docbrown

docbrown is an empirical progress library. It determines the overall duration and progression of a process by looking at the time it took in the past.

docbrown might be an option for you, if you have one long running task which parts take very different amounts of time. It is also helpful when used in environments where you can’t inform consumers about the progress of a task out-of-band (like a WSGI request/response cycle vs a WebSocket).

Example Usage

Record progress:

import time
from docbrown import record_progress

with record_progress('process_name', ident='my_ident') as record:
    # outputs "my_ident", but normally would return a 
    # unique id if not set manually
    print(record.ident)
    # do some stuff that takes time
    record('loading_data')
    time.sleep(4)
    record('calculating_matrices')
    time.sleep(9)
    record('rendering_structures')
    time.sleep(23)
    record('uploading_models')
    time.sleep(15)

As docbrown determines progression by looking at the past every process needs to run at least once before. The following get_progress call will therefore only work if you’ve executed the code above at least once!

Get the progression of our process in another process:

from docbrown import get_progress
print(get_progress('my_ident'))

Future

There are some things that would be nice, but have not been implemented yet.

• additional storage backends apart from SQLite
• configurable strategy for aggregating phase durations apart from the simple arithmetic average like median
• code path detection that takes optional phases into account and updates the expected duration and progres on-the-fly