CSV format

Starting with version 2.0.0, the schedule is read from a CSV file rather than fetched from an ecolle instance. This page describes the expected file.

Where to put the file

By default the script reads, inside the container:

input/colles.csv

The script runs with /app as its working directory, so that is /app/input/colles.csv. The provided docker-compose.yml mounts the input/ folder that sits next to it onto /app/input:

volumes:
  - ./input:/app/input:ro

So on your server you only have to create an input folder next to docker-compose.yml and drop your export in it:

my-ecolle/
├── docker-compose.yml
├── .env
└── input/
    └── colles.csv      <- your export

COLLES_CSV_PATH is a path inside the container. These two are equivalent and both work:

COLLES_CSV_PATH=input/colles.csv
COLLES_CSV_PATH=/app/input/colles.csv

If you want a single file elsewhere, mount it explicitly instead:

volumes:
  - /srv/colles/colles.csv:/app/data/colles.csv:ro

with COLLES_CSV_PATH=/app/data/colles.csv.

The file must be world-readable (chmod a+r input/colles.csv): a file that is only readable by its owner is not always readable inside the container.

Expected content

The file must be comma separated and start with a header row:

Date,Matière,Colleur,Jour,Salle,Début,Fin
2026-09-24,Anglais,Mme Hatri,jeudi,L037,17:00,18:00
2026-09-30,Physique,M Blain,mercredi,L124,15:00,16:00
2026-10-06,Mathématiques,M Jouve,mardi,L005,16:00,17:00

Columns

Column	Required	Meaning
Date	yes	Date of the colle, YYYY-MM-DD
Matière	yes	Subject, e.g. Mathématiques
Colleur	yes	Who examines you, shown by the {colleur} field
Salle	yes	Room, shown by the {salle} field
Début	yes	Start time, HH:MM
Fin	no	End time, used by the {fin} and {heure_fin} fields
Jour	no	Day name. Stored but never used: the day is computed from Date, so an inconsistent value cannot mislead you

Date and time

Date is parsed on its own, so a wrong Jour cannot shift your schedule. Accepted layouts are YYYY-MM-DD (recommended), DD/MM/YYYY, DD-MM-YYYY and DD.MM.YYYY.

Début accepts HH:MM, HH:MM:SS, HHhMM and HHh.

Tolerated variations

The reader is deliberately forgiving:

• accents are optional: Matiere works as well as Matière
• capitalisation, surrounding spaces, _ and - in headers are ignored
• a UTF-8 BOM (as produced by Excel) is stripped
• extra columns are ignored: any header the tool does not recognise is simply skipped
• blank lines are skipped
• a duplicate column is refused, rather than letting the last one silently win

Rows with an unreadable date or time are skipped with a warning, and the rest of the file is still used. A row of data may not span several lines.

Encodings

The file must be UTF-8. Excel's "CSV UTF-16" and "ANSI" exports are refused with an explicit message; re-export as CSV UTF-8. A UTF-16 byte-order mark is recognised and named in the error.

Fetching from ecolle

By default, when the CSV file is missing the script falls back to logging in to your ecolle instance, as it did before 2.0.0. That fallback needs BASE_URL, COLLES_USERNAME and COLLES_PASSWORD, and fails with an explanatory message when they are not set. Since the instance is usually unreachable, it is better to switch it off:

DISABLE_ECOLLE_FETCH=true

The CSV file then becomes the only source. If it is missing, the run stops immediately with a message explaining what was looked for, instead of trying to reach ecolle. This is a good way to be sure the container is really using your file.

What happens on an error

The two situations behave differently, on purpose:

Situation	What happens
CSV file missing	the scraper fallback is tried; if it is disabled or unavailable, the run stops with an explanation and exit code 1. The cache is not used.
CSV file present but unusable (wrong separator, missing column, no usable row, bad encoding)	the run continues from the agenda cached by an earlier run, so you keep receiving notifications, and prints why. Exit code 1 only if there is no usable cache either.

Sample file

An input/colles.example.csv sample is included in the repository. Copy it and replace the rows with your own:

cp input/colles.example.csv input/colles.csv