first way: take a message text as a sequence of the not-white space character, split the message and take first word as start_time, second word as end_time, third as project_name.
drawback the message "12 to 13 pencil" will not work
second way: recognize the sequence of part of speech (English) and take number (NUM) second NUM and the nearest NOUN or ADJ as a project_name.
advantage: "12 to something 13 work" will work
drawback: problem with spicy library install on docker. not possible via reuqirements.txt but only via RUN python -m spacy download en_core_web_sm
for example, I have a message: "12 to 13 pencil"
How should the app recognize the timelog message?