Documentation.md

Documentation 📰📃😃

Table of Contents

• Documentation 📰📃😃
  • Table of Contents
    • Create and start a python virtual env
    • The Procedure
      • Step 1. Run record_batch.py to create a new batch of phrases & words
      • Step 2. Cleanup your new batch
    • Structure of the data
    • To-Do
      • Completed
      • Pending
    • Requirements (.txt)

---

Create and start a python virtual env

# Starting a virtual environment
py -m venv ll_env

# Activate the virtual environment
ll_env/scripts/activate

---

The Procedure

Step 1. Run record_batch.py to create a new batch of phrases & words

You can look at a list of words or phrases and generate an audio recording of the word or phrase while holding down the space bar on your keyboard.

Each word or phrase recorded, gets a newline in the words.txt file, and the audio file is saved as a .wav file in the new-audios/ directory.

Example:

assurance (new-audios\assurance.wav)
deep-seated (new-audios\deep-seated.wav)
burden (new-audios\burden.wav)
load (new-audios\load.wav)
accomplishments (new-audios\accomplishments.wav)
...

Step 2. Cleanup your new batch

Trim any dead/silent audio in the .wav audio file.

Structure of the data

Note: all of the data is in the data/ directory.

• data/
  • audio/ (folder)
  • backed-up/ (folder)
  • dict/ (folder)
  • sentences/ (folder)
  • videos/ (folder)
  • markdown_links.txt (text file)
  • mistakes.json (json file)
  • words.json (json file)

---

• data/audio/ (audio) brendan's recordings of pronunciations in .wav audio files
  • audio/phrases/ (phrases)
  • audio/words/ (words)

---

• data/dict/ (dict) stands for dictionary contains data saved from the webster dictionary API
  • dict/audio/ (.wav audio) downloaded from dictionary API's media endpoint
    • dict/audio/phrases/ (phrases)
    • dict/audio/words/ (words)
  • dict/json/ (json) json response from dictionary API for {searched word} + .json
    • dict/json/phrases/ (phrases) phrases are camel case, example: a_thrill_swept.json
    • dict/json/words/ (words)

---

To-Do

Completed

• Only pages 5 and 6 are left to complete
• Currently at: Harsh (id = 124) as of 10/07/2024
• Fix recordings that are too long with blank/no noise. Write script to scan for and output all files with duration > 3 seconds.
• pip install the package for env variables - then fix the api_req file to remove from .gitignore
• If the md file will say List of words: then write a message like This word or phrase doesn't seem to exist in the English dictionary - maybe I misunderstood the text in your image
• Mark the edited audio files w/ a flag to not have the volumex(2) used again during the video creation
• Need to replace all json objects in the words.json file with their new durations from after.json
• write script to seperate these into two categories: 1. Phrases and 2. Words.
• Update the long-entries/before.json file

Pending

• Currently at: Harsh (id = 198 (125-198)) as of 10/13/2024
• Update the Requirements.txt
• Finish the MoviePy script to prompt to record using each word in a sentence
• Write script to separate into batches by date

---

Requirements (.txt)

• certifi==2024.8.30
• charset-normalizer==3.3.2
• comtypes==1.4.7
• idna==3.9
• PyAudio==0.2.14
• pydub==0.25.1
• pypiwin32==223
• pyttsx3==2.97
• pywin32==306
• requests==2.32.3
• setuptools==74.1.2
• SpeechRecognition==3.10.4
• typing_extensions==4.12.2
• urllib3==2.2.3