timlev/read-story

Read Text Out Loud

read-story

Read Text Out Loud

Read-Story is a command line program takes the text from an html document, downloads sound files for each word, and creates a clickable html file that plays the pronunciation of each word that is clicked.

The purpose of this app is to support students learning to read in English.

Requires: python2.7 python-html5 ffmpeg or avconv (libav-tools) LibreOffice (for best results) SoX (optional, for using record_missing_words.py)

Steps:

1. Export document as html from LibreOffice to html_input
2. Type command to create clickable html file in html_output/new_NAMEOFFILE.html: python parse_html.py html_input/NAMEOFFILE.html
3. Record missing sounds as mp3 files in sounds/ a. Words are listed under missing_words/NAMEOFFILE.html_missing_words.txt b. Recordings can be made using command: python record_missing_words.py missing_words/NAMEOFFILE.html_missing_words.txt c. Hit Ctrl-C after saying each word to stop recording
4. Open html_output/new_NAMEOFFILE.html in preferred web browser.

read-story's Issues

Dashes are not rendering correctly

This is an issue with going from unicode to ascii. I am temporarily displaying dashes as --

Another problem is that dashes are not combining words that should not be combined.

A better name for this bird might be Gary--I think we've got a rooster here.

Ideally

Gary--I

should be two separate, clickable words.

Longer passages freeze Chrome on Ubuntu 16.10 (Awww Snap)

Passages do not freeze in Firefox.

This might be an issue with preloading too many audio files

Quotation marks are not rendering correctly

From new_The One That Got Away:
Hey, where did you come from? he asked,