upload_suggestions.py

import os
import sqlite3
import json
import csv
import hashlib
import subprocess
from datetime import datetime
from huggingface_hub import HfApi, hf_hub_download

# Settings
HF_TOKEN = os.environ.get("HF_TOKEN")
REPO_ID = "boffire/libretranslate-en-kab-suggestions"
DEST_JSON_PATH_IN_REPO = "suggestions.json"
DEST_CSV_PATH_IN_REPO = "suggestions.csv"
REPO_TYPE = "dataset"

JSON_OUTPUT_PATH = "/tmp/suggestions.json"
CSV_OUTPUT_PATH = "/tmp/suggestions.csv"
CHECKSUM_FILE_JSON = "/tmp/.last_suggestions_json_checksum"
CHECKSUM_FILE_CSV = "/tmp/.last_suggestions_csv_checksum"

# If start.sh found the DB and symlinked it, this will hit first.
# Otherwise we fall back to a dynamic search.
possible_paths = [
    "/app/db/suggestions.db",                       # symlink created by start.sh
    "/app/.local/share/db/suggestions.db",          # when HOME=/app
    "/app/suggestions.db",
    "/root/.local/share/db/suggestions.db",
    "/home/libretranslate/.local/share/db/suggestions.db"
]

def find_db():
    print(f"Running in CWD: {os.getcwd()}")

    # 1. Try known paths
    for path in possible_paths:
        if os.path.exists(path):
            print(f"Found suggestions.db at {path}")
            return path

    # 2. Dynamic fallback: search the whole filesystem
    print("Known paths missed — running dynamic search...")
    try:
        result = subprocess.run(
            ["find", "/", "-name", "suggestions.db", "2>/dev/null"],
            capture_output=True, text=True, timeout=15
        )
        # The stderr redirect in the shell command won't work via subprocess,
        # so we filter empty lines.
        lines = [l for l in result.stdout.splitlines() if l.strip()]
        if lines:
            print(f"Found suggestions.db via search at {lines[0]}")
            return lines[0]
    except Exception as e:
        print(f"Dynamic search failed: {e}")

    print("suggestions.db not found anywhere.")
    return None

def extract_suggestions(db_path):
    suggestions = []
    try:
        conn = sqlite3.connect(db_path)
        cursor = conn.cursor()
        cursor.execute("SELECT q, s, source, target FROM suggestions")
        rows = cursor.fetchall()
        conn.close()

        for row in rows:
            unique_id = hashlib.md5((row[0] + row[1] + row[2] + row[3]).encode()).hexdigest()
            suggestions.append({
                "id": unique_id,
                "source_text": row[0],
                "suggested_text": row[1],
                "source_lang": row[2],
                "target_lang": row[3],
                "timestamp": datetime.now().isoformat()
            })
    except sqlite3.Error as e:
        print(f"SQLite error: {e}")
    return suggestions

def download_existing_json():
    try:
        path = hf_hub_download(
            repo_id=REPO_ID,
            repo_type=REPO_TYPE,
            filename=DEST_JSON_PATH_IN_REPO,
            token=HF_TOKEN,
            local_dir="/tmp"
        )
        print("Downloaded existing suggestions from Hugging Face.")
        return path
    except Exception as e:
        print(f"Could not fetch existing suggestions from HF: {e}")
        return None

def merge_with_existing(suggestions, existing_json_path):
    existing = {}

    if existing_json_path and os.path.exists(existing_json_path):
        try:
            with open(existing_json_path, "r", encoding="utf-8") as f:
                for item in json.load(f):
                    existing[item["id"]] = {
                        "source_text": item["source_text"],
                        "suggested_text": item["suggested_text"],
                        "source_lang": item["source_lang"],
                        "target_lang": item["target_lang"],
                        "timestamp": item.get("timestamp", datetime.now().isoformat())
                    }
        except Exception as e:
            print(f"Failed to read existing JSON: {e}")

    changed = False
    for s in suggestions:
        s_clean = {
            "source_text": s["source_text"],
            "suggested_text": s["suggested_text"],
            "source_lang": s["source_lang"],
            "target_lang": s["target_lang"],
        }

        existing_entry = existing.get(s["id"])
        if not existing_entry:
            changed = True
            existing[s["id"]] = {**s_clean, "timestamp": datetime.now().isoformat()}

    if not changed:
        print("No new suggestions — skipping write/upload.")
        return None

    # Write merged JSON
    final = []
    for id_, data in existing.items():
        final.append({**data, "id": id_})

    with open(JSON_OUTPUT_PATH, "w", encoding="utf-8") as f:
        json.dump(final, f, indent=2, ensure_ascii=False)

    # Also write CSV
    write_csv(final, CSV_OUTPUT_PATH)

    return JSON_OUTPUT_PATH

def write_csv(suggestions, csv_path):
    with open(csv_path, "w", newline="", encoding="utf-8") as csvfile:
        writer = csv.DictWriter(csvfile, fieldnames=[
            "id", "source_text", "suggested_text", "source_lang", "target_lang", "timestamp"
        ])
        writer.writeheader()
        for item in suggestions:
            writer.writerow(item)

def get_checksum(filepath):
    if not os.path.exists(filepath):
        return None
    with open(filepath, "rb") as f:
        return hashlib.md5(f.read()).hexdigest()

def upload_if_updated(filepath, dest_path, checksum_file):
    if not filepath or not os.path.exists(filepath):
        return

    new_checksum = get_checksum(filepath)
    old_checksum = None

    if os.path.exists(checksum_file):
        with open(checksum_file, "r") as f:
            old_checksum = f.read().strip()

    if new_checksum != old_checksum:
        print(f"Uploading updated {os.path.basename(dest_path)} to Hugging Face...")
        try:
            api = HfApi()
            api.upload_file(
                path_or_fileobj=filepath,
                path_in_repo=dest_path,
                repo_id=REPO_ID,
                repo_type=REPO_TYPE,
                token=HF_TOKEN
            )
            with open(checksum_file, "w") as f:
                f.write(new_checksum)
            print(f"Upload successful: {dest_path} at {datetime.now().isoformat()}")
        except Exception as e:
            print(f"Upload failed for {dest_path}:", e)
    else:
        print(f"No changes in {os.path.basename(dest_path)} — skipping upload.")

def main():
    print(f"===== Application Startup at {datetime.now().isoformat()} =====")

    if not HF_TOKEN:
        print("HF_TOKEN not set — skipping upload.")
        return

    db_path = find_db()
    if not db_path:
        return

    suggestions = extract_suggestions(db_path)
    if not suggestions:
        print("No suggestions found — skipping.")
        return

    existing_path = download_existing_json()
    merged_json = merge_with_existing(suggestions, existing_path)

    if merged_json:
        upload_if_updated(JSON_OUTPUT_PATH, DEST_JSON_PATH_IN_REPO, CHECKSUM_FILE_JSON)
        upload_if_updated(CSV_OUTPUT_PATH, DEST_CSV_PATH_IN_REPO, CHECKSUM_FILE_CSV)

if __name__ == "__main__":
    main()