app.py

#######################################################################
#
# This code shows how to use mediainfo from AWS Lambda. 
#
# PREREQUISITES:
#   Set region in AWS_REGION environment variable in the Lamdba environment
#
# USAGE:
#   See https://www.bigendiandata.com/2019-12-10-MediaInfo_AWS_Lambda/.
# 
# USEFUL REFERENCES:
#   https://github.com/iandow/mediainfo_aws_lambda
#   https://aws.amazon.com/blogs/compute/extracting-video-metadata-using-lambda-and-mediainfo/
#
#######################################################################


import json
import logging
import boto3
import botocore
import os
from pymediainfo import MediaInfo
from botocore.config import Config

region = os.environ['AWS_REGION']

def get_signed_url(expires_in, bucket, obj):
    """
    Generate a signed URL
    :param expires_in:  URL Expiration time in seconds
    :param bucket:
    :param obj:         S3 Key name
    :return:            Signed URL
    """
    s3_cli = boto3.client("s3", region_name=region, config = Config(signature_version = 's3v4', s3={'addressing_style': 'virtual'}))
    presigned_url = s3_cli.generate_presigned_url('get_object', Params={'Bucket': bucket, 'Key': obj}, ExpiresIn=expires_in)
    return presigned_url

def lambda_handler(event, context):
    logger = logging.getLogger()
    logger.setLevel(logging.INFO)
    logger.info('event parameter: {}'.format(event))
    tmp_filename='/tmp/my_video.mp4'

    s3 = boto3.resource('s3')
    BUCKET_NAME = os.environ.get("BUCKET_NAME")
    S3_KEY = os.environ.get("S3_KEY")

    ##########################################################################
    # I'm going to show two ways to invoke MediaInfo:
    #
    # FIRST WAY: download the vidoe file to local storage and provide the local file path as input to MediaInfo. 
    # Disadvantage here is that Lambda only provides 512MB of disk space to save that file.
    #
    # SECOND WAY: Provide an S3 URL as the input to MediaInfo. 
    # Disadvantage here is you must have compiled libcurl support into the MediaInfo library, which I did in https://github.com/iandow/mediainfo_aws_lambda/blob/master/Dockerfile
    ##########################################################################
    
    ##########################################################################
    # FIRST WAY: download the vidoe file to local storage and provide the local file path as input to MediaInfo.
    ##########################################################################

    try:
        s3.Bucket(BUCKET_NAME).download_file(S3_KEY, tmp_filename)
    except botocore.exceptions.ClientError as e:
        if e.response['Error']['Code'] == "404":
            print("The object does not exist: s3://" + BUCKET_NAME + S3_KEY)
        else:
            raise

    media_info = MediaInfo.parse(tmp_filename, library_file='/opt/libmediainfo.so.0')

    print(str(media_info.to_json()))

    for track in media_info.tracks:
        if track.track_type == 'Video':
            print("track info: " + str(track.bit_rate) + " " + str(track.bit_rate_mode)  + " " + str(track.codec))

    ##########################################################################
    # SECOND WAY: Provide an S3 URL as the input to MediaInfo.
    ##########################################################################

    SIGNED_URL_EXPIRATION = 300     # The number of seconds that the Signed URL is valid

    # Generate a signed URL for the uploaded asset
    signed_url = get_signed_url(SIGNED_URL_EXPIRATION, BUCKET_NAME, S3_KEY)
    # Launch MediaInfo
    media_info = MediaInfo.parse(signed_url, library_file='/opt/libmediainfo.so.0')
    for track in media_info.tracks:
        if track.track_type == 'Video':
            print("track info: " + str(track.bit_rate) + " " + str(track.bit_rate_mode)  + " " + str(track.codec))

    # Finish the Lambda function with an HTTP 200 status code:
    return {
        "statusCode": 200,
        "body": json.dumps({
            "message": str(media_info.to_json())
        }),
    }