Parsing RPM Spec files to generate json data
The main.py generates 2 json files (one pretty print, one minimized) from the folder ../fedora-spec/**/*.spec.
The fields (parsed from the RPM spec) are be:
- name
- version
- url (of the source project)
- license
- summary
- description
- packages (list of packages names and summary)
For the "Tags", they are added as comment in the spec file:
...
# tags: foo, bar
# category: Synthesiser
# type: ..
...
Other tags are used, but not yet parsed:
# GUIToolkit: Qt4, Fltk, GTK3, SWT ...
# LastSourceUpdate: 2020, 2005
Might be replaced by a tag: Legacy or tag: Obsolete
Run python main.py <path to spec files> a glob is used to scan sub-folders **/*.spec.
For test, the application can read the tests directory, which contains a few folders. Run python main.py .
To run on the full list of spec, clone the fedora-spec spec repo next to this repo. Run python main.py ../fedora-spec
- Output more statistical analysis
- Add a mass modification feature
This repo embedded a modified copy of https://github.com/bkircher/python-rpm-spec/ that adds support
for %description (submitted upstream)
and # Tags: ... that we keep local to this repo, because these are non-standard.