fix(spellcheck): use input highlighter's word segmentation to find word (#6822)

We used to use `QTextCursor`'s definition of a word when right-clicking.
This doesn't necessarily align with ours (e.g.
#6821 (comment)).
Furthermore, we didn't consider the character after a word as part of
it. Since Qt does some rounding, it might determine that you clicked on
the character after a word if you clicked on the right side of it. With
this PR, that's fixed.

Here's a visualization of which characters we consider as part of a word
(separated by even/odd):

```
     ███     █████  ██████   ████
this is some text#I#write in here
█████   █████     ██      ███
```

Fixes #6819

Reported-by: pajlada <rasmus.karlsson@pajlada.com>
Reviewed-by: pajlada <rasmus.karlsson@pajlada.com>
Parent-pr: #6446

Author: Nerixyz

CHANGELOG.md

@@ -95,7 +95,7 @@
 - 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, #6780)
+- Dev: Added experimental spell checker support. (#6446, #6703, #6722, #6730, #6731, #6779, #6780, #6822)
 - 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

@@ -137,6 +137,41 @@ std::vector<QString> InputHighlighter::getSpellCheckedWords(const QString &text)
     return words;
 }
 
+QStringView InputHighlighter::getWordAt(QStringView text, qsizetype pos)
+{
+    auto tokenIt = this->tokenRegex.globalMatchView(text);
+    QString token;
+    qsizetype posInWord = 0;
+    qsizetype tokenStart = 0;
+    while (tokenIt.hasNext())
+    {
+        auto match = tokenIt.next();
+        // using '<= end' to include the word left to the cursor if it's at the end
+        if (match.capturedStart() <= pos && pos <= match.capturedEnd())
+        {
+            token = match.captured();
+            tokenStart = match.capturedStart();
+            posInWord = pos - tokenStart;
+            break;
+        }
+    }
+    if (token.isEmpty())
+    {
+        return {};
+    }
+
+    QStringView word;
+    this->visitWords(token, [&](const QString & /*curWord*/, qsizetype start,
+                                qsizetype count) {
+        if (start <= posInWord && posInWord <= start + count)
+        {
+            assert(word.isEmpty());
+            word = text.sliced(tokenStart + start, count);
+        }
+    });
+    return word;
+}
+
 void InputHighlighter::highlightBlock(const QString &text)
 {
     if (!this->spellChecker.isLoaded())

src/widgets/splits/InputHighlighter.hpp

@@ -43,6 +43,10 @@ class InputHighlighter : public QSyntaxHighlighter
     /// checked by the spell checker.
     std::vector<QString> getSpellCheckedWords(const QString &text);
 
+    /// Get the word in \p text at \p pos. If there isn't any word, returns an
+    /// empty view.
+    QStringView getWordAt(QStringView text, qsizetype pos);
+
 protected:
     void highlightBlock(const QString &text) override;
 

src/widgets/splits/SplitInput.cpp

@@ -801,13 +801,27 @@ void SplitInput::installTextEditEvents()
                              });
             menu->addAction(spellcheckAction);
 
-            auto cursor = this->ui_.textEdit->cursorForPosition(pos);
-            cursor.select(QTextCursor::WordUnderCursor);
-            auto word = cursor.selectedText();
+            if (!this->inputHighlighter)
+            {
+                return;
+            }
+
+            auto cursorAtPos = this->ui_.textEdit->cursorForPosition(pos);
+            QString text = this->ui_.textEdit->toPlainText();
+            QStringView word =
+                this->inputHighlighter->getWordAt(text, cursorAtPos.position());
             if (!word.isEmpty())
             {
+                auto cursor = this->ui_.textEdit->textCursor();
+                // Select `word`. `word` is a view into `text`, so we can use
+                // the offsets of `word` from the start of `text`.
+                cursor.setPosition(
+                    static_cast<int>(word.begin() - text.begin()));
+                cursor.setPosition(static_cast<int>(word.end() - text.begin()),
+                                   QTextCursor::KeepAnchor);
+
                 auto suggestions =
-                    getApp()->getSpellChecker()->suggestions(word);
+                    getApp()->getSpellChecker()->suggestions(word.toString());
                 for (const auto &sugg : suggestions)
                 {
                     auto qSugg = QString::fromStdString(sugg);

tests/src/InputHighlighter.cpp

@@ -287,3 +287,88 @@ TEST(InputHighlight, wordRegex)
         ASSERT_EQ(got, c.words) << "index=" << i;
     }
 }
+
+TEST_F(InputHighlighterTest, getWordAt)
+{
+    struct Case {
+        std::pair<QStringView, qsizetype> input;
+        QStringView word;
+    };
+
+    std::vector<Case> cases{
+        {.input = {u"", 0}, .word = {}},
+        {.input = {u"", -1}, .word = {}},
+        {.input = {u"", -2}, .word = {}},
+        {.input = {u"", 1}, .word = {}},
+        {.input = {u"", 2}, .word = {}},
+        {.input = {u"word", -1}, .word = {}},
+        {.input = {u"word", 0}, .word = u"word"},
+        {.input = {u"word", 1}, .word = u"word"},
+        {.input = {u"word", 2}, .word = u"word"},
+        {.input = {u"word", 3}, .word = u"word"},
+        {.input = {u"word", 4}, .word = u"word"},
+        {.input = {u"word", 5}, .word = {}},
+        {.input = {u"a word", -1}, .word = {}},
+        {.input = {u"a word", 0}, .word = u"a"},
+        {.input = {u"a word", 1}, .word = u"a"},
+        {.input = {u"a word", 2}, .word = u"word"},
+        {.input = {u"a word", 3}, .word = u"word"},
+        {.input = {u"a word", 4}, .word = u"word"},
+        {.input = {u"a word", 5}, .word = u"word"},
+        {.input = {u"a word", 6}, .word = u"word"},
+        {.input = {u"a word", 7}, .word = {}},
+        {.input = {u"a word!", 6}, .word = u"word"},
+        {.input = {u"a word!", 7}, .word = {}},
+        {.input = {u"a word!", 8}, .word = {}},
+        {.input = {u"three words are", 0}, .word = u"three"},
+        {.input = {u"three words are", 4}, .word = u"three"},
+        {.input = {u"three words are", 5}, .word = u"three"},
+        {.input = {u"three words are", 6}, .word = u"words"},
+        {.input = {u"three words are", 10}, .word = u"words"},
+        {.input = {u"three words are", 11}, .word = u"words"},
+        {.input = {u"three words are", 12}, .word = u"are"},
+        {.input = {u"three words are", 14}, .word = u"are"},
+        {.input = {u"three words are", 15}, .word = u"are"},
+        {.input = {u"three words are", 16}, .word = {}},
+        {.input = {u"a#text#is#here!", 0}, .word = u"a"},
+        {.input = {u"a#text#is#here!", 1}, .word = u"a"},
+        {.input = {u"a#text#is#here!", 2}, .word = u"text"},
+        {.input = {u"a#text#is#here!", 5}, .word = u"text"},
+        {.input = {u"a#text#is#here!", 6}, .word = u"text"},
+        {.input = {u"a#text#is#here!", 7}, .word = u"is"},
+        {.input = {u"a#text#is#here!", 8}, .word = u"is"},
+        {.input = {u"a#text#is#here!", 9}, .word = u"is"},
+        {.input = {u"a#text#is#here!", 10}, .word = u"here"},
+        {.input = {u"a#text#is#here!", 13}, .word = u"here"},
+        {.input = {u"a#text#is#here!", 14}, .word = u"here"},
+        {.input = {u"a#text#is#here!", 15}, .word = {}},
+        {.input = {u"here! there!", 3}, .word = u"here"},
+        {.input = {u"here! there!", 4}, .word = u"here"},
+        {.input = {u"here! there!", 5}, .word = {}},
+        {.input = {u"here! there!", 6}, .word = u"there"},
+        {.input = {u"/my-command wow", 0}, .word = {}},
+        {.input = {u"/my-command wow", 1}, .word = {}},
+        {.input = {u"/my-command wow", 2}, .word = {}},
+        {.input = {u"/my-command wow", 10}, .word = {}},
+        {.input = {u"/my-command wow", 11}, .word = {}},
+        {.input = {u"/my-command wow", 12}, .word = u"wow"},
+        {.input = {u"a 7TVGlobal b", 0}, .word = u"a"},
+        {.input = {u"a 7TVGlobal b", 1}, .word = u"a"},
+        {.input = {u"a 7TVGlobal b", 2}, .word = {}},
+        {.input = {u"a 7TVGlobal b", 10}, .word = {}},
+        {.input = {u"a 7TVGlobal b", 11}, .word = {}},
+        {.input = {u"a 7TVGlobal b", 12}, .word = u"b"},
+        {.input = {u"a 7TVGlobal b", 13}, .word = u"b"},
+        {.input = {u"a 7TVGlobal b", 14}, .word = {}},
+    };
+
+    SpellChecker nullSpellChecker;
+    InputHighlighter highlighter(nullSpellChecker, nullptr);
+    highlighter.setChannel(this->channel);
+    for (size_t i = 0; i < cases.size(); i++)
+    {
+        const auto &c = cases[i];
+        auto got = highlighter.getWordAt(c.input.first, c.input.second);
+        ASSERT_EQ(got, c.word) << "index=" << i;
+    }
+}