Issue 3143 - Filter elements with :-abp-has()

chrome/content/elemHideEmulation.js

 /*
  * This file is part of Adblock Plus <https://adblockplus.org/>,
  * Copyright (C) 2006-2017 eyeo GmbH
  *
  * Adblock Plus is free software: you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 3 as
  * published by the Free Software Foundation.
  *
  * Adblock Plus is distributed in the hope that it will be useful,
  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  * GNU General Public License for more details.
  *
  * You should have received a copy of the GNU General Public License
  * along with Adblock Plus.  If not, see <http://www.gnu.org/licenses/>.
  */
 
 /* globals filterToRegExp */
 
 "use strict";
 
+const abpSelectorRegexp = /:-abp-([\w-]+)\(/i;
+
+let reportError = () => {};
+
 function splitSelector(selector)
 {
   if (selector.indexOf(",") == -1)
     return [selector];
 
   let selectors = [];
   let start = 0;
   let level = 0;
   let sep = "";
 
@@ skipping 18 matching lines @@
         selectors.push(selector.substring(start, i));
         start = i + 1;
       }
     }
   }
 
   selectors.push(selector.substring(start));
   return selectors;
 }
 
-// Return position of node from parent.
-// 1 base index like for :nth-child()
+/** Return position of node from parent.
+ * @param {Node} node the node to find the position of.
+ * @return {number} One-based index like for :nth-child(), or 0 on error.
+ */
 function positionInParent(node)
 {
-  let parentNode = node ? node.parentNode : null;
-  if (parentNode == null)
-    return 0;
-
-  let {children} = parentNode;
-  if (!children)
-    return 0;
-  let i = 0;
-  for (i = 0; i < children.length; i++)
+  let {children} = node.parentNode;
+  for (let i = 0; i < children.length; i++)
     if (children[i] == node)
-      break;
-  return i + 1;
+      return i + 1;
+  return 0;
 }
 
 function makeSelector(node, selector)
 {
-  if (node && node.id && node.id != "")

[hub, 2017/05/12 17:31:18]

I need to check that node.id is a valid CSS identifier. I have seen cases in
real life where the id start with a number causing querySelector() to fail. I
already have a patch for that. Will update review with it.

-  {
-    let newSelector = "#" + node.id;
-    if (selector != "")
-      newSelector += " > ";
-    return newSelector + selector;
+  if (!node.parentElement)
+  {
+    let newSelector = ":root";
+    if (selector)
+      newSelector += " > " + selector;
+    return newSelector;
   }
   let idx = positionInParent(node);
   if (idx > 0)
   {
-    let newSelector = `${node.tagName}:nth-child(${idx}) `;
-    if (selector != "")
-      newSelector += "> ";
-    return makeSelector(node.parentNode, newSelector + selector);
+    let newSelector = `${node.tagName}:nth-child(${idx})`;
+    if (selector)
+      newSelector += " > " + selector;
+    return makeSelector(node.parentElement, newSelector);
   }
 
   return selector;
 }
 
-// return the regexString for the properties
-function parsePropSelPattern(propertyExpression)
+function parseSelectorContent(content, startIndex)
+{
+  let parens = 1;
+  let quote = null;
+  let i = startIndex;
+  for (; i < content.length; i++)
+  {
+    let c = content[i];
+    if (c == "\\")
+    {
+      // Ignore escaped characters
+      i++;
+    }
+    else if (quote)
+    {
+      if (c == quote)
+        quote = null;
+    }
+    else if (c == "'" || c == '"')
+      quote = c;
+    else if (c == "(")
+      parens++;
+    else if (c == ")")
+    {
+      parens--;
+      if (parens == 0)
+        break;
+    }
+  }
+
+  if (parens > 0)
+    return null;
+  return {text: content.substring(startIndex, i), end: i};
+}
+
+/** Parse the selector
+ * @param {string} selector the selector to parse
+ * @return {Object} selectors is an array of objects,
+ * or null in case of errors. hide is true if we'll hide
+ * elements instead of styles..
+ */
+function parseSelector(selector)
+{
+  if (selector.length == 0)
+    return [];
+
+  let match = abpSelectorRegexp.exec(selector);
+  if (!match)
+    return [new PlainSelector(selector)];
+
+  let selectors = [];
+  if (match.index > 0)
+    selectors.push(new PlainSelector(selector.substr(0, match.index)));
+
+  let startIndex = match.index + match[0].length;
+  let content = parseSelectorContent(selector, startIndex);
+  if (!content)
+  {
+    reportError(new SyntaxError("Failed to parse Adblock Plus " +
+                                `selector ${selector}, ` +
+                                "due to unmatched parentheses."));
+    return null;
+  }
+  if (match[1] == "properties")
+    selectors.push(new PropsSelector(content.text));
+  else if (match[1] == "has")
+  {
+    let hasSelector = new HasSelector(content.text);
+    if (!hasSelector.valid())
+      return null;
+    selectors.push(hasSelector);
+  }
+  else
+  {
+    // this is an error, can't parse selector.
+    reportError(new SyntaxError("Failed to parse Adblock Plus " +
+                                `selector ${selector}, invalid ` +
+                                `pseudo-class :-abp-${match[1]}().`));
+    return null;
+  }
+
+  let suffix = parseSelector(selector.substr(content.end + 1));
+  if (suffix == null)
+    return null;
+
+  selectors.push(...suffix);
+
+  return selectors;
+}
+
+/** Stringified style objects
+ * @typedef {Object} StringifiedStyle
+ * @property {string} style CSS style represented by a string.
+ * @property {string[]} subSelectors selectors the CSS properties apply to.
+ */
+
+/**
+ * Produce a string representation of the stylesheet entry.
+ * @param {CSSStyleRule} rule the CSS style rule.
+ * @return {StringifiedStyle} the stringified style.
+ */
+function stringifyStyle(rule)
+{
+  let styles = [];
+  for (let i = 0; i < rule.style.length; i++)
+  {
+    let property = rule.style.item(i);
+    let value = rule.style.getPropertyValue(property);
+    let priority = rule.style.getPropertyPriority(property);
+    styles.push(`${property}: ${value}${priority ? " !" + priority : ""};`);
+  }
+  styles.sort();
+  return {
+    style: styles.join(" "),
+    subSelectors: splitSelector(rule.selectorText)
+  };
+}
+
+function* evaluate(chain, index, prefix, subtree, styles)
+{
+  if (index >= chain.length)
+  {
+    yield prefix;
+    return;
+  }
+  for (let [selector, element] of
+       chain[index].getSelectors(prefix, subtree, styles))
+    yield* evaluate(chain, index + 1, selector, element, styles);
+}
+
+function PlainSelector(selector)
+{
+  this._selector = selector;
+}
+
+PlainSelector.prototype = {
+  /**
+   * Generator function returning a pair of selector
+   * string and subtree.
+   * @param {string} prefix the prefix for the selector.
+   * @param {Node} subtree the subtree we work on.
+   * @param {StringifiedStyle[]} styles the stringified style objects.
+   */
+  *getSelectors(prefix, subtree, styles)
+  {
+    yield [prefix + this._selector, subtree];
+  }
+};
+
+const incompletePrefixRegexp = /[\s>+~]$/;
+
+function HasSelector(selector)
+{
+  this._innerSelectors = parseSelector(selector);
+}
+
+HasSelector.prototype = {
+  requiresHiding: true,
+
+  valid()
+  {
+    return this._innerSelectors != null;
+  },
+
+  *getSelectors(prefix, subtree, styles)
+  {
+    for (let element of this.getElements(prefix, subtree, styles))
+      yield [makeSelector(element, ""), element];
+  },
+
+  /**
+   * Generator function returning selected elements.
+   * @param {string} prefix the prefix for the selector.
+   * @param {Node} subtree the subtree we work on.
+   * @param {StringifiedStyle[]} styles the stringified style objects.
+   */
+  *getElements(prefix, subtree, styles)
+  {
+    let actualPrefix = (!prefix || incompletePrefixRegexp.test(prefix)) ?
+        prefix + "*" : prefix;
+    let elements = subtree.querySelectorAll(actualPrefix);
+    for (let element of elements)
+    {
+      let newPrefix = makeSelector(element, "");
+      let iter = evaluate(this._innerSelectors, 0, newPrefix + " ",
+                          element, styles);
+      for (let selector of iter)
+        // we insert a space between the two. It becomes a no-op if selector
+        // doesn't have a combinator
+        if (subtree.querySelector(selector))
+          yield element;
+    }
+  }
+};
+
+function PropsSelector(propertyExpression)
 {
   let regexpString;
   if (propertyExpression.length >= 2 && propertyExpression[0] == "/" &&
       propertyExpression[propertyExpression.length - 1] == "/")
+  {
     regexpString = propertyExpression.slice(1, -1)
-    .replace("\\x7B ", "{").replace("\\x7D ", "}");
+      .replace("\\x7B ", "{").replace("\\x7D ", "}");
+  }
   else
     regexpString = filterToRegExp(propertyExpression);
-  return regexpString;
-}
-
-function parseSelector(selector)
-{
-  if (selector.length == 0)
-    return [];
-
-  let abpSelectorIndex = selector.indexOf(":-abp-");
-  if (abpSelectorIndex == -1)
-    return [new PlainSelector(selector)];
-
-  let selectors = [];
-  if (abpSelectorIndex > 0)
-    selectors.push(new PlainSelector(selector.substr(0, abpSelectorIndex)));
-
-  let suffixStart = abpSelectorIndex;
-
-  if (selector.indexOf(":-abp-properties(", abpSelectorIndex) ==
-      abpSelectorIndex)
-  {
-    let startIndex = abpSelectorIndex + 17;
-    let endquoteIndex = selector.indexOf(selector[startIndex], startIndex + 1);
-    if ((endquoteIndex == -1) || (selector[endquoteIndex + 1] != ")"))
-      return null;
-
-    selectors.push(new PropsSelector(
-      selector.substr(startIndex + 1, endquoteIndex - startIndex - 1)));
-    suffixStart = endquoteIndex + 2;
-  }
-  else if (selector.indexOf(":-abp-has(", abpSelectorIndex) ==
-      abpSelectorIndex)
-  {
-    let startIndex = abpSelectorIndex + 10;
-    let parens = 1;
-    let i;
-    for (i = startIndex; i < selector.length; i++)
-    {
-      if (selector[i] == "(")
-        parens++;
-      else if (selector[i] == ")")
-        parens--;
-
-      if (parens == 0)
-        break;
-    }
-    if (parens != 0)
-      return null;
-    selectors.push(new HasSelector(
-      selector.substr(startIndex, i - startIndex)));
-    suffixStart = i + 1;
-  }
-
-  let suffix = parseSelector(selector.substr(suffixStart));

[hub, 2017/05/12 19:48:24]

here, if we had been parsing something matching "-abp-*" then we'd infinite
recursion. I do have a patch too.

-  if (suffix)
-    selectors.push(...suffix);
-
-  return selectors;
-}
-
-function matchStyleProps(style, rule, pattern, selectors, filters)
-{
-  if (pattern.regexp.test(style))
-  {
-    let subSelectors = splitSelector(rule.selectorText);
-    for (let i = 0; i < subSelectors.length; i++)
-    {
-      let subSelector = subSelectors[i];
-      selectors.push(pattern.prefix + subSelector + pattern.suffix);
-      filters.push(pattern.text);
-    }
-  }
-}
-
-function findPropsSelectors(stylesheet, pattern, selectors, filters)
-{
-  let rules = stylesheet.cssRules;
-  if (!rules)
-    return;
-
-  for (let rule of rules)
-  {
-    if (rule.type != rule.STYLE_RULE)
-      continue;
-
-    let style = stringifyStyle(rule.style);
-    matchStyleProps(style, rule, pattern, selectors, filters);
-  }
-}
-
-function stringifyStyle(style)
-{
-  let styles = [];
-  for (let i = 0; i < style.length; i++)
-  {
-    let property = style.item(i);
-    let value = style.getPropertyValue(property);
-    let priority = style.getPropertyPriority(property);
-    styles.push(property + ": " + value + (priority ? " !" + priority : "") +
-                ";");
-  }
-  styles.sort();
-  return styles.join(" ");
-}
-
-function* evaluate(chain, index, prefix, subtree, stylesheet)
-{
-  if (index >= chain.length)
-  {
-    yield prefix;
-    return;
-  }
-  for (let [selector, element] of
-       chain[index].getSelectors(prefix, subtree, stylesheet))
-    yield* evaluate(chain, index + 1, selector, element, stylesheet);
-}
-
-/*
- * getSelector() is a generator function returning a pair of selector
- * string and subtree.
- * getElements() is a generator function returning elements selected.
- */
-function PlainSelector(selector)
-{
-  this._selector = selector;
-}
-
-PlainSelector.prototype = {
-  *getSelectors(prefix, subtree, stylesheet)
-  {
-    yield [prefix + this._selector, subtree];
-  },
-
-  *getElements(prefix, subtree, stylesheet)
-  {
-    for (let selector of this.getSelectors(prefix, subtree, stylesheet))
-      for (let element of subtree.querySelectorAll(selector[0]))
-        yield element;
+
+  this._regexp = new RegExp(regexpString, "i");
+}
+
+PropsSelector.prototype = {
+  *findPropsSelectors(styles, prefix, regexp)
+  {
+    for (let style of styles)
+      if (regexp.test(style.style))
+        for (let subSelector of style.subSelectors)
+          yield prefix + subSelector;
+  },
+
+  *getSelectors(prefix, subtree, styles)
+  {
+    for (let selector of this.findPropsSelectors(styles, prefix, this._regexp))
+      yield [selector, subtree];
   }
 };
 
-function HasSelector(selector)
-{
-  this._innerSelectors = parseSelector(selector);
-}
-
-HasSelector.prototype = {
-  *getSelectors(prefix, subtree, stylesheet)
-  {
-    for (let element of this.getElements(prefix, subtree, stylesheet))
-      yield [prefix + makeSelector(element, ""), subtree];
-  },
-
-  *getElements(prefix, subtree, stylesheet)
-  {
-    let elements = subtree.querySelectorAll(prefix ? prefix : "*");
-    for (let element of elements)
-    {
-      let newPrefix = makeSelector(element, "");
-      let iter = evaluate(this._innerSelectors, 0, "", element, stylesheet);
-      for (let selector of iter)
-        // we insert a space between the two. It becomes a no-op if selector
-        // doesn't have a combinator
-        if (subtree.querySelector(newPrefix + " " + selector))
-          yield element;
-    }
-  }
-};
-
-function PropsSelector(selector)
-{
-  this._regexp = new RegExp(parsePropSelPattern(selector), "i");
-}
-
-PropsSelector.prototype = {
-  *getSelectors(prefix, subtree, stylesheet)
-  {
-    let selectors = [];
-    let filters = [];
-    let selPattern = {
-      prefix,
-      suffix: "",
-      regexp: this._regexp
-    };
-
-    findPropsSelectors(stylesheet, selPattern, selectors, filters);
-    for (let selector of selectors)
-      yield [selector, subtree];
-  },
-
-  *getElements(prefix, subtree, stylesheet)
-  {
-    for (let [selector, element] of
-         this.getSelectors(prefix, subtree, stylesheet))
-      for (let subElement of element.querySelectorAll(selector))
-        yield subElement;
-  }
-};
-
-function ElemHideEmulation(window, getFiltersFunc, addSelectorsFunc)
+function ElemHideEmulation(window, getFiltersFunc, addSelectorsFunc,
+                           hideElemsFunc)
 {
   this.window = window;
   this.getFiltersFunc = getFiltersFunc;
   this.addSelectorsFunc = addSelectorsFunc;
+  this.hideElemsFunc = hideElemsFunc;
 }
 
 ElemHideEmulation.prototype = {
-
   isSameOrigin(stylesheet)
   {
     try
     {
       return new URL(stylesheet.href).origin == this.window.location.origin;
     }
     catch (e)
     {
       // Invalid URL, assume that it is first-party.
       return true;
     }
   },
 
-  addSelectors(stylesheet)
+  addSelectors(stylesheets)
   {
     let selectors = [];
-    let filters = [];
-
-    // Explicitly ignore third-party stylesheets to ensure consistent behavior
-    // between Firefox and Chrome.
-    if (!this.isSameOrigin(stylesheet))
-      return;
-
-    for (let patterns of this.selPatterns)
-      for (let selector of
-           evaluate(patterns.selectors, 0, "", document, stylesheet))
+    let selectorFilters = [];
+
+    let elements = [];
+    let elementFilters = [];
+
+    let cssStyles = [];
+
+    for (let stylesheet of stylesheets)
+    {
+      // Explicitly ignore third-party stylesheets to ensure consistent behavior
+      // between Firefox and Chrome.
+      if (!this.isSameOrigin(stylesheet))
+        continue;
+
+      let rules = stylesheet.cssRules;
+      if (!rules)
+        continue;
+
+      for (let rule of rules)
       {
-        selectors.push(selector);
-        filters.push(patterns.text);
+        if (rule.type != rule.STYLE_RULE)
+          continue;
+
+        cssStyles.push(stringifyStyle(rule));
       }
-
-    this.addSelectorsFunc(selectors, filters);
+    }
+
+    let {document} = this.window;
+    for (let pattern of this.patterns)
+    {
+      for (let selector of evaluate(pattern.selectors,
+                                    0, "", document, cssStyles))
+      {
+        if (!pattern.selectors.some(s => s.requiresHiding))
+        {
+          selectors.push(selector);
+          selectorFilters.push(pattern.text);
+        }
+        else
+        {
+          for (let element of document.querySelectorAll(selector))
+          {
+            elements.push(element);
+            elementFilters.push(pattern.text);
+          }
+        }
+      }
+    }
+
+    this.addSelectorsFunc(selectors, selectorFilters);
+    this.hideElemsFunc(elements, elementFilters);
   },
 
   onLoad(event)
   {
     let stylesheet = event.target.sheet;
     if (stylesheet)
-      this.addSelectors(stylesheet);
+      this.addSelectors([stylesheet]);
   },
 
   apply()
   {
     this.getFiltersFunc(patterns =>
     {
-      this.selPatterns = [];
-
+      let oldReportError = reportError;
+      reportError = error => this.window.console.error(error);
+
+      this.patterns = [];
       for (let pattern of patterns)
       {
         let selectors = parseSelector(pattern.selector);
         if (selectors != null && selectors.length > 0)
-          this.selPatterns.push({selectors, text: pattern.text});
+          this.patterns.push({selectors, text: pattern.text});
       }
 
-      if (this.selPatterns.length > 0)
+      if (this.patterns.length > 0)
       {
         let {document} = this.window;
-        for (let stylesheet of document.styleSheets)
-          this.addSelectors(stylesheet);
+        this.addSelectors(document.styleSheets);
         document.addEventListener("load", this.onLoad.bind(this), true);
       }
+      reportError = oldReportError;
     });
   }
 };