From f7ed63179fa4f99322d6c7716e17466ec4e3e4ce Mon Sep 17 00:00:00 2001 From: Aqua-sama Date: Sat, 23 Dec 2017 16:45:31 +0100 Subject: Request filter now properly takes hostlists - hostslist directory is set in browser.filterPath --- src/formats/adblockrule.cpp | 198 -------------------------------------------- src/formats/adblockrule.h | 71 ---------------- 2 files changed, 269 deletions(-) delete mode 100644 src/formats/adblockrule.cpp delete mode 100644 src/formats/adblockrule.h (limited to 'src/formats') diff --git a/src/formats/adblockrule.cpp b/src/formats/adblockrule.cpp deleted file mode 100644 index fa9d93a..0000000 --- a/src/formats/adblockrule.cpp +++ /dev/null @@ -1,198 +0,0 @@ -/* - * This file is part of smolbote. It's copyrighted by the contributors recorded - * in the version control history of the file, available from its original - * location: git://neueland.iserlohn-fortress.net/smolbote.git - * - * SPDX-License-Identifier: GPL-3.0 - */ - -#include "blockerrule.h" - -/* AdBlock filter reference - * https://adblockplus.org/en/filters - * https://adblockplus.org/en/filter-cheatsheet - */ - -BlockerRule::BlockerRule(QString rule, QObject *parent) : - QObject(parent) -{ - m_filter = rule; - QString pattern = rule; - - // Empty rule or comment - if(pattern.trimmed().isEmpty() || pattern.startsWith("!")) { - m_valid = false; - return; - } - - // Exception - if(pattern.startsWith("@@")) { - m_exception = true; - pattern.remove(0, 2); - } - - // Ignore element hiding rules for now - if(pattern.contains("##") || pattern.contains("#@#")) { - m_valid = false; - return; - } - - // Options - if(pattern.contains("$")) { - QString opts = pattern.mid(pattern.indexOf("$")+1); - pattern.remove(pattern.indexOf("$"), pattern.length()); - - const QStringList optList = opts.split(','); - QStringList::const_iterator i; - for(i = optList.constBegin(); i != optList.constEnd(); ++i) { - parseOption(*i); - } - } - - // Regular expression - if(rule.startsWith("/") && rule.endsWith("/")) { - m_valid = true; - ruleExpression.setPattern(pattern); - return; - } - - // Domain rules - if(pattern.startsWith("||")) { - pattern.remove(0, 2); - // find the end point for the domain - int end = pattern.indexOf(QRegularExpression("(?:[^\\w\\d\\_\\-\\.\\%]|$)"), 0); - domainExpression.setPattern(pattern.mid(0, end)); - pattern.remove(0, end+1); - } else if(pattern.startsWith("|") && pattern.endsWith("|")) { - pattern.remove(0, 1); - pattern.chop(1); - domainExpression.setPattern(pattern); - } else { - domainExpression.setPattern(".*"); - } - - // Regular rule - ruleExpression.setWildcardPattern(pattern); - m_valid = true; -} - -bool BlockerRule::match(const QWebEngineUrlRequestInfo &info) -{ - if(!m_valid) { - return false; - } - - // if both domain and rule match - if(domainExpression.match(info.requestUrl().host()) && ruleExpression.match(info.requestUrl().toString())) { - - // option explicitly allows - if(matchOptions(info, m_whitelistOptions)) { - return false; - } - - // option explicitly bans - if(matchOptions(info, m_blacklistOptions)) { - return true; - } - - // no options, but both domain and rule match --> rule matches - return true; - } - - // domain and/or rule do not match - return false; -} - -bool BlockerRule::isValid() -{ - return m_valid; -} -bool BlockerRule::isException() -{ - return m_exception; -} - -QString BlockerRule::filter() const -{ - return m_filter; -} - -void BlockerRule::parseOption(const QString &opt) -{ - if(opt.startsWith("script")) { - m_blacklistOptions.setFlag(RuleOption::script, true); - } else if(opt.startsWith("~script")) { - m_whitelistOptions.setFlag(RuleOption::script, true); - - } else if(opt.startsWith("image")) { - m_blacklistOptions.setFlag(RuleOption::image, true); - } else if(opt.startsWith("~image")) { - m_whitelistOptions.setFlag(RuleOption::image, true); - - } else if(opt.startsWith("stylesheet")) { - m_blacklistOptions.setFlag(RuleOption::stylesheet, true); - } else if(opt.startsWith("~stylesheet")) { - m_whitelistOptions.setFlag(RuleOption::stylesheet, true); - - } else if(opt.startsWith("object")) { - m_blacklistOptions.setFlag(RuleOption::object, true); - } else if(opt.startsWith("~object")) { - m_whitelistOptions.setFlag(RuleOption::object, true); - - } else if(opt.startsWith("object-subrequest")) { - m_blacklistOptions.setFlag(RuleOption::objectsubrequest, true); - } else if(opt.startsWith("~object-subrequest")) { - m_whitelistOptions.setFlag(RuleOption::objectsubrequest, true); - - } else if(opt.startsWith("subdocument")) { - m_blacklistOptions.setFlag(RuleOption::subdocument, true); - } else if(opt.startsWith("~subdocument")) { - m_whitelistOptions.setFlag(RuleOption::subdocument, true); - } -} - -bool BlockerRule::matchOptions(const QWebEngineUrlRequestInfo &info, const RuleOptions &options) -{ - // no options are defined - if(options == 0) { - return false; - } - - bool hasOption = false; - switch (info.resourceType()) { - case QWebEngineUrlRequestInfo::ResourceTypeScript: - if(options.testFlag(RuleOption::script)) { - hasOption = true; - } - break; - case QWebEngineUrlRequestInfo::ResourceTypeImage: - if(options.testFlag(RuleOption::image)) { - hasOption = true; - } - break; - case QWebEngineUrlRequestInfo::ResourceTypeStylesheet: - if(options.testFlag(RuleOption::stylesheet)) { - hasOption = true; - } - break; - case QWebEngineUrlRequestInfo::ResourceTypeObject: - if(options.testFlag(RuleOption::object)) { - hasOption = true; - } - break; - case QWebEngineUrlRequestInfo::ResourceTypePluginResource: - if(options.testFlag(RuleOption::objectsubrequest)) { - hasOption = true; - } - break; - case QWebEngineUrlRequestInfo::ResourceTypeSubFrame: - if(options.testFlag(RuleOption::subdocument)) { - hasOption = true; - } - break; - default: - break; - } - - return hasOption; -} diff --git a/src/formats/adblockrule.h b/src/formats/adblockrule.h deleted file mode 100644 index a56be2a..0000000 --- a/src/formats/adblockrule.h +++ /dev/null @@ -1,71 +0,0 @@ -/* - * This file is part of smolbote. It's copyrighted by the contributors recorded - * in the version control history of the file, available from its original - * location: git://neueland.iserlohn-fortress.net/smolbote.git - * - * SPDX-License-Identifier: GPL-3.0 - */ - -#ifndef ADBLOCKRULE_H -#define ADBLOCKRULE_H - -#include -#include -#include "regexp.h" -#include - -class BlockerRule : public QObject -{ - Q_OBJECT -public: - - // https://adblockplus.org/en/filters#options - enum RuleOption { - script = 1, - image = 2, - stylesheet = 4, - object = 8, - xmlhttprequest = 16, - objectsubrequest = 32, - subdocument = 64, - ping = 128, - websocket = 256, - document = 512, - elemhide = 1024, - generichide = 2048, - genericblock = 4096, - other = 8192 - }; - Q_DECLARE_FLAGS(RuleOptions, RuleOption) - - explicit BlockerRule(QString rule, QObject *parent = 0); - - bool match(const QWebEngineUrlRequestInfo &info); - bool isValid(); - bool isException(); - QString filter() const; - -signals: - -public slots: - -private: - void parseOption(const QString &opt); - bool matchOptions(const QWebEngineUrlRequestInfo &info, const RuleOptions &options); - - QString m_filter; - - bool m_valid; - bool m_exception = false; - bool m_elementRule; - - QStringList hostsBlacklist; - QStringList hostsWhitelist; - RegExp domainExpression; - RegExp ruleExpression; - - RuleOptions m_blacklistOptions; - RuleOptions m_whitelistOptions; -}; - -#endif // ADBLOCKRULE_H -- cgit v1.2.1