Lots of changes in regex and plugin and adjust langregex
This commit is contained in:
parent
10eadbee43
commit
6ad62d95c2
4 changed files with 86 additions and 71 deletions
|
|
@ -75,11 +75,8 @@ MODULES = [
|
|||
|
||||
# Info Message
|
||||
tqdm.write(
|
||||
Fore.LIGHTYELLOW_EX
|
||||
+ "WARNING: Once the translation starts do not close it unless you want to lose your \
|
||||
translated data. If a file fails or gets stuck, translated lines will remain translated so you don't have \
|
||||
to worry about being charged twice. You can simply copy the file generated in /translations back over to \
|
||||
/files and start the script again. It will skip over any translated text."
|
||||
Fore.CYAN
|
||||
+ "-Dazed MTL Tool -"
|
||||
+ Fore.RESET,
|
||||
end="\n\n",
|
||||
)
|
||||
|
|
|
|||
|
|
@ -214,7 +214,7 @@ def translateRegex(data, filename, translatedList):
|
|||
while i < len(data):
|
||||
voice = False
|
||||
lineRegexText = r'●(?:「| |（)+(.+?)(?:[」）]|$)+'
|
||||
lineRegexSpeaker = r"●.+●([^ 「（].+)"
|
||||
lineRegexSpeaker = r"text\(\"(.+)\"\)"
|
||||
choiceRegex = r"\$menu_item.+?,(.*?),"
|
||||
titleRegex = r"title\s'(.*)'$"
|
||||
speaker = ""
|
||||
|
|
@ -240,41 +240,34 @@ def translateRegex(data, filename, translatedList):
|
|||
# Speaker
|
||||
match = re.search(lineRegexSpeaker, data[i])
|
||||
if match:
|
||||
if match.group(1):
|
||||
if match.group(1) == "主人公":
|
||||
speaker = "Protagonist"
|
||||
else:
|
||||
response = getSpeaker(match.group(1))
|
||||
speaker = response[0]
|
||||
tokens[0] += response[1][0]
|
||||
tokens[1] += response[1][1]
|
||||
if translatedList:
|
||||
data[i] = data[i].replace(match.group(1), speaker)
|
||||
save_progress_lines(data, filename)
|
||||
i += 3
|
||||
else:
|
||||
speaker = None
|
||||
elif data[i] == "Protagonist\n":
|
||||
speaker = "Protagonist"
|
||||
i += 1
|
||||
response = getSpeaker(match.group(1))
|
||||
speaker = response[0]
|
||||
tokens[0] += response[1][0]
|
||||
tokens[1] += response[1][1]
|
||||
data[i] = data[i].replace(match.group(1), speaker)
|
||||
|
||||
# Dialogue
|
||||
match = re.search(lineRegexText, data[i])
|
||||
jaString = None
|
||||
if match and match.group(0).replace("\u3000", "") != '\n':
|
||||
# Set String
|
||||
jaString = match.group(1)
|
||||
|
||||
# Save Original String
|
||||
originalString = jaString
|
||||
# Grab multi-line text
|
||||
if "\\text" in data[i].strip():
|
||||
lines = []
|
||||
i += 1
|
||||
|
||||
# Pass 1
|
||||
if not translatedList:
|
||||
while "\\endtext" not in data[i].strip():
|
||||
lines.append(data[i])
|
||||
i += 1
|
||||
if lines:
|
||||
jaString = "".join(lines).replace("\n", "")
|
||||
|
||||
# Save Original String
|
||||
originalString = jaString
|
||||
|
||||
# Strip Spaces
|
||||
jaString = jaString.strip()
|
||||
|
||||
# Remove Textwrap
|
||||
jaString = jaString.replace('\\n', ' ')
|
||||
jaString = jaString.replace('\n', ' ')
|
||||
|
||||
if jaString:
|
||||
if speaker:
|
||||
|
|
@ -305,13 +298,12 @@ def translateRegex(data, filename, translatedList):
|
|||
translatedText = translatedText.replace(">", ")")
|
||||
|
||||
# Textwrap
|
||||
translatedText = dazedwrap.wrapText(translatedText, width=WIDTH).replace("\n", "\\n")
|
||||
translatedText = dazedwrap.wrapText(translatedText, width=WIDTH)
|
||||
|
||||
# Set Data
|
||||
if "「" in data[i-1] and "」" not in translatedText:
|
||||
data[i] = data[i].replace(originalString, f"「{translatedText}」")
|
||||
else:
|
||||
data[i] = data[i].replace(originalString, f"{translatedText}")
|
||||
while "\\endtext" not in data[i].strip():
|
||||
del data[i]
|
||||
data.insert(i, f"{translatedText}\n")
|
||||
save_progress_lines(data, filename)
|
||||
|
||||
# Choices
|
||||
|
|
|
|||
|
|
@ -52,7 +52,7 @@ _speakerCacheLock = threading.Lock()
|
|||
SPEAKER_COLLECTED = [] # Original speaker names collected during parse mode (untranslated)
|
||||
|
||||
# Regex - Need to change this if you want to translate from/to other languages. Default is Japanese Regex
|
||||
LANGREGEX = r"[一-龠ぁ-ゔァ-ヴーａ-ｚＡ-Ｚ０-９\uFF61-\uFF9F]+"
|
||||
LANGREGEX = r"[\u3000-\u303F\u3040-\u309F\u30A0-\u30FF\u31F0-\u31FF\u3400-\u4DBF\u4E00-\u9FFF\uF900-\uFAFF\uFF00-\uFFEF]+"
|
||||
|
||||
# Get pricing configuration based on the model
|
||||
PRICING_CONFIG = getPricingConfig(MODEL)
|
||||
|
|
@ -1450,6 +1450,7 @@ def searchCodes(page, pbar, jobList, filename):
|
|||
jaString = jaString.replace("･", ".")
|
||||
jaString = jaString.replace("「", '"')
|
||||
jaString = jaString.replace("」", '"')
|
||||
jaString = jaString.replace("～", '~')
|
||||
|
||||
# Check if there is text to translate
|
||||
if not re.search(r"\w+", jaString):
|
||||
|
|
|
|||
|
|
@ -101,17 +101,21 @@ def handlePlugin(filename, estimate):
|
|||
|
||||
else:
|
||||
try:
|
||||
with open("translated/" + filename, "w", encoding="utf_8", errors="ignore") as outFile:
|
||||
start = time.time()
|
||||
translatedData = openFiles(filename)
|
||||
# Perform translation first; incremental progress writes happen during translation.
|
||||
# We no longer keep the destination file open simultaneously to avoid Windows locking issues
|
||||
# during atomic os.replace operations inside save_progress_lines.
|
||||
start = time.time()
|
||||
translatedData = openFiles(filename)
|
||||
|
||||
# Print Result
|
||||
end = time.time()
|
||||
outFile.writelines(translatedData[0])
|
||||
tqdm.write(getResultString(translatedData, end - start, filename))
|
||||
with LOCK:
|
||||
TOKENS[0] += translatedData[1][0]
|
||||
TOKENS[1] += translatedData[1][1]
|
||||
# Ensure final state is flushed (in case no incremental writes occurred for some reason).
|
||||
save_progress_lines(translatedData[0], filename)
|
||||
|
||||
# Print Result
|
||||
end = time.time()
|
||||
tqdm.write(getResultString(translatedData, end - start, filename))
|
||||
with LOCK:
|
||||
TOKENS[0] += translatedData[1][0]
|
||||
TOKENS[1] += translatedData[1][1]
|
||||
except Exception:
|
||||
traceback.print_exc()
|
||||
return "Fail"
|
||||
|
|
@ -184,18 +188,31 @@ def save_progress_lines(lines, filename, encoding="utf_8"):
|
|||
try:
|
||||
if ESTIMATE:
|
||||
return
|
||||
global LOCK
|
||||
os.makedirs("translated", exist_ok=True)
|
||||
tmp_fd, tmp_path = tempfile.mkstemp(prefix=f"{filename}.", suffix=".tmp", dir="translated")
|
||||
try:
|
||||
with os.fdopen(tmp_fd, "w", encoding=encoding, newline="\n", errors="ignore") as tmp_file:
|
||||
tmp_file.writelines(lines)
|
||||
os.replace(tmp_path, os.path.join("translated", filename))
|
||||
finally:
|
||||
if os.path.exists(tmp_path):
|
||||
try:
|
||||
os.remove(tmp_path)
|
||||
except OSError:
|
||||
pass
|
||||
# Use a single lock to prevent concurrent replace attempts on Windows (which causes PermissionError)
|
||||
with LOCK:
|
||||
tmp_fd, tmp_path = tempfile.mkstemp(prefix=f"{filename}.", suffix=".tmp", dir="translated")
|
||||
try:
|
||||
with os.fdopen(tmp_fd, "w", encoding=encoding, newline="\n", errors="ignore") as tmp_file:
|
||||
tmp_file.writelines(lines)
|
||||
|
||||
dest_path = os.path.join("translated", filename)
|
||||
# Retry a few times in case another process/thread still has the file open momentarily.
|
||||
for attempt in range(5):
|
||||
try:
|
||||
os.replace(tmp_path, dest_path)
|
||||
break
|
||||
except PermissionError:
|
||||
if attempt == 4:
|
||||
raise
|
||||
time.sleep(0.1 * (attempt + 1))
|
||||
finally:
|
||||
if os.path.exists(tmp_path):
|
||||
try:
|
||||
os.remove(tmp_path)
|
||||
except OSError:
|
||||
pass
|
||||
except Exception:
|
||||
traceback.print_exc()
|
||||
|
||||
|
|
@ -235,37 +252,44 @@ def translatePlugin(data, pbar, filename, translatedList):
|
|||
while i < len(data):
|
||||
voice = False
|
||||
speaker = ""
|
||||
newline = r"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\n"
|
||||
colorCode = r"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\c"
|
||||
|
||||
# Custom
|
||||
# Useful Regex's
|
||||
# r'"Text[\\]+":[\\]+"(.+?)[\\]+",'
|
||||
# r'"HelpText[\\]+":[\\]+"(.+?)[\\]+",'
|
||||
# r"this.drawTextEx\(\\'(.+?)\',"
|
||||
regex = r"this.drawTextEx\(\\'(.+?)\',"
|
||||
regex = r'Text[\\]+":[\\]+"(.*?)[\\]+?[\\]'
|
||||
matchList = re.findall(regex, data[i])
|
||||
if len(matchList) > 0:
|
||||
for match in matchList:
|
||||
# Save Original String
|
||||
jaString = match
|
||||
originalString = jaString
|
||||
newline = None
|
||||
colorCode = None
|
||||
|
||||
# Make sure it contains Japanese
|
||||
if not re.search(LANGREGEX, jaString):
|
||||
continue
|
||||
|
||||
# Make sure didn't grab \\
|
||||
if re.search(r"^[\\]+$", jaString):
|
||||
i += 1
|
||||
continue
|
||||
|
||||
# Replace \n and \c
|
||||
jaString = re.sub(r"\\+n", r"\\n", jaString)
|
||||
jaString = re.sub(r"\\+C", r"\\C", jaString)
|
||||
if re.search(r"\\+n", jaString):
|
||||
newline = re.search(r"\\+n", jaString).group(0)
|
||||
jaString = re.sub(r"\\+n", r"\\n", jaString)
|
||||
if re.search(r"\\+C", jaString):
|
||||
colorCode = re.search(r"\\+C", jaString).group(0)
|
||||
jaString = re.sub(r"\\+C", r"\\C", jaString)
|
||||
|
||||
# Remove any textwrap
|
||||
jaString = jaString.replace("\\n", " ")
|
||||
|
||||
if jaString.replace("\u3000", "") and jaString:
|
||||
# Pass 1
|
||||
if setData == False:
|
||||
if setData == False and jaString.strip():
|
||||
custom.append(jaString.strip())
|
||||
|
||||
# Pass 2
|
||||
|
|
@ -283,12 +307,13 @@ def translatePlugin(data, pbar, filename, translatedList):
|
|||
translatedText = re.sub(r"([^\\'])'", r"\1՚", translatedText)
|
||||
translatedText = re.sub(r"([^\\'])\"", r"\1՚", translatedText)
|
||||
|
||||
# Textwrap
|
||||
translatedText = dazedwrap.wrapText(translatedText, WIDTH)
|
||||
|
||||
# Replace \n and \c
|
||||
translatedText = re.sub(r"\n", re.escape(newline), translatedText)
|
||||
translatedText = re.sub(r"\n", re.escape(colorCode), translatedText)
|
||||
if newline:
|
||||
# Textwrap
|
||||
# translatedText = dazedwrap.wrapText(translatedText, WIDTH)
|
||||
translatedText = re.sub(r"\n", re.escape(newline), translatedText)
|
||||
if colorCode:
|
||||
translatedText = re.sub(r"\n", re.escape(colorCode), translatedText)
|
||||
|
||||
# Set Data
|
||||
with open("log/translations.txt", "a+", encoding="utf-8") as tlFile:
|
||||
|
|
|
|||
Loading…
Reference in a new issue