Rietveld Code Review Tool
Help | Bug tracker | Discussion group | Source code

Side by Side Diff: lib/contentBlockerList.js

Issue 29336787: Issue 3670 - Make rules case sensitive where possible (Closed)
Patch Set: Improved documentation of toRegexp function and switched to JSDoc syntax Created Feb. 25, 2016, 3:21 p.m.
Left:
Right:
Use n/p to move between diff chunks; N/P to move between comments.
Jump to:
View unified diff | Download patch
« no previous file with comments | « no previous file | no next file » | no next file with comments »
Toggle Intra-line Diffs ('i') | Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
OLDNEW
1 /* 1 /*
2 * This file is part of Adblock Plus <https://adblockplus.org/>, 2 * This file is part of Adblock Plus <https://adblockplus.org/>,
3 * Copyright (C) 2006-2016 Eyeo GmbH 3 * Copyright (C) 2006-2016 Eyeo GmbH
4 * 4 *
5 * Adblock Plus is free software: you can redistribute it and/or modify 5 * Adblock Plus is free software: you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License version 3 as 6 * it under the terms of the GNU General Public License version 3 as
7 * published by the Free Software Foundation. 7 * published by the Free Software Foundation.
8 * 8 *
9 * Adblock Plus is distributed in the hope that it will be useful, 9 * Adblock Plus is distributed in the hope that it will be useful,
10 * but WITHOUT ANY WARRANTY; without even the implied warranty of 10 * but WITHOUT ANY WARRANTY; without even the implied warranty of
(...skipping 32 matching lines...) Expand 10 before | Expand all | Expand 10 after
43 } 43 }
44 } 44 }
45 45
46 function escapeRegExp(s) 46 function escapeRegExp(s)
47 { 47 {
48 return s.replace(/[.*+?^${}()|[\]\\]/g, "\\$&"); 48 return s.replace(/[.*+?^${}()|[\]\\]/g, "\\$&");
49 } 49 }
50 50
51 function matchDomain(domain) 51 function matchDomain(domain)
52 { 52 {
53 return "^https?://([^/:]*\\.)?" + escapeRegExp(domain) + "[/:]"; 53 return "^https?://([^/:]*\\.)?" + escapeRegExp(domain).toLowerCase() + "[/:]";
54 } 54 }
55 55
56 function convertElemHideFilter(filter, elemhideSelectorExceptions) 56 function convertElemHideFilter(filter, elemhideSelectorExceptions)
57 { 57 {
58 let included = []; 58 let included = [];
59 let excluded = []; 59 let excluded = [];
60 let rules = []; 60 let rules = [];
61 61
62 parseDomains(filter.domains, included, excluded); 62 parseDomains(filter.domains, included, excluded);
63 63
64 if (excluded.length == 0 && !(filter.selector in elemhideSelectorExceptions)) 64 if (excluded.length == 0 && !(filter.selector in elemhideSelectorExceptions))
65 return {matchDomains: included.map(matchDomain), selector: filter.selector}; 65 return {matchDomains: included.map(matchDomain), selector: filter.selector};
66 } 66 }
67 67
68 /**
69 * Convert the given filter "regexpSource" string into a regular expression.
70 * (Also deciding if the regular expression can be safely converted to and
71 * matched as lower case or not.)
72 *
73 * @param {string} text regexpSource property of a filter
74 * @returns {object} An object containing a regular expression string and a bool
75 * indicating if the filter can be safely matched as lower
76 * case: {regexp: "...", caseSenstive: true/false}
77 */
68 function toRegExp(text) 78 function toRegExp(text)
69 { 79 {
70 let result = []; 80 let result = [];
71 let lastIndex = text.length - 1; 81 let lastIndex = text.length - 1;
82 let hostnameStarted = false;
83 let hostnameFinished = false;
84 let caseSensitive = false;
72 85
73 for (let i = 0; i < text.length; i++) 86 for (let i = 0; i < text.length; i++)
74 { 87 {
75 let c = text[i]; 88 let c = text[i];
76 89
77 switch (c) 90 switch (c)
78 { 91 {
79 case "*": 92 case "*":
93 if (hostnameStarted)
94 hostnameFinished = true;
80 if (result.length > 0 && i < lastIndex && text[i + 1] != "*") 95 if (result.length > 0 && i < lastIndex && text[i + 1] != "*")
81 result.push(".*"); 96 result.push(".*");
82 break; 97 break;
83 case "^": 98 case "^":
99 if (hostnameStarted)
100 hostnameFinished = true;
84 if (i < lastIndex) 101 if (i < lastIndex)
85 result.push("."); 102 result.push(".");
86 break; 103 break;
87 case "|": 104 case "|":
88 if (i == 0) 105 if (i == 0)
89 { 106 {
90 result.push("^"); 107 result.push("^");
91 break; 108 break;
92 } 109 }
93 if (i == lastIndex) 110 if (i == lastIndex)
94 { 111 {
95 result.push("$"); 112 result.push("$");
96 break; 113 break;
97 } 114 }
98 if (i == 1 && text[0] == "|") 115 if (i == 1 && text[0] == "|")
99 { 116 {
117 hostnameStarted = caseSensitive = true;
100 result.push("https?://"); 118 result.push("https?://");
101 break; 119 break;
102 } 120 }
103 case ".": case "+": case "?": case "$":
104 case "{": case "}": case "(": case ")":
105 case "[": case "]": case "\\":
106 result.push("\\", c); 121 result.push("\\", c);
107 break; 122 break;
123 case "?":
124 if (hostnameStarted)
125 hostnameFinished = true;
126 case ".": case "+": case "$": case "{": case "}":
127 case "(": case ")": case "[": case "]": case "\\":
128 result.push("\\", c);
129 break;
130 case "/":
131 if (hostnameStarted)
132 hostnameFinished = true;
133 else if (text.charAt(i-2) == ":" && text.charAt(i-1) == "/")
134 hostnameStarted = caseSensitive = true;
108 default: 135 default:
136 if (hostnameFinished && (c >= "a" && c <= "z" ||
137 c >= "A" && c <= "Z"))
138 caseSensitive = false;
109 result.push(c); 139 result.push(c);
110 } 140 }
111 } 141 }
112 142
113 return result.join(""); 143 return {regexp: result.join(""), caseSensitive: caseSensitive};
114 } 144 }
115 145
116 function getRegExpSource(filter) 146 function getRegExpTrigger(filter)
117 { 147 {
118 let source = toRegExp(filter.regexpSource.replace( 148 let result = toRegExp(filter.regexpSource.replace(
119 // Safari expects punycode, filter lists use unicode 149 // Safari expects punycode, filter lists use unicode
120 /^(\|\||\|?https?:\/\/)([\w\-.*\u0080-\uFFFF]+)/i, 150 /^(\|\||\|?https?:\/\/)([\w\-.*\u0080-\uFFFF]+)/i,
121 function (match, prefix, domain) 151 function (match, prefix, domain)
122 { 152 {
123 return prefix + punycode.toASCII(domain); 153 return prefix + punycode.toASCII(domain);
124 } 154 }
125 )); 155 ));
126 156
157 let trigger = {"url-filter": result.regexp};
158
127 // Limit rules to to HTTP(S) URLs 159 // Limit rules to to HTTP(S) URLs
128 if (!/^(\^|http)/i.test(source)) 160 if (!/^(\^|http)/i.test(trigger["url-filter"]))
129 source = "^https?://.*" + source; 161 trigger["url-filter"] = "^https?://.*" + trigger["url-filter"];
130 162
131 return source; 163 // For rules containing only a hostname we know that we're matching against
164 // a lowercase string unless the matchCase option was passed.
165 if (result.caseSensitive && !filter.matchCase)
166 trigger["url-filter"] = trigger["url-filter"].toLowerCase();
167
168 if (result.caseSensitive || filter.matchCase)
169 trigger["url-filter-is-case-sensitive"] = true;
170
171 return trigger;
132 } 172 }
133 173
134 function getResourceTypes(filter) 174 function getResourceTypes(filter)
135 { 175 {
136 let types = []; 176 let types = [];
137 177
138 if (filter.contentType & typeMap.IMAGE) 178 if (filter.contentType & typeMap.IMAGE)
139 types.push("image"); 179 types.push("image");
140 if (filter.contentType & typeMap.STYLESHEET) 180 if (filter.contentType & typeMap.STYLESHEET)
141 types.push("style-sheet"); 181 types.push("style-sheet");
(...skipping 26 matching lines...) Expand all
168 208
169 if (tldjs.getDomain(domain) == domain) 209 if (tldjs.getDomain(domain) == domain)
170 result.push("www." + domain); 210 result.push("www." + domain);
171 } 211 }
172 212
173 return result; 213 return result;
174 } 214 }
175 215
176 function convertFilter(filter, action, withResourceTypes) 216 function convertFilter(filter, action, withResourceTypes)
177 { 217 {
178 let trigger = {"url-filter": getRegExpSource(filter)}; 218 let trigger = getRegExpTrigger(filter);
179 let included = []; 219 let included = [];
180 let excluded = []; 220 let excluded = [];
181 221
182 parseDomains(filter.domains, included, excluded); 222 parseDomains(filter.domains, included, excluded);
183 223
184 if (withResourceTypes) 224 if (withResourceTypes)
185 trigger["resource-type"] = getResourceTypes(filter); 225 trigger["resource-type"] = getResourceTypes(filter);
186 if (filter.thirdParty != null) 226 if (filter.thirdParty != null)
187 trigger["load-type"] = [filter.thirdParty ? "third-party" : "first-party"]; 227 trigger["load-type"] = [filter.thirdParty ? "third-party" : "first-party"];
188 228
(...skipping 181 matching lines...) Expand 10 before | Expand all | Expand 10 after
370 { 410 {
371 while (selectors.length) 411 while (selectors.length)
372 { 412 {
373 let selector = selectors.splice(0, selectorLimit).join(", "); 413 let selector = selectors.splice(0, selectorLimit).join(", ");
374 414
375 // As of Safari 9.0 element IDs are matched as lowercase. We work around 415 // As of Safari 9.0 element IDs are matched as lowercase. We work around
376 // this by converting to the attribute format [id="elementID"] 416 // this by converting to the attribute format [id="elementID"]
377 selector = convertIDSelectorsToAttributeSelectors(selector); 417 selector = convertIDSelectorsToAttributeSelectors(selector);
378 418
379 addRule({ 419 addRule({
380 trigger: {"url-filter": matchDomain}, 420 trigger: {"url-filter": matchDomain,
421 "url-filter-is-case-sensitive": true},
381 action: {type: "css-display-none", 422 action: {type: "css-display-none",
382 selector: selector} 423 selector: selector}
383 }); 424 });
384 } 425 }
385 }); 426 });
386 427
387 for (let filter of this.elemhideExceptions) 428 for (let filter of this.elemhideExceptions)
388 addRule(convertFilter(filter, "ignore-previous-rules", false)); 429 addRule(convertFilter(filter, "ignore-previous-rules", false));
389 for (let filter of this.requestFilters) 430 for (let filter of this.requestFilters)
390 addRule(convertFilter(filter, "block", true)); 431 addRule(convertFilter(filter, "block", true));
391 for (let filter of this.requestExceptions) 432 for (let filter of this.requestExceptions)
392 addRule(convertFilter(filter, "ignore-previous-rules", true)); 433 addRule(convertFilter(filter, "ignore-previous-rules", true));
393 434
394 return rules; 435 return rules;
395 }; 436 };
OLDNEW
« no previous file with comments | « no previous file | no next file » | no next file with comments »

Powered by Google App Engine
This is Rietveld