Generalize tag translation using emojis heck yeah

Auto Trans/README.md

@@ -1,30 +1,32 @@
-This was mostly made using the GPL chatbot, enjoy!
+This was mostly made using the OpenAI chatbot, enjoy!
 
 It requires googletrans and colorama (For le pretty colors)
 
-`pip install googletrans==3.1.0a0`
+```
+pip install googletrans==3.1.0a0
+```
 
 and
 
-`pip install colorama`
+```
+pip install colorama
+```
 
-It expects a dialogue.txt generated from renpy. I personally only include dialogue (That is, no screen text) but it should work regardless if you choose it.
+respectively.
-Including the tags is NECESSARY, otherwise the applier.py will fail.
 
-They are meant to be run in this order
+It expects a `dialogue.txt` generated from renpy. I personally only include dialogue (that is, no screen text), but it should work regardless if you choose to include it. Including the tags is NECESSARY, otherwise `applier.py` will fail.
 
-- (Place your scripts to translate in the scripts folder)
+They are meant to be run in this order:
-- (Place dialogue.txt on the same folder as the python script)
-- Parser.py
-- Translator.py
-- Applier.py
 
-And your translated scripts should be in the output folder.
+- Place your scripts to translate in the `scripts` folder.
-Doing this to snootgame took around an hour to autotranslate, googletrans also doesn't use the official Google API, so don't tell em!
+- Place `dialogue.txt` in the same folder as the python script.
+- Run `parser.py`
+- Run `translator.py`
+- Run `applier.py`
 
-# This is not a general use tool, you WILL need to modify it to adjust to your needs, and due the nature of the googletrans library it might break any moment
+Your translated scripts should be in the `output` folder. Translating snootgame took around an hour using this method.
-# I wanted to do this the legit way but the jews at Deppl will NOT give out free tier API keys to south americans (smh, fucking racists).
+Keep in mind that googletrans does not use the official Google API, so don't tell 'em!
 
-If you need help or something just open up an issue ig.
+# This is not a general use tool, you WILL need to modify it to fit your needs. Due to the nature of the googletrans library, it may break at any moment. I wanted to do this the legitimate way, but the juice at Deepl will NOT give out free tier API keys to South Americans (smh, fucking racists).
 
-Also, Ñ.
+If you need help or have any questions, just open an issue.

Auto Trans/Translator.py

@@ -1,9 +1,10 @@
-#Translates the game using good ol fashioned google translator. Be sure to replace the dialogue code in the translator and the tags in es_tags
+# Translates the game using good ol fashioned google translator. Be sure to replace the dialogue code in the translator.
-#Also this breaks a tiny bit if the dialogue contains \n, but oh well. I don't want to open that can of worms.
+# Also this breaks a tiny bit if the dialogue contains \n, but oh well. I don't want to open that can of worms.
-#Also by break I mean that it may leave some text in english, is not that it will implode... I think
+# Also by break I mean that it may leave some text in english, is not that it will implode... I think.
 
 import csv
 import os
+import re
 from colorama import init
 from colorama import Fore
 from googletrans import Translator
@@ -16,12 +17,30 @@ init()
 
 times = []
 
-# Good ol google doesn't know what renpy tags are, so it translates them. This was the cheapest solution.
+def translate(string):
-es_tags = {
+  # Split the string by it's tags
-  "{rápido}":"{fast}",
+  tokens = re.findall(r'\b\w+\b|{[^}]*}|[^\w\s]|\s+', string)
-  "{w =": "{w=",
+
-  "{alfa":"{alpha"
+  # We encode this bitch up so that google trans doesn't botch the tags
-}
+  to_restore = []
+  for idx, token in enumerate(tokens):
+    if token[0] == "{":
+      to_restore.append(token)
+      # Emojis aren't touched by the translator and retain their position 😎
+      tokens[idx] = "🔠"
+  encoded_string = "".join(tokens)
+
+  # Translate the encoded string
+  trans = translator.translate(encoded_string, dest='es')
+  temp = list(trans.text)
+
+  # Restore the original tags
+  for tag in to_restore:
+    for idx, char in enumerate(temp):
+      if char == "🔠":
+        temp[idx] = tag
+        break
+  return "".join(temp)
 
 def avrg(nums):
   average = sum(nums)/len(nums)
@@ -32,6 +51,7 @@ with open('original.csv', 'r', encoding='utf-8') as input_file, \
      open('trans.csv', 'w', encoding='utf-8',newline='') as output_file:
 
   length = len(input_file.readlines())
+  # Wacky hacks to calculate the time left
   input_file.seek(0)
 
   # Create readers and writers for the input and output files
@@ -43,13 +63,10 @@ with open('original.csv', 'r', encoding='utf-8') as input_file, \
     start_time = datetime.now().replace(microsecond=0)
     try:
       # Translate the string
-      translation = translator.translate(row[0], dest='es')
+      translation = translate(row[0])
 
-      for tag_es, tag in es_tags.items():
+      #Hope this works for escaping, lol
-        translation.text = translation.text.replace(tag_es,tag)
+      translation = translation.replace('"',r'\"')
-
-      #Hope this works for escaping lol
-      translation.text = translation.text.replace('"',r'\"')
 
     except Exception as e:
       print(f'{Fore.RESET}An error occurred: {e}')