Replaces existing SED and PY file

[ctiijima]

Still needs the newline after every sentence function worked in to finish a complete replacement.

[yhwang]

one question: where can we use this csv file?

[ctiijima]

It will write the CSV in the folder where the RTF files were sourced. If you would prefer to continue to use individual TXT files, that can be modified into the script functionality. I am more or less providing this to you because it was part of the preliminary work to bring the ezDI dataset to the Tensorflow/Datasets project before I drop from the group.

[tonanhngo]

Hi Chris, thanks for putting this script together. It appears the csv usage along with the listing of wav file name is really for the TF Datasets for ezDI that we are working on now. You might want to check into my fork instead: https://github.com/tonanhngo/datasets.git
For the code pattern, it's also good to consolidate some of the shell script commands into the python script as you mention, but we would want the output to be a regular text file without the wav filename. If you want to make another version for this usage, that would be good. In this case, please update the README file also to reflect the new way to process the data.