[CFRG] Re: Where should test vectors live?

[Carsten Bormann <[e-mail]>]

On 20. Sep 2024, at 02:17, Christopher Patton <cpatton=[e-mail]> wrote:
> 
> In theory you could write a script that downloads https://www.rfc-editor.org/rfc/rfc9180.txt and write a parser to pull out the test vectors, but does anyone really do this?

It is actually much better to download https://www.rfc-editor.org/rfc/rfc9180.xml and use a tool such as kramdown-rfc-extract-sourcecode [1] to extract the <sourcecode snippets in the document.
If the document authors have used the type= and name= attributes, the extraction result can go right into CI scripts and other processing.
(In RFC 9180, the type attribute is used wisely, but the name attributes are all empty, so kramdown-rfc-extract-sourcecode will guess file names from the context.)

Maintaining your own test vector blob of course gives you more ways to evolve the test vectors (and to handle those pesky errata), but I wanted to point out that it is easy to get at the actual test vectors in the RFCXML document without writing fragile heuristic text file parsers.

Grüße, Carsten

[1]: https://github.com/cabo/kramdown-rfc/wiki/Extracting-sourcecode