import sys
from unittest.mock import MagicMock

# Comprehensive workaround for missing _lzma in some Python builds
try:
    import lzma
except ImportError:
    mock_lzma = MagicMock()
    mock_lzma.FORMAT_XZ = 1
    mock_lzma.FORMAT_ALONE = 2
    mock_lzma.FORMAT_RAW = 3
    mock_lzma.CHECK_NONE = 0
    mock_lzma.CHECK_CRC32 = 1
    mock_lzma.CHECK_CRC64 = 4
    mock_lzma.CHECK_SHA256 = 10
    sys.modules["_lzma"] = MagicMock()
    sys.modules["lzma"] = mock_lzma

import torch
from transformers import MarianMTModel, MarianTokenizer
import re
import argparse
import multiprocessing
import time

TARGET_LANGS = {
    "es": "Helsinki-NLP/opus-mt-en-es",
    "fr": "Helsinki-NLP/opus-mt-en-fr",
    "ar": "Helsinki-NLP/opus-mt-en-ar"
}

def split_text_for_translation(text, max_chars=250):
    normalized = " ".join(text.split()).strip()
    if not normalized or len(normalized) <= max_chars:
        return [normalized] if normalized else []

    chunks = []
    current = ""
    sentences = [s for s in re.split(r"(?<=[.!?])\s+", normalized) if s]

    for sentence in sentences:
        if len(sentence) > max_chars:
            words = sentence.split()
            word_chunk = ""
            for word in words:
                candidate = f"{word_chunk} {word}".strip()
                if len(candidate) <= max_chars:
                    word_chunk = candidate
                else:
                    if word_chunk: chunks.append(word_chunk)
                    word_chunk = word
            if word_chunk: chunks.append(word_chunk)
            continue

        candidate = f"{current} {sentence}".strip()
        if candidate and len(candidate) <= max_chars:
            current = candidate
        else:
            if current: chunks.append(current)
            current = sentence

    if current: chunks.append(current)
    return chunks

def run_translation(in_queue, out_queue, args):
    device = "mps" if torch.backends.mps.is_available() else "cpu"
    print(f"[Translate] Loading translation models on {device}...")
    
    translation_engines = {}
    for lang_key, model_id in TARGET_LANGS.items():
        if getattr(args, lang_key, False):
            print(f"[Translate] Loading {lang_key} model...")
            tokenizer = MarianTokenizer.from_pretrained(model_id)
            model = MarianMTModel.from_pretrained(model_id).to(device)
            translation_engines[lang_key] = (model, tokenizer)

    paragraph_buffer = []
    
    print("[Translate] Ready.")
    
    while True:
        try:
            item = in_queue.get()
            if item is None: break
            
            text = item.get("original", "").strip()
            detected_lang = item.get("detected_lang", "en")
            
            # Simple paragraph logic: 
            # If the segment ends with sentence-terminal punctuation, flush the paragraph.
            paragraph_buffer.append(text)
            
            if text.endswith((".", "?", "!")):
                full_text = " ".join(paragraph_buffer)
                payload = {"original": full_text, "en": full_text if detected_lang == "en" else None}
                
                # If detected language is not English, we'd normally bridge to English first.
                # For now, let's assume direct translation for simplicity or bridge if needed.
                # (Refining bridge logic can come later)
                
                for lang_key, (model, tokenizer) in translation_engines.items():
                    chunks = split_text_for_translation(full_text)
                    translated_parts = []
                    for chunk in chunks:
                        inputs = tokenizer(chunk, return_tensors="pt", padding=True).to(device)
                        with torch.no_grad():
                            translated_tokens = model.generate(**inputs, max_new_tokens=150)
                        translated_parts.append(tokenizer.decode(translated_tokens[0], skip_special_tokens=True).strip())
                    
                    translated_text = " ".join(part for part in translated_parts if part).strip()
                    payload[lang_key] = translated_text
                    print(f"[Translate] {lang_key.upper()}: {translated_text}")

                out_queue.put(payload)
                paragraph_buffer = []
                
        except Exception as e:
            print(f"[Translate] Error: {e}")

if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument("-es", action="store_true")
    parser.add_argument("-fr", action="store_true")
    parser.add_argument("-ar", action="store_true")
    args = parser.parse_args()
    
    # Dummy queues for testing
    in_q = multiprocessing.Queue()
    out_q = multiprocessing.Queue()
    run_translation(in_q, out_q, args)