From ee6eee131b95cb430d548a5ed6cee3534b7077ff Mon Sep 17 00:00:00 2001
From: Shivan Kaul Sahib <shivankaulsahib@gmail.com>
Date: Thu, 18 Aug 2022 19:54:25 -0700
Subject: [PATCH] Allow multiple capture groups for debounce regex action
 (#14687)

---
 components/debounce/browser/debounce_rule.cc  | 31 +++++++--
 .../browser/test/debounce_rule_unittest.cc    | 68 +++++++++++++++++++
 2 files changed, 94 insertions(+), 5 deletions(-)

diff --git a/components/debounce/browser/debounce_rule.cc b/components/debounce/browser/debounce_rule.cc
index bf7bb18b60f..e57d3e1fd0c 100644
--- a/components/debounce/browser/debounce_rule.cc
+++ b/components/debounce/browser/debounce_rule.cc
@@ -187,17 +187,36 @@ bool DebounceRule::ValidateAndParsePatternRegex(
             << " which is an invalid regex pattern";
     return false;
   }
-  if (pattern_regex.NumberOfCapturingGroups() != 1) {
+  if (pattern_regex.NumberOfCapturingGroups() < 1) {
     VLOG(1) << "Debounce rule has param: " << pattern
-            << " which captures != 1 groups";
+            << " which captures < 1 groups";
     return false;
   }
 
-  if (!RE2::PartialMatch(path, pattern_regex, parsed_value)) {
+  // Get matching capture groups by applying regex to the path
+  size_t number_of_capturing_groups =
+      pattern_regex.NumberOfCapturingGroups() + 1;
+  std::vector<re2::StringPiece> match_results(number_of_capturing_groups);
+
+  if (!pattern_regex.Match(path, 0, path.size(), RE2::UNANCHORED,
+                           match_results.data(), match_results.size())) {
     VLOG(1) << "Debounce rule with param: " << param_
             << " was unable to capture string";
     return false;
   }
+
+  // This will always be at least 2: the first one is the full match
+  DCHECK_GT(match_results.size(), 1u);
+
+  // Build parsed_value string by appending matches, ignoring the first match
+  // which will be the whole match
+  std::for_each(std::begin(match_results) + 1, std::end(match_results),
+                [parsed_value](re2::StringPiece matched_string) {
+                  if (!matched_string.empty()) {
+                    matched_string.AppendToString(parsed_value);
+                  }
+                });
+
   return true;
 }
 
@@ -212,12 +231,14 @@ bool DebounceRule::Apply(const GURL& original_url,
       action_ != kDebounceRegexPath)
     return false;
   // If URL matches an explicitly excluded pattern, this rule does not apply.
-  if (exclude_pattern_set_.MatchesURL(original_url))
+  if (exclude_pattern_set_.MatchesURL(original_url)) {
     return false;
+  }
   // If URL does not match an explicitly included pattern, this rule does not
   // apply.
-  if (!include_pattern_set_.MatchesURL(original_url))
+  if (!include_pattern_set_.MatchesURL(original_url)) {
     return false;
+  }
 
   if (!DebounceRule::CheckPrefForRule(prefs)) {
     return false;
diff --git a/components/debounce/browser/test/debounce_rule_unittest.cc b/components/debounce/browser/test/debounce_rule_unittest.cc
index 59448c8df17..4ba28178c80 100644
--- a/components/debounce/browser/test/debounce_rule_unittest.cc
+++ b/components/debounce/browser/test/debounce_rule_unittest.cc
@@ -183,6 +183,74 @@ TEST(DebounceRuleUnitTest, ParamCapturesNonURLWithPrependScheme) {
   }
 }
 
+TEST(DebounceRuleUnitTest, TwoCaptureGroups) {
+  const std::string contents = R"json(
+
+      [{
+          "include": [
+              "*://test.com/*"
+          ],
+          "exclude": [
+          ],
+          "action": "regex-path",
+          "prepend_scheme": "https",
+          "param": "^/([^/]+)/xyz(/.*)$"
+      }]
+      
+      )json";
+  std::vector<std::unique_ptr<DebounceRule>> rules = StringToRules(contents);
+
+  for (const std::unique_ptr<DebounceRule>& rule : rules) {
+    CheckApplyResult(rule.get(), GURL("https://test.com/brave.com/xyz/abc.jpg"),
+                     "https://brave.com/abc.jpg", false);
+  }
+}
+
+TEST(DebounceRuleUnitTest, CurlyBracesInRegexGetParseError) {
+  const std::string contents = R"json(
+
+      [{
+          "include": [
+              "*://test.com/*"
+          ],
+          "exclude": [
+          ],
+          "action": "regex-path",
+          "prepend_scheme": "https",
+          "param": "^/turbo/([^/]+)/xyz(/\d{4})/xyzzy(/.*)$"
+      }]
+      
+      )json";
+
+  auto parsed = DebounceRule::ParseRules(contents);
+  EXPECT_FALSE(parsed.has_value());
+}
+
+TEST(DebounceRuleUnitTest, ThreeCaptureGroups) {
+  const std::string contents = R"json(
+
+      [{
+          "include": [
+              "*://test.com/*"
+          ],
+          "exclude": [
+          ],
+          "action": "regex-path",
+          "prepend_scheme": "https",
+          "param": "^/turbo/([^/]+)/xyz(/[0-9]+)/xyzzy(/.*)$"
+      }]
+      
+      )json";
+  std::vector<std::unique_ptr<DebounceRule>> rules = StringToRules(contents);
+
+  for (const std::unique_ptr<DebounceRule>& rule : rules) {
+    CheckApplyResult(
+        rule.get(),
+        GURL("https://test.com/turbo/brave.com/xyz/2022/xyzzy/abc.jpg"),
+        "https://brave.com/2022/abc.jpg", false);
+  }
+}
+
 TEST(DebounceRuleUnitTest, ParamCapturesURLWithPrependScheme) {
   const std::string contents = R"json(