ls-s3-logs 0.1.2

A library to parse S3 log files.

A python library to parse S3 log files.

Warning

Unit tests currently require actual S3 credentials (and a bucket with logs) and can therefore only be run manually. Mocking the relevant parts of boto.s3 is on the roadmap. Contributions are welcome :)

Purpose

Download S3 logs from a bucket, and parse them.

This application does not store the log objects generated and leaves that to other applications.

Install

Eventually from pypi …

Use

To get logs, simply use the Downloader class:

from itertools import chain
from pprint import pprint
from lss3logs.download import Downloader

MY_ACCESS_KEY_ID = 'XXX'
MY_KEY_SECRET = 'XXX'
MY_S3_BUCKET_NAME = 'XXX'

downloader = Downloader(
    connection=None,
    aws_access_key_id=MY_ACCESS_KEY_ID,
    aws_key_secret=MY_KEY_SECRET,
)

# download 10 logs
logs = downloader.download_files(
    MY_S3_BUCKET_NAME,
    prefix='logs/',
    max_logs=1)

entries = [ log.entries for log in logs ]
entries = list(chain.from_iterable(entries))
[pprint(entry.__dict__) for entry in entries]

Running tests

First you need to specify the test config, which contains the AWS credentials and details of bucket tot test with. python-testconfig is used to manage the test configuration.

Copy test_config.ini.sample to test_config.ini (in the same directory) and set correct values:

export NOSE_TESTCONFIG_AUTOLOAD_INI=`pwd`/test_config.ini

To test with nose:

python setup.py nosetests

or running nosetests directly:

nosetests -s --exe

Directly and with coverage:

nosetests -s --exe --with-coverage --cover-package=lss3logs

(Note: the –exe includes python files whoch are executable, so it’s optional if you don’t have any.)

Code quality

Checking code with pylint:

pylint lss3logs

TODO

• mock boto output (see https://github.com/eykd/duo/blob/master/test_duo.py for an example)

• fix as many errors as possbile reported by pylint

• bring test coverage to 100%

• use Sphinx for docs

Credits

The regular expression for parsing the log lines is copied from a script by “kkowalczyk” located at http://code.google.com/p/kjk/source/browse/trunk/scripts/test_parse_s3_log.py