You are not logged in.
I develop pajman. a boosted bash based package manager.
What I found is that it would be easier to understand and implement pacman alternatives if variable names where in a concrete file to be able to check them.
Pacman should have a file in which we have all the variables that sit in configuration or information files (/etc/pacman.conf, /var/lib/pacman/local/*/depends, etc).
Right now if we look for %DEPENDS% we get that it is inlined in the code:
[root@patito src]# busca %DEPENDS%
gensync
129: echo "%DEPENDS%" >>depends
updatesync
132: echo "%DEPENDS%" >>depends
convertdb.c
102: fputs("%DEPENDS%nn", fp);
db.c
450: if(!strcmp(line, "%DEPENDS%")) {
581: fputs("%DEPENDS%n", fp);
[root@patito src]#
if there was a file with all variables, pajman would be able to check if there is something new to keep in mind and tell me.
I could check if the vesion of pacman changed and then using cvs download this new version of the variable definitions.
This file could also implement a clear explanation of what is what every variable means and what it does. This way I could also check if the explanation changed and know that i have to update pajman, because the few things that it does may not work anymore.
i suggest having something like:
definitions.c
23: VARNAME_DEPENDS="%DEPENDS%";
24: // DEPENDS CONTAINS A LIST OF PACKAGES THAT ARE NEEDED FOR THIS PACKAGE TO RUN
25: // ASSOCIATED TO EACH PACKAGE YOU CAN FIND REQUIRED VERSION INFORMATION (ex: kdelibs>=3.2.3)
I post this in the forum first to see other opinions about it.
Offline
you can probably easilly parse it the "pacman way" - a line that matches %*% is the "heading", then you read until an empty line is found.
Offline
fputs("%DEPENDS%n", fp);
%*% = %DEPENDS%
read until end of line:
n", fp);
what's the point of that?
what i want to do is to be able to know when pacman has new/modified features.
Offline