Single map file
This commit is contained in:
parent
e91001ea66
commit
3e5c093800
1 changed files with 381 additions and 199 deletions
|
|
@ -1,4 +1,5 @@
|
||||||
from concurrent.futures import ThreadPoolExecutor, as_completed
|
from concurrent.futures import ThreadPoolExecutor, as_completed
|
||||||
|
import json
|
||||||
import os
|
import os
|
||||||
from pathlib import Path
|
from pathlib import Path
|
||||||
import re
|
import re
|
||||||
|
|
@ -8,17 +9,13 @@ import threading
|
||||||
import time
|
import time
|
||||||
import traceback
|
import traceback
|
||||||
import tiktoken
|
import tiktoken
|
||||||
|
from ruamel.yaml import YAML
|
||||||
|
|
||||||
from colorama import Fore
|
from colorama import Fore
|
||||||
from dotenv import load_dotenv
|
from dotenv import load_dotenv
|
||||||
import openai
|
import openai
|
||||||
from retry import retry
|
from retry import retry
|
||||||
from tqdm import tqdm
|
from tqdm import tqdm
|
||||||
from ruamel.yaml import YAML
|
|
||||||
|
|
||||||
#Yaml
|
|
||||||
yaml = YAML()
|
|
||||||
yaml.preserve_quotes = True
|
|
||||||
|
|
||||||
#Globals
|
#Globals
|
||||||
load_dotenv()
|
load_dotenv()
|
||||||
|
|
@ -30,7 +27,7 @@ PROMPT = Path('prompt.txt').read_text(encoding='utf-8')
|
||||||
THREADS = 20
|
THREADS = 20
|
||||||
LOCK = threading.Lock()
|
LOCK = threading.Lock()
|
||||||
WIDTH = 70
|
WIDTH = 70
|
||||||
LISTWIDTH = 75
|
LISTWIDTH = 80
|
||||||
MAXHISTORY = 10
|
MAXHISTORY = 10
|
||||||
ESTIMATE = ''
|
ESTIMATE = ''
|
||||||
TOTALCOST = 0
|
TOTALCOST = 0
|
||||||
|
|
@ -44,14 +41,20 @@ LEAVE=False
|
||||||
|
|
||||||
# Flags
|
# Flags
|
||||||
CODE401 = True
|
CODE401 = True
|
||||||
CODE102 = True
|
CODE405 = False
|
||||||
|
CODE102 = False
|
||||||
CODE122 = False
|
CODE122 = False
|
||||||
CODE101 = False
|
CODE101 = False
|
||||||
CODE355655 = False
|
CODE355655 = False
|
||||||
CODE357 = False
|
CODE357 = False
|
||||||
CODE356 = False
|
CODE657 = False
|
||||||
|
CODE356 = True
|
||||||
CODE320 = False
|
CODE320 = False
|
||||||
|
CODE324 = False
|
||||||
CODE111 = False
|
CODE111 = False
|
||||||
|
CODE408 = False
|
||||||
|
|
||||||
|
yaml=YAML() # default, if not specfied, is 'rt' (round-trip)
|
||||||
|
|
||||||
def handleACE(filename, estimate):
|
def handleACE(filename, estimate):
|
||||||
global ESTIMATE, TOKENS, TOTALTOKENS, TOTALCOST
|
global ESTIMATE, TOKENS, TOTALTOKENS, TOTALCOST
|
||||||
|
|
@ -172,25 +175,23 @@ def parseMap(data, filename):
|
||||||
events = data['events']
|
events = data['events']
|
||||||
global LOCK
|
global LOCK
|
||||||
|
|
||||||
# Translate displayName for Map files
|
|
||||||
# if 'Map' in filename:
|
|
||||||
# response = translateGPT(data['displayName'], 'Reply with only the english translated name', False)
|
|
||||||
# totalTokens += response[1]
|
|
||||||
# data['displayName'] = response[0].strip('.\"')
|
|
||||||
|
|
||||||
# Get total for progress bar
|
# Get total for progress bar
|
||||||
for eventKey in events:
|
for key in events:
|
||||||
if eventKey is not None:
|
if key is not None:
|
||||||
for page in events[eventKey]['pages']:
|
for page in events[key]['pages']:
|
||||||
totalLines += len(page['list'])
|
totalLines += len(page['list'])
|
||||||
|
|
||||||
with tqdm(bar_format=BAR_FORMAT, position=POSITION, total=totalLines, leave=LEAVE) as pbar:
|
with tqdm(bar_format=BAR_FORMAT, position=POSITION, total=totalLines, leave=LEAVE) as pbar:
|
||||||
pbar.desc=filename
|
pbar.desc=filename
|
||||||
pbar.total=totalLines
|
pbar.total=totalLines
|
||||||
with ThreadPoolExecutor(max_workers=THREADS) as executor:
|
with ThreadPoolExecutor(max_workers=THREADS) as executor:
|
||||||
for eventKey in events:
|
for key in events:
|
||||||
if eventKey is not None:
|
if key is not None:
|
||||||
futures = [executor.submit(searchCodes, page, pbar) for page in events[eventKey]['pages'] if page is not None]
|
# This translates text above items on the map.
|
||||||
|
# if 'LB:' in event['note']:
|
||||||
|
# totalTokens += translateNote(event, r'(?<=LB:)[^u0000-u0080]+')
|
||||||
|
|
||||||
|
futures = [executor.submit(searchCodes, page, pbar) for page in events[key]['pages'] if page is not None]
|
||||||
for future in as_completed(futures):
|
for future in as_completed(futures):
|
||||||
try:
|
try:
|
||||||
totalTokens += future.result()
|
totalTokens += future.result()
|
||||||
|
|
@ -198,6 +199,31 @@ def parseMap(data, filename):
|
||||||
return [data, totalTokens, e]
|
return [data, totalTokens, e]
|
||||||
return [data, totalTokens, None]
|
return [data, totalTokens, None]
|
||||||
|
|
||||||
|
def translateNote(event, regex):
|
||||||
|
# Regex that only matches text inside LB.
|
||||||
|
jaString = event['note']
|
||||||
|
|
||||||
|
match = re.search(regex, jaString)
|
||||||
|
if match:
|
||||||
|
jaString = match.group(1)
|
||||||
|
# Need to remove outside code
|
||||||
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】]+', '', jaString)
|
||||||
|
jaString = re.sub(r'[^一-龠ぁ-ゔァ-ヴー\<\>【】。!?]+$', '', jaString)
|
||||||
|
oldjaString = jaString
|
||||||
|
|
||||||
|
# Remove any textwrap
|
||||||
|
jaString = re.sub(r'\n', ' ', jaString)
|
||||||
|
|
||||||
|
response = translateGPT(jaString, '', True)
|
||||||
|
translatedText = response[0]
|
||||||
|
|
||||||
|
# Textwrap
|
||||||
|
translatedText = textwrap.fill(translatedText, width=LISTWIDTH)
|
||||||
|
|
||||||
|
event['note'] = event['note'].replace(oldjaString, translatedText)
|
||||||
|
return response[1]
|
||||||
|
return 0
|
||||||
|
|
||||||
def parseCommonEvents(data, filename):
|
def parseCommonEvents(data, filename):
|
||||||
totalTokens = 0
|
totalTokens = 0
|
||||||
totalLines = 0
|
totalLines = 0
|
||||||
|
|
@ -304,10 +330,12 @@ def parseSystem(data, filename):
|
||||||
for term in data['terms']:
|
for term in data['terms']:
|
||||||
termList = data['terms'][term]
|
termList = data['terms'][term]
|
||||||
totalLines += len(termList)
|
totalLines += len(termList)
|
||||||
totalLines += len(data['game_title'])
|
totalLines += len(data['gameTitle'])
|
||||||
totalLines += len(data['armor_types'])
|
totalLines += len(data['terms']['messages'])
|
||||||
totalLines += len(data['skill_types'])
|
totalLines += len(data['variables'])
|
||||||
totalLines += len(data['weapon_types'])
|
totalLines += len(data['equipTypes'])
|
||||||
|
totalLines += len(data['armorTypes'])
|
||||||
|
totalLines += len(data['skillTypes'])
|
||||||
|
|
||||||
with tqdm(bar_format=BAR_FORMAT, position=POSITION, total=totalLines, leave=LEAVE) as pbar:
|
with tqdm(bar_format=BAR_FORMAT, position=POSITION, total=totalLines, leave=LEAVE) as pbar:
|
||||||
pbar.desc=filename
|
pbar.desc=filename
|
||||||
|
|
@ -324,9 +352,11 @@ def searchThings(name, pbar):
|
||||||
|
|
||||||
# Set the context of what we are translating
|
# Set the context of what we are translating
|
||||||
responseList = []
|
responseList = []
|
||||||
responseList.append(translateGPT(name['name'], 'Reply with only the english translated menu item name.', False))
|
responseList.append(translateGPT(name['name'], 'Reply with only the English translation of the RPG Item name.', False))
|
||||||
responseList.append(translateGPT(name['description'], 'Reply with only the english translated description.', True))
|
responseList.append(translateGPT(name['description'], 'Reply with only the English translation of the description.', False))
|
||||||
# responseList.append(translateGPT(name['note'], 'Reply with only the english translated note.', False))
|
|
||||||
|
# if '<SG説明:' in name['note']:
|
||||||
|
# tokens += translateNote(name, r'<SG説明:([\s\S]*?)>')
|
||||||
|
|
||||||
# Extract all our translations in a list from response
|
# Extract all our translations in a list from response
|
||||||
for i in range(len(responseList)):
|
for i in range(len(responseList)):
|
||||||
|
|
@ -347,27 +377,38 @@ def searchNames(name, pbar, context):
|
||||||
|
|
||||||
# Set the context of what we are translating
|
# Set the context of what we are translating
|
||||||
if 'Actors' in context:
|
if 'Actors' in context:
|
||||||
newContext = 'Reply with only the english translation. The original text is a menu item.'
|
newContext = 'Reply with only the english translation of the NPC name'
|
||||||
if 'Armors' in context:
|
if 'Armors' in context:
|
||||||
newContext = 'Reply with only the english translation.'
|
newContext = 'Reply with only the english translation of the RPG armor/clothing name'
|
||||||
if 'Classes' in context:
|
if 'Classes' in context:
|
||||||
newContext = 'Reply with only the english translated class name'
|
newContext = 'Reply with only the english translation of the RPG class name'
|
||||||
if 'MapInfos' in context:
|
if 'MapInfos' in context:
|
||||||
newContext = 'Reply with only the english translated map name'
|
newContext = 'Reply with only the english translation of the location name'
|
||||||
if 'Enemies' in context:
|
if 'Enemies' in context:
|
||||||
newContext = 'Reply with only the english translated enemy'
|
newContext = 'Reply with only the english translation of the enemy NPC name'
|
||||||
if 'Weapons' in context:
|
if 'Weapons' in context:
|
||||||
newContext = 'Reply with only the english translated weapon name'
|
newContext = 'Reply with only the english translation of the RPG weapon name'
|
||||||
|
|
||||||
# Extract Data
|
# Extract Data
|
||||||
responseList = []
|
responseList = []
|
||||||
responseList.append(translateGPT(name['name'], newContext, True))
|
responseList.append(translateGPT(name['name'], newContext, False))
|
||||||
if 'Actors' in context:
|
if 'Actors' in context:
|
||||||
responseList.append(translateGPT(name['profile'], '', True))
|
responseList.append(translateGPT(name['profile'], '', True))
|
||||||
|
responseList.append(translateGPT(name['nickname'], 'Reply with ONLY the english translation of the NPC nickname', False))
|
||||||
|
|
||||||
if 'Armors' in context or 'Weapons' in context:
|
if 'Armors' in context or 'Weapons' in context:
|
||||||
responseList.append(translateGPT(name['description'], '', True))
|
responseList.append(translateGPT(name['description'], '', True))
|
||||||
|
|
||||||
|
if 'Enemies' in context:
|
||||||
|
if 'desc1' in name['note']:
|
||||||
|
tokens += translateNote(name, r'<desc1:([^>]*)>')
|
||||||
|
|
||||||
|
if 'desc2' in name['note']:
|
||||||
|
tokens += translateNote(name, r'<desc2:([^>]*)>')
|
||||||
|
|
||||||
|
if 'desc3' in name['note']:
|
||||||
|
tokens += translateNote(name, r'<desc3:([^>]*)>')
|
||||||
|
|
||||||
# Extract all our translations in a list from response
|
# Extract all our translations in a list from response
|
||||||
for i in range(len(responseList)):
|
for i in range(len(responseList)):
|
||||||
tokens += responseList[i][1]
|
tokens += responseList[i][1]
|
||||||
|
|
@ -378,10 +419,14 @@ def searchNames(name, pbar, context):
|
||||||
if 'Actors' in context:
|
if 'Actors' in context:
|
||||||
translatedText = textwrap.fill(responseList[1], LISTWIDTH)
|
translatedText = textwrap.fill(responseList[1], LISTWIDTH)
|
||||||
name['profile'] = translatedText.strip('\"')
|
name['profile'] = translatedText.strip('\"')
|
||||||
|
translatedText = textwrap.fill(responseList[2], LISTWIDTH)
|
||||||
|
name['nickname'] = translatedText.strip('\"')
|
||||||
|
|
||||||
if 'Armors' in context or 'Weapons' in context:
|
if 'Armors' in context or 'Weapons' in context:
|
||||||
translatedText = textwrap.fill(responseList[1], LISTWIDTH)
|
translatedText = textwrap.fill(responseList[1], LISTWIDTH)
|
||||||
name['description'] = translatedText.strip('\"')
|
name['description'] = translatedText.strip('\"')
|
||||||
|
if '<SG説明:' in name['note']:
|
||||||
|
tokens += translateNote(name, r'<SG説明:([^>]*)>')
|
||||||
pbar.update(1)
|
pbar.update(1)
|
||||||
|
|
||||||
return tokens
|
return tokens
|
||||||
|
|
@ -394,13 +439,8 @@ def searchCodes(page, pbar):
|
||||||
tokens = 0
|
tokens = 0
|
||||||
speaker = ''
|
speaker = ''
|
||||||
match = []
|
match = []
|
||||||
speakerCaught = False
|
|
||||||
global LOCK
|
global LOCK
|
||||||
|
|
||||||
# Regex
|
|
||||||
subVarRegex = r'(\\+[a-zA-Z]+)\[([a-zA-Z0-9一-龠ぁ-ゔァ-ヴー\s]+)\]'
|
|
||||||
reSubVarRegex = r'\<([\\a-zA-Z]+)([a-zA-Z0-9一-龠ぁ-ゔァ-ヴー\s]+)\>'
|
|
||||||
|
|
||||||
try:
|
try:
|
||||||
for i in range(len(page['list'])):
|
for i in range(len(page['list'])):
|
||||||
with LOCK:
|
with LOCK:
|
||||||
|
|
@ -410,22 +450,36 @@ def searchCodes(page, pbar):
|
||||||
### IF these crash or fail your game will do the same. Use the flags to skip codes.
|
### IF these crash or fail your game will do the same. Use the flags to skip codes.
|
||||||
|
|
||||||
## Event Code: 401 Show Text
|
## Event Code: 401 Show Text
|
||||||
if page['list'][i]['c'] == 401 and CODE401 == True:
|
if page['list'][i]['c'] == 401 and CODE401 == True or page['list'][i]['c'] == 405 and CODE405:
|
||||||
jaString = page['list'][i]['p'][0]
|
jaString = page['list'][i]['p'][0]
|
||||||
|
if "peek inside" in jaString:
|
||||||
|
print('hi')
|
||||||
oldjaString = jaString
|
oldjaString = jaString
|
||||||
jaString = jaString.replace('゙', '')
|
jaString = jaString.replace('゙', '')
|
||||||
jaString = jaString.replace('。', '.')
|
jaString = jaString.replace('。', '.')
|
||||||
jaString = re.sub(r'([\u3000-\uffef])\1{1,}', r'\1', jaString)
|
jaString = jaString.replace('・', '.')
|
||||||
|
jaString = jaString.replace('‶', '')
|
||||||
|
jaString = jaString.replace('”', '')
|
||||||
|
jaString = jaString.replace('ー', '-')
|
||||||
|
jaString = jaString.replace('―', '-')
|
||||||
|
jaString = jaString.replace('…', '...')
|
||||||
|
jaString = re.sub(r'([\u3000-\uffef])\1{3,}', r'\1\1\1', jaString)
|
||||||
|
|
||||||
# Using this to keep track of 401's in a row. Throws IndexError at EndOfList (Expected Behavior)
|
# Using this to keep track of 401's in a row. Throws IndexError at EndOfList (Expected Behavior)
|
||||||
currentGroup.append(jaString)
|
currentGroup.append(jaString)
|
||||||
|
|
||||||
while (page['list'][i+1]['c'] == 401):
|
while (page['list'][i+1]['c'] == 401 or page['list'][i+1]['c'] == 405):
|
||||||
del page['list'][i]
|
del page['list'][i]
|
||||||
jaString = page['list'][i]['p'][0]
|
jaString = page['list'][i]['p'][0]
|
||||||
jaString = jaString.replace('゙', '')
|
jaString = jaString.replace('゙', '')
|
||||||
jaString = jaString.replace('。', '.')
|
jaString = jaString.replace('。', '.')
|
||||||
jaString = re.sub(r'([\u3000-\uffef])\1{1,}', r'\1', jaString)
|
jaString = jaString.replace('・', '.')
|
||||||
|
jaString = jaString.replace('‶', '')
|
||||||
|
jaString = jaString.replace('”', '')
|
||||||
|
jaString = jaString.replace('ー', '-')
|
||||||
|
jaString = jaString.replace('―', '-')
|
||||||
|
jaString = jaString.replace('…', '...')
|
||||||
|
jaString = re.sub(r'([\u3000-\uffef])\1{3,}', r'\1\1\1', jaString)
|
||||||
currentGroup.append(jaString)
|
currentGroup.append(jaString)
|
||||||
|
|
||||||
# Join up 401 groups for better translation.
|
# Join up 401 groups for better translation.
|
||||||
|
|
@ -433,51 +487,32 @@ def searchCodes(page, pbar):
|
||||||
finalJAString = ' '.join(currentGroup)
|
finalJAString = ' '.join(currentGroup)
|
||||||
|
|
||||||
# Check for speaker
|
# Check for speaker
|
||||||
if '\\nw' in finalJAString:
|
if '\\N' in finalJAString:
|
||||||
match = re.findall(r'([\\]+nw\[([a-zA-Z0-9一-龠ぁ-ゔァ-ヴー\s]+)\])', finalJAString)
|
match = re.findall(r'[\\]+N<([一-龠ぁ-ゔァ-ヴー]+)>', finalJAString)
|
||||||
if len(match) != 0:
|
if len(match) != 0:
|
||||||
response = translateGPT(match[0][1], 'Reply with only the english translated actor', False)
|
response = translateGPT(match[0], 'Reply with only the english translation of the NPC name', False)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
speaker = response[0].strip('.')
|
speaker = response[0].strip('.')
|
||||||
|
|
||||||
finalJAString = re.sub(r'([\\]+nw\[[a-zA-Z0-9一-龠ぁ-ゔァ-ヴー\s]+\])', '', finalJAString)
|
finalJAString = finalJAString.replace(match[0], speaker)
|
||||||
|
|
||||||
# Need to remove outside code and put it back later
|
# Need to remove outside code and put it back later
|
||||||
startString = re.search(r'^[^ぁ-んァ-ン一-龯【】()「」a-zA-ZA-Z0-9\\]+', finalJAString)
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー【】()「」『』a-zA-Z0-9A-Z0-9\\]+', finalJAString)
|
||||||
finalJAString = re.sub(r'^[^ぁ-んァ-ン一-龯【】()「」a-zA-ZA-Z0-9\\]+', '', finalJAString)
|
finalJAString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー【】()「」『』a-zA-Z0-9A-Z0-9\\]+', '', finalJAString)
|
||||||
if startString is None: startString = ''
|
if startString is None: startString = ''
|
||||||
else: startString = startString.group()
|
else: startString = startString.group()
|
||||||
|
|
||||||
# Sub Vars
|
|
||||||
finalJAString = re.sub(subVarRegex, r'<\1\2>', finalJAString)
|
|
||||||
|
|
||||||
# Remove any textwrap
|
# Remove any textwrap
|
||||||
finalJAString = re.sub(r'\n', ' ', finalJAString)
|
finalJAString = re.sub(r'\n', ' ', finalJAString)
|
||||||
|
|
||||||
# Translate
|
# Translate
|
||||||
if speaker != '':
|
response = translateGPT(finalJAString, 'Previous Text for Context: ' + '\n\n'.join(textHistory), True)
|
||||||
response = translateGPT(finalJAString, 'Previously Translated Text for Context: ' + ' '.join(textHistory) \
|
|
||||||
+ '\n\n\n###\n\n\nCurrent Speaker: ' + speaker, True)
|
|
||||||
else:
|
|
||||||
response = translateGPT(finalJAString, 'Previous Translated Text for Context: ' + ' '.join(textHistory), True)
|
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
translatedText = response[0]
|
translatedText = response[0]
|
||||||
|
|
||||||
# ReSub Vars
|
|
||||||
translatedText = re.sub(reSubVarRegex, r'\1[\2]', translatedText)
|
|
||||||
|
|
||||||
# TextHistory is what we use to give GPT Context, so thats appended here.
|
# TextHistory is what we use to give GPT Context, so thats appended here.
|
||||||
# rawTranslatedText = re.sub(r'[\\<>]+[a-zA-Z]+\[[a-zA-Z0-9]+\]', '', translatedText)
|
# rawTranslatedText = re.sub(r'[\\<>]+[a-zA-Z]+\[[a-zA-Z0-9]+\]', '', translatedText)
|
||||||
if speaker != '':
|
textHistory.append('\"' + translatedText + '\"')
|
||||||
textHistory.append(speaker + ': ' + translatedText)
|
|
||||||
else:
|
|
||||||
textHistory.append('\"' + translatedText + '\"')
|
|
||||||
|
|
||||||
# Name Handling
|
|
||||||
if len(match) != 0:
|
|
||||||
name = '\\nw[' + speaker + ']'
|
|
||||||
if name not in translatedText:
|
|
||||||
translatedText = translatedText + '\\nw[' + speaker + ']'
|
|
||||||
|
|
||||||
# if speakerCaught == True:
|
# if speakerCaught == True:
|
||||||
# translatedText = speakerRaw + ':\n' + translatedText
|
# translatedText = speakerRaw + ':\n' + translatedText
|
||||||
|
|
@ -490,7 +525,10 @@ def searchCodes(page, pbar):
|
||||||
translatedText = startString + translatedText
|
translatedText = startString + translatedText
|
||||||
|
|
||||||
# Set Data
|
# Set Data
|
||||||
page['list'][i]['p'][0] = translatedText.replace('\"', '')
|
translatedText = translatedText.replace('ッ', '')
|
||||||
|
translatedText = translatedText.replace('っ', '')
|
||||||
|
translatedText = translatedText.replace('\"', '')
|
||||||
|
page['list'][i]['p'][0] = translatedText
|
||||||
speaker = ''
|
speaker = ''
|
||||||
match = []
|
match = []
|
||||||
|
|
||||||
|
|
@ -506,53 +544,49 @@ def searchCodes(page, pbar):
|
||||||
continue
|
continue
|
||||||
|
|
||||||
# Definitely don't want to mess with files
|
# Definitely don't want to mess with files
|
||||||
if '_' in jaString:
|
if '■' in jaString or '_' in jaString:
|
||||||
continue
|
|
||||||
|
|
||||||
# If there isn't any Japanese in the text just skip
|
|
||||||
if re.search(r'[a-zA-Z0-9]+', jaString):
|
|
||||||
continue
|
continue
|
||||||
|
|
||||||
# If there isn't any Japanese in the text just skip
|
# If there isn't any Japanese in the text just skip
|
||||||
if not re.search(r'[一-龠]+|[ぁ-ゔ]+|[ァ-ヴー]+', jaString):
|
if not re.search(r'[一-龠]+|[ぁ-ゔ]+|[ァ-ヴー]+', jaString):
|
||||||
continue
|
continue
|
||||||
|
|
||||||
|
# Definitely don't want to mess with files
|
||||||
|
if '\"' not in jaString:
|
||||||
|
continue
|
||||||
|
|
||||||
# Remove repeating characters because it confuses ChatGPT
|
# Remove outside text
|
||||||
jaString = re.sub(r'([\u3000-\uffef])\1{2,}', r'\1\1', jaString)
|
oldjaString = jaString
|
||||||
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】\\]+', jaString)
|
||||||
# Need to remove outside code and put it back later
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】\\]+', '', jaString)
|
||||||
startString = re.search(r'^[^ぁ-んァ-ン一-龯\<\>【】]+', jaString)
|
endString = re.search(r'[^一-龠ぁ-ゔァ-ヴー\<\>【】。!?\\]+$', jaString)
|
||||||
jaString = re.sub(r'^[^ぁ-んァ-ン一-龯\<\>【】]+', '', jaString)
|
jaString = re.sub(r'[^一-龠ぁ-ゔァ-ヴー\<\>【】。!?\\]+$', '', jaString)
|
||||||
endString = re.search(r'[^ぁ-んァ-ン一-龯\<\>【】 。!?]+$', jaString)
|
|
||||||
jaString = re.sub(r'[^ぁ-んァ-ン一-龯\<\>【】 。!?]+$', '', jaString)
|
|
||||||
if startString is None: startString = ''
|
if startString is None: startString = ''
|
||||||
else: startString = startString.group()
|
else: startString = startString.group()
|
||||||
if endString is None: endString = ''
|
if endString is None: endString = ''
|
||||||
else: endString = endString.group()
|
else: endString = endString.group()
|
||||||
|
|
||||||
# Sub Vars
|
|
||||||
jaString = re.sub(subVarRegex, r'<\1\2>', jaString)
|
|
||||||
|
|
||||||
# Translate
|
# Translate
|
||||||
response = translateGPT(jaString, 'Reply with only the english translation', False)
|
response = translateGPT(jaString, '', True)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
translatedText = response[0]
|
translatedText = response[0]
|
||||||
|
|
||||||
# Remove characters that may break scripts
|
# Remove characters that may break scripts
|
||||||
charList = ['.', '\"', '\\n', '\\']
|
charList = ['.', '\"', "\'"]
|
||||||
for char in charList:
|
for char in charList:
|
||||||
translatedText = translatedText.replace(char, '')
|
translatedText = translatedText.replace(char, '')
|
||||||
|
|
||||||
# ReSub Vars
|
# Proper Formatting
|
||||||
translatedText = re.sub(reSubVarRegex, r'\1[\2]', translatedText)
|
translatedText = translatedText.replace('"', '\"')
|
||||||
|
|
||||||
# Set Data
|
# Set Data
|
||||||
page['list'][i]['p'][4] = startString + translatedText + endString
|
translatedText = startString + translatedText + endString
|
||||||
|
page['list'][i]['p'][4] = translatedText
|
||||||
|
|
||||||
## Event Code: 357 [Picture Text] [Optional]
|
## Event Code: 357 [Picture Text] [Optional]
|
||||||
if page['list'][i]['c'] == 357 and CODE357 == True:
|
if page['list'][i]['c'] == 357 and CODE357 == True:
|
||||||
if 'text' in page['list'][i]['p'][3]:
|
if 'message' in page['list'][i]['p'][3]:
|
||||||
jaString = page['list'][i]['p'][3]['text']
|
jaString = page['list'][i]['p'][3]['message']
|
||||||
if type(jaString) != str:
|
if type(jaString) != str:
|
||||||
continue
|
continue
|
||||||
|
|
||||||
|
|
@ -564,14 +598,54 @@ def searchCodes(page, pbar):
|
||||||
if not re.search(r'[一-龠]+|[ぁ-ゔ]+|[ァ-ヴー]+', jaString):
|
if not re.search(r'[一-龠]+|[ぁ-ゔ]+|[ァ-ヴー]+', jaString):
|
||||||
continue
|
continue
|
||||||
|
|
||||||
# Need to remove outside non-japanese text and put it back later
|
# Need to remove outside code and put it back later
|
||||||
startString = re.search(r'^[^ぁ-んァ-ン一-龯\<\>【】]+', jaString)
|
oldjaString = jaString
|
||||||
jaString = re.sub(r'^[^ぁ-んァ-ン一-龯\<\>【】]+', '', jaString)
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー【】()「」a-zA-ZA-Z0-9\\]+', jaString)
|
||||||
|
finalJAString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー【】()「」a-zA-ZA-Z0-9\\]+', '', jaString)
|
||||||
if startString is None: startString = ''
|
if startString is None: startString = ''
|
||||||
else: startString = startString.group()
|
else: startString = startString.group()
|
||||||
|
|
||||||
# Sub Vars
|
# Remove any textwrap
|
||||||
jaString = re.sub(r'\\+([a-zA-Z]+)\[([0-9]+)\]', r'[\1\2]', jaString)
|
finalJAString = re.sub(r'\n', ' ', finalJAString)
|
||||||
|
|
||||||
|
# Translate
|
||||||
|
response = translateGPT(finalJAString, '', True)
|
||||||
|
tokens += response[1]
|
||||||
|
translatedText = response[0]
|
||||||
|
|
||||||
|
# Textwrap
|
||||||
|
translatedText = textwrap.fill(translatedText, width=WIDTH)
|
||||||
|
|
||||||
|
# Set Data
|
||||||
|
page['list'][i]['p'][3]['message'] = startString + translatedText
|
||||||
|
|
||||||
|
## Event Code: 657 [Picture Text] [Optional]
|
||||||
|
if page['list'][i]['c'] == 657 and CODE657 == True:
|
||||||
|
if 'text' in page['list'][i]['p'][0]:
|
||||||
|
jaString = page['list'][i]['p'][0]
|
||||||
|
if type(jaString) != str:
|
||||||
|
continue
|
||||||
|
|
||||||
|
# Definitely don't want to mess with files
|
||||||
|
if '_' in jaString:
|
||||||
|
continue
|
||||||
|
|
||||||
|
# If there isn't any Japanese in the text just skip
|
||||||
|
if not re.search(r'[一-龠]+|[ぁ-ゔ]+|[ァ-ヴー]+', jaString):
|
||||||
|
continue
|
||||||
|
|
||||||
|
# Remove outside text
|
||||||
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】\\]+', jaString)
|
||||||
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】\\]+', '', jaString)
|
||||||
|
endString = re.search(r'[^一-龠ぁ-ゔァ-ヴー\<\>【】。!?\\]+$', jaString)
|
||||||
|
jaString = re.sub(r'[^一-龠ぁ-ゔァ-ヴー\<\>【】。!?\\]+$', '', jaString)
|
||||||
|
if startString is None: startString = ''
|
||||||
|
else: startString = startString.group()
|
||||||
|
if endString is None: endString = ''
|
||||||
|
else: endString = endString.group()
|
||||||
|
|
||||||
|
# Remove any textwrap
|
||||||
|
jaString = re.sub(r'\n', ' ', jaString)
|
||||||
|
|
||||||
# Translate
|
# Translate
|
||||||
response = translateGPT(jaString, '', True)
|
response = translateGPT(jaString, '', True)
|
||||||
|
|
@ -579,18 +653,18 @@ def searchCodes(page, pbar):
|
||||||
translatedText = response[0]
|
translatedText = response[0]
|
||||||
|
|
||||||
# Remove characters that may break scripts
|
# Remove characters that may break scripts
|
||||||
charList = ['\"', '\\', '\\n']
|
charList = ['.', '\"', "'"]
|
||||||
for char in charList:
|
for char in charList:
|
||||||
translatedText = translatedText.replace(char, '')
|
translatedText = translatedText.replace(char, '')
|
||||||
|
|
||||||
# Textwrap
|
# Textwrap
|
||||||
translatedText = textwrap.fill(translatedText, width=50)
|
translatedText = textwrap.fill(translatedText, width=WIDTH)
|
||||||
|
translatedText = startString + translatedText + endString
|
||||||
# ReSub Vars
|
|
||||||
translatedText = re.sub(r'\[([a-zA-Z]+)([0-9]+)]', r'\\\\\1[\2]', translatedText)
|
|
||||||
|
|
||||||
# Set Data
|
# Set Data
|
||||||
page['list'][i]['p'][3]['text'] = startString + translatedText
|
if '\\' in jaString:
|
||||||
|
print('Hi')
|
||||||
|
page['list'][i]['p'][0] = translatedText
|
||||||
|
|
||||||
## Event Code: 101 [Name] [Optional]
|
## Event Code: 101 [Name] [Optional]
|
||||||
if page['list'][i]['c'] == 101 and CODE101 == True:
|
if page['list'][i]['c'] == 101 and CODE101 == True:
|
||||||
|
|
@ -607,22 +681,34 @@ def searchCodes(page, pbar):
|
||||||
speaker = jaString
|
speaker = jaString
|
||||||
continue
|
continue
|
||||||
|
|
||||||
|
# Need to remove outside code and put it back later
|
||||||
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】]+', jaString)
|
||||||
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】]+', '', jaString)
|
||||||
|
endString = re.search(r'[^一-龠ぁ-ゔァ-ヴー\<\>【】。!?]+$', jaString)
|
||||||
|
jaString = re.sub(r'[^一-龠ぁ-ゔァ-ヴー\<\>【】。!?]+$', '', jaString)
|
||||||
|
if startString is None: startString = ''
|
||||||
|
else: startString = startString.group() + ' '
|
||||||
|
if endString is None: endString = ''
|
||||||
|
else: endString = endString.group()
|
||||||
|
|
||||||
# Translate
|
# Translate
|
||||||
response = translateGPT(jaString, 'Reply with only the english translation. NEVER reply in anything other than English. I repeat, only reply with the english translation of the original text.', False)
|
response = translateGPT(jaString, 'Reply with only the english translation of the NPC name.', False)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
translatedText = response[0]
|
translatedText = response[0]
|
||||||
|
|
||||||
# Remove characters that may break scripts
|
# Remove characters that may break scripts
|
||||||
charList = ['.', '\"', '\\n']
|
charList = ['.', '\"']
|
||||||
for char in charList:
|
for char in charList:
|
||||||
translatedText = translatedText.replace(char, '')
|
translatedText = translatedText.replace(char, '')
|
||||||
|
|
||||||
|
translatedText = startString + translatedText + endString
|
||||||
|
|
||||||
# Set Data
|
# Set Data
|
||||||
speaker = translatedText
|
speaker = translatedText
|
||||||
page['list'][i]['p'][4] = translatedText
|
page['list'][i]['p'][4] = translatedText
|
||||||
|
|
||||||
## Event Code: 355 or 655 Scripts [Optional]
|
## Event Code: 355 or 655 Scripts [Optional]
|
||||||
if (page['list'][i]['c'] == 355 or page['list'][i]['c'] == 655) and CODE355655 == True:
|
if (page['list'][i]['c'] == 355) and CODE355655 == True:
|
||||||
jaString = page['list'][i]['p'][0]
|
jaString = page['list'][i]['p'][0]
|
||||||
|
|
||||||
# If there isn't any Japanese in the text just skip
|
# If there isn't any Japanese in the text just skip
|
||||||
|
|
@ -630,38 +716,39 @@ def searchCodes(page, pbar):
|
||||||
continue
|
continue
|
||||||
|
|
||||||
# Want to translate this script
|
# Want to translate this script
|
||||||
if page['list'][i]['c'] == 355 and '.setName' not in jaString:
|
if page['list'][i]['c'] == 355 and '_logWindow.push' not in jaString:
|
||||||
continue
|
continue
|
||||||
|
|
||||||
# Don't want to touch certain scripts
|
# Don't want to touch certain scripts
|
||||||
if page['list'][i]['c'] == 655 and 'this.' in jaString:
|
if page['list'][i]['c'] == 655 and '.' in jaString:
|
||||||
continue
|
continue
|
||||||
|
|
||||||
# Need to remove outside code and put it back later
|
# Need to remove outside code and put it back later
|
||||||
startString = re.search(r'^[^ぁ-んァ-ン一-龯\<\>【】]+', jaString)
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】()「」『』]+', jaString)
|
||||||
jaString = re.sub(r'^[^ぁ-んァ-ン一-龯\<\>【】]+', '', jaString)
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】()「」『』]+', '', jaString)
|
||||||
endString = re.search(r'[^ぁ-んァ-ン一-龯\<\>【】 。!?]+$', jaString)
|
endString = re.search(r'[^一-龠ぁ-ゔァ-ヴー\<\>【】()「」『』。!?]+$', jaString)
|
||||||
jaString = re.sub(r'[^ぁ-んァ-ン一-龯\<\>【】 。!?]+$', '', jaString)
|
jaString = re.sub(r'[^一-龠ぁ-ゔァ-ヴー\<\>【】()「」『』。!?]+$', '', jaString)
|
||||||
if startString is None: startString = ''
|
if startString is None: startString = ''
|
||||||
else: startString = startString.group()
|
else: startString = startString.group()
|
||||||
if endString is None: endString = ''
|
if endString is None: endString = ''
|
||||||
else: endString = endString.group()
|
else: endString = endString.group()
|
||||||
|
|
||||||
# Translate
|
# Translate
|
||||||
response = translateGPT(jaString, 'Reply with only the english translation.', True)
|
response = translateGPT(jaString, 'Reply with the English Translation of the text.', True)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
translatedText = response[0]
|
translatedText = response[0]
|
||||||
|
|
||||||
# Remove characters that may break scripts
|
# Remove characters that may break scripts
|
||||||
charList = ['.', '\"', '\\n']
|
charList = ['\"', "\'"]
|
||||||
for char in charList:
|
for char in charList:
|
||||||
translatedText = translatedText.replace(char, '')
|
translatedText = translatedText.replace(char, '')
|
||||||
|
|
||||||
# Set Data
|
# Set Data
|
||||||
page['list'][i]['p'][0] = startString + translatedText + endString
|
translatedText = startString + translatedText + endString
|
||||||
|
page['list'][i]['p'][0] = translatedText
|
||||||
|
|
||||||
## Event Code: 356 D_TEXT
|
## Event Code: 408 (Script)
|
||||||
if page['list'][i]['c'] == 356 and CODE356 == True:
|
if (page['list'][i]['c'] == 408) and CODE408 == True:
|
||||||
jaString = page['list'][i]['p'][0]
|
jaString = page['list'][i]['p'][0]
|
||||||
|
|
||||||
# If there isn't any Japanese in the text just skip
|
# If there isn't any Japanese in the text just skip
|
||||||
|
|
@ -669,21 +756,61 @@ def searchCodes(page, pbar):
|
||||||
continue
|
continue
|
||||||
|
|
||||||
# Want to translate this script
|
# Want to translate this script
|
||||||
if 'PSM_SHOW_POPUP' not in jaString:
|
if page['list'][i]['c'] == 408 and '\\>' not in jaString:
|
||||||
continue
|
continue
|
||||||
|
|
||||||
# Need to remove outside code and put it back later
|
# Need to remove outside code and put it back later
|
||||||
startString = re.search(r'^[^ぁ-んァ-ン一-龯【】()「」]+-1 ', jaString)
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー【】]+', jaString)
|
||||||
jaString = re.sub(r'^[^ぁ-んァ-ン一-龯【】()「」]+-1 ', '', jaString)
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー【】]+', '', jaString)
|
||||||
endString = re.search(r' [^ぁ-んァ-ン一-龯\<\>【】 。!?]+$', jaString)
|
endString = re.search(r'[^一-龠ぁ-ゔァ-ヴー【】。!?]+$', jaString)
|
||||||
jaString = re.sub(r' [^ぁ-んァ-ン一-龯\<\>【】 。!?]+$', '', jaString)
|
jaString = re.sub(r'[^一-龠ぁ-ゔァ-ヴー【】。!?]+$', '', jaString)
|
||||||
if startString is None: startString = ''
|
if startString is None: startString = ''
|
||||||
else: startString = startString.group()
|
else: startString = startString.group()
|
||||||
if endString is None: endString = ''
|
if endString is None: endString = ''
|
||||||
else: endString = endString.group()
|
else: endString = endString.group()
|
||||||
|
|
||||||
# Translate
|
# Translate
|
||||||
response = translateGPT(jaString, 'Reply with only the English Translation.', True)
|
response = translateGPT(jaString, '', True)
|
||||||
|
tokens += response[1]
|
||||||
|
translatedText = response[0]
|
||||||
|
|
||||||
|
# Remove characters that may break scripts
|
||||||
|
charList = ['.', '\"']
|
||||||
|
for char in charList:
|
||||||
|
translatedText = translatedText.replace(char, '')
|
||||||
|
|
||||||
|
translatedText = startString + translatedText + endString
|
||||||
|
|
||||||
|
translatedText = translatedText.replace('"', '\"')
|
||||||
|
|
||||||
|
# Set Data
|
||||||
|
page['list'][i]['p'][0] = translatedText
|
||||||
|
|
||||||
|
## Event Code: 356 D_TEXT
|
||||||
|
if page['list'][i]['c'] == 356 and CODE356 == True:
|
||||||
|
jaString = page['list'][i]['p'][0]
|
||||||
|
oldjaString = jaString
|
||||||
|
|
||||||
|
# If there isn't any Japanese in the text just skip
|
||||||
|
if not re.search(r'[一-龠]+|[ぁ-ゔ]+|[ァ-ヴー]+', jaString):
|
||||||
|
continue
|
||||||
|
|
||||||
|
# Want to translate this script
|
||||||
|
if 'addLog' not in jaString:
|
||||||
|
continue
|
||||||
|
|
||||||
|
# Need to remove outside code and put it back later
|
||||||
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー【】()「」『』]+', jaString)
|
||||||
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー【】()「」『』]+', '', jaString)
|
||||||
|
endString = re.search(r' [^一-龠ぁ-ゔァ-ヴー\<\>【】()「」『』 。!?]+$', jaString)
|
||||||
|
jaString = re.sub(r' [^一-龠ぁ-ゔァ-ヴー\<\>【】()「」『』 。!?]+$', '', jaString)
|
||||||
|
if startString is None: startString = ''
|
||||||
|
else: startString = startString.group()
|
||||||
|
if endString is None: endString = ''
|
||||||
|
else: endString = endString.group()
|
||||||
|
|
||||||
|
# Translate
|
||||||
|
response = translateGPT(jaString, 'Reply with only the English Translation of the text.', True)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
translatedText = response[0]
|
translatedText = response[0]
|
||||||
|
|
||||||
|
|
@ -695,6 +822,9 @@ def searchCodes(page, pbar):
|
||||||
# Cant have spaces?
|
# Cant have spaces?
|
||||||
translatedText = translatedText.replace(' ', ' ')
|
translatedText = translatedText.replace(' ', ' ')
|
||||||
|
|
||||||
|
# Textwrap
|
||||||
|
translatedText = textwrap.fill(translatedText, width=1000)
|
||||||
|
|
||||||
# Set Data
|
# Set Data
|
||||||
page['list'][i]['p'][0] = startString + translatedText + endString
|
page['list'][i]['p'][0] = startString + translatedText + endString
|
||||||
|
|
||||||
|
|
@ -702,20 +832,24 @@ def searchCodes(page, pbar):
|
||||||
if page['list'][i]['c'] == 102 and CODE102 == True:
|
if page['list'][i]['c'] == 102 and CODE102 == True:
|
||||||
for choice in range(len(page['list'][i]['p'][0])):
|
for choice in range(len(page['list'][i]['p'][0])):
|
||||||
jaString = page['list'][i]['p'][0][choice]
|
jaString = page['list'][i]['p'][0][choice]
|
||||||
translatedText = translatedText.replace(' 。', '.')
|
jaString = jaString.replace(' 。', '.')
|
||||||
|
|
||||||
# Need to remove outside code and put it back later
|
# Need to remove outside code and put it back later
|
||||||
startString = re.search(r'^[^ぁ-んァ-ン一-龯\<\>【】()A-Z0-9]+', jaString)
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー【】()A-Z0-9]+', jaString)
|
||||||
jaString = re.sub(r'^[^ぁ-んァ-ン一-龯\<\>【】()A-Z0-9]+', '', jaString)
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー【】()A-Z0-9]+', '', jaString)
|
||||||
endString = re.search(r'[^ぁ-んァ-ン一-龯【】 。!?()A-Z0-9]+$', jaString)
|
endString = re.search(r'[^一-龠ぁ-ゔァ-ヴー【】()A-Z0-9 。!?]+$', jaString)
|
||||||
jaString = re.sub(r'[^ぁ-んァ-ン一-龯【】 。!?()A-Z0-9]+$', '', jaString)
|
jaString = re.sub(r'[^一-龠ぁ-ゔァ-ヴー【】()A-Z0-9 。!?]+$', '', jaString)
|
||||||
if startString is None: startString = ''
|
if startString is None: startString = ''
|
||||||
else: startString = startString.group()
|
else: startString = startString.group()
|
||||||
if endString is None: endString = ''
|
if endString is None: endString = ''
|
||||||
else: endString = endString.group()
|
else: endString = endString.group()
|
||||||
|
|
||||||
response = translateGPT(jaString, 'Keep your reply prompt.', True)
|
if len(textHistory) > 0:
|
||||||
translatedText = response[0]
|
response = translateGPT(jaString, 'Previous text for context: ' + textHistory[len(textHistory)-1], False)
|
||||||
|
translatedText = response[0]
|
||||||
|
else:
|
||||||
|
response = translateGPT(jaString, '', False)
|
||||||
|
translatedText = response[0]
|
||||||
|
|
||||||
# Remove characters that may break scripts
|
# Remove characters that may break scripts
|
||||||
charList = ['.', '\"', '\\n']
|
charList = ['.', '\"', '\\n']
|
||||||
|
|
@ -736,10 +870,10 @@ def searchCodes(page, pbar):
|
||||||
continue
|
continue
|
||||||
|
|
||||||
# Need to remove outside code and put it back later
|
# Need to remove outside code and put it back later
|
||||||
startString = re.search(r'^[^ぁ-んァ-ン一-龯\<\>【】]+', jaString)
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】]+', jaString)
|
||||||
jaString = re.sub(r'^[^ぁ-んァ-ン一-龯\<\>【】]+', '', jaString)
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー\<\>【】]+', '', jaString)
|
||||||
endString = re.search(r'[^ぁ-んァ-ン一-龯【】 。!?]+$', jaString)
|
endString = re.search(r'[^一-龠ぁ-ゔァ-ヴー【】 。!?]+$', jaString)
|
||||||
jaString = re.sub(r'[^ぁ-んァ-ン一-龯【】 。!?]+$', '', jaString)
|
jaString = re.sub(r'[^一-龠ぁ-ゔァ-ヴー【】 。!?]+$', '', jaString)
|
||||||
if startString is None: startString = ''
|
if startString is None: startString = ''
|
||||||
else: startString = startString.group()
|
else: startString = startString.group()
|
||||||
if endString is None: endString = ''
|
if endString is None: endString = ''
|
||||||
|
|
@ -758,33 +892,34 @@ def searchCodes(page, pbar):
|
||||||
page['list'][i]['p'][j] = startString + translatedText + endString
|
page['list'][i]['p'][j] = startString + translatedText + endString
|
||||||
|
|
||||||
### Event Code: 320 Set Variable
|
### Event Code: 320 Set Variable
|
||||||
if page['list'][i]['c'] == 320 and CODE320 == True:
|
if page['list'][i]['c'] == 320 and CODE320 == True or page['list'][i]['c'] == 324 and CODE324 == True:
|
||||||
jaString = page['list'][i]['p'][1]
|
jaString = page['list'][i]['p'][1]
|
||||||
translatedText = translatedText.replace(' 。', '.')
|
|
||||||
|
|
||||||
# Need to remove outside code and put it back later
|
# Need to remove outside code and put it back later
|
||||||
startString = re.search(r'^[^ぁ-んァ-ン一-龯【】a-zA-Z\\]+', jaString)
|
startString = re.search(r'^[^一-龠ぁ-ゔァ-ヴー【】a-zA-Z\\]+', jaString)
|
||||||
jaString = re.sub(r'^[^ぁ-んァ-ン一-龯【】a-zA-Z\\]+', '', jaString)
|
jaString = re.sub(r'^[^一-龠ぁ-ゔァ-ヴー【】a-zA-Z\\]+', '', jaString)
|
||||||
endString = re.search(r'[^ぁ-んァ-ン一-龯【】 。!?]+$', jaString)
|
endString = re.search(r'[^一-龠ぁ-ゔァ-ヴー【】。!?]+$', jaString)
|
||||||
jaString = re.sub(r'[^ぁ-んァ-ン一-龯【】 。!?]+$', '', jaString)
|
jaString = re.sub(r'[^一-龠ぁ-ゔァ-ヴー【】。!?]+$', '', jaString)
|
||||||
if startString is None: startString = ''
|
if startString is None: startString = ''
|
||||||
else: startString = startString.group()
|
else: startString = startString.group()
|
||||||
if endString is None: endString = ''
|
if endString is None: endString = ''
|
||||||
else: endString = endString.group()
|
else: endString = endString.group()
|
||||||
|
|
||||||
response = translateGPT(jaString, 'Reply with only the english translation.', True)
|
response = translateGPT(jaString, 'Reply with only the english translation of the npc nickname.', False)
|
||||||
translatedText = response[0]
|
translatedText = response[0]
|
||||||
|
|
||||||
# Remove characters that may break scripts
|
# Remove characters that may break scripts
|
||||||
charList = ['.', '\"', '\\n']
|
charList = ['\"']
|
||||||
for char in charList:
|
for char in charList:
|
||||||
translatedText = translatedText.replace(char, '')
|
translatedText = translatedText.replace(char, '')
|
||||||
|
|
||||||
|
translatedText = translatedText.strip('.')
|
||||||
|
|
||||||
# Set Data
|
# Set Data
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
page['list'][i]['p'][1] = startString + translatedText + endString
|
page['list'][i]['p'][1] = startString + translatedText + endString
|
||||||
|
|
||||||
except IndexError:
|
except IndexError as e:
|
||||||
# This is part of the logic so we just pass it.
|
# This is part of the logic so we just pass it.
|
||||||
pass
|
pass
|
||||||
except Exception as e:
|
except Exception as e:
|
||||||
|
|
@ -794,29 +929,13 @@ def searchCodes(page, pbar):
|
||||||
# Append leftover groups in 401
|
# Append leftover groups in 401
|
||||||
if len(currentGroup) > 0:
|
if len(currentGroup) > 0:
|
||||||
# Translate
|
# Translate
|
||||||
if speaker != '':
|
response = translateGPT(finalJAString, 'Previous Translated Text for Context: ' + '\n\n'.join(textHistory), True)
|
||||||
response = translateGPT(finalJAString, 'Previous text for context: ' + ' '.join(textHistory) \
|
|
||||||
+ '\n\n\n###\n\n\nCurrent Speaker: ' + speaker, True)
|
|
||||||
else:
|
|
||||||
response = translateGPT(finalJAString, 'Previous text for context: ' + ' '.join(textHistory), True)
|
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
translatedText = response[0]
|
translatedText = response[0]
|
||||||
|
|
||||||
# ReSub Vars
|
|
||||||
translatedText = re.sub(reSubVarRegex, r'\1[\2]', translatedText)
|
|
||||||
|
|
||||||
# TextHistory is what we use to give GPT Context, so thats appended here.
|
# TextHistory is what we use to give GPT Context, so thats appended here.
|
||||||
rawTranslatedText = re.sub(r'[\\<>]+[a-zA-Z]+\[[a-zA-Z0-9]+\]', '', translatedText)
|
# rawTranslatedText = re.sub(r'[\\<>]+[a-zA-Z]+\[[a-zA-Z0-9]+\]', '', translatedText)
|
||||||
if speaker != '':
|
textHistory.append('\"' + translatedText + '\"')
|
||||||
textHistory.append(speaker + ': ' + rawTranslatedText)
|
|
||||||
else:
|
|
||||||
textHistory.append('\"' + rawTranslatedText + '\"')
|
|
||||||
|
|
||||||
# Name Handling
|
|
||||||
if len(match) != 0:
|
|
||||||
name = '\\nw[' + speaker + ']'
|
|
||||||
if name not in translatedText:
|
|
||||||
translatedText = translatedText + '\\nw[' + speaker + ']'
|
|
||||||
|
|
||||||
# if speakerCaught == True:
|
# if speakerCaught == True:
|
||||||
# translatedText = speakerRaw + ':\n' + translatedText
|
# translatedText = speakerRaw + ':\n' + translatedText
|
||||||
|
|
@ -829,30 +948,36 @@ def searchCodes(page, pbar):
|
||||||
translatedText = startString + translatedText
|
translatedText = startString + translatedText
|
||||||
|
|
||||||
# Set Data
|
# Set Data
|
||||||
page['list'][i]['p'][0] = translatedText.replace('\"', '')
|
translatedText = translatedText.replace('ッ', '')
|
||||||
|
translatedText = translatedText.replace('っ', '')
|
||||||
|
translatedText = translatedText.replace('\"', '')
|
||||||
|
page['list'][i]['p'][0] = translatedText
|
||||||
speaker = ''
|
speaker = ''
|
||||||
match = []
|
match = []
|
||||||
|
|
||||||
# Keep textHistory list at length maxHistory
|
# Keep textHistory list at length maxHistory
|
||||||
if len(textHistory) > maxHistory:
|
if len(textHistory) > maxHistory:
|
||||||
textHistory.pop(0)
|
textHistory.pop(0)
|
||||||
currentGroup = []
|
currentGroup = []
|
||||||
|
|
||||||
return tokens
|
return tokens
|
||||||
|
|
||||||
def searchSS(state, pbar):
|
def searchSS(state, pbar):
|
||||||
'''Searches skills and states yaml files'''
|
'''Searches skills and states json files'''
|
||||||
tokens = 0
|
tokens = 0
|
||||||
responseList = [0] * 7
|
responseList = [0] * 7
|
||||||
|
|
||||||
responseList[0] = (translateGPT(state['message1'], 'Reply with the english translated Action being performed and no subject.', False))
|
responseList[0] = (translateGPT(state['message1'], 'reply with only the english translation of the text.', False))
|
||||||
responseList[1] = (translateGPT(state['message2'], 'Reply with the english translated Action being performed and no subject.', False))
|
responseList[1] = (translateGPT(state['message2'], 'reply with only the english translation of the text.', False))
|
||||||
responseList[2] = (translateGPT(state.get('message3', ''), 'Reply with the english translated Action being performed and no subject..', False))
|
responseList[2] = (translateGPT(state.get('message3', ''), 'reply with only the english translation of the text.', False))
|
||||||
responseList[3] = (translateGPT(state.get('message4', ''), 'Reply with the english translated Action being performed and no subject..', False))
|
responseList[3] = (translateGPT(state.get('message4', ''), 'reply with only the english translation of the text.', False))
|
||||||
responseList[4] = (translateGPT(state['name'], 'Reply with only the english translation', True))
|
responseList[4] = (translateGPT(state['name'], 'Reply with only the english translation of the RPG item name.', False))
|
||||||
# responseList[5] = (translateGPT(state['note'], 'Reply with only the translated english note.', False))
|
|
||||||
if 'description' in state:
|
if 'description' in state:
|
||||||
responseList[6] = (translateGPT(state['description'], 'Reply with the english translated description.', True))
|
responseList[6] = (translateGPT(state['description'], 'reply with only the english translation of the description.', False))
|
||||||
|
|
||||||
|
# if 'note' in state:
|
||||||
|
# if 'raceDesc' in state['note']:
|
||||||
|
# tokens += translateNote(state, r'<raceDesc:([^>]*)>')
|
||||||
|
|
||||||
# Put all our translations in a list
|
# Put all our translations in a list
|
||||||
for i in range(len(responseList)):
|
for i in range(len(responseList)):
|
||||||
|
|
@ -881,12 +1006,12 @@ def searchSS(state, pbar):
|
||||||
|
|
||||||
def searchSystem(data, pbar):
|
def searchSystem(data, pbar):
|
||||||
tokens = 0
|
tokens = 0
|
||||||
context = 'Reply with only the english translated menu item.'
|
context = 'Reply with only the english translation of the UI textbox'
|
||||||
|
|
||||||
# Title
|
# Title
|
||||||
response = translateGPT(data['game_title'], context, True)
|
response = translateGPT(data['gameTitle'], ' Reply with the English translation of the game title name', False)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
data['game_title'] = response[0].strip('.')
|
data['gameTitle'] = response[0].strip('.')
|
||||||
pbar.update(1)
|
pbar.update(1)
|
||||||
|
|
||||||
# Terms
|
# Terms
|
||||||
|
|
@ -895,34 +1020,78 @@ def searchSystem(data, pbar):
|
||||||
termList = data['terms'][term]
|
termList = data['terms'][term]
|
||||||
for i in range(len(termList)): # Last item is a messages object
|
for i in range(len(termList)): # Last item is a messages object
|
||||||
if termList[i] is not None:
|
if termList[i] is not None:
|
||||||
response = translateGPT(termList[i], context, True)
|
response = translateGPT(termList[i], context, False)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
termList[i] = response[0].strip('.\"')
|
termList[i] = response[0].strip('.\"')
|
||||||
pbar.update(1)
|
pbar.update(1)
|
||||||
|
|
||||||
# Armor Types
|
# Armor Types
|
||||||
for i in range(len(data['armor_types'])):
|
for i in range(len(data['armorTypes'])):
|
||||||
response = translateGPT(data['armor_types'][i], 'Reply with only the english translated armor type', False)
|
response = translateGPT(data['armorTypes'][i], 'Reply with only the english translation of the armor type', False)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
data['armor_types'][i] = response[0].strip('.\"')
|
data['armorTypes'][i] = response[0].strip('.\"')
|
||||||
pbar.update(1)
|
pbar.update(1)
|
||||||
|
|
||||||
# Skill Types
|
# Skill Types
|
||||||
for i in range(len(data['skill_types'])):
|
for i in range(len(data['skillTypes'])):
|
||||||
response = translateGPT(data['skill_types'][i], 'Reply with only the english translation', False)
|
response = translateGPT(data['skillTypes'][i], 'Reply with only the english translation', False)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
data['skill_types'][i] = response[0].strip('.\"')
|
data['skillTypes'][i] = response[0].strip('.\"')
|
||||||
pbar.update(1)
|
pbar.update(1)
|
||||||
|
|
||||||
# Weapon Types
|
# Equip Types
|
||||||
for i in range(len(data['weapon_types'])):
|
for i in range(len(data['equipTypes'])):
|
||||||
response = translateGPT(data['weapon_types'][i], 'Reply with only the english translated equipment type. No disclaimers.', False)
|
response = translateGPT(data['equipTypes'][i], 'Reply with only the english translation of the equipment type. No disclaimers.', False)
|
||||||
tokens += response[1]
|
tokens += response[1]
|
||||||
data['weapon_types'][i] = response[0].strip('.\"')
|
data['equipTypes'][i] = response[0].strip('.\"')
|
||||||
pbar.update(1)
|
pbar.update(1)
|
||||||
|
|
||||||
|
# Variables
|
||||||
|
for i in range(len(data['variables'])):
|
||||||
|
response = translateGPT(data['variables'][i], 'Reply with only the english translation of the variable name.', False)
|
||||||
|
tokens += response[1]
|
||||||
|
data['variables'][i] = response[0].strip('.\"')
|
||||||
|
pbar.update(1)
|
||||||
|
|
||||||
|
# Messages
|
||||||
|
messages = (data['terms']['messages'])
|
||||||
|
for key, value in messages.items():
|
||||||
|
response = translateGPT(value, 'Reply with only the english translation of the text.', False)
|
||||||
|
translatedText = response[0]
|
||||||
|
|
||||||
|
# Remove characters that may break scripts
|
||||||
|
charList = ['.', '\"', '\\n']
|
||||||
|
for char in charList:
|
||||||
|
translatedText = translatedText.replace(char, '')
|
||||||
|
|
||||||
|
tokens += response[1]
|
||||||
|
messages[key] = translatedText
|
||||||
|
pbar.update(1)
|
||||||
|
|
||||||
return tokens
|
return tokens
|
||||||
|
|
||||||
|
def subVars(jaString):
|
||||||
|
varRegex = r'\\+[a-zA-Z]+\[[0-9a-zA-Z\\\[\]]+\]|[\\]+[#a-zA-Z]'
|
||||||
|
count = 0
|
||||||
|
|
||||||
|
varList = re.findall(varRegex, jaString)
|
||||||
|
if len(varList) != 0:
|
||||||
|
for var in varList:
|
||||||
|
jaString = jaString.replace(var, '[' + str(count) + ']')
|
||||||
|
count += 1
|
||||||
|
|
||||||
|
return [jaString, varList]
|
||||||
|
|
||||||
|
def resubVars(translatedText, varList):
|
||||||
|
count = 0
|
||||||
|
|
||||||
|
if len(varList) != 0:
|
||||||
|
for var in varList:
|
||||||
|
translatedText = translatedText.replace('[' + str(count) + ']', var)
|
||||||
|
count += 1
|
||||||
|
|
||||||
|
return translatedText
|
||||||
|
|
||||||
@retry(exceptions=Exception, tries=5, delay=5)
|
@retry(exceptions=Exception, tries=5, delay=5)
|
||||||
def translateGPT(t, history, fullPromptFlag):
|
def translateGPT(t, history, fullPromptFlag):
|
||||||
with LOCK:
|
with LOCK:
|
||||||
|
|
@ -933,31 +1102,44 @@ def translateGPT(t, history, fullPromptFlag):
|
||||||
TOKENS += len(enc.encode(t)) * 2 + len(enc.encode(history)) + len(enc.encode(PROMPT))
|
TOKENS += len(enc.encode(t)) * 2 + len(enc.encode(history)) + len(enc.encode(PROMPT))
|
||||||
return (t, 0)
|
return (t, 0)
|
||||||
|
|
||||||
|
# Sub Vars
|
||||||
|
varResponse = subVars(t)
|
||||||
|
subbedT = varResponse[0]
|
||||||
|
|
||||||
# If there isn't any Japanese in the text just skip
|
# If there isn't any Japanese in the text just skip
|
||||||
if not re.search(r'[一-龠]+|[ぁ-ゔ]+|[ァ-ヴ]+', t):
|
if not re.search(r'[一-龠]+|[ぁ-ゔ]+|[ァ-ヴー]+', subbedT):
|
||||||
return(t, 0)
|
return(t, 0)
|
||||||
|
|
||||||
"""Translate text using GPT"""
|
"""Translate text using GPT"""
|
||||||
if fullPromptFlag:
|
if fullPromptFlag:
|
||||||
system = "###\n" + history + PROMPT
|
system = PROMPT
|
||||||
|
user = 'Reply with only the English Translation of the following text maintaining any code: ' + subbedT
|
||||||
else:
|
else:
|
||||||
system = 'You are going to pretend to be Japanese visual novel translator, \
|
system = 'Reply with only the English translation of the text.'
|
||||||
editor, and localizer. ' + history
|
user = 'Reply with only the English translation of this dialogue menu option: ' + subbedT
|
||||||
response = openai.ChatCompletion.create(
|
response = openai.ChatCompletion.create(
|
||||||
temperature=0,
|
temperature=0,
|
||||||
model="gpt-3.5-turbo",
|
model="gpt-3.5-turbo-16k",
|
||||||
messages=[
|
messages=[
|
||||||
{"role": "system", "content": system},
|
{"role": "system", "content": system},
|
||||||
{"role": "user", "content": t}
|
{"role": "user", "content": history},
|
||||||
|
{"role": "user", "content": user}
|
||||||
],
|
],
|
||||||
request_timeout=30,
|
request_timeout=30,
|
||||||
)
|
)
|
||||||
|
|
||||||
|
translatedText = response.choices[0].message.content
|
||||||
|
tokens = response.usage.total_tokens
|
||||||
|
|
||||||
# Make sure translation didn't wonk out
|
# Make sure translation didn't wonk out
|
||||||
mlen=len(response.choices[0].message.content)
|
mlen=len(response.choices[0].message.content)
|
||||||
elnt=10*len(t)
|
elnt=10*len(subbedT)
|
||||||
if len(response.choices[0].message.content) > 9 * len(t):
|
|
||||||
|
#Resub Vars
|
||||||
|
translatedText = resubVars(translatedText, varResponse[1])
|
||||||
|
|
||||||
|
if len(response.choices[0].message.content) > 10 * len(t):
|
||||||
return [t, response.usage.total_tokens]
|
return [t, response.usage.total_tokens]
|
||||||
else:
|
else:
|
||||||
return [response.choices[0].message.content, response.usage.total_tokens]
|
return [translatedText, tokens]
|
||||||
|
|
||||||
Loading…
Reference in a new issue