ParserMarkdown: Inherit from interface, not ParserLegacy

.gitmodules

@@ -7,3 +7,6 @@
 [submodule "submodules/qssb.h"]
 	path = submodules/qssb.h
 	url = https://gitea.quitesimple.org/crtxcr/qssb.h.git
+[submodule "submodules/qsmaddy"]
+	path = submodules/qsmaddy
+	url = https://gitea.quitesimple.org/crtxcr/qsmaddy

Makefile

@@ -2,8 +2,8 @@
 
 CXXFLAGS=-std=c++17 -O0 -g -no-pie -pipe -MMD -Wall -Wextra
 RELEASE_CXXFLAGS=-std=c++17 -O3 -pipe -MMD -Wall -Wextra
-LDFLAGS=-lsqlite3 -lpthread -lcrypto -lstdc++fs
-INCLUDEFLAGS=-I submodules/sqlitemoderncpp/hdr -I submodules/cpp-httplib -I submodules/qssb.h
+LDFLAGS=-lsqlite3 -lpthread -lcrypto -lstdc++fs -lseccomp
+INCLUDEFLAGS=-I submodules/sqlitemoderncpp/hdr -I submodules/cpp-httplib -I submodules/qssb.h -I submodules/qsmaddy/include/
 
 CXX=g++
 

config.cpp

@@ -71,6 +71,7 @@ Config::Config(const std::map<std::string, std::string> &map)
 
 	this->configmap = map;
 	this->wikipath = optional("wikipath", "/");
+	this->parser = optional("parser", "markdown");
 	this->handlersConfig.anon_username = optional("anon_username", "anonymouse");
 	this->handlersConfig.wikiname = required("wikiname");
 	this->logfile = required("logfile");

config.h

@@ -86,6 +86,7 @@ class Config
 	std::string templateprefix;
 	std::string logfile;
 	std::string connectionstring;
+	std::string parser;
 	int session_max_lifetime;
 	int threadscount;
 

handlers/handler.h

@@ -9,10 +9,13 @@
 #include "../database/queryoption.h"
 #include "../logger.h"
 #include "../cache/icache.h"
+#include "../iparser.h"
+
 class Handler
 {
   protected:
 	ICache *cache;
+	IParser *parser;
 	Template *templ;
 	Database *database;
 	Session *userSession;
@@ -25,7 +28,7 @@ class Handler
 
   public:
 	Handler(HandlerConfig &handlersConfig, Template &templ, Database &db, Session &userSession, UrlProvider &provider,
-			ICache &cache)
+			ICache &cache, IParser &parser)
 	{
 		this->handlersConfig = &handlersConfig;
 		this->templ = &templ;
@@ -33,6 +36,7 @@ class Handler
 		this->userSession = &userSession;
 		this->urlProvider = &provider;
 		this->cache = &cache;
+		this->parser = &parser;
 	}
 
 	virtual Response handle(const Request &r);

handlers/handlerfactory.h

@@ -10,15 +10,16 @@ class HandlerFactory
 	Database &db;
 	UrlProvider &urlProvider;
 	ICache &cache;
+	IParser &parser;
 
 	template <class T> inline std::unique_ptr<T> produce(Session &userSession)
 	{
-		return std::make_unique<T>(handlerConfig, templ, db, userSession, urlProvider, cache);
+		return std::make_unique<T>(handlerConfig, templ, db, userSession, urlProvider, cache, parser);
 	}
 
   public:
-	HandlerFactory(HandlerConfig &handlerConfig, Template &templ, Database &db, UrlProvider &urlprovider, ICache &cache)
-		: handlerConfig(handlerConfig), templ(templ), db(db), urlProvider(urlprovider), cache(cache)
+	HandlerFactory(HandlerConfig &handlerConfig, Template &templ, Database &db, UrlProvider &urlprovider, ICache &cache, IParser &parser)
+		: handlerConfig(handlerConfig), templ(templ), db(db), urlProvider(urlprovider), cache(cache), parser(parser)
 	{
 	}
 	std::unique_ptr<Handler> createHandler(const std::string &action, Session &userSession);

handlers/handlerpageedit.cpp

@@ -21,8 +21,8 @@ SOFTWARE.
 #include "handlerpageedit.h"
 #include "../database/exceptions.h"
 #include "../request.h"
+#include "../parserlegacy.h"
 
-#include "../parser.h"
 bool HandlerPageEdit::canAccess(std::string page)
 {
 	return this->userSession->user.permissions.canEdit();
@@ -59,7 +59,7 @@ Response HandlerPageEdit::handleRequest(PageDao &pageDao, std::string pagename,
 
 			// TODO: must check, whether categories differ, and perhaps don't allow every user
 			// to set categories
-			Parser parser;
+			ParserLegacy parser;
 			std::vector<std::string> cats = parser.extractCategories(newContent);
 			try
 			{
@@ -107,7 +107,7 @@ Response HandlerPageEdit::handleRequest(PageDao &pageDao, std::string pagename,
 		if(r.post("do") == "preview")
 		{
 			std::string newContent = r.post("content");
-			Parser parser;
+			ParserLegacy parser;
 			TemplatePage templatePage = this->templ->getPage("page_creation_preview");
 			templatePage.setVar("actionurl", urlProvider->editPage(pagename));
 			templatePage.setVar("preview_content", parser.parse(pageDao, *this->urlProvider, newContent));

handlers/handlerpageview.cpp

@@ -21,7 +21,7 @@ SOFTWARE.
 #include "handlerpageview.h"
 #include "../database/exceptions.h"
 #include "../logger.h"
-#include "../parser.h"
+#include "../parserlegacy.h"
 #include "../htmllink.h"
 
 bool HandlerPageView::canAccess(std::string page)
@@ -130,7 +130,7 @@ Response HandlerPageView::handleRequest(PageDao &pageDao, std::string pagename,
 
 	TemplatePage &page = this->templ->getPage(templatepartname);
 
-	Parser parser;
+
 	Response result;
 	result.setStatus(200);
 	std::string indexcontent;
@@ -138,8 +138,8 @@ Response HandlerPageView::handleRequest(PageDao &pageDao, std::string pagename,
 
 	if(revisionid > 0)
 	{
-		indexcontent = createIndexContent(parser, revision->content);
-		parsedcontent = parser.parse(pageDao, *this->urlProvider, revision->content);
+		indexcontent = createIndexContent(*parser, revision->content);
+		parsedcontent = parser->parse(pageDao, *this->urlProvider, revision->content);
 	}
 	else
 	{
@@ -153,7 +153,7 @@ Response HandlerPageView::handleRequest(PageDao &pageDao, std::string pagename,
 		}
 		else
 		{
-			indexcontent = createIndexContent(parser, revision->content);
+			indexcontent = createIndexContent(*parser, revision->content);
 			this->cache->put(cachekeyindexcontent, indexcontent);
 		}
 		if(cachedparsedcontent)
@@ -162,7 +162,7 @@ Response HandlerPageView::handleRequest(PageDao &pageDao, std::string pagename,
 		}
 		else
 		{
-			parsedcontent = parser.parse(pageDao, *this->urlProvider, revision->content);
+			parsedcontent = parser->parse(pageDao, *this->urlProvider, revision->content);
 			this->cache->put(cachekeyparsedcontent, parsedcontent);
 		}
 	}

iparser.h

@@ -5,6 +5,8 @@
 #include "headline.h"
 #include "database/pagedao.h"
 #include "urlprovider.h"
+
+
 class IParser
 {
   public:
@@ -16,4 +18,6 @@ class IParser
 	virtual ~IParser(){};
 };
 
+
+
 #endif // PARSER_H

parserlegacy.cpp

@@ -24,29 +24,22 @@ SOFTWARE.
 #include <vector>
 #include <algorithm>
 #include <iterator>
-#include "parser.h"
+#include "parserlegacy.h"
 #include "utils.h"
 #include "htmllink.h"
-std::vector<Headline> Parser::extractHeadlines(std::string content) const
+std::vector<Headline> ParserLegacy::extractHeadlines(std::string content) const
 {
 	std::vector<Headline> result;
-	std::string reg = R"(\[h(1|2|3)\](.*?)\[/h\1\])";
-	std::regex headerfinder(reg);
-	auto begin = std::sregex_iterator(content.begin(), content.end(), headerfinder);
-	auto end = std::sregex_iterator();
-
-	for(auto it = begin; it != end; it++)
-	{
-		auto smatch = *it;
+	utils::regex_callback_extractor(std::regex(R"(\[h(1|2|3)\](.*?)\[/h\1\])"), content, [&](std::smatch &smatch) {
 		Headline h;
 		h.level = utils::toUInt(smatch.str(1));
 		h.title = smatch.str(2);
 		result.push_back(h);
-	}
+	});
 	return result;
 }
 
-std::vector<std::string> Parser::extractCategories(std::string content) const
+std::vector<std::string> ParserLegacy::extractCategories(std::string content) const
 {
 	std::vector<std::string> result;
 	std::string reg = R"(\[category\](.*?)\[/category\])";
@@ -62,7 +55,7 @@ std::vector<std::string> Parser::extractCategories(std::string content) const
 	return result;
 }
 
-std::string Parser::extractCommand(std::string cmdname, std::string content) const
+std::string ParserLegacy::extractCommand(std::string cmdname, std::string content) const
 {
 	std::string cmd = "[cmd:" + cmdname + "]";
 	std::string cmdend = "[/cmd:" + cmdname + "]";
@@ -81,7 +74,7 @@ std::string Parser::extractCommand(std::string cmdname, std::string content) con
 	}
 	return "";
 }
-std::string Parser::processLink(const PageDao &pageDao, UrlProvider &urlProvider, std::smatch &match) const
+std::string ParserLegacy::processLink(const PageDao &pageDao, UrlProvider &urlProvider, std::smatch &match) const
 {
 	std::string linktag = match.str(1);
 	std::string inside = match.str(2);
@@ -116,7 +109,7 @@ std::string Parser::processLink(const PageDao &pageDao, UrlProvider &urlProvider
 	return htmllink.render();
 }
 
-std::string Parser::parse(const PageDao &pagedao, UrlProvider &provider, std::string content) const
+std::string ParserLegacy::parse(const PageDao &pagedao, UrlProvider &provider, std::string content) const
 {
 	std::string result;
 	// we don't care about commands, but we nevertheless replace them with empty strings

parserlegacy.h

@@ -2,7 +2,7 @@
 #define PARSER_H
 #include "iparser.h"
 
-class Parser : public IParser
+class ParserLegacy : public IParser
 {
   private:
 	std::string processLink(const PageDao &pageDao, UrlProvider &urlProvider, std::smatch &match) const;
@@ -13,7 +13,7 @@ class Parser : public IParser
 	std::vector<std::string> extractCategories(std::string content) const override;
 	std::string parse(const PageDao &pagedao, UrlProvider &provider, std::string content) const override;
 	using IParser::IParser;
-	~Parser(){};
+	~ParserLegacy(){};
 };
 
 #endif // PARSER_H

parsermarkdown.cpp

@@ -0,0 +1,71 @@
+#include "parsermarkdown.h"
+#include "logger.h"
+#include "htmllink.h"
+
+std::string ParserMarkdown::processLink(const PageDao &pageDao, UrlProvider &urlProvider, std::smatch &match) const
+{
+	std::string inner = match.str(1);
+	std::string link = match.str(2);
+	HtmlLink htmllink;
+	htmllink.href = link;
+	htmllink.innervalue = inner;
+
+	if(link.find("http://") == 0 || link.find("https://") == 0)
+	{
+		return htmllink.render();
+	}
+
+	if(pageDao.exists(link))
+	{
+		htmllink.cssclass = "exists";
+	}
+	else
+	{
+		htmllink.cssclass = "notexists";
+	}
+
+	htmllink.href = urlProvider.page(htmllink.href);
+
+	return htmllink.render();
+}
+
+ParserMarkdown::ParserMarkdown()
+{
+}
+
+std::vector<Headline> ParserMarkdown::extractHeadlines(std::string content) const
+{
+	std::vector<Headline> result;
+	utils::regex_callback_extractor(std::regex(R"((#{1,6}) (.*))"), content, [&](std::smatch &smatch) {
+		Headline h;
+		h.level = smatch.str(1).length();
+		h.title = smatch.str(2);
+		result.push_back(h);
+	});
+	return result;
+}
+
+std::string ParserMarkdown::parse(const PageDao &pagedao, UrlProvider &provider, std::string content) const
+{
+	std::shared_ptr<maddy::ParserConfig> config = std::make_shared<maddy::ParserConfig>();
+	auto maddy = std::make_shared<maddy::Parser>(config);
+
+	auto linkParser = std::make_shared<maddy::LinkParser>();
+	linkParser->setCallback([&](std::smatch &match) { return processLink(pagedao, provider, match); });
+	maddy->setLinkParser(linkParser);
+	// TODO: hack because the parser breaks if there is an \r
+	content = utils::strreplace(content, "\r", "");
+	std::stringstream s{content};
+	std::string result = maddy->Parse(s);
+	return result;
+}
+
+std::string ParserMarkdown::extractCommand(std::string cmdname, std::string content) const
+{
+	return "";
+}
+
+std::vector<std::string> ParserMarkdown::extractCategories(std::string content) const
+{
+	return { };
+}

parsermarkdown.h

@@ -0,0 +1,21 @@
+#ifndef PARSER_MARKDOWN_H
+#define PARSER_MARKDOWN_H
+
+#include "iparser.h"
+#include "maddy/parser.h"
+
+class ParserMarkdown : public IParser
+{
+private:
+	std::string processLink(const PageDao &pageDao, UrlProvider &urlProvider, std::smatch &match) const;
+
+public:
+	ParserMarkdown();
+	std::vector<Headline> extractHeadlines(std::string content) const;
+	std::string parse(const PageDao &pagedao, UrlProvider &provider, std::string content) const;
+
+	std::string extractCommand(std::string cmdname, std::string content) const;
+	std::vector<std::string> extractCategories(std::string content) const;
+};
+
+#endif // PARSER_MARKDOWN_H

qswiki.cpp

@@ -37,6 +37,10 @@ SOFTWARE.
 #include "requestworker.h"
 #include "cache/fscache.h"
 #include "sandbox/sandboxfactory.h"
+#include "iparser.h"
+#include "parserlegacy.h"
+#include "parsermarkdown.h"
+
 void sigterm_handler(int arg)
 {
 	// TODO: proper shutdown.
@@ -63,6 +67,17 @@ std::unique_ptr<ICache> createCache(const ConfigVariableResolver &resolver)
 
 	return std::make_unique<FsCache>(path);
 }
+
+std::unique_ptr<IParser> createParser(const ConfigVariableResolver &resolver)
+{
+	std::string parser = resolver.getConfig("parser");
+	if(parser == "legacy")
+	{
+		return std::make_unique<ParserLegacy>();
+	}
+	return std::make_unique<ParserMarkdown>();
+}
+
 int main(int argc, char **argv)
 {
 	if(geteuid() == 0)
@@ -77,21 +92,21 @@ int main(int argc, char **argv)
 		Logger::error() << "Sandbox is not supported, exiting";
 		exit(EXIT_FAILURE);
 	}
-	if(argc < 2)
-	{
-		std::cerr << "no path to config file provided" << std::endl;
-		return 1;
-	}
-	std::string configpath = std::filesystem::absolute(argv[1]).string();
 	if(!sandbox->enableForInit())
 	{
 		Logger::error() << "Sandboxing for init mode could not be activated.";
 		exit(EXIT_FAILURE);
 	}
 
+	if(argc < 2)
+	{
+		std::cerr << "no path to config file provided" << std::endl;
+		return 1;
+	}
+
 	try
 	{
-		ConfigReader configreader(configpath);
+		ConfigReader configreader(argv[1]);
 		Config config = configreader.readConfig();
 
 		// TODO: config.connectiontring only works as long as we only support sqlite of course
@@ -135,7 +150,10 @@ int main(int argc, char **argv)
 		auto cache = createCache(config.configVarResolver);
 		cache->clear();
 
-		HandlerFactory handlerFactory{config.handlersConfig, siteTemplate, *database.get(), urlProvider, *cache.get()};
+		auto parser = createParser(config.configVarResolver);
+
+
+		HandlerFactory handlerFactory{config.handlersConfig, siteTemplate, *database.get(), urlProvider, *cache.get(), *parser.get()};
 		RequestWorker requestWorker{handlerFactory, database->createSessionDao(), siteTemplate};
 
 		auto interface = createGateway(config);

sandbox/sandbox-linux.cpp

@@ -26,26 +26,16 @@
 bool SandboxLinux::enableForInit()
 {
 	umask(0027);
-	struct qssb_policy *policy = qssb_init_policy();
-	if(policy == NULL)
-	{
-		Logger::error() << "Failed to init sandboxing policy (init)";
-		return false;
-	}
-	policy->namespace_options = QSSB_UNSHARE_USER;
-	policy->drop_caps = 0;
-	qssb_append_syscall_policy(policy, QSSB_SYSCALL_DENY_KILL_PROCESS, QSSB_SYS(execveat));
-	qssb_append_syscall_policy(policy, QSSB_SYSCALL_DENY_KILL_PROCESS, QSSB_SYS(execve));
-	qssb_append_syscall_default_policy(policy, QSSB_SYSCALL_ALLOW);
-
-	int result = qssb_enable_policy(policy);
+	struct qssb_policy policy = {0};
+	int blacklisted_syscalls[] = {QSSB_SYS(execveat), QSSB_SYS(execve), -1};
+	policy.blacklisted_syscalls = blacklisted_syscalls;
+	policy.no_new_privs = 1;
+	int result = qssb_enable_policy(&policy);
 	if(result != 0)
 	{
-		Logger::error() << "Failed to enable sandboxing policy (init): " << result;
-		qssb_free_policy(policy);
+		Logger::error() << "Failed to install sandboxing policy (init): " << result;
 		return false;
 	}
-	qssb_free_policy(policy);
 	return true;
 }
 
@@ -54,34 +44,28 @@ bool SandboxLinux::enablePreWorker(std::vector<std::string> fsPaths)
 	std::sort(fsPaths.begin(), fsPaths.end(),
 			  [](const std::string &a, const std::string &b) { return a.length() < b.length(); });
 
-	struct qssb_policy *policy = qssb_init_policy();
-	if(policy == NULL)
-	{
-		Logger::error() << "Failed to init sandboxing policy (pre)";
-		return false;
-	}
-
+	struct qssb_path_policy *policies = new qssb_path_policy[fsPaths.size()];
 	for(unsigned int i = 0; i < fsPaths.size(); i++)
 	{
-		qssb_append_path_policy(policy, QSSB_FS_ALLOW_READ | QSSB_FS_ALLOW_WRITE, fsPaths[i].c_str());
+		policies[i].next = policies + (i + 1);
+		policies[i].mountpoint = fsPaths[i].c_str();
+		policies[i].policy = QSSB_MOUNT_ALLOW_READ | QSSB_MOUNT_ALLOW_WRITE;
 	}
+	policies[fsPaths.size() - 1].next = NULL;
 
-	policy->namespace_options = QSSB_UNSHARE_MOUNT;
-	policy->drop_caps = 0;
-	policy->mount_path_policies_to_chroot = 1;
-
-	qssb_append_syscall_policy(policy, QSSB_SYSCALL_DENY_KILL_PROCESS, QSSB_SYS(execveat));
-	qssb_append_syscall_policy(policy, QSSB_SYSCALL_DENY_KILL_PROCESS, QSSB_SYS(execve));
-	qssb_append_syscall_default_policy(policy, QSSB_SYSCALL_ALLOW);
-
-	int result = qssb_enable_policy(policy);
+	struct qssb_policy policy = {0};
+	policy.path_policies = policies;
+	policy.namespace_options |= QSSB_UNSHARE_MOUNT;
+	policy.namespace_options |= QSSB_UNSHARE_USER;
+	int blacklisted_syscalls[] = {QSSB_SYS(execveat), QSSB_SYS(execve), -1};
+	policy.blacklisted_syscalls = blacklisted_syscalls;
+	int result = qssb_enable_policy(&policy);
 	if(result != 0)
 	{
 		Logger::error() << "Failed to install sandboxing policy (preworker): %i" << result;
-		qssb_free_policy(policy);
 		return false;
 	}
-	qssb_free_policy(policy);
+	delete[] policies;
 	return true;
 }
 
@@ -104,35 +88,30 @@ bool SandboxLinux::supported()
 }
 bool SandboxLinux::enableForWorker()
 {
-	struct qssb_policy *policy = qssb_init_policy();
-	if(policy == NULL)
-	{
-		Logger::error() << "Failed to init sandboxing policy (worker) ";
-		return false;
-	}
-	policy->drop_caps = 1;
-	policy->not_dumpable = 1;
-	policy->no_new_privs = 1;
-	policy->namespace_options = 0;
+	struct qssb_policy policy = {0};
+	policy.drop_caps = 1;
+	policy.not_dumpable = 1;
+	policy.no_new_privs = 1;
 
 	/* TODO: as said, a whitelist approach is better. As such, this list is bound to be incomplete in the
 	 * sense that more could be listed here and some critical ones are probably missing */
-
-	/* TODO: use qssb groups */
-	long blacklisted_syscalls[] = {QSSB_SYS(setuid),	  QSSB_SYS(connect), QSSB_SYS(chroot),	QSSB_SYS(pivot_root),
-								   QSSB_SYS(mount),		  QSSB_SYS(setns),	 QSSB_SYS(unshare), QSSB_SYS(ptrace),
-								   QSSB_SYS(personality), QSSB_SYS(prctl)};
-
-	qssb_append_syscalls_policy(policy, QSSB_SYSCALL_DENY_KILL_PROCESS, blacklisted_syscalls,
-								sizeof(blacklisted_syscalls) / sizeof(blacklisted_syscalls[0]));
-	qssb_append_syscall_default_policy(policy, QSSB_SYSCALL_ALLOW);
-
-	if(qssb_enable_policy(policy) != 0)
+	int blacklisted_syscalls[] = {QSSB_SYS(setuid),
+								  QSSB_SYS(connect),
+								  QSSB_SYS(chroot),
+								  QSSB_SYS(pivot_root),
+								  QSSB_SYS(mount),
+								  QSSB_SYS(setns),
+								  QSSB_SYS(unshare),
+								  QSSB_SYS(ptrace),
+								  QSSB_SYS(personality),
+								  QSSB_SYS(prctl),
+								  -1};
+	policy.blacklisted_syscalls = blacklisted_syscalls;
+	if(qssb_enable_policy(&policy) != 0)
 	{
 		Logger::error() << "Sandbox: Activation of seccomp blacklist failed!";
-		qssb_free_policy(policy);
 		return false;
 	}
-	qssb_free_policy(policy);
+
 	return true;
 }

setup/sqlite.sql

@@ -25,13 +25,24 @@ count integer
 CREATE TABLE category(id INTEGER PRIMARY KEY, name varchar(255));
 CREATE TABLE categorymember(id INTEGER PRIMARY KEY, category REFERENCES category(id), page REFERENCES page (id));
 CREATE INDEX revisionid ON revision (revisionid DESC);
-CREATE INDEX pagename ON page (name);
-CREATE INDEX token ON session (token);
-CREATE VIRTUAL TABLE search USING fts5(content, page UNINDEXED, content=revision,content_rowid=id);
+CREATE INDEX pagename ON page (name)
+;
+CREATE INDEX token ON session (token)
+;
+CREATE TRIGGER search_ai AFTER INSERT ON revision BEGIN
+  DELETE FROM search WHERE page = new.page;
+  INSERT INTO search(rowid, content, page) VALUES (new.id, new.content, new.page);
+END;
+CREATE TRIGGER search_au AFTER UPDATE ON revision BEGIN
+  DELETE FROM search WHERE page = old.page;
+  INSERT INTO search(rowid, content, page) VALUES (new.id, new.content, new.page);
+END;
+CREATE VIRTUAL TABLE search USING fts5(content, page UNINDEXED, content=revision,content_rowid=id)
+/* search(content,page) */;
+CREATE TABLE IF NOT EXISTS 'search_data'(id INTEGER PRIMARY KEY, block BLOB);
+CREATE TABLE IF NOT EXISTS 'search_idx'(segid, term, pgno, PRIMARY KEY(segid, term)) WITHOUT ROWID;
+CREATE TABLE IF NOT EXISTS 'search_docsize'(id INTEGER PRIMARY KEY, sz BLOB);
+CREATE TABLE IF NOT EXISTS 'search_config'(k PRIMARY KEY, v) WITHOUT ROWID;
 CREATE TRIGGER search_ad AFTER DELETE ON revision BEGIN
   INSERT INTO search(search, rowid, content, page) VALUES('delete', old.id, old.content, old.page);
 END;
-CREATE TRIGGER search_ai AFTER INSERT ON revision BEGIN
-  INSERT INTO search(search, rowid, content, page) SELECT 'delete', id, content, page FROM revision WHERE page = new.page AND revisionid = new.revisionid - 1; 
-  INSERT INTO search(rowid, content, page) VALUES (new.id, new.content, new.page);
-END;

utils.cpp

@@ -46,12 +46,6 @@ std::string utils::html_xss(std::string_view str)
 		case '%':
 			result += "%";
 			break;
-		case '\'':
-			result += "'";
-			break;
-		case '&':
-			result += "&";
-			break;
 		default:
 			result += c;
 		}
@@ -99,7 +93,7 @@ std::vector<std::string> utils::split(const std::string &str, char delim)
 // TODO: can easily break if we pass a regex here
 std::vector<std::string> utils::split(const std::string &str, const std::string &delim)
 {
-	std::regex regex{delim + "+"};
+	std::regex regex { delim + "+" };
 	return split(str, regex);
 }
 
@@ -181,3 +175,14 @@ std::string utils::toISODate(time_t t)
 	}
 	return std::string{result};
 }
+
+void utils::regex_callback_extractor(std::regex regex, const std::string &input, std::function<void (std::smatch &)> callback)
+{
+	auto begin = std::sregex_iterator(input.begin(), input.end(), regex);
+	auto end = std::sregex_iterator();
+	for(auto it = begin; it != end; it++)
+	{
+		std::smatch smatch = *it;
+		callback(smatch);
+	}
+}

utils.h

@@ -60,6 +60,7 @@ template <class T, class U> std::vector<U> getAll(std::multimap<T, U> map, T key
 std::string regex_callback_replacer(std::regex regex, const std::string &input,
 									std::function<std::string(std::smatch &)> callback);
 
+void regex_callback_extractor(std::regex regex, const std::string &input, std::function<void(std::smatch &)> callback);
 std::string readCompleteFile(std::string_view filepath);
 
 inline std::string nz(const char *s)