On Wednesday 16 December 2020 07:12:32 pm Slávek Banko via tde-users wrote:
On Friday 02 of October 2020 17:37:02 Michael via
tde-users wrote:
On Thursday 01 October 2020 06:37:00 pm Slávek
Banko via tde-users wrote:
Is there any automated way to get a list of all the programs?
Hi Michael,
I seem to have a fairly successful procedure. During the preparation of the
PHP routing script for creating new issues in TGW, I needed to get a table
that would contain all the binaries (including games and KCM modules) and
information about the git repository to which it belongs.
I have used Contents files that are accessible in the apt repository to
find all the files that are of interest. From this I found information
about the corresponding deb package name. And by searching in
the "control" files in the git repository tde-packaging, I found the
corresponding git repository for each binary (each application).
It seems to me that this table, or more precisely the procedure for
obtaining it, could be useful for automating the retrieval of information
and creating the initial look of wiki pages for applications. In that
procedure, finding descriptions from the corresponding deb package could
be added, which could serve as a basis for descriptions on the wiki pages.
At the same time, information about the git repository will be useful for
users to search for and create issues for applications.
What do you think about that?
Note: Thank you Leslie for your suggestion.
I think that’s definitely the way to go. It’s going to probably add a few too
many individual application pages that are actually slaves, launchers and
whatnot, but, really that’s not that big a deal.
To me the only way to remove the ‘excess’ would be to take the auto-created
binaries table and then manually remove those that aren’t actually
applications. The downside is it’s now manual, and if some new application
gets added to TDE, then someone has to remember to add it to the app pages
script.
# # #
Although that last sentence made me realize my assumption; that this is going
to be run on a scheduled basis? If this is a run once then it’d be a
different workflow and manually removing all the ‘cruft’ would be a good
idea.
# # #
In either option (run once, run scheduled) I can do the automate ‘load into
wiki’ part. I’ll need:
- Copy of the current wiki to install in a dev environment.
- All the deb packages, full apt repository?, complete git repository?, or
whatever to extract ‘info’ from.
- That table (, the scripts to build it, or the .bash_history of the commands
you used to create it).
- A finished example wiki individual application page to work against.*
* Anyone good at UIX? Want to create a wiki page for KWrite?
Best,
Michael