The program should parse the file and output the statistics of the top 10 packages that have the most files associated with them.
An example output could be:
python package_statistics.py --arch amd64
1. <package name 1> <number of files>
2. <package name 2> <number of files>
......
10. <package name 10> <number of files>
You can use the following Debian mirror
To install locally you need to have
- Pyton 3.7
- pipenv
Once above is done install requirements:
pipenv --python 3.7
pipenv shell
pipenv instal
When there are issues with using pipenv you can install with pip below packages
nose = "==1.3.7"
mock = "*"
ipython = "*"
requests = "*"
nose-progressive = "*"
To get help with script usage just do as follows:
python package_statistics.py -h
Usage: package_statistics.py [options]
Options:
-h, --help show this help message and exit
--mirror=MIRROR
--arch=ARCH
--level=LEVEL
--cache Use local cache
Example usage (downloading and analuzing amd64)
python package_statistics.py --cache --arch arm64
Using option cache means that if you rerun script it will use donloaded file from mirron in local /var/tmp folder.
If you want to get more verbose output use --level=debug
To test script just use nose, ie.
nosetests tests
That is going to give you overview of number of tests performed and potential fails :)
Basically I used barebone Python 3.7 with a bit of support of requets (to make my life easier).
- Script does not use parallel processing to analyze and sort results.
- Instead of lame and slow loop Regexp is being used:
- getting only repeated package names in Contents file
- counting occurrences of same package name in list
- desc sorting number of repetitions
Total time spend on exercise ~ 2h