Sync for validator cpp engine and cpp htmlparser (#37510)

* Add C++ tests on whether validator rules make sense

PiperOrigin-RevId: 423415201

* Remove allowlist of tags which have no descriptive name

PiperOrigin-RevId: 423430953

* Fix a bug by no longer using string_view of block scope string

PiperOrigin-RevId: 423900419

* Test whether attribute rule makes sense

PiperOrigin-RevId: 424432911

* Test whether satisfies match up with requires and excludes

PiperOrigin-RevId: 424454520

* Test whether error codes are unique

PiperOrigin-RevId: 424454591

* Fix compile errors in github unit test

PiperOrigin-RevId: 424701686

Co-authored-by: Boxiao Cao <caoboxiao@google.com>

validator/cpp/engine/BUILD

@@ -294,6 +294,7 @@ cc_library(
     copts = ["-std=c++17"],
     deps = [
         ":validator-internal",
+        "//cpp/htmlparser/css:parse-css",
         "//:validator_cc_proto",
     ],
     alwayslink = 1,
@@ -329,6 +330,8 @@ cc_test(
         ":testing-utils",
         ":validator",
         ":validator_pb",
+        "@com_google_absl//absl/container:flat_hash_map",
+        "@com_google_absl//absl/container:flat_hash_set",
         "@com_google_absl//absl/strings",
         "@com_google_googletest//:gtest_main",
         "//cpp/htmlparser/css:parse_css_cc_proto",

validator/cpp/engine/validator-internal.cc

@@ -127,6 +127,22 @@ static const LazyRE2 kRuntimeScriptPathRe = {
 static const LazyRE2 kExtensionPathRe = {
     R"re((?:lts/)?v0/(amp-[a-z0-9-]*)-([a-z0-9.]*)\.(?:m)?js(?:\?f=sxg)?)re"};
 
+// Generates a htmlparser::css::CssParsingConfig.
+CssParsingConfig GenCssParsingConfig() {
+  CssParsingConfig config;
+  // If other @ rule types are added to the rules, their block parsing types
+  // will need to be added here as well.
+  config.at_rule_spec["font-face"] = BlockType::PARSE_AS_DECLARATIONS;
+  config.at_rule_spec["keyframes"] = BlockType::PARSE_AS_RULES;
+  config.at_rule_spec["media"] = BlockType::PARSE_AS_RULES;
+  config.at_rule_spec["page"] = BlockType::PARSE_AS_DECLARATIONS;
+  config.at_rule_spec["supports"] = BlockType::PARSE_AS_RULES;
+  config.at_rule_spec["-moz-document"] = BlockType::PARSE_AS_RULES;
+  // Note that ignoring still generates an error.
+  config.default_spec = BlockType::PARSE_AS_IGNORE;
+  return config;
+}
+
 namespace {
 
 #define CHECK_NOTNULL(x) (x)
@@ -845,22 +861,6 @@ class ParsedAttrSpecs {
   unordered_map<std::string, vector<const ParsedAttrSpec*>> attr_lists_by_name_;
 };
 
-// Generates a htmlparser::css::CssParsingConfig.
-CssParsingConfig GenCssParsingConfig() {
-  CssParsingConfig config;
-  // If other @ rule types are added to the rules, their block parsing types
-  // will need to be added here as well.
-  config.at_rule_spec["font-face"] = BlockType::PARSE_AS_DECLARATIONS;
-  config.at_rule_spec["keyframes"] = BlockType::PARSE_AS_RULES;
-  config.at_rule_spec["media"] = BlockType::PARSE_AS_RULES;
-  config.at_rule_spec["page"] = BlockType::PARSE_AS_DECLARATIONS;
-  config.at_rule_spec["supports"] = BlockType::PARSE_AS_RULES;
-  config.at_rule_spec["-moz-document"] = BlockType::PARSE_AS_RULES;
-  // Note that ignoring still generates an error.
-  config.default_spec = BlockType::PARSE_AS_IGNORE;
-  return config;
-}
-
 // Instances of this class precompute the regular expressions for a particular
 // Cdata specification.
 class ParsedCdataSpec {

validator/cpp/engine/validator.h

@@ -23,6 +23,7 @@
 #ifndef CPP_ENGINE_VALIDATOR_H_
 #define CPP_ENGINE_VALIDATOR_H_
 
+#include "cpp/htmlparser/css/parse-css.h"
 #include "cpp/htmlparser/document.h"
 #include "validator.pb.h"
 
@@ -38,6 +39,7 @@ ValidationResult Validate(std::string_view html,
 
 int RulesSpecVersion();
 int ValidatorVersion();
+htmlparser::css::CssParsingConfig GenCssParsingConfig();
 
 }  // namespace amp::validator