Data structures for Safe Browsing subresource filtering rules.

components/subresource_filter/core/common/proto/rules.proto

Index: components/subresource_filter/core/common/proto/rules.proto
diff --git a/components/subresource_filter/core/common/proto/rules.proto b/components/subresource_filter/core/common/proto/rules.proto
new file mode 100644
index 0000000000000000000000000000000000000000..66ac444b88f8a2eb813f53d3bed67196c65f0e77
--- /dev/null
+++ b/components/subresource_filter/core/common/proto/rules.proto
@@ -0,0 +1,196 @@
+// Copyright 2016 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+syntax = "proto2";
+option optimize_for = LITE_RUNTIME;
+
+package subresource_filter.proto;
+option java_package = "org.chromium.components.subresource_filter.proto";
+
+// The type of a subresource filtering rule.
+enum RuleType {
+  RULE_TYPE_UNSPECIFIED = 0;
+
+  RULE_TYPE_COMMENT = 1;  // Comment rule.
+  RULE_TYPE_URL = 2;      // Network level filtering rule based on URL pattern.
+  RULE_TYPE_CSS = 3;      // Element hiding rule based on a CSS selector.
+};
+
+// The format of a URL pattern.
+enum UrlPatternType {
+  URL_PATTERN_TYPE_UNSPECIFIED = 0;
+
+  // A pattern without special characters, e.g. "example.com".
+  URL_PATTERN_TYPE_SUBSTRING = 1;
+
+  // The pattern contains one or more wildcards, namely '*' and/or '^'
+  // characters. The '*' matches any sequence of characters, while the '^'
+  // matches a separator, i.e. anything but a letter, a digit, or one of [-._%].
+  URL_PATTERN_TYPE_WILDCARDED = 2;
+
+  // The pattern is a regular expression.
+  URL_PATTERN_TYPE_REGEXP = 3;
+};
+
+// Types of anchors that can be used to constrain where a URL pattern must
+// begin/end in the URL in order to be considered a match.
+enum AnchorType {
+  ANCHOR_TYPE_UNSPECIFIED = 0;
+
+  // Acts like a '*' wildcard at the respective end of a pattern.
+  ANCHOR_TYPE_NONE = 1;
+  // The pattern must match from the start/until the end of the URL.
+  ANCHOR_TYPE_BOUNDARY = 2;
+  // The pattern must match starting with the TLD+n of the URL's domain, but the
+  // scheme and subdomains (if any) can be arbitrary.
+  ANCHOR_TYPE_SUBDOMAIN = 3;
+};
+
+// The types of subresource requests that a URL rule should be applied to.
+// Note: Values are used as flags in a bitmask.
+enum ElementType {
+  option allow_alias = true;
+  ELEMENT_TYPE_UNSPECIFIED = 0;
+
+  ELEMENT_TYPE_OTHER = 1;
+  ELEMENT_TYPE_SCRIPT = 2;
+  ELEMENT_TYPE_IMAGE = 4;
+  ELEMENT_TYPE_STYLESHEET = 8;
+  ELEMENT_TYPE_OBJECT = 16;
+  ELEMENT_TYPE_XMLHTTPREQUEST = 32;
+  ELEMENT_TYPE_OBJECT_SUBREQUEST = 64;
+  ELEMENT_TYPE_SUBDOCUMENT = 128;
+  ELEMENT_TYPE_PING = 256;
+  ELEMENT_TYPE_MEDIA = 512;
+  ELEMENT_TYPE_FONT = 1024;
+  ELEMENT_TYPE_POPUP = 2048;
+
+  // NOTE: Keep these two values consistent with the values above.
+  ELEMENT_TYPE_MAX = 2048;
+  ELEMENT_TYPE_ALL = 4095;
+};
+
+// The options controlling whether or not to activate filtering for subresources
+// of documents that match the URL pattern of the rule.
+// Note: Values are used as flags in a bitmask.
+enum ActivationType {
+  option allow_alias = true;
+  ACTIVATION_TYPE_UNSPECIFIED = 0;
+
+  ACTIVATION_TYPE_DOCUMENT = 1;      // Disable all rules on the page.
+  ACTIVATION_TYPE_ELEMHIDE = 2;      // Disable CSS rules on the page.
+  ACTIVATION_TYPE_GENERICHIDE = 4;   // Disable generic CSS rules on the page.
+  ACTIVATION_TYPE_GENERICBLOCK = 8;  // Disable generic URL rules on the page.
+
+  // NOTE: Keep these two values consistent with the values above.
+  ACTIVATION_TYPE_MAX = 8;
+  ACTIVATION_TYPE_ALL = 15;
+};
+
+// The semantics of a rule. Defines how the rule relates to other rules and how
+// it influences the filtering decision.
+enum RuleSemantics {
+  RULE_SEMANTICS_UNSPECIFIED = 0;
+
+  // Matching subresource requests should be aborted, matching elements should
+  // be hidden.
+  RULE_SEMANTICS_BLACKLIST = 1;
+  // If the rule matches, it suppresses any matching RULE_SEMANTICS_BLACKLIST
+  // rule.
+  RULE_SEMANTICS_WHITELIST = 2;
+}
+
+// The type of relation between the source of the requested subresource and that
+// of the document.
+enum SourceType {
+  SOURCE_TYPE_UNSPECIFIED = 0;
+
+  SOURCE_TYPE_ANY = 1;          // Doesn't matter.
+  SOURCE_TYPE_THIRD_PARTY = 2;  // Requesting a trird-party resource.
+  SOURCE_TYPE_FIRST_PARTY = 3;  // Requesting a first-party resource.
+}
+
+// An item of the domain list.
+message DomainListItem {
+  // The UTF-8 representation of the domain, e.g. "subdomain.example.com".
+  optional string domain = 1;
+
+  // Defines whether the domain is excluded from the set of domains.
+  optional bool exclude = 2;
+}
+
+// A network level filtering rule based on a URL pattern. Corresponds to
+// RULE_TYPE_URL.
+message UrlRule {
+  // The semantics of the rule.
+  optional RuleSemantics semantics = 1;
+
+  // Restricts application of the rule to first-party/third-party requests.
+  optional SourceType source_type = 2;
+
+  // Stores the ElementTypes that the rule applies to as a bitwise OR of the
+  // corresponding ElementType values.
+  optional int32 element_types = 3;
+
+  // Stores the ActivationTypes associated with the rule as a bitwise OR of the
+  // corresponding ActivationType values.
+  optional int32 activation_types = 4;
+
+  // The list of domains to be included/excluded from the filter's affected set.
+
+  // The rule applies only to subresources of documents loaded from included
+  // domains (or subdomains thereof). If the list is empty, the rule is applied
+  // on documents from all domains.
+  // If |domains| is empty or has exceptions only, the rule is called generic.
+  // Otherwise is it called domain specific, i.e. applies to a limited number of
+  // domains.
+  repeated DomainListItem domains = 5;
+
+  // The format of |url_pattern|.
+  optional UrlPatternType url_pattern_type = 6;
+
+  // Defines where the URL pattern must start in the URL in order to be
+  // considered a match. Never used with REGEXP patterns.
+  optional AnchorType anchor_left = 7;
+
+  // Defines where the URL pattern must end in the URL in order to be
+  // considered a match. Never used with REGEXP patterns. Never equals to
+  // ANCHOR_TYPE_SUBDOMAIN.
+  optional AnchorType anchor_right = 8;
+
+  // When set, the rule applies only to URLs that match |url_pattern| in a
+  // case-sensitive way.
+  optional bool match_case = 9;
+
+  // The URL pattern of the format prescribed by |url_pattern_type|.
+  optional string url_pattern = 10;
+}
+
+// Element hiding rule based on a CSS selector. Corresponds to RULE_TYPE_CSS.
+message CssRule {
+  // The semantics of the rule.
+  optional RuleSemantics semantics = 1;
+
+  // The list of domains, same as UrlRule::domains.
+  repeated DomainListItem domains = 2;
+
+  // A CSS selector as specified in http://www.w3.org/TR/css3-selectors.
+  optional string css_selector = 3;
+}
+
+// A comment line.
+message Comment {
+  // Comment text.
+  optional string text = 1;
+
+  // For special key-value comments, if any.
+  optional string key = 2;
+  optional string value = 3;
+}
+
+// A container for lists of non-comment rules collated by RuleType.
+message FilteringRules {
+  repeated UrlRule url_rules = 1;
+  repeated CssRule css_rules = 2;
+}