Implemented the app for slicing text given start and end time

bohJiang12 commented 1 month ago

Implementations

The sliced text results are tested via:

Verify if additional view is added into the input .mmif file
Compare the sliced text is the same as the one returned by text_document_helper.slice_text of clams-python=1.2.4

Should the TextDocument be one of the inputs in app's metadata rather than only TimeFrame (while using add_input function)?

bohJiang12 commented 3 weeks ago

Updated version of clams-python to 1.2.5
Removed auxiliary comment blocks
Added an option/argument for the app, so that users are able to select TimeFrames containing wanted labels, for example, "slate", "credits", "bars", etc.

Using the .mmif file exported by the whisper-wrapper that I run at the very first time, an edge case for TimeFrame was its start and end properties were the same, therefore an assertion error would be raised since our get_annotations_between_time feature assumed start < end. So probably we need to fix this assertion to tolerate this edge case so that start <= end?
In terms of input format that users select their wanted labels contained in TimeFrame, I implemented its type as str. Then, it's fine for users choose only one label, that is, 'credits' for example. However, if multiple labels are selected, like "credits, slate, chyron", I need to tokenized this parameter into a list of str as ['credits', 'slate', 'chyron']. Since my initial and ideal thought towards this input format is to ask users to give a list/array, however, it seems that the app has no such type option. Do you think the str type input is enough for the app?

bohJiang12 commented 2 weeks ago

This is the updated app implementations that