clean some aws ecr docker image

ecr-clean.py

#!/usr/bin/env python
# encoding: utf-8

import subprocess
import json
import logging
import os
import sys

logger = logging.getLogger('docker')
handler = logging.FileHandler('/tmp/ecr-clean.log')
# handler = logging.StreamHandler(sys.stdout)
handler.setLevel(logging.INFO)
handler.setFormatter(logging.Formatter('%(asctime)s - %(message)s'))
logger.setLevel(logging.INFO)
logger.addHandler(handler)

# aws_cli = 'aws'
aws_cli = 'docker run --rm -tiv %s/.aws:/root/.aws mikesir87/aws-cli aws' % (os.environ.get('HOME', ''))


def list_images(repository):
    cmd = '%s ecr list-images --no-cli-pager --repository-name %s' % (aws_cli, repository)
    logger.info('list_images: %s' % (cmd))
    filename = '/tmp/images-%s.json' % (repository)
    with open(filename, 'w') as f:
        p = subprocess.Popen(
            cmd,
            shell=True,
            stdout=f,
            stderr=sys.stderr
        )
        p.wait()
    if p.returncode != 0:
        logger.error('list_images error: %s' % (p.stderr.read()))
        return []
    with open(filename, 'r') as f:
        ret = json.load(f)
    os.unlink(filename)
    return ret.get('imageIds', [])


def delete_images(repository, all_tags):
    if not all_tags:
        return
    count = len(all_tags)
    limit = 100
    i = 0
    while i < count:
        end = i + limit
        tags = all_tags[i:end]
        i = end
        ids = " ".join(map(lambda t: 'imageTag=%s' % t, tags))
        logger.info('delete images: %s %s' % (repository, " ".join(tags)))
        p = subprocess.Popen(
            '%s ecr batch-delete-image --no-cli-pager --repository-name %s --image-ids %s' % (aws_cli, repository, ids),
            shell=True,
            stdout=subprocess.PIPE
        )
        p.wait()
        if p.returncode != 0:
            logger.error(p.stdout.read())


def group_images(images):
    tags = list(map(lambda t: (t['imageTag']), images))
    tags.sort()
    images = { 'dev': [], 'prod': [], }
    for tag in tags:
        if tag.startswith('develop'):
            images['dev'].append(tag)
        if tag.startswith('production'):
            images['prod'].append(tag)
    images['dev'] = images['dev'][0:-20]
    images['prod'] = images['prod'][0:-20]
    return images


def main():
    repository_name = ['project1']
    for r in repository_name:
        images = list_images(r)
        tags = group_images(images)
        delete_images(r, tags['dev'])
        delete_images(r, tags['prod'])

if __name__ == '__main__':
    main()