OLD | NEW |
| (Empty) |
1 // Copyright 2014 The Chromium Authors. All rights reserved. | |
2 // Use of this source code is governed by a BSD-style license that can be | |
3 // found in the LICENSE file. | |
4 | |
5 #include "chrome/browser/ui/elide_url.h" | |
6 | |
7 #include "base/logging.h" | |
8 #include "base/strings/string_split.h" | |
9 #include "base/strings/utf_string_conversions.h" | |
10 #include "net/base/escape.h" | |
11 #include "net/base/net_util.h" | |
12 #include "net/base/registry_controlled_domains/registry_controlled_domain.h" | |
13 #include "ui/gfx/text_elider.h" | |
14 #include "ui/gfx/text_utils.h" | |
15 #include "url/gurl.h" | |
16 #include "url/url_constants.h" | |
17 | |
18 using base::UTF8ToUTF16; | |
19 using gfx::ElideText; | |
20 using gfx::GetStringWidthF; | |
21 using gfx::kEllipsisUTF16; | |
22 using gfx::kForwardSlash; | |
23 | |
24 namespace { | |
25 | |
26 const base::char16 kDot = '.'; | |
27 | |
28 // Build a path from the first |num_components| elements in |path_elements|. | |
29 // Prepends |path_prefix|, appends |filename|, inserts ellipsis if appropriate. | |
30 base::string16 BuildPathFromComponents( | |
31 const base::string16& path_prefix, | |
32 const std::vector<base::string16>& path_elements, | |
33 const base::string16& filename, | |
34 size_t num_components) { | |
35 // Add the initial elements of the path. | |
36 base::string16 path = path_prefix; | |
37 | |
38 // Build path from first |num_components| elements. | |
39 for (size_t j = 0; j < num_components; ++j) | |
40 path += path_elements[j] + kForwardSlash; | |
41 | |
42 // Add |filename|, ellipsis if necessary. | |
43 if (num_components != (path_elements.size() - 1)) | |
44 path += base::string16(kEllipsisUTF16) + kForwardSlash; | |
45 path += filename; | |
46 | |
47 return path; | |
48 } | |
49 | |
50 // Takes a prefix (Domain, or Domain+subdomain) and a collection of path | |
51 // components and elides if possible. Returns a string containing the longest | |
52 // possible elided path, or an empty string if elision is not possible. | |
53 base::string16 ElideComponentizedPath( | |
54 const base::string16& url_path_prefix, | |
55 const std::vector<base::string16>& url_path_elements, | |
56 const base::string16& url_filename, | |
57 const base::string16& url_query, | |
58 const gfx::FontList& font_list, | |
59 float available_pixel_width) { | |
60 const size_t url_path_number_of_elements = url_path_elements.size(); | |
61 | |
62 CHECK(url_path_number_of_elements); | |
63 for (size_t i = url_path_number_of_elements - 1; i > 0; --i) { | |
64 base::string16 elided_path = BuildPathFromComponents(url_path_prefix, | |
65 url_path_elements, url_filename, i); | |
66 if (available_pixel_width >= GetStringWidthF(elided_path, font_list)) | |
67 return ElideText(elided_path + url_query, font_list, | |
68 available_pixel_width, gfx::ELIDE_TAIL); | |
69 } | |
70 | |
71 return base::string16(); | |
72 } | |
73 | |
74 // Splits the hostname in the |url| into sub-strings for the full hostname, | |
75 // the domain (TLD+1), and the subdomain (everything leading the domain). | |
76 void SplitHost(const GURL& url, | |
77 base::string16* url_host, | |
78 base::string16* url_domain, | |
79 base::string16* url_subdomain) { | |
80 // Get Host. | |
81 *url_host = UTF8ToUTF16(url.host()); | |
82 | |
83 // Get domain and registry information from the URL. | |
84 *url_domain = UTF8ToUTF16( | |
85 net::registry_controlled_domains::GetDomainAndRegistry( | |
86 url, net::registry_controlled_domains::EXCLUDE_PRIVATE_REGISTRIES)); | |
87 if (url_domain->empty()) | |
88 *url_domain = *url_host; | |
89 | |
90 // Add port if required. | |
91 if (!url.port().empty()) { | |
92 *url_host += UTF8ToUTF16(":" + url.port()); | |
93 *url_domain += UTF8ToUTF16(":" + url.port()); | |
94 } | |
95 | |
96 // Get sub domain. | |
97 const size_t domain_start_index = url_host->find(*url_domain); | |
98 base::string16 kWwwPrefix = UTF8ToUTF16("www."); | |
99 if (domain_start_index != base::string16::npos) | |
100 *url_subdomain = url_host->substr(0, domain_start_index); | |
101 if ((*url_subdomain == kWwwPrefix || url_subdomain->empty() || | |
102 url.SchemeIsFile())) { | |
103 url_subdomain->clear(); | |
104 } | |
105 } | |
106 | |
107 } // namespace | |
108 | |
109 // TODO(pkasting): http://crbug.com/77883 This whole function gets | |
110 // kerning/ligatures/etc. issues potentially wrong by assuming that the width of | |
111 // a rendered string is always the sum of the widths of its substrings. Also I | |
112 // suspect it could be made simpler. | |
113 base::string16 ElideUrl(const GURL& url, | |
114 const gfx::FontList& font_list, | |
115 float available_pixel_width, | |
116 const std::string& languages) { | |
117 // Get a formatted string and corresponding parsing of the url. | |
118 url::Parsed parsed; | |
119 const base::string16 url_string = | |
120 net::FormatUrl(url, languages, net::kFormatUrlOmitAll, | |
121 net::UnescapeRule::SPACES, &parsed, NULL, NULL); | |
122 if (available_pixel_width <= 0) | |
123 return url_string; | |
124 | |
125 // If non-standard, return plain eliding. | |
126 if (!url.IsStandard()) | |
127 return ElideText(url_string, font_list, available_pixel_width, | |
128 gfx::ELIDE_TAIL); | |
129 | |
130 // Now start eliding url_string to fit within available pixel width. | |
131 // Fist pass - check to see whether entire url_string fits. | |
132 const float pixel_width_url_string = GetStringWidthF(url_string, font_list); | |
133 if (available_pixel_width >= pixel_width_url_string) | |
134 return url_string; | |
135 | |
136 // Get the path substring, including query and reference. | |
137 const size_t path_start_index = parsed.path.begin; | |
138 const size_t path_len = parsed.path.len; | |
139 base::string16 url_path_query_etc = url_string.substr(path_start_index); | |
140 base::string16 url_path = url_string.substr(path_start_index, path_len); | |
141 | |
142 // Return general elided text if url minus the query fits. | |
143 const base::string16 url_minus_query = | |
144 url_string.substr(0, path_start_index + path_len); | |
145 if (available_pixel_width >= GetStringWidthF(url_minus_query, font_list)) | |
146 return ElideText(url_string, font_list, available_pixel_width, | |
147 gfx::ELIDE_TAIL); | |
148 | |
149 base::string16 url_host; | |
150 base::string16 url_domain; | |
151 base::string16 url_subdomain; | |
152 SplitHost(url, &url_host, &url_domain, &url_subdomain); | |
153 | |
154 // If this is a file type, the path is now defined as everything after ":". | |
155 // For example, "C:/aa/aa/bb", the path is "/aa/bb/cc". Interesting, the | |
156 // domain is now C: - this is a nice hack for eliding to work pleasantly. | |
157 if (url.SchemeIsFile()) { | |
158 // Split the path string using ":" | |
159 std::vector<base::string16> file_path_split; | |
160 base::SplitString(url_path, ':', &file_path_split); | |
161 if (file_path_split.size() > 1) { // File is of type "file:///C:/.." | |
162 url_host.clear(); | |
163 url_domain.clear(); | |
164 url_subdomain.clear(); | |
165 | |
166 const base::string16 kColon = UTF8ToUTF16(":"); | |
167 url_host = url_domain = file_path_split.at(0).substr(1) + kColon; | |
168 url_path_query_etc = url_path = file_path_split.at(1); | |
169 } | |
170 } | |
171 | |
172 // Second Pass - remove scheme - the rest fits. | |
173 const float pixel_width_url_host = GetStringWidthF(url_host, font_list); | |
174 const float pixel_width_url_path = GetStringWidthF(url_path_query_etc, | |
175 font_list); | |
176 if (available_pixel_width >= | |
177 pixel_width_url_host + pixel_width_url_path) | |
178 return url_host + url_path_query_etc; | |
179 | |
180 // Third Pass: Subdomain, domain and entire path fits. | |
181 const float pixel_width_url_domain = GetStringWidthF(url_domain, font_list); | |
182 const float pixel_width_url_subdomain = | |
183 GetStringWidthF(url_subdomain, font_list); | |
184 if (available_pixel_width >= | |
185 pixel_width_url_subdomain + pixel_width_url_domain + | |
186 pixel_width_url_path) | |
187 return url_subdomain + url_domain + url_path_query_etc; | |
188 | |
189 // Query element. | |
190 base::string16 url_query; | |
191 const float kPixelWidthDotsTrailer = GetStringWidthF( | |
192 base::string16(kEllipsisUTF16), font_list); | |
193 if (parsed.query.is_nonempty()) { | |
194 url_query = UTF8ToUTF16("?") + url_string.substr(parsed.query.begin); | |
195 if (available_pixel_width >= | |
196 (pixel_width_url_subdomain + pixel_width_url_domain + | |
197 pixel_width_url_path - GetStringWidthF(url_query, font_list))) { | |
198 return ElideText(url_subdomain + url_domain + url_path_query_etc, | |
199 font_list, available_pixel_width, gfx::ELIDE_TAIL); | |
200 } | |
201 } | |
202 | |
203 // Parse url_path using '/'. | |
204 std::vector<base::string16> url_path_elements; | |
205 base::SplitString(url_path, kForwardSlash, &url_path_elements); | |
206 | |
207 // Get filename - note that for a path ending with / | |
208 // such as www.google.com/intl/ads/, the file name is ads/. | |
209 base::string16 url_filename( | |
210 url_path_elements.empty() ? base::string16() : url_path_elements.back()); | |
211 size_t url_path_number_of_elements = url_path_elements.size(); | |
212 if (url_filename.empty() && (url_path_number_of_elements > 1)) { | |
213 // Path ends with a '/'. | |
214 --url_path_number_of_elements; | |
215 url_filename = url_path_elements[url_path_number_of_elements - 1] + | |
216 kForwardSlash; | |
217 } | |
218 | |
219 const size_t kMaxNumberOfUrlPathElementsAllowed = 1024; | |
220 if (url_path_number_of_elements <= 1 || | |
221 url_path_number_of_elements > kMaxNumberOfUrlPathElementsAllowed) { | |
222 // No path to elide, or too long of a path (could overflow in loop below) | |
223 // Just elide this as a text string. | |
224 return ElideText(url_subdomain + url_domain + url_path_query_etc, font_list, | |
225 available_pixel_width, gfx::ELIDE_TAIL); | |
226 } | |
227 | |
228 // Start eliding the path and replacing elements by ".../". | |
229 const base::string16 kEllipsisAndSlash = | |
230 base::string16(kEllipsisUTF16) + kForwardSlash; | |
231 const float pixel_width_ellipsis_slash = | |
232 GetStringWidthF(kEllipsisAndSlash, font_list); | |
233 | |
234 // Check with both subdomain and domain. | |
235 base::string16 elided_path = | |
236 ElideComponentizedPath(url_subdomain + url_domain, url_path_elements, | |
237 url_filename, url_query, font_list, | |
238 available_pixel_width); | |
239 if (!elided_path.empty()) | |
240 return elided_path; | |
241 | |
242 // Check with only domain. | |
243 // If a subdomain is present, add an ellipsis before domain. | |
244 // This is added only if the subdomain pixel width is larger than | |
245 // the pixel width of kEllipsis. Otherwise, subdomain remains, | |
246 // which means that this case has been resolved earlier. | |
247 base::string16 url_elided_domain = url_subdomain + url_domain; | |
248 if (pixel_width_url_subdomain > kPixelWidthDotsTrailer) { | |
249 if (!url_subdomain.empty()) | |
250 url_elided_domain = kEllipsisAndSlash[0] + url_domain; | |
251 else | |
252 url_elided_domain = url_domain; | |
253 | |
254 elided_path = ElideComponentizedPath(url_elided_domain, url_path_elements, | |
255 url_filename, url_query, font_list, | |
256 available_pixel_width); | |
257 | |
258 if (!elided_path.empty()) | |
259 return elided_path; | |
260 } | |
261 | |
262 // Return elided domain/.../filename anyway. | |
263 base::string16 final_elided_url_string(url_elided_domain); | |
264 const float url_elided_domain_width = GetStringWidthF(url_elided_domain, | |
265 font_list); | |
266 | |
267 // A hack to prevent trailing ".../...". | |
268 if ((available_pixel_width - url_elided_domain_width) > | |
269 pixel_width_ellipsis_slash + kPixelWidthDotsTrailer + | |
270 GetStringWidthF(base::ASCIIToUTF16("UV"), font_list)) { | |
271 final_elided_url_string += BuildPathFromComponents(base::string16(), | |
272 url_path_elements, url_filename, 1); | |
273 } else { | |
274 final_elided_url_string += url_path; | |
275 } | |
276 | |
277 return ElideText(final_elided_url_string, font_list, available_pixel_width, | |
278 gfx::ELIDE_TAIL); | |
279 } | |
280 | |
281 base::string16 ElideHost(const GURL& url, | |
282 const gfx::FontList& font_list, | |
283 float available_pixel_width) { | |
284 base::string16 url_host; | |
285 base::string16 url_domain; | |
286 base::string16 url_subdomain; | |
287 SplitHost(url, &url_host, &url_domain, &url_subdomain); | |
288 | |
289 const float pixel_width_url_host = GetStringWidthF(url_host, font_list); | |
290 if (available_pixel_width >= pixel_width_url_host) | |
291 return url_host; | |
292 | |
293 if (url_subdomain.empty()) | |
294 return url_domain; | |
295 | |
296 const float pixel_width_url_domain = GetStringWidthF(url_domain, font_list); | |
297 float subdomain_width = available_pixel_width - pixel_width_url_domain; | |
298 if (subdomain_width <= 0) | |
299 return base::string16(kEllipsisUTF16) + kDot + url_domain; | |
300 | |
301 const base::string16 elided_subdomain = ElideText( | |
302 url_subdomain, font_list, subdomain_width, gfx::ELIDE_HEAD); | |
303 return elided_subdomain + url_domain; | |
304 } | |
305 | |
306 base::string16 FormatUrlForSecurityDisplay(const GURL& url, | |
307 const std::string& languages) { | |
308 if (!url.is_valid() || url.is_empty() || !url.IsStandard()) | |
309 return net::FormatUrl(url, languages); | |
310 | |
311 const base::string16 colon(base::ASCIIToUTF16(":")); | |
312 const base::string16 scheme_separator( | |
313 base::ASCIIToUTF16(url::kStandardSchemeSeparator)); | |
314 | |
315 if (url.SchemeIsFile()) { | |
316 return base::ASCIIToUTF16(url::kFileScheme) + scheme_separator + | |
317 base::UTF8ToUTF16(url.path()); | |
318 } | |
319 | |
320 if (url.SchemeIsFileSystem()) { | |
321 const GURL* inner_url = url.inner_url(); | |
322 if (inner_url->SchemeIsFile()) { | |
323 return base::ASCIIToUTF16(url::kFileSystemScheme) + colon + | |
324 FormatUrlForSecurityDisplay(*inner_url, languages) + | |
325 base::UTF8ToUTF16(url.path()); | |
326 } | |
327 return base::ASCIIToUTF16(url::kFileSystemScheme) + colon + | |
328 FormatUrlForSecurityDisplay(*inner_url, languages); | |
329 } | |
330 | |
331 const GURL origin = url.GetOrigin(); | |
332 const std::string& scheme = origin.scheme(); | |
333 const std::string& host = origin.host(); | |
334 | |
335 base::string16 result = base::UTF8ToUTF16(scheme); | |
336 result += scheme_separator; | |
337 result += base::UTF8ToUTF16(host); | |
338 | |
339 const int port = origin.IntPort(); | |
340 const int default_port = url::DefaultPortForScheme(origin.scheme().c_str(), | |
341 origin.scheme().length()); | |
342 if (port != url::PORT_UNSPECIFIED && port != default_port) | |
343 result += colon + base::UTF8ToUTF16(origin.port()); | |
344 | |
345 return result; | |
346 } | |
OLD | NEW |