better splitting of selectors (rrweb-io#1440)

* better splitting of selectors - overlapping with rrweb-io#1401 * Add test from example at PostHog/posthog#21427 * ignore brackets inside selector strings * Add another test as noticed that it's possible to escape strings * Ensure we are ignoring commas within strings Co-authored-by: Eoghan Murray <[email protected]>
getsentry · Apr 19, 2024 · 8053361 · 8053361
1 parent 928280e
commit 8053361
Show file tree

Hide file tree

Showing 3 changed files with 183 additions and 5 deletions.
diff --git a/.changeset/modern-doors-watch.md b/.changeset/modern-doors-watch.md
@@ -0,0 +1,5 @@
+---
+'rrweb-snapshot': patch
+---
+
+better nested css selector splitting when commas or brackets happen to be in quoted text
diff --git a/packages/rrweb-snapshot/src/css.ts b/packages/rrweb-snapshot/src/css.ts
@@ -435,17 +435,67 @@ export function parse(css: string, options: ParserOptions = {}) {
     if (!m) {
       return;
     }
+
     /* @fix Remove all comments from selectors
      * http://ostermiller.org/findcomment.html */
-    return trim(m[0])
+    const cleanedInput = m[0]
+      .trim()
       .replace(/\/\*([^*]|[\r\n]|(\*+([^*/]|[\r\n])))*\*\/+/g, '')
+      // Handle strings by replacing commas inside them
       .replace(/"(?:\\"|[^"])*"|'(?:\\'|[^'])*'/g, (m) => {
         return m.replace(/,/g, '\u200C');
-      })
-      .split(/\s*(?![^(]*\)),\s*/)
-      .map((s) => {
-        return s.replace(/\u200C/g, ',');
       });
+
+    // Split using a custom function and restore commas in strings
+    return customSplit(cleanedInput).map((s) =>
+      s.replace(/\u200C/g, ',').trim(),
+    );
+  }
+
+  /**
+   * Split selector correctly, ensuring not to split on comma if inside ().
+   */
+
+  function customSplit(input: string) {
+    const result = [];
+    let currentSegment = '';
+    let depthParentheses = 0; // Track depth of parentheses
+    let depthBrackets = 0; // Track depth of square brackets
+    let currentStringChar = null;
+
+    for (const char of input) {
+      const hasStringEscape = currentSegment.endsWith('\\');
+
+      if (currentStringChar) {
+        if (currentStringChar === char && !hasStringEscape) {
+          currentStringChar = null;
+        }
+      } else if (char === '(') {
+        depthParentheses++;
+      } else if (char === ')') {
+        depthParentheses--;
+      } else if (char === '[') {
+        depthBrackets++;
+      } else if (char === ']') {
+        depthBrackets--;
+      } else if ('\'"'.includes(char)) {
+        currentStringChar = char;
+      }
+
+      // Split point is a comma that is not inside parentheses or square brackets
+      if (char === ',' && depthParentheses === 0 && depthBrackets === 0) {
+        result.push(currentSegment);
+        currentSegment = '';
+      } else {
+        currentSegment += char;
+      }
+    }
+
+    // Add the last segment
+    if (currentSegment) {
+      result.push(currentSegment);
+    }
+    return result;
   }
 
   /**

diff --git a/packages/rrweb-snapshot/test/css.test.ts b/packages/rrweb-snapshot/test/css.test.ts
@@ -119,6 +119,129 @@ describe('css parser', () => {
     expect(out3).toEqual('[data-aa\\:other] { color: red; }');
   });
 
+  it('parses nested commas in selectors correctly', () => {
+    const result = parse(
+      `
+body > ul :is(li:not(:first-of-type) a:hover, li:not(:first-of-type).active a) {
+  background: red;
+}
+`,
+    );
+    expect((result.stylesheet!.rules[0] as Rule)!.selectors!.length).toEqual(1);
+
+    const trickresult = parse(
+      `
+li[attr="weirdly("] a:hover, li[attr="weirdly)"] a {
+  background-color: red;
+}
+`,
+    );
+    expect(
+      (trickresult.stylesheet!.rules[0] as Rule)!.selectors!.length,
+    ).toEqual(2);
+
+    const weirderresult = parse(
+      `
+li[attr="weirder\\"("] a:hover, li[attr="weirder\\")"] a {
+  background-color: red;
+}
+`,
+    );
+    expect(
+      (weirderresult.stylesheet!.rules[0] as Rule)!.selectors!.length,
+    ).toEqual(2);
+
+    const commainstrresult = parse(
+      `
+li[attr="has,comma"] a:hover {
+  background-color: red;
+}
+`,
+    );
+    expect(
+      (commainstrresult.stylesheet!.rules[0] as Rule)!.selectors!.length,
+    ).toEqual(1);
+  });
+
+  it.each([
+    ['.foo,.bar {}', ['.foo', '.bar']],
+    ['.bar:has(:disabled) {}', ['.bar:has(:disabled)']],
+    ['.bar:has(input, button) {}', ['.bar:has(input, button)']],
+    [
+      '.bar:has(input:is(:disabled),button:has(:disabled)) {}',
+      ['.bar:has(input:is(:disabled),button:has(:disabled))'],
+    ],
+    [
+      '.bar:has(div, input:is(:disabled), button) {}',
+      ['.bar:has(div, input:is(:disabled), button)'],
+    ],
+    [
+      '.bar:has(div, input:is(:disabled),button:has(:disabled,.baz)) {}',
+      ['.bar:has(div, input:is(:disabled),button:has(:disabled,.baz))'],
+    ],
+    [
+      '.bar:has(input), .foo:has(input, button), .baz {}',
+      ['.bar:has(input)', '.foo:has(input, button)', '.baz'],
+    ],
+    [
+      '.bar:has(input:is(:disabled),button:has(:disabled,.baz), div:has(:disabled,.baz)){color: red;}',
+      [
+        '.bar:has(input:is(:disabled),button:has(:disabled,.baz), div:has(:disabled,.baz))',
+      ],
+    ],
+    [
+      '.bar:has(:has(:has(a), :has(:has(:has(b, :has(a), c), e))), input:is(:disabled), button) {}',
+      [
+        '.bar:has(:has(:has(a), :has(:has(:has(b, :has(a), c), e))), input:is(:disabled), button)',
+      ],
+    ],
+    [
+      '.foo,.bar:has(input:is(:disabled)){color: red;}',
+      ['.foo', '.bar:has(input:is(:disabled))'],
+    ],
+    [
+      '.foo,.bar:has(input:is(:disabled),button:has(:disabled,.baz)){color: red;}',
+      ['.foo', '.bar:has(input:is(:disabled),button:has(:disabled,.baz))'],
+    ],
+    [
+      '.foo,.bar:has(input:is(:disabled),button:has(:disabled), div:has(:disabled,.baz)){color: red;}',
+      [
+        '.foo',
+        '.bar:has(input:is(:disabled),button:has(:disabled), div:has(:disabled,.baz))',
+      ],
+    ],
+    [
+      '.foo,.bar:has(input:is(:disabled),button:has(:disabled,.baz), div:has(:disabled,.baz)){color: red;}',
+      [
+        '.foo',
+        '.bar:has(input:is(:disabled),button:has(:disabled,.baz), div:has(:disabled,.baz))',
+      ],
+    ],
+    ['.bar:has(:disabled), .foo {}', ['.bar:has(:disabled)', '.foo']],
+    [
+      '.bar:has(input:is(:disabled),.foo,button:is(:disabled)), .foo {}',
+      ['.bar:has(input:is(:disabled),.foo,button:is(:disabled))', '.foo'],
+    ],
+    [
+      '.bar:has(input:is(:disabled),.foo,button:is(:disabled)), .foo:has(input, button), .baz,  {}',
+      [
+        '.bar:has(input:is(:disabled),.foo,button:is(:disabled))',
+        '.foo:has(input, button)',
+        '.baz',
+      ],
+    ],
+  ])(
+    'can parse selector(s) with functional pseudo classes: %s',
+    (cssText, expected) => {
+      expect(
+        parse(
+          cssText,
+          // @ts-ignore
+        ).stylesheet?.rules[0].selectors,
+      ).toEqual(expected);
+    },
+  );
+
   it('parses imports with quotes correctly', () => {
     const out1 = escapeImportStatement({
       cssText: `@import url("/foo.css;900;800"");`,