TODOS.txt

__v1__  ===============  DONE
Given I have a source language phrase saved in a text file
And I don't have a translation file with that phrase yet

When I run the program
And the phrase is read
And a translation API is contacted

Then I receive a translation
And the translation is saved to a text file

__v0.1__  ===============
Given I have more than 1 source language phrase saved in a text file
And I don't have a translation file with that phrase yet

When I run the program

Then the phrases are read in bulk
And a translation API is contacted only once


__v0.2__  ===============
Given I have a target language phrase saved in a text file
And I don't have an mp3 file with that phrase yet

When I run the program
And the phrase is read
And a text-to-speech API is contacted

Then I receive a stream of audio data
And the audio data is saved to an mp3 file

__v0.3__  ===============
Given I have a couple of mp3 files with target language phrases
And I have a couple of mp3 files with the corresponding source language phrases

When I run the program
And I specify which phrase __pairs__ (!) are to be concatenated
And I specify the length of pause between phrases in the pair
And I specify the length of pause between pairs

Then I receive a stream of audio data
And the audio data is saved to an mp3 file
And the source phrase always comes first in the pair


TODO:
1) Use argument parser in create_recording.py
2) Always delete output folder within mp3_outputs if an error occurs while connecting to Bing API.
Alternatively, add the ability to dump remaining CSV lines to a file and resume from where you left off (less wasteful) - this
would require:
 a) using a different data structure (ability to pop things off stack),
 b) Checking for number of items in output folder,
rather than the presence of the output folder.