speechmarkdown / speechmarkdown-test-files

Test data files to assist those writing parsers, formatters, libraries, and other software.

Speech Markdown Test Files

The open source version of the Speech Markdown Test Files

The files contained in this project are to assist those writing parsers, formatters, libraries, and other software to have a consistent set of test data files.

These files are not for a specific testing framework, rather they contain Speech Markdown formatted text and what the corresponding text would be after a formatter has converted it. Samples are shown for Plain Text as well as platform-specifice Speech Synthesis Markup Language (SSML).

Each group of test files are in their own folder and file naming follows this convention:

extension	description
.smd	Speech Markdown
.txt	Converted to Plain Text
.alexa.ssml	SSML - Amazon Alexa
.google.ssml	SSML - Google Assistant

All data files are under the test-data folder.

We welcome your contributions to improving this project. All submissions will be reviewed for technical accuracy and possibly reworded to comply with style guidelines required by our editorial team.

The goal of this project is to support voice designers, developers, and tool creators in adoption and enhancement of Speech Markdown. We want you to be successful. If this project is not helping you achieve your goals, then please let us know where we can do a better job. Thank you for your contributions!

License Summary

The documentation is made available under the Creative Commons Attribution-ShareAlike 4.0 International License. See the LICENSE file.

The sample code within this documentation is made available under a modified MIT license. See the LICENSE-SAMPLECODE file.

speechmarkdown-test-files's Issues

Inconsistency in test case

I found an inconsistency in this test

Is [break:none] expected to be valid syntax? e.g. an attribute value without quotes?

One source of truth for all test cases?

I see that the JS implementation holds way more test cases than this repository.

How should implementers in other languages go about this? Ideally there is one "source of truth" for test cases, independently of any implementation.

I have written a script that extracts all test cases from the Typescript repository (using an AST walker); would you be willing to accept a PR that adds those tests into this repository and also a PR for the Typescript repository to use the reference tests as defined here?