✓ Branch 1 taken 2 times.
✗ Branch 2 not taken.
|
2 | prefix_regex_ = CompileRegularExpression(regex); |
189 | 2 | prefix_regex_compiled_ = true; | |
190 | 2 | } | |
191 | |||
192 | 11 | return prefix_regex_; | |
193 | } | ||
194 | |||
195 | 279 | regex_t* Pathspec::GetRelaxedRegularExpression() const { | |
196 |
2/2✓ Branch 0 taken 28 times.
✓ Branch 1 taken 251 times.
|
279 | if (!relaxed_regex_compiled_) { |
197 | 28 | const bool is_relaxed = true; | |
198 |
1/2✓ Branch 1 taken 28 times.
✗ Branch 2 not taken.
|
28 | const std::string regex = GenerateRegularExpression(is_relaxed); |
199 |
1/2✓ Branch 2 taken 28 times.
✗ Branch 3 not taken.
|
28 | LogCvmfs(kLogPathspec, kLogDebug, "compiled relaxed regex: %s", |
200 | regex.c_str()); | ||
201 | |||
202 |
1/2✓ Branch 1 taken 28 times.
✗ Branch 2 not taken.
|
28 | relaxed_regex_ = CompileRegularExpression(regex); |
203 | 28 | relaxed_regex_compiled_ = true; | |
204 | 28 | } | |
205 | |||
206 | 279 | return relaxed_regex_; | |
207 | } | ||
208 | |||
209 | 105 | std::string Pathspec::GenerateRegularExpression( | |
210 | const bool is_relaxed, | ||
211 | const bool is_prefix) const | ||
212 | { | ||
213 | // start matching at the first character | ||
214 |
1/2✓ Branch 2 taken 105 times.
✗ Branch 3 not taken.
|
105 | std::string regex = "^"; |
215 | |||
216 | // absolute paths require a / in the beginning | ||
217 |
2/2✓ Branch 1 taken 86 times.
✓ Branch 2 taken 19 times.
|
105 | if (IsAbsolute()) { |
218 |
1/2✓ Branch 1 taken 86 times.
✗ Branch 2 not taken.
|
86 | regex += kSeparator; |
219 | } | ||
220 | |||
221 | // concatenate the regular expressions of the compiled path elements | ||
222 | 105 | ElementPatterns::const_iterator i = patterns_.begin(); | |
223 | 105 | const ElementPatterns::const_iterator iend = patterns_.end(); | |
224 |
2/2✓ Branch 2 taken 227 times.
✓ Branch 3 taken 105 times.
|
332 | for (; i != iend; ++i) { |
225 |
2/4✓ Branch 2 taken 227 times.
✗ Branch 3 not taken.
✓ Branch 5 taken 227 times.
✗ Branch 6 not taken.
|
227 | regex += i->GenerateRegularExpression(is_relaxed); |
226 |
2/2✓ Branch 2 taken 122 times.
✓ Branch 3 taken 105 times.
|
227 | if (i + 1 != iend) { |
227 |
1/2✓ Branch 1 taken 122 times.
✗ Branch 2 not taken.
|
122 | regex += kSeparator; |
228 | } | ||
229 | } | ||
230 | |||
231 |
2/2✓ Branch 0 taken 2 times.
✓ Branch 1 taken 103 times.
|
105 | if (is_prefix) { |
232 |
1/2✓ Branch 1 taken 2 times.
✗ Branch 2 not taken.
|
2 | regex += "($|"; |
233 |
1/2✓ Branch 1 taken 2 times.
✗ Branch 2 not taken.
|
2 | regex += kSeparator; |
234 |
1/2✓ Branch 1 taken 2 times.
✗ Branch 2 not taken.
|
2 | regex += ".*$)"; |
235 | } else { | ||
236 | // a path might end with a trailing slash | ||
237 | // (pathspec does not distinguish files and directories) | ||
238 |
1/2✓ Branch 1 taken 103 times.
✗ Branch 2 not taken.
|
103 | regex += kSeparator; |
239 |
1/2✓ Branch 1 taken 103 times.
✗ Branch 2 not taken.
|
103 | regex += "?$"; |
240 | } | ||
241 | |||
242 | 210 | return regex; | |
243 | } | ||
244 | |||
245 | 105 | regex_t* Pathspec::CompileRegularExpression(const std::string ®ex) const { | |
246 | 105 | regex_t *result = reinterpret_cast<regex_t *>(smalloc(sizeof(regex_t))); | |
247 | 105 | const int flags = REG_NOSUB | REG_NEWLINE | REG_EXTENDED; | |
248 | 105 | const int retval = regcomp(result, regex.c_str(), flags); | |
249 | |||
250 |
1/2✗ Branch 0 not taken.
✓ Branch 1 taken 105 times.
|
105 | if (retval != 0) { |
251 | ✗ | PrintRegularExpressionError(retval); | |
252 | ✗ | assert(false && "failed to compile regex"); | |
253 | } | ||
254 | |||
255 | 105 | return result; | |
256 | } | ||
257 | |||
258 | 349 | void Pathspec::DestroyRegularExpressions() { | |
259 |
2/2✓ Branch 0 taken 75 times.
✓ Branch 1 taken 274 times.
|
349 | if (regex_compiled_) { |
260 |
1/2✗ Branch 0 not taken.
✓ Branch 1 taken 75 times.
|
75 | assert(regex_ != NULL); |
261 | 75 | regfree(regex_); | |
262 | 75 | regex_ = NULL; | |
263 | 75 | regex_compiled_ = false; | |
264 | } | ||
265 | |||
266 |
2/2✓ Branch 0 taken 28 times.
✓ Branch 1 taken 321 times.
|
349 | if (relaxed_regex_compiled_) { |
267 |
1/2✗ Branch 0 not taken.
✓ Branch 1 taken 28 times.
|
28 | assert(relaxed_regex_ != NULL); |
268 | 28 | regfree(relaxed_regex_); | |
269 | 28 | relaxed_regex_ = NULL; | |
270 | 28 | relaxed_regex_compiled_ = false; | |
271 | } | ||
272 | 349 | } | |
273 | |||
274 | 90 | bool Pathspec::operator==(const Pathspec &other) const { | |
275 |
1/2✓ Branch 2 taken 54 times.
✗ Branch 3 not taken.
|
144 | if (patterns_.size() != other.patterns_.size() || |
276 |
6/6✓ Branch 0 taken 54 times.
✓ Branch 1 taken 36 times.
✓ Branch 4 taken 5 times.
✓ Branch 5 taken 49 times.
✓ Branch 6 taken 41 times.
✓ Branch 7 taken 49 times.
|
144 | IsValid() != other.IsValid() || |
277 | 54 | IsAbsolute() != other.IsAbsolute()) { | |
278 | 41 | return false; | |
279 | } | ||
280 | |||
281 | 49 | ElementPatterns::const_iterator i = patterns_.begin(); | |
282 | 49 | const ElementPatterns::const_iterator iend = patterns_.end(); | |
283 | 49 | ElementPatterns::const_iterator j = other.patterns_.begin(); | |
284 | 49 | const ElementPatterns::const_iterator jend = other.patterns_.end(); | |
285 | |||
286 |
5/6✓ Branch 3 taken 102 times.
✓ Branch 4 taken 16 times.
✓ Branch 6 taken 102 times.
✗ Branch 7 not taken.
✓ Branch 8 taken 102 times.
✓ Branch 9 taken 16 times.
|
118 | for (; i != iend && j != jend; ++i, ++j) { |
287 |
3/4✓ Branch 3 taken 102 times.
✗ Branch 4 not taken.
✓ Branch 5 taken 33 times.
✓ Branch 6 taken 69 times.
|
102 | if (*i != *j) { |
288 | 33 | return false; | |
289 | } | ||
290 | } | ||
291 | |||
292 | 16 | return true; | |
293 | } | ||
294 | |||
295 | ✗ | void Pathspec::PrintRegularExpressionError(const int error_code) const { | |
296 | ✗ | assert(regex_compiled_); | |
297 | ✗ | const size_t errbuf_size = 1024; | |
298 | char error[errbuf_size]; | ||
299 | ✗ | regerror(error_code, regex_, error, errbuf_size); | |
300 | ✗ | LogCvmfs(kLogPathspec, kLogStderr, "RegEx Error: %d - %s", error_code, error); | |
301 | } | ||
302 | |||
303 | 17 | const Pathspec::GlobStringSequence& Pathspec::GetGlobStringSequence() const { | |
304 |
1/2✓ Branch 0 taken 17 times.
✗ Branch 1 not taken.
|
17 | if (!glob_string_sequence_compiled_) { |
305 | 17 | GenerateGlobStringSequence(); | |
306 | 17 | glob_string_sequence_compiled_ = true; | |
307 | } | ||
308 | 17 | return glob_string_sequence_; | |
309 | } | ||
310 | |||
311 | |||
312 | 17 | void Pathspec::GenerateGlobStringSequence() const { | |
313 |
1/2✗ Branch 1 not taken.
✓ Branch 2 taken 17 times.
|
17 | assert(glob_string_sequence_.empty()); |
314 | 17 | ElementPatterns::const_iterator i = patterns_.begin(); | |
315 | 17 | const ElementPatterns::const_iterator iend = patterns_.end(); | |
316 |
2/2✓ Branch 2 taken 34 times.
✓ Branch 3 taken 17 times.
|
51 | for (; i != iend; ++i) { |
317 |
1/2✓ Branch 2 taken 34 times.
✗ Branch 3 not taken.
|
34 | const std::string glob_string = i->GenerateGlobString(); |
318 |
1/2✓ Branch 1 taken 34 times.
✗ Branch 2 not taken.
|
34 | glob_string_sequence_.push_back(glob_string); |
319 | 34 | } | |
320 | 17 | } | |
321 | |||
322 | |||
323 | 23 | const std::string& Pathspec::GetGlobString() const { | |
324 |
2/2✓ Branch 0 taken 17 times.
✓ Branch 1 taken 6 times.
|
23 | if (!glob_string_compiled_) { |
325 | 17 | GenerateGlobString(); | |
326 | 17 | glob_string_compiled_ = true; | |
327 | } | ||
328 | 23 | return glob_string_; | |
329 | } | ||
330 | |||
331 | |||
332 | 17 | void Pathspec::GenerateGlobString() const { | |
333 |
1/2✗ Branch 1 not taken.
✓ Branch 2 taken 17 times.
|
17 | assert(glob_string_.empty()); |
334 | |||
335 | 17 | bool is_first = true; | |
336 |
1/2✓ Branch 1 taken 17 times.
✗ Branch 2 not taken.
|
17 | const GlobStringSequence &seq = GetGlobStringSequence(); |
337 | 17 | GlobStringSequence::const_iterator i = seq.begin(); | |
338 | 17 | const GlobStringSequence::const_iterator iend = seq.end(); | |
339 |
2/2✓ Branch 1 taken 34 times.
✓ Branch 2 taken 17 times.
|
51 | for (; i != iend; ++i) { |
340 |
6/6✓ Branch 0 taken 17 times.
✓ Branch 1 taken 17 times.
✓ Branch 3 taken 11 times.
✓ Branch 4 taken 6 times.
✓ Branch 5 taken 28 times.
✓ Branch 6 taken 6 times.
|
34 | if (!is_first || IsAbsolute()) { |
341 |
1/2✓ Branch 1 taken 28 times.
✗ Branch 2 not taken.
|
28 | glob_string_ += kSeparator; |
342 | } | ||
343 |
1/2✓ Branch 2 taken 34 times.
✗ Branch 3 not taken.
|
34 | glob_string_ += *i; |
344 | 34 | is_first = false; | |
345 | } | ||
346 | 17 | } | |
347 |