celcat_to_ics.py, the tiny python parser for getting rid of this hiddeous and unusable CELCAT Timetable from the 90's and that does even have .ics export (at least not on edt.univ-tlse3.fr)

You just have to find the CELCAT url of the .html for your group/course, replace the ending .html with .xml and run it.

Example

curl https://edt.univ-tlse3.fr/FSI/FSImentionL/Math/g29617.xml\
| ./celcat_to_ics.py - -r "L1 Chimie s1 - TPA12:Info+L1 CUPGE s1 - TPA12:Info+\
    L2 Info s1 - TPA41,L2 Info s1 - TPA42,L2 Info s1 - TPA52:Logique" > calendar.ics

In this example, with want to get the .ics from the .xml CELCAT Timetable. We first fetch the .xml and pass it to the parser.

FILTER is a string of the form "G1[,G2...]:C1[,C2...][+...]" which translates to "select events that have G1 or G2 in their group name AND that have C1 or C2 in their course name. The "+" separates multiple of these filters and behaves like a OR.

Note: "+",":" and "," are reserved keywords. You can use spaces. You don't need to enter the full group/course name (only needs to be a part of it).

To be able to import the resulting ics file into Google Calendar (or iCloud), you need to push it to a place that serves HTML pages. The scheme would be

curl cal.xml  ->  celcat_to_ics.py  ->  cal.ics on apache server  -> google calendar
<------------ python server ------->   <------- html server ----->

My configuration is:

• I run the python script on private server at IMT; it runs as a cron job every once in a while (2h)
• it then pushes the .ics files to a public server

TODO

• accept multiple XML/CELCAT file in input for building one unique .ics for all your courses
• allow the spaces in the filters for groups and courses
• allow a way to change the names of the courses for friendlier/handier reading on your smartphone when the original title isn't nice enough
• make sure that the user won't miss any "special" events, because filtering implies that the other events won't be shown

Dependencies

Three pip packages are used: icalendar, lxml and docopt. To install them:

pip3 install icalendar lxml docopt

Performance

WARNING: these tests have been made on a previous version, but the performance is essentially the same on the newest version.

Parsing one average .xml CELCAT file takes about 100ms for a 3460 lines XML file. Example with two filters (more filters => heavier computation)

./celcat_to_ics.py g29617.xml -g TPA11 -c Info  
    -> 0,13s user 0,02s system 94% cpu 0,156 total

When parsing with a little more filters, the execution is slightly longer:

./celcat_to_ics.py g29617.xml -g TPA11,TD,TP -c Info,Math  
    -> 0,15s user 0,03s system 91% cpu 0,197 total

Detailed usage

See usage in celcat_to_ics.py.