1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174
|
// Copyright 2017 The Chromium Authors
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.
namespace url_pattern_index.flat;
// NOTE: Increment url_pattern_index::kUrlPatternIndexFormatVersion whenever
// making a breaking change to this schema.
// Corresponds to url_pattern_index::proto::UrlPatternType.
enum UrlPatternType : ubyte {
SUBSTRING,
WILDCARDED,
REGEXP,
}
// Corresponds to url_pattern_index::proto::AnchorType.
enum AnchorType : ubyte {
NONE,
BOUNDARY,
SUBDOMAIN,
}
// URL rule matching options. These correspond to multiple fields of
// url_pattern_index::proto::UrlRule, but here, they are represented as flags
// of the same bitmask to allow for compact storage.
enum OptionFlag : ubyte (bit_flags) {
IS_ALLOWLIST,
APPLIES_TO_FIRST_PARTY,
APPLIES_TO_THIRD_PARTY,
IS_MATCH_CASE,
}
// The options controlling whether or not to activate filtering for subresources
// of documents that match the URL pattern of the rule.
// Corresponds to url_pattern_index::proto::ActivationType.
enum ActivationType : ubyte (bit_flags) {
DOCUMENT, // Disable all rules on the page.
GENERIC_BLOCK, // Disable generic URL rules on the page.
}
// The types of subresource requests that a URL rule should be applied to.
enum ElementType : ushort (bit_flags) {
OTHER,
SCRIPT,
IMAGE,
STYLESHEET,
OBJECT,
XMLHTTPREQUEST,
// TODO(crbug.com/40516600): Remove OBJECT_SUBREQUEST type once
// url_pattern_index no longer has a dependency on proto::UrlRule.
OBJECT_SUBREQUEST,
SUBDOCUMENT,
PING,
MEDIA,
FONT,
WEBSOCKET,
WEBTRANSPORT,
WEBBUNDLE,
CSP_REPORT,
MAIN_FRAME,
// Note: Update the default value for |element_types| field in UrlRule, on
// adding/removing values from this enum.
}
// The request methods that a URL rule should be applied to.
enum RequestMethod : ushort (bit_flags) {
CONNECT,
DELETE,
GET,
HEAD,
OPTIONS,
PATCH,
POST,
PUT,
OTHER_HTTP, // Matches HTTP(s) request methods not listed above.
NON_HTTP, // Matches non-HTTP(s) requests.
// Note: Update the default value for `request_methods` field in UrlRule, on
// adding/removing values from this enum.
}
// The flat representation of a single URL rule. For more details regarding the
// fields please see the comments to url_pattern_index::proto::UrlRule.
table UrlRule {
// Rule matching options, a bitmask consisting of OptionFlags.
options : ubyte;
// A bitmask of ElementType. Equals ElementType_ANY & ~ElementType_MAIN_FRAME
// by default for compactness. We expect most rules to not use
// ElementType_MAIN_FRAME. Keep this in sync with
// url_pattern_index::kDefaultFlatElementTypesMask.
element_types : ushort = 32767;
// A bitmask of RequestMethod. Equals RequestMethod_ANY by default.
request_methods : ushort = 1023;
// A bitmask of ActivationType. Disables all activation types by default.
activation_types : ubyte = 0;
// Use SUBSTRING as default, since it's the most used pattern type. Same as
// the corresponding proto::UrlRule::url_pattern_type.
url_pattern_type : UrlPatternType = SUBSTRING;
// Use NONE as default, since most of the rules are not anchored.
anchor_left : AnchorType = NONE;
anchor_right : AnchorType = NONE;
// The list of initiator and request domains to be included/excluded from the
// filter's affected set. Each should either be null or have at least a single
// element. The domains should be in lower-case and kept sorted as defined by
// url_pattern_index::CompareDomains. The entries must consist of only ascii
// characters. Use punycode encoding for internationalized domains.
initiator_domains_included : [string];
initiator_domains_excluded : [string];
request_domains_included : [string];
request_domains_excluded : [string];
// A URL pattern in the format defined by |url_pattern_type|. This should
// only consist of ascii characters, since it's matched against a url where
// the host is encoded in the punycode format (in case of internationalized
// domains) and any other non-ascii characters are percent-escaped in utf-8.
// This should be in lower case if the rule is case-insensitive.
url_pattern : string;
// An id which uniquely identifies the rule. Clients must ensure uniqueness if
// they use this field.
id : uint;
// Priority of the rule. Larger the value, greater the priority.
priority : uint64;
// Binary blob for any embedder specific rule conditions. When specified,
// these should be interpreted by the url_pattern_index embedder.
embedder_conditions : [ubyte];
}
// Contains an N-gram (acting as a key in a hash table) and a list of URL rules
// associated with that N-gram.
table NGramToRules {
// A string consisting of N (up to 8) ascii-only non-special characters, which
// are stored in the lowest N non-zero bytes, lower bytes corresponding to
// later symbols. These are lower-cased to support case-insensitive matching.
ngram : ulong;
// The list of rules containing |ngram| as a substring of their URL pattern.
// Sorted in descending order of rule priority.
rule_list : [UrlRule];
}
// A data structure used to select only a handful of URL rule candidates that
// need to be matched against a certain resource URL.
table UrlPatternIndex {
// The N of an N-gram index. Note: |n| should be between 1 and 8.
n : uint;
// A hash table with open addressing. The keys of the table are N-grams.
ngram_index : [NGramToRules];
// The slot that is pointed to by all empty slots of |ngram_index| hash table.
// Note: This is a workaround needed because null offsets are not allowed as
// elements of FlatBuffer arrays.
ngram_index_empty_slot : NGramToRules;
// A list storing the rules that doesn't contain any valid N-grams in their
// URL patterns. Contains all the REGEXP rules as well. Sorted in descending
// order of rule priority.
// TODO(pkalinnikov): Think about better implementation for the fallback
// index. Possibly make it a hash map and maybe merge it with the N-gram
// index, since we can treat any sequence of characters shorter than N as an
// N-gram with zero bytes used for padding.
fallback_rules : [UrlRule];
}
root_type UrlPatternIndex;
|