netwerk/base/public/nsIRequest.idl

Thu, 22 Jan 2015 13:21:57 +0100

author
Michael Schloh von Bennewitz <michael@schloh.com>
date
Thu, 22 Jan 2015 13:21:57 +0100
branch
TOR_BUG_9701
changeset 15
b8a032363ba2
permissions
-rw-r--r--

Incorporate requested changes from Mozilla in review:
https://bugzilla.mozilla.org/show_bug.cgi?id=1123480#c6

michael@0 1 /* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
michael@0 2 /* This Source Code Form is subject to the terms of the Mozilla Public
michael@0 3 * License, v. 2.0. If a copy of the MPL was not distributed with this
michael@0 4 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */
michael@0 5
michael@0 6 #include "nsISupports.idl"
michael@0 7
michael@0 8 interface nsILoadGroup;
michael@0 9
michael@0 10 typedef unsigned long nsLoadFlags;
michael@0 11
michael@0 12 /**
michael@0 13 * nsIRequest
michael@0 14 */
michael@0 15 [scriptable, uuid(ef6bfbd2-fd46-48d8-96b7-9f8f0fd387fe)]
michael@0 16 interface nsIRequest : nsISupports
michael@0 17 {
michael@0 18 /**
michael@0 19 * The name of the request. Often this is the URI of the request.
michael@0 20 */
michael@0 21 readonly attribute AUTF8String name;
michael@0 22
michael@0 23 /**
michael@0 24 * Indicates whether the request is pending. nsIRequest::isPending is
michael@0 25 * true when there is an outstanding asynchronous event that will make
michael@0 26 * the request no longer be pending. Requests do not necessarily start
michael@0 27 * out pending; in some cases, requests have to be explicitly initiated
michael@0 28 * (e.g. nsIChannel implementations are only pending once asyncOpen
michael@0 29 * returns successfully).
michael@0 30 *
michael@0 31 * Requests can become pending multiple times during their lifetime.
michael@0 32 *
michael@0 33 * @return TRUE if the request has yet to reach completion.
michael@0 34 * @return FALSE if the request has reached completion (e.g., after
michael@0 35 * OnStopRequest has fired).
michael@0 36 * @note Suspended requests are still considered pending.
michael@0 37 */
michael@0 38 boolean isPending();
michael@0 39
michael@0 40 /**
michael@0 41 * The error status associated with the request.
michael@0 42 */
michael@0 43 readonly attribute nsresult status;
michael@0 44
michael@0 45 /**
michael@0 46 * Cancels the current request. This will close any open input or
michael@0 47 * output streams and terminate any async requests. Users should
michael@0 48 * normally pass NS_BINDING_ABORTED, although other errors may also
michael@0 49 * be passed. The error passed in will become the value of the
michael@0 50 * status attribute.
michael@0 51 *
michael@0 52 * Implementations must not send any notifications (e.g. via
michael@0 53 * nsIRequestObserver) synchronously from this function. Similarly,
michael@0 54 * removal from the load group (if any) must also happen asynchronously.
michael@0 55 *
michael@0 56 * Requests that use nsIStreamListener must not call onDataAvailable
michael@0 57 * anymore after cancel has been called.
michael@0 58 *
michael@0 59 * @param aStatus the reason for canceling this request.
michael@0 60 *
michael@0 61 * NOTE: most nsIRequest implementations expect aStatus to be a
michael@0 62 * failure code; however, some implementations may allow aStatus to
michael@0 63 * be a success code such as NS_OK. In general, aStatus should be
michael@0 64 * a failure code.
michael@0 65 */
michael@0 66 void cancel(in nsresult aStatus);
michael@0 67
michael@0 68 /**
michael@0 69 * Suspends the current request. This may have the effect of closing
michael@0 70 * any underlying transport (in order to free up resources), although
michael@0 71 * any open streams remain logically opened and will continue delivering
michael@0 72 * data when the transport is resumed.
michael@0 73 *
michael@0 74 * Calling cancel() on a suspended request must not send any
michael@0 75 * notifications (such as onstopRequest) until the request is resumed.
michael@0 76 *
michael@0 77 * NOTE: some implementations are unable to immediately suspend, and
michael@0 78 * may continue to deliver events already posted to an event queue. In
michael@0 79 * general, callers should be capable of handling events even after
michael@0 80 * suspending a request.
michael@0 81 */
michael@0 82 void suspend();
michael@0 83
michael@0 84 /**
michael@0 85 * Resumes the current request. This may have the effect of re-opening
michael@0 86 * any underlying transport and will resume the delivery of data to
michael@0 87 * any open streams.
michael@0 88 */
michael@0 89 void resume();
michael@0 90
michael@0 91 /**
michael@0 92 * The load group of this request. While pending, the request is a
michael@0 93 * member of the load group. It is the responsibility of the request
michael@0 94 * to implement this policy.
michael@0 95 */
michael@0 96 attribute nsILoadGroup loadGroup;
michael@0 97
michael@0 98 /**
michael@0 99 * The load flags of this request. Bits 0-15 are reserved.
michael@0 100 *
michael@0 101 * When added to a load group, this request's load flags are merged with
michael@0 102 * the load flags of the load group.
michael@0 103 */
michael@0 104 attribute nsLoadFlags loadFlags;
michael@0 105
michael@0 106 /**
michael@0 107 * Mask defining the bits reserved for nsIRequest LoadFlags
michael@0 108 */
michael@0 109 const unsigned long LOAD_REQUESTMASK = 0xFFFF;
michael@0 110
michael@0 111 /**************************************************************************
michael@0 112 * Listed below are the various load flags which may be or'd together.
michael@0 113 */
michael@0 114
michael@0 115 /**
michael@0 116 * No special load flags:
michael@0 117 */
michael@0 118 const unsigned long LOAD_NORMAL = 0;
michael@0 119
michael@0 120 /**
michael@0 121 * Don't deliver status notifications to the nsIProgressEventSink, or keep
michael@0 122 * this load from completing the nsILoadGroup it may belong to.
michael@0 123 */
michael@0 124 const unsigned long LOAD_BACKGROUND = 1 << 0;
michael@0 125
michael@0 126 /**************************************************************************
michael@0 127 * The following flags control the flow of data into the cache.
michael@0 128 */
michael@0 129
michael@0 130 /**
michael@0 131 * This flag prevents loading of the request with an HTTP pipeline.
michael@0 132 * Generally this is because the resource is expected to take a
michael@0 133 * while to load and may cause head of line blocking problems.
michael@0 134 */
michael@0 135 const unsigned long INHIBIT_PIPELINE = 1 << 6;
michael@0 136
michael@0 137 /**
michael@0 138 * This flag prevents caching of any kind. It does not, however, prevent
michael@0 139 * cached content from being used to satisfy this request.
michael@0 140 */
michael@0 141 const unsigned long INHIBIT_CACHING = 1 << 7;
michael@0 142
michael@0 143 /**
michael@0 144 * This flag prevents caching on disk (or other persistent media), which
michael@0 145 * may be needed to preserve privacy. For HTTPS, this flag is set auto-
michael@0 146 * matically.
michael@0 147 */
michael@0 148 const unsigned long INHIBIT_PERSISTENT_CACHING = 1 << 8;
michael@0 149
michael@0 150 /**************************************************************************
michael@0 151 * The following flags control what happens when the cache contains data
michael@0 152 * that could perhaps satisfy this request. They are listed in descending
michael@0 153 * order of precidence.
michael@0 154 */
michael@0 155
michael@0 156 /**
michael@0 157 * Force an end-to-end download of content data from the origin server.
michael@0 158 * This flag is used for a shift-reload.
michael@0 159 */
michael@0 160 const unsigned long LOAD_BYPASS_CACHE = 1 << 9;
michael@0 161
michael@0 162 /**
michael@0 163 * Attempt to force a load from the cache, bypassing ALL validation logic
michael@0 164 * (note: this is stronger than VALIDATE_NEVER, which still validates for
michael@0 165 * certain conditions).
michael@0 166 *
michael@0 167 * If the resource is not present in cache, it will be loaded from the
michael@0 168 * network. Combine this flag with LOAD_ONLY_FROM_CACHE if you wish to
michael@0 169 * perform cache-only loads without validation checks.
michael@0 170 *
michael@0 171 * This flag is used when browsing via history. It is not recommended for
michael@0 172 * normal browsing as it may likely violate reasonable assumptions made by
michael@0 173 * the server and confuse users.
michael@0 174 */
michael@0 175 const unsigned long LOAD_FROM_CACHE = 1 << 10;
michael@0 176
michael@0 177 /**
michael@0 178 * The following flags control the frequency of cached content validation
michael@0 179 * when neither LOAD_BYPASS_CACHE or LOAD_FROM_CACHE are set. By default,
michael@0 180 * cached content is automatically validated if necessary before reuse.
michael@0 181 *
michael@0 182 * VALIDATE_ALWAYS forces validation of any cached content independent of
michael@0 183 * its expiration time.
michael@0 184 *
michael@0 185 * VALIDATE_NEVER disables validation of cached content, unless it arrived
michael@0 186 * with the "Cache: no-store" header, or arrived via HTTPS with the
michael@0 187 * "Cache: no-cache" header.
michael@0 188 *
michael@0 189 * VALIDATE_ONCE_PER_SESSION disables validation of expired content,
michael@0 190 * provided it has already been validated (at least once) since the start
michael@0 191 * of this session.
michael@0 192 *
michael@0 193 * NOTE TO IMPLEMENTORS:
michael@0 194 * These flags are intended for normal browsing, and they should therefore
michael@0 195 * not apply to content that must be validated before each use. Consider,
michael@0 196 * for example, a HTTP response with a "Cache-control: no-cache" header.
michael@0 197 * According to RFC2616, this response must be validated before it can
michael@0 198 * be taken from a cache. Breaking this requirement could result in
michael@0 199 * incorrect and potentially undesirable side-effects.
michael@0 200 */
michael@0 201 const unsigned long VALIDATE_ALWAYS = 1 << 11;
michael@0 202 const unsigned long VALIDATE_NEVER = 1 << 12;
michael@0 203 const unsigned long VALIDATE_ONCE_PER_SESSION = 1 << 13;
michael@0 204
michael@0 205 /**
michael@0 206 * When set, this flag indicates that no user-specific data should be added
michael@0 207 * to the request when opened. This means that things like authorization
michael@0 208 * tokens or cookie headers should not be added.
michael@0 209 */
michael@0 210 const unsigned long LOAD_ANONYMOUS = 1 << 14;
michael@0 211
michael@0 212 /**
michael@0 213 * When set, this flag indicates that caches of network connections,
michael@0 214 * particularly HTTP persistent connections, should not be used.
michael@0 215 */
michael@0 216 const unsigned long LOAD_FRESH_CONNECTION = 1 << 15;
michael@0 217 };

mercurial