Improve spell checking (#2030)

novelwriter/core/coretools.py

@@ -343,7 +343,7 @@ class DocSearch:
         count = 0
         capped = False
         results = []
-        for match in re.finditer(self._regEx, text):
+        for match in self._regEx.finditer(text):
             pos = match.start(0)
             num = len(match.group(0))
             lim = text[:pos].rfind("\n") + 1

novelwriter/core/spellcheck.py

@@ -125,21 +125,16 @@ class NWSpellEnchant:
         except Exception:
             return []
 
-    def addWord(self, word: str) -> bool:
+    def addWord(self, word: str, save: bool = True) -> None:
         """Add a word to the project dictionary."""
-        word = word.strip()
-        if not word:
-            return False
-        try:
-            self._enchant.add_to_session(word)
-        except Exception:
-            return False
-
-        added = self._userDict.add(word)
-        if added:
-            self._userDict.save()
-
-        return added
+        if word := word.strip():
+            try:
+                self._enchant.add_to_session(word)
+            except Exception:
+                return
+            if save and self._userDict.add(word):
+                self._userDict.save()
+        return
 
     def listDictionaries(self) -> list[tuple[str, str]]:
         """List available dictionaries."""

novelwriter/core/tokenizer.py

@@ -1109,14 +1109,14 @@ class Tokenizer(ABC):
 
         # Match Markdown
         for regEx, fmts in self._rxMarkdown:
-            for match in re.finditer(regEx, text):
+            for match in regEx.finditer(text):
                 temp.extend(
                     (match.start(n), match.end(n), fmt, "")
                     for n, fmt in enumerate(fmts) if fmt > 0
                 )
 
         # Match Shortcodes
-        for match in re.finditer(REGEX_PATTERNS.shortcodePlain, text):
+        for match in REGEX_PATTERNS.shortcodePlain.finditer(text):
             temp.append((
                 match.start(1), match.end(1),
                 self._shortCodeFmt.get(match.group(1).lower(), 0),
@@ -1125,7 +1125,7 @@ class Tokenizer(ABC):
 
         # Match Shortcode w/Values
         tHandle = self._handle or ""
-        for match in re.finditer(REGEX_PATTERNS.shortcodeValue, text):
+        for match in REGEX_PATTERNS.shortcodeValue.finditer(text):
             kind = self._shortCodeVals.get(match.group(1).lower(), 0)
             temp.append((
                 match.start(0), match.end(0),
@@ -1136,7 +1136,7 @@ class Tokenizer(ABC):
         # Match Dialogue
         if self._rxDialogue and hDialog:
             for regEx, fmtB, fmtE in self._rxDialogue:
-                for match in re.finditer(regEx, text):
+                for match in regEx.finditer(text):
                     temp.append((match.start(0), 0, fmtB, ""))
                     temp.append((match.end(0), 0, fmtE, ""))
 

novelwriter/gui/doceditor.py

@@ -1164,8 +1164,10 @@ class GuiDocEditor(QPlainTextEdit):
                     ctxMenu.addAction(f"{nwUnicode.U_ENDASH} {trNone}")
 
                 ctxMenu.addSeparator()
+                action = ctxMenu.addAction(self.tr("Ignore Word"))
+                action.triggered.connect(lambda: self._addWord(word, block, False))
                 action = ctxMenu.addAction(self.tr("Add Word to Dictionary"))
-                action.triggered.connect(lambda: self._addWord(word, block))
+                action.triggered.connect(lambda: self._addWord(word, block, True))
 
         # Execute the context menu
         ctxMenu.exec(self.viewport().mapToGlobal(pos))
@@ -1173,30 +1175,6 @@ class GuiDocEditor(QPlainTextEdit):
 
         return
 
-    @pyqtSlot("QTextCursor", str)
-    def _correctWord(self, cursor: QTextCursor, word: str) -> None:
-        """Slot for the spell check context menu triggering the
-        replacement of a word with the word from the dictionary.
-        """
-        pos = cursor.selectionStart()
-        cursor.beginEditBlock()
-        cursor.removeSelectedText()
-        cursor.insertText(word)
-        cursor.endEditBlock()
-        cursor.setPosition(pos)
-        self.setTextCursor(cursor)
-        return
-
-    @pyqtSlot(str, "QTextBlock")
-    def _addWord(self, word: str, block: QTextBlock) -> None:
-        """Slot for the spell check context menu triggered when the user
-        wants to add a word to the project dictionary.
-        """
-        logger.debug("Added '%s' to project dictionary", word)
-        SHARED.spelling.addWord(word)
-        self._qDocument.syntaxHighlighter.rehighlightBlock(block)
-        return
-
     @pyqtSlot()
     def _runDocumentTasks(self) -> None:
         """Run timer document tasks."""
@@ -1875,6 +1853,28 @@ class GuiDocEditor(QPlainTextEdit):
     #  Internal Functions
     ##
 
+    def _correctWord(self, cursor: QTextCursor, word: str) -> None:
+        """Slot for the spell check context menu triggering the
+        replacement of a word with the word from the dictionary.
+        """
+        pos = cursor.selectionStart()
+        cursor.beginEditBlock()
+        cursor.removeSelectedText()
+        cursor.insertText(word)
+        cursor.endEditBlock()
+        cursor.setPosition(pos)
+        self.setTextCursor(cursor)
+        return
+
+    def _addWord(self, word: str, block: QTextBlock, save: bool) -> None:
+        """Slot for the spell check context menu triggered when the user
+        wants to add a word to the project dictionary.
+        """
+        logger.debug("Added '%s' to project dictionary, %s", word, "saved" if save else "unsaved")
+        SHARED.spelling.addWord(word, save=save)
+        self._qDocument.syntaxHighlighter.rehighlightBlock(block)
+        return
+
     def _processTag(self, cursor: QTextCursor | None = None,
                     follow: bool = True, create: bool = False) -> nwTrinary:
         """Activated by Ctrl+Enter. Checks that we're in a block

novelwriter/gui/dochighlight.py

@@ -483,20 +483,19 @@ class TextBlockData(QTextBlockUserData):
         """
         if "[" in text:
             # Strip shortcodes
-            for rX in [RX_FMT_SC, RX_FMT_SV]:
-                for match in re.finditer(rX, text[offset:]):
-                    iS = match.start(0) + offset
-                    iE = match.end(0) + offset
-                    if iS >= 0 and iE >= 0:
-                        text = text[:iS] + " "*(iE - iS) + text[iE:]
+            for regEx in [RX_FMT_SC, RX_FMT_SV]:
+                for match in regEx.finditer(text, offset):
+                    if (s := match.start(0)) >= 0 and (e := match.end(0)) >= 0:
+                        pad = " "*(e - s)
+                        text = f"{text[:s]}{pad}{text[e:]}"
 
         self._spellErrors = []
         checker = SHARED.spelling
-        for match in re.finditer(RX_WORDS, text[offset:].replace("_", " ")):
+        for match in RX_WORDS.finditer(text.replace("_", " ")):
             if (
                 (word := match.group(0))
                 and not (word.isnumeric() or word.isupper() or checker.checkWord(word))
             ):
-                self._spellErrors.append((match.start(0) + offset, match.end(0) + offset))
+                self._spellErrors.append((match.start(0), match.end(0)))
 
         return self._spellErrors

tests/test_core/test_core_spellcheck.py

@@ -158,15 +158,21 @@ def testCoreSpell_Enchant(monkeypatch, mockGUI, fncPath):
         assert isinstance(spChk._enchant, FakeEnchant)
         assert spChk.checkWord("word") is True
         assert spChk.suggestWords("word") == []
-        assert spChk.addWord("word") is True
+        spChk.addWord("word")
+        assert "word" in spChk._userDict
 
     # Set the dict to None, and check enchant error handling
     spChk = NWSpellEnchant(project)
     spChk._enchant = None  # type: ignore
     assert spChk.checkWord("word") is True
     assert spChk.suggestWords("word") == []
-    assert spChk.addWord("word") is False
-    assert spChk.addWord("\n\t ") is False
+
+    spChk.addWord("word")
+    assert "word" not in spChk._userDict
+
+    spChk.addWord("\n\t ")
+    assert "\n\t " not in spChk._userDict
+
     assert spChk.describeDict() == ("", "")
 
     # Load the proper enchant package (twice)

tests/test_gui/test_gui_doceditor.py

@@ -453,9 +453,13 @@ def testGuiEditor_SpellChecking(qtbot, monkeypatch, nwGUI, projPath, ipsumText,
         ctxMenu = getMenuForPos(docEditor, 16)
         assert ctxMenu is not None
         actions = [x.text() for x in ctxMenu.actions() if x.text()]
+        assert "Ignore Word" in actions
         assert "Add Word to Dictionary" in actions
+
         assert "Lorax" not in SHARED.spelling._userDict
-        ctxMenu.actions()[7].trigger()
+        ctxMenu.actions()[7].trigger()  # Ignore
+        assert "Lorax" not in SHARED.spelling._userDict
+        ctxMenu.actions()[8].trigger()  # Add
         assert "Lorax" in SHARED.spelling._userDict
         ctxMenu.setObjectName("")
         ctxMenu.deleteLater()

tests/test_text/test_text_patterns.py

@@ -32,7 +32,7 @@ from novelwriter.text.patterns import REGEX_PATTERNS
 def allMatches(regEx: re.Pattern, text: str) -> list[list[str]]:
     """Get all matches for a regex."""
     result = []
-    for match in re.finditer(regEx, text):
+    for match in regEx.finditer(text):
         result.append([
             (match.group(n), match.start(n), match.end(n))
             for n in range((match.lastindex or 0) + 1)