|
1 /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 4 -*- */ |
|
2 /* This Source Code Form is subject to the terms of the Mozilla Public |
|
3 * License, v. 2.0. If a copy of the MPL was not distributed with this |
|
4 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */ |
|
5 |
|
6 #include "nsISupports.idl" |
|
7 |
|
8 /** |
|
9 * URIs are essentially structured names for things -- anything. This interface |
|
10 * provides accessors to set and query the most basic components of an URI. |
|
11 * Subclasses, including nsIURL, impose greater structure on the URI. |
|
12 * |
|
13 * This interface follows Tim Berners-Lee's URI spec (RFC2396) [1], where the |
|
14 * basic URI components are defined as such: |
|
15 * <pre> |
|
16 * ftp://username:password@hostname:portnumber/pathname#ref |
|
17 * \ / \ / \ / \ /\ \ / |
|
18 * - --------------- ------ -------- | - |
|
19 * | | | | | | |
|
20 * | | | | | Ref |
|
21 * | | | Port \ / |
|
22 * | | Host / -------- |
|
23 * | UserPass / | |
|
24 * Scheme / Path |
|
25 * \ / |
|
26 * -------------------------------- |
|
27 * | |
|
28 * PrePath |
|
29 * </pre> |
|
30 * The definition of the URI components has been extended to allow for |
|
31 * internationalized domain names [2] and the more generic IRI structure [3]. |
|
32 * |
|
33 * Note also that the RFC defines #-separated fragment identifiers as being |
|
34 * "not part of the URI". Despite this, we bundle them as part of the URI, for |
|
35 * convenience. |
|
36 * |
|
37 * [1] http://www.ietf.org/rfc/rfc2396.txt |
|
38 * [2] http://www.ietf.org/internet-drafts/draft-ietf-idn-idna-06.txt |
|
39 * [3] http://www.ietf.org/internet-drafts/draft-masinter-url-i18n-08.txt |
|
40 */ |
|
41 |
|
42 %{C++ |
|
43 #undef GetPort // XXX Windows! |
|
44 #undef SetPort // XXX Windows! |
|
45 %} |
|
46 |
|
47 /** |
|
48 * nsIURI - interface for an uniform resource identifier w/ i18n support. |
|
49 * |
|
50 * AUTF8String attributes may contain unescaped UTF-8 characters. |
|
51 * Consumers should be careful to escape the UTF-8 strings as necessary, but |
|
52 * should always try to "display" the UTF-8 version as provided by this |
|
53 * interface. |
|
54 * |
|
55 * AUTF8String attributes may also contain escaped characters. |
|
56 * |
|
57 * Unescaping URI segments is unadvised unless there is intimate |
|
58 * knowledge of the underlying charset or there is no plan to display (or |
|
59 * otherwise enforce a charset on) the resulting URI substring. |
|
60 * |
|
61 * The correct way to create an nsIURI from a string is via |
|
62 * nsIIOService.newURI. |
|
63 * |
|
64 * NOTE: nsBinaryInputStream::ReadObject contains a hackaround to intercept the |
|
65 * old (pre-gecko6) nsIURI IID and swap in the current IID instead, in order |
|
66 * for sessionstore to work after an upgrade. If this IID is revved further, |
|
67 * we will need to add additional checks there for all intermediate IIDs, until |
|
68 * nsPrincipal is fixed to serialize its URIs as nsISupports (bug 662693). |
|
69 */ |
|
70 [scriptable, uuid(395fe045-7d18-4adb-a3fd-af98c8a1af11)] |
|
71 interface nsIURI : nsISupports |
|
72 { |
|
73 /************************************************************************ |
|
74 * The URI is broken down into the following principal components: |
|
75 */ |
|
76 |
|
77 /** |
|
78 * Returns a string representation of the URI. Setting the spec causes |
|
79 * the new spec to be parsed per the rules for the scheme the URI |
|
80 * currently has. In particular, setting the spec to a URI string with a |
|
81 * different scheme will generally produce incorrect results; no one |
|
82 * outside of a protocol handler implementation should be doing that. If |
|
83 * the URI stores information from the nsIIOService.newURI call used to |
|
84 * create it other than just the parsed string, then behavior of this |
|
85 * information on setting the spec attribute is undefined. |
|
86 * |
|
87 * Some characters may be escaped. |
|
88 */ |
|
89 attribute AUTF8String spec; |
|
90 |
|
91 /** |
|
92 * The prePath (eg. scheme://user:password@host:port) returns the string |
|
93 * before the path. This is useful for authentication or managing sessions. |
|
94 * |
|
95 * Some characters may be escaped. |
|
96 */ |
|
97 readonly attribute AUTF8String prePath; |
|
98 |
|
99 /** |
|
100 * The Scheme is the protocol to which this URI refers. The scheme is |
|
101 * restricted to the US-ASCII charset per RFC2396. Setting this is |
|
102 * highly discouraged outside of a protocol handler implementation, since |
|
103 * that will generally lead to incorrect results. |
|
104 */ |
|
105 attribute ACString scheme; |
|
106 |
|
107 /** |
|
108 * The username:password (or username only if value doesn't contain a ':') |
|
109 * |
|
110 * Some characters may be escaped. |
|
111 */ |
|
112 attribute AUTF8String userPass; |
|
113 |
|
114 /** |
|
115 * The optional username and password, assuming the preHost consists of |
|
116 * username:password. |
|
117 * |
|
118 * Some characters may be escaped. |
|
119 */ |
|
120 attribute AUTF8String username; |
|
121 attribute AUTF8String password; |
|
122 |
|
123 /** |
|
124 * The host:port (or simply the host, if port == -1). |
|
125 * |
|
126 * Characters are NOT escaped. |
|
127 */ |
|
128 attribute AUTF8String hostPort; |
|
129 |
|
130 /** |
|
131 * The host is the internet domain name to which this URI refers. It could |
|
132 * be an IPv4 (or IPv6) address literal. If supported, it could be a |
|
133 * non-ASCII internationalized domain name. |
|
134 * |
|
135 * Characters are NOT escaped. |
|
136 */ |
|
137 attribute AUTF8String host; |
|
138 |
|
139 /** |
|
140 * A port value of -1 corresponds to the protocol's default port (eg. -1 |
|
141 * implies port 80 for http URIs). |
|
142 */ |
|
143 attribute long port; |
|
144 |
|
145 /** |
|
146 * The path, typically including at least a leading '/' (but may also be |
|
147 * empty, depending on the protocol). |
|
148 * |
|
149 * Some characters may be escaped. |
|
150 */ |
|
151 attribute AUTF8String path; |
|
152 |
|
153 |
|
154 /************************************************************************ |
|
155 * An URI supports the following methods: |
|
156 */ |
|
157 |
|
158 /** |
|
159 * URI equivalence test (not a strict string comparison). |
|
160 * |
|
161 * eg. http://foo.com:80/ == http://foo.com/ |
|
162 */ |
|
163 boolean equals(in nsIURI other); |
|
164 |
|
165 /** |
|
166 * An optimization to do scheme checks without requiring the users of nsIURI |
|
167 * to GetScheme, thereby saving extra allocating and freeing. Returns true if |
|
168 * the schemes match (case ignored). |
|
169 */ |
|
170 boolean schemeIs(in string scheme); |
|
171 |
|
172 /** |
|
173 * Clones the current URI. |
|
174 */ |
|
175 nsIURI clone(); |
|
176 |
|
177 /** |
|
178 * This method resolves a relative string into an absolute URI string, |
|
179 * using this URI as the base. |
|
180 * |
|
181 * NOTE: some implementations may have no concept of a relative URI. |
|
182 */ |
|
183 AUTF8String resolve(in AUTF8String relativePath); |
|
184 |
|
185 |
|
186 /************************************************************************ |
|
187 * Additional attributes: |
|
188 */ |
|
189 |
|
190 /** |
|
191 * The URI spec with an ASCII compatible encoding. Host portion follows |
|
192 * the IDNA draft spec. Other parts are URL-escaped per the rules of |
|
193 * RFC2396. The result is strictly ASCII. |
|
194 */ |
|
195 readonly attribute ACString asciiSpec; |
|
196 |
|
197 /** |
|
198 * The URI host with an ASCII compatible encoding. Follows the IDNA |
|
199 * draft spec for converting internationalized domain names (UTF-8) to |
|
200 * ASCII for compatibility with existing internet infrasture. |
|
201 */ |
|
202 readonly attribute ACString asciiHost; |
|
203 |
|
204 /** |
|
205 * The charset of the document from which this URI originated. An empty |
|
206 * value implies UTF-8. |
|
207 * |
|
208 * If this value is something other than UTF-8 then the URI components |
|
209 * (e.g., spec, prePath, username, etc.) will all be fully URL-escaped. |
|
210 * Otherwise, the URI components may contain unescaped multibyte UTF-8 |
|
211 * characters. |
|
212 */ |
|
213 readonly attribute ACString originCharset; |
|
214 |
|
215 /************************************************************************ |
|
216 * Additional attribute & methods added for .ref support: |
|
217 */ |
|
218 |
|
219 /** |
|
220 * Returns the reference portion (the part after the "#") of the URI. |
|
221 * If there isn't one, an empty string is returned. |
|
222 * |
|
223 * Some characters may be escaped. |
|
224 */ |
|
225 attribute AUTF8String ref; |
|
226 |
|
227 /** |
|
228 * URI equivalence test (not a strict string comparison), ignoring |
|
229 * the value of the .ref member. |
|
230 * |
|
231 * eg. http://foo.com/# == http://foo.com/ |
|
232 * http://foo.com/#aaa == http://foo.com/#bbb |
|
233 */ |
|
234 boolean equalsExceptRef(in nsIURI other); |
|
235 |
|
236 /** |
|
237 * Clones the current URI, clearing the 'ref' attribute in the clone. |
|
238 */ |
|
239 nsIURI cloneIgnoringRef(); |
|
240 |
|
241 /** |
|
242 * returns a string for the current URI with the ref element cleared. |
|
243 */ |
|
244 readonly attribute AUTF8String specIgnoringRef; |
|
245 |
|
246 /** |
|
247 * Returns if there is a reference portion (the part after the "#") of the URI. |
|
248 */ |
|
249 readonly attribute boolean hasRef; |
|
250 }; |