From 00abc6bc1b460da78b0221a0166d1f2f71a6df32 Mon Sep 17 00:00:00 2001
From: Albert S <mail@quitesimple.org>
Date: Sat, 6 Aug 2022 08:57:37 +0200
Subject: [PATCH] gui, shared: Fix and simplify word extraction regexes

They did not work for chars like '-', causing errors.

We can actually just extract non-space chars for these cases.
---
 gui/mainwindow.cpp    | 2 +-
 shared/looqsquery.cpp | 5 ++---
 2 files changed, 3 insertions(+), 4 deletions(-)

diff --git a/gui/mainwindow.cpp b/gui/mainwindow.cpp
index e7a73b1..c93e9a1 100644
--- a/gui/mainwindow.cpp
+++ b/gui/mainwindow.cpp
@@ -669,7 +669,7 @@ void MainWindow::makePreviews(int page)
 	processedPdfPreviews = 0;
 
 	QVector<QString> wordsToHighlight;
-	QRegularExpression extractor(R"#("([^"]*)"|((\p{L}|\p{N})+))#");
+	QRegularExpression extractor(R"#("([^"]*)"|([^\s]+))#");
 	for(const Token &token : this->contentSearchQuery.getTokens())
 	{
 		if(token.type == FILTER_CONTENT_CONTAINS)
diff --git a/shared/looqsquery.cpp b/shared/looqsquery.cpp
index c606b54..29fc187 100644
--- a/shared/looqsquery.cpp
+++ b/shared/looqsquery.cpp
@@ -180,9 +180,8 @@ LooqsQuery LooqsQuery::build(QString expression, TokenType loneWordsTokenType, b
 
 	QStringList loneWords;
 	LooqsQuery result;
-	QRegularExpression rx(
-		"((?<filtername>(\\.|\\w)+):(?<args>\\((?<innerargs>[^\\)]+)\\)|([\\p{L}\\p{N},])+)|(?<boolean>AND|OR)"
-		"|(?<negation>!)|(?<bracket>\\(|\\))|(?<loneword>[\"\\p{L}\\p{N}]+))");
+	QRegularExpression rx("((?<filtername>(\\.|\\w)+):(?<args>\\((?<innerargs>[^\\)]+)\\)|([^\\s])+)|(?<boolean>AND|OR)"
+						  "|(?<negation>!)|(?<bracket>\\(|\\))|(?<loneword>[^\\s]+))");
 	QRegularExpressionMatchIterator i = rx.globalMatch(expression);
 	auto previousWasBool = [&result] { return !result.tokens.empty() && ((result.tokens.last().type & BOOL) == BOOL); };
 	auto previousWas = [&result](TokenType t) { return !result.tokens.empty() && (result.tokens.last().type == t); };