import boto3
from botocore.client import Config
import subprocess, os, logging, json

logging.basicConfig(level=logging.DEBUG)
logger = logging.getLogger()

os.environ['HOME'] = '/tmp'
os.environ.setdefault('PYTHONHOME', '/usr')
os.environ.setdefault('LD_LIBRARY_PATH', '/opt/libreoffice25.2/program')

s3 = boto3.client(
    's3',
    region_name='us-east-1',
    config=Config(signature_version='s3v4'),
    endpoint_url='https://s3.us-east-1.amazonaws.com'
)

def download_file_from_s3(bucket: str, key: str, dest: str):
    logger.debug("Downloading s3://%s/%s → %s", bucket, key, dest)
    s3.download_file(bucket, key, dest)

def convert_pdf_to_docx(input_pdf: str, output_dir: str = '/tmp') -> str:
    logger.debug("Converting PDF to DOCX: %s", input_pdf)
    subprocess.run([
        'libreoffice', '--headless',
        '--infilter=writer_pdf_import',
        '--convert-to', 'docx:MS Word 2007 XML',
        '--outdir', output_dir,
        input_pdf
    ], check=True)
    base = os.path.splitext(os.path.basename(input_pdf))[0]
    return os.path.join(output_dir, f"{base}.docx")

def cleanup_files(*paths):
    for p in paths:
        try:
            if os.path.exists(p):
                os.remove(p)
                logger.debug("Removed %s", p)
        except Exception as e:
            logger.warning("Error deleting %s: %s", p, e)

def handler(event, context):
    bucket = event.get('s3_bucket')
    key    = event.get('s3_key')
    if not bucket or not key:
        return {'statusCode': 400, 'body': "Missing s3_bucket or s3_key"}

    pdf_name  = os.path.basename(key)
    local_pdf = f"/tmp/{pdf_name}"

    try:
        download_file_from_s3(bucket, key, local_pdf)
        docx_path = convert_pdf_to_docx(local_pdf)

        converted_key = f"converted-docs/{os.path.basename(docx_path)}"
        s3.upload_file(docx_path, bucket, converted_key)
        logger.debug("Uploaded to s3://%s/%s", bucket, converted_key)

        presign = s3.generate_presigned_url(
            'get_object',
            Params={'Bucket': bucket, 'Key': converted_key},
            ExpiresIn=3600
        )
    except subprocess.CalledProcessError:
        logger.exception("LibreOffice conversion failed")
        return {'statusCode': 500, 'body': "Conversion failed"}
    except Exception as e:
        logger.exception("Error in handler")
        return {'statusCode': 500, 'body': f"Error: {e}"}
    finally:
        cleanup_files(local_pdf, locals().get('docx_path', ''))

    return {
        'statusCode': 200,
        'headers': {'Content-Type': 'application/json'},
        'body': json.dumps({'download_url': presign})
    }