chatterino2
Code Issues

fix(spellcheck): some chatters and emotes not ignored (#6780)

Browse Source 
Reviewed-by: Nerixyz <nerixdev@outlook.de>
Reviewed-by: pajlada <rasmus.karlsson@pajlada.com>
This commit is contained in:
Arne
2026-01-31 15:45:29 +01:00
committed by GitHub
parent cebe7e415c
commit 8f660b6de5
4 changed files with 84 additions and 39 deletions
Download Patch File Download Diff File Expand all files Collapse all files
CHANGELOG.md
+1 -1
View File
@@ -86,7 +86,7 @@
- Dev: Unwrapped `LimitedQueueSnapshot` to `std::vector`. (#6606)
- Dev: Simplified uses of `getMessageSnapshot`. (#6607)
- Dev: Disabled `llvm-prefer-static-over-anonymous-namespace` in clang-tidy. (#6610)
- Dev: Added experimental spell checker support. (#6446, #6703, #6722, #6730, #6731, #6779)
- Dev: Added experimental spell checker support. (#6446, #6703, #6722, #6730, #6731, #6779, #6780)
- Dev: Added Clazy linting in CI. (#6623)
- Dev: Added custom clang-tidy module linting in CI. (#6626)
- Dev: CMake option `USE_ALTERNATE_LINKER` now errors if the given linker can't be found. (#6692)
src/widgets/splits/InputHighlighter.cpp
+59 -20
View File
@@ -15,6 +15,7 @@
#include "providers/seventv/SeventvEmotes.hpp"
#include "providers/twitch/TwitchAccount.hpp"
#include "providers/twitch/TwitchChannel.hpp"
#include "singletons/Settings.hpp"
#include <QTextCharFormat>
#include <QTextDocument>
@@ -23,7 +24,7 @@ namespace {
using namespace chatterino;
bool isIgnoredWord(TwitchChannel *twitch, const QString &word)
bool isEmote(TwitchChannel *twitch, const QString &word)
{
EmoteName name{word};
if (twitch)
@@ -38,11 +39,6 @@ bool isIgnoredWord(TwitchChannel *twitch, const QString &word)
{
return true;
}
if (twitch->accessChatters()->contains(word))
{
return true;
}
}
if (getApp()->getBttvEmotes()->emote(name) ||
getApp()->getFfzEmotes()->emote(name) ||
@@ -60,22 +56,52 @@ bool isIgnoredWord(TwitchChannel *twitch, const QString &word)
return true;
}
return false;
}
bool isChatter(TwitchChannel *twitch, const QString &word)
{
if (twitch)
{
if (twitch->accessChatters()->contains(word) ||
(getSettings()->alwaysIncludeBroadcasterInUserCompletions &&
word.compare(twitch->getName(), Qt::CaseInsensitive) == 0))
{
return true;
}
}
return false;
}
bool isLink(const QString &token)
{
// TODO: Replace this with a link parser variant that doesn't return the parsed data
auto link = linkparser::parse(word);
auto link = linkparser::parse(token);
return link.has_value();
}
bool isIgnoredWord(TwitchChannel *twitch, const QString &word)
{
return isEmote(twitch, word) || isChatter(twitch, word);
}
bool isIgnoredToken(TwitchChannel *twitch, const QString &token)
{
return isEmote(twitch, token) || isLink(token);
}
} // namespace
namespace chatterino {
namespace inputhighlight::detail {
// FIXME: this also matches URLs - this probably needs to be some function like Firefox' mozEnglishWordUtils::FindNextWord
// A word is a string of unicode letters. Words are seperated by whitespace
// (tokenRegex) or, inside a token, by punctuation characters (except '_')
QRegularExpression wordRegex()
{
static QRegularExpression regex{
R"(\p{L}(?:\P{Z}+\p{L}+)*)",
R"((?<=^|(?!_)\p{P})\p{L}+(?=$|(?!_)\p{P}))",
QRegularExpression::PatternOption::UseUnicodePropertiesOption,
};
return regex;
@@ -87,6 +113,7 @@ InputHighlighter::InputHighlighter(SpellChecker &spellChecker, QObject *parent)
: QSyntaxHighlighter(parent)
, spellChecker(spellChecker)
, wordRegex(inputhighlight::detail::wordRegex())
, tokenRegex(R"(\S+)")
{
this->spellFmt.setUnderlineStyle(QTextCharFormat::SpellCheckUnderline);
this->spellFmt.setUnderlineColor(Qt::red);
@@ -138,20 +165,32 @@ void InputHighlighter::visitWords(
auto cmdTriggerLen = getApp()->getCommands()->commandTriggerLen(textView);
textView = textView.sliced(cmdTriggerLen);
#if QT_VERSION >= QT_VERSION_CHECK(6, 5, 0)
auto it = this->wordRegex.globalMatchView(textView);
#else
auto it = this->wordRegex.globalMatch(textView);
#endif
auto tokenIt = this->tokenRegex.globalMatchView(textView);
while (it.hasNext())
// iterate over whitespace-delimited tokens
while (tokenIt.hasNext())
{
auto match = it.next();
auto text = match.captured();
if (!isIgnoredWord(channel, text))
auto tokenMatch = tokenIt.next();
auto token = tokenMatch.captured();
if (isIgnoredToken(channel, token))
{
cb(text, static_cast<int>(match.capturedStart() + cmdTriggerLen),
static_cast<int>(text.size()));
continue;
}
auto wordIt = this->wordRegex.globalMatchView(token);
while (wordIt.hasNext())
{
auto wordMatch = wordIt.next();
auto word = wordMatch.captured();
if (!isIgnoredWord(channel, word))
{
cb(word,
static_cast<int>(cmdTriggerLen + tokenMatch.capturedStart() +
wordMatch.capturedStart()),
static_cast<int>(word.size()));
}
}
}
}
src/widgets/splits/InputHighlighter.hpp
+1
View File
@@ -59,6 +59,7 @@ private:
std::weak_ptr<TwitchChannel> channel;
QRegularExpression wordRegex;
QRegularExpression tokenRegex;
};
} // namespace chatterino
tests/src/InputHighlighter.cpp
+23 -18
View File
@@ -185,17 +185,17 @@ TEST_F(InputHighlighterTest, getSpellCheckedWords)
{.input = "word word", .words = {"word", "word"}},
{.input = " word word ", .words = {"word", "word"}},
{.input = "word?", .words = {"word"}},
{.input = "word?word", .words = {"word?word"}},
{.input = "word-word", .words = {"word-word"}},
{.input = "word?word", .words = {"word", "word"}},
// FIXME: should be "word-word"
{.input = "word-word", .words = {"word", "word"}},
{
.input = "channel emotes 7TVEmote a BTTVEmote b FFZEmote c",
.words = {"channel", "emotes", "TVEmote", "a", "b", "c"},
.words = {"channel", "emotes", "a", "b", "c"},
},
{
.input = "global emotes 7TVGlobal a BTTVGlobal b FFZGlobal c "
"MyCoolTwitchEmote d",
// FIXME: TVGlobal shouldn't show up
.words = {"global", "emotes", "TVGlobal", "a", "b", "c", "d"},
.words = {"global", "emotes", "a", "b", "c", "d"},
},
{
.input = "/ban a user",
@@ -228,9 +228,14 @@ TEST_F(InputHighlighterTest, getSpellCheckedWords)
},
},
{
.input = "Hey, @userchatter a 123kappa123 b MyUser42 c",
// FIXME: 'kappa' and 'MyUser' shouldn't show up
.words = {"Hey", "a", "kappa", "b", "MyUser", "c"},
.input = "Hey, @userchatter, a 123kappa123 b MyUser42 c",
.words = {"Hey", "a", "b", "c"},
},
{
.input =
"twitch.tv ignore "
"https://wiki.chatterino.com/Help/#basic-troubleshooting links",
.words = {"ignore", "links"},
},
};
@@ -255,18 +260,18 @@ TEST(InputHighlight, wordRegex)
std::vector<Case> cases{
{.input = u"", .words = {}},
{.input = u"word", .words = {u"word"}},
{.input = u"word word", .words = {u"word", u"word"}},
{.input = u" word word ", .words = {u"word", u"word"}},
{.input = u"test/man", .words = {u"test", u"man"}},
{.input = u"word?", .words = {u"word"}},
{.input = u"word?word", .words = {u"word?word"}},
{.input = u"a12word", .words = {u"a12word"}},
{.input = u"word?word", .words = {u"word", u"word"}},
{.input = u"sentence.", .words = {u"sentence"}},
{.input = u"#hashtag", .words = {u"hashtag"}},
{.input = u"inogre123numbers", .words = {}},
{.input = u"under_score", .words = {}},
{.input = u"äwördü", .words = {u"äwördü"}},
// FIXME: should be the whole input
{.input = u"123kappa123", .words = {u"kappa"}},
{.input = u"123kappa", .words = {u"kappa"}},
{.input = u"kappa123", .words = {u"kappa"}},
{.input = u"abc @foo bar@baz@", .words = {u"abc", u"foo", u"bar@baz"}},
{.input = u"1234567 word a123", .words = {u"word", u"a"}},
{.input = u"abc!@foo#bar&(baz]",
.words = {u"abc", u"foo", u"bar", u"baz"}},
{.input = u"1234567,word/a123", .words = {u"word"}},
{.input = u"'quotes\"", .words = {u"quotes"}},
};
auto re = inputhighlight::detail::wordRegex();