Use Python regex for spell checking

novelwriter/constants.py

@@ -23,6 +23,8 @@ along with this program. If not, see <https://www.gnu.org/licenses/>.
 """
 from __future__ import annotations
 
+from re import UNICODE, compile
+
 from PyQt5.QtCore import QT_TRANSLATE_NOOP, QCoreApplication
 
 from novelwriter.enum import (
@@ -66,6 +68,10 @@ class nwRegEx:
     FMT_SC = r"(?i)(?<!\\)(\[[\/\!]?(?:b|i|s|u|m|sup|sub)\])"
     FMT_SV = r"(?i)(?<!\\)(\[(?:footnote):)(.+?)(?<!\\)(\])"
 
+    RX_WORDS = compile(r"\b([^\s\-\+\/–—\[\]:]+)\b", UNICODE)
+    RX_FMT_SC = compile(r"(?i)(?<!\\)(\[[\/\!]?(?:b|i|s|u|m|sup|sub)\])", UNICODE)
+    RX_FMT_SV = compile(r"(?i)(?<!\\)(\[(?:footnote):)(.+?)(?<!\\)(\])", UNICODE)
+
 
 class nwShortcode:
 

novelwriter/gui/dochighlight.py

@@ -25,6 +25,7 @@ along with this program. If not, see <https://www.gnu.org/licenses/>.
 from __future__ import annotations
 
 import logging
+import re
 
 from time import time
 
@@ -44,13 +45,6 @@ from novelwriter.types import QRegExUnicode
 
 logger = logging.getLogger(__name__)
 
-SPELLRX = QRegularExpression(r"\b[^\s\-\+\/–—\[\]:]+\b")
-SPELLRX.setPatternOptions(QRegExUnicode)
-SPELLSC = QRegularExpression(nwRegEx.FMT_SC)
-SPELLSC.setPatternOptions(QRegExUnicode)
-SPELLSV = QRegularExpression(nwRegEx.FMT_SV)
-SPELLSV.setPatternOptions(QRegExUnicode)
-
 BLOCK_NONE  = 0
 BLOCK_TEXT  = 1
 BLOCK_META  = 2
@@ -427,8 +421,8 @@ class GuiDocHighlighter(QSyntaxHighlighter):
             self.setCurrentBlockUserData(data)
 
         if self._spellCheck:
-            for xPos, xLen in data.spellCheck(text, xOff):
+            for xPos, xEnd in data.spellCheck(text, xOff):
-                for x in range(xPos, xPos+xLen):
+                for x in range(xPos, xEnd):
                     cFmt = self.format(x)
                     cFmt.merge(self._spellErr)
                     self.setFormat(x, 1, cFmt)
@@ -492,22 +486,22 @@ class TextBlockData(QTextBlockUserData):
         """
         if "[" in text:
             # Strip shortcodes
-            for rX in [SPELLSC, SPELLSV]:
+            for rX in [nwRegEx.RX_FMT_SC, nwRegEx.RX_FMT_SV]:
-                rxItt = rX.globalMatch(text, offset)
+                for match in re.finditer(rX, text[offset:]):
-                while rxItt.hasNext():
+                    iS = match.start(0) + offset
-                    rxMatch = rxItt.next()
+                    iE = match.end(0) + offset
-                    xPos = rxMatch.capturedStart(0)
+                    if iS >= 0 and iE >= 0:
-                    xLen = rxMatch.capturedLength(0)
+                        text = text[:iS] + " "*(iE - iS) + text[iE:]
-                    xEnd = rxMatch.capturedEnd(0)
-                    text = text[:xPos] + " "*xLen + text[xEnd:]
 
         self._spellErrors = []
-        rxSpell = SPELLRX.globalMatch(text.replace("_", " "), offset)
+        checker = SHARED.spelling
-        while rxSpell.hasNext():
+        for match in re.finditer(nwRegEx.RX_WORDS, text[offset:].replace("_", " ")):
-            rxMatch = rxSpell.next()
+            if (
-            if not SHARED.spelling.checkWord(rxMatch.captured(0)):
+                (word := match.group(0))
-                if not rxMatch.captured(0).isnumeric() and not rxMatch.captured(0).isupper():
+                and (iS := match.start(0)) >= 0
-                    self._spellErrors.append(
+                and (iE := match.end(0)) >= 0
-                        (rxMatch.capturedStart(0), rxMatch.capturedLength(0))
+                and not (word.isnumeric() or word.isupper() or checker.checkWord(word))
-                    )
+            ):
+                self._spellErrors.append((iS + offset, iE + offset))
+
         return self._spellErrors

novelwriter/gui/editordocument.py

@@ -107,8 +107,8 @@ class GuiTextDocument(QTextDocument):
             text = block.text()
             check = pos - block.position()
             if check >= 0:
-                for cPos, cLen in data.spellErrors:
+                for cPos, cEnd in data.spellErrors:
-                    cEnd = cPos + cLen
+                    cLen = cEnd - cPos
                     if cPos <= check <= cEnd:
                         word = text[cPos:cEnd]
                         return word, cPos, cLen, SHARED.spelling.suggestWords(word)