homeclaw/.smallclaw/databases/translate_api.py

#!/usr/bin/env python3
"""Batch translate PCSP article titles to Korean using the homeclaw gateway API."""

import sqlite3
import json
import time
import sys
import urllib.request
import urllib.parse

DB_PATH = "/home/kim/homeclaw/.smallclaw/databases/psychotherapy_cases.db"
API_URL = "http://localhost:18789/v1/chat/completions"

def translate_batch(titles, model="exa-one-3.0:free"):
    """Translate a batch of titles using the gateway API."""
    prompt = """Translate the following psychology/psychotherapy article titles from English to Korean.
Rules:
- Keep person names in quotes as-is (e.g., "Fran", "Jane", "Tommy")
- Keep acronyms as-is (CBT, DBT, AEDP, OCD, etc.)
- Translate therapy names: Cognitive Behavioral Therapy→인지행동치료, Dialectical Behavior Therapy→변증법적행동치료, Psychodynamic→정신역동, Schema Therapy→도식치료, Exposure Therapy→노출치료, etc.
- Keep the overall meaning and tone professional
- Each line has format: NUMBER|ENGLISH_TITLE
- Return ONLY the translations in format: NUMBER|KOREAN_TITLE
- Do not add any explanations or notes

"""
    for id_, title, cat in titles:
        prompt += f"{id_}|{title}\n"

    data = json.dumps({
        "model": model,
        "messages": [
            {"role": "system", "content": "You are a professional Korean-English translator specializing in psychology and psychotherapy terminology. Translate accurately and naturally."},
            {"role": "user", "content": prompt}
        ],
        "temperature": 0.3,
        "max_tokens": 4096,
    }).encode("utf-8")

    req = urllib.request.Request(
        API_URL,
        data=data,
        headers={"Content-Type": "application/json"},
    )

    try:
        with urllib.request.urlopen(req, timeout=120) as resp:
            result = json.loads(resp.read().decode("utf-8"))
            return result["choices"][0]["message"]["content"]
    except Exception as e:
        print(f"API error: {e}", file=sys.stderr)
        return None

def main():
    conn = sqlite3.connect(DB_PATH)
    c = conn.cursor()

    # Get untranslated titles
    c.execute("SELECT id, title_en, category_en FROM cases WHERE title_ko = '' OR title_ko IS NULL ORDER BY id")
    rows = c.fetchall()

    if not rows:
        print("All titles translated!")
        return

    print(f"Translating {len(rows)} remaining titles...")

    batch_size = 20
    total_translated = 0

    for i in range(0, len(rows), batch_size):
        batch = rows[i:i+batch_size]
        print(f"\nBatch {i//batch_size + 1}/{(len(rows) + batch_size - 1)//batch_size} ({len(batch)} titles)...")

        result = translate_batch(batch)
        if not result:
            print("Translation failed, skipping batch")
            continue

        # Parse translations
        for line in result.strip().split("\n"):
            line = line.strip()
            if "|" in line:
                parts = line.split("|", 1)
                try:
                    id_ = int(parts[0].strip())
                    title_ko = parts[1].strip()
                    # Remove markdown formatting if present
                    title_ko = title_ko.strip("*").strip()
                    c.execute("UPDATE cases SET title_ko = ? WHERE id = ?", (title_ko, id_))
                    total_translated += 1
                except (ValueError, IndexError):
                    print(f"  Skipping malformed line: {line}")
                    continue

        conn.commit()
        print(f"  Translated so far: {total_translated}")
        time.sleep(2)  # Rate limiting

    # Final stats
    c.execute("SELECT COUNT(*) FROM cases WHERE title_ko != '' AND title_ko IS NOT NULL")
    translated = c.fetchone()[0]
    c.execute("SELECT COUNT(*) FROM cases WHERE title_ko = '' OR title_ko IS NULL")
    remaining = c.fetchone()[0]

    print(f"\nDone! Translated: {translated}, Remaining: {remaining}")

    # Sample
    print("\n--- Sample translations ---")
    c.execute("SELECT title_en, title_ko FROM cases WHERE title_ko != '' AND title_ko IS NOT NULL ORDER BY RANDOM() LIMIT 10")
    for row in c.fetchall():
        en = row[0][:60]
        ko = row[1][:60]
        print(f"  EN: {en}...")
        print(f"  KO: {ko}...")
        print()

    conn.close()

if __name__ == "__main__":
    main()