[RFC] readonly branch and repository tests

[Robert Collins]

On Fri, 2007-06-22 at 13:38 +1000, Martin Pool wrote:

> That sounds reasonable.
> 
> In testing it seems that we'll typically want to put some information
> into the repository while it's mutable, then examine it in the
> readonly form.  With bundles it would be straightforward to write into
> eg the default format then convert to a repository, but if we want to
> thoroughly test a foreign repository and we can't convert from bazaar
> into that format we might need to give a more abstract description of
> what data should be in there.

I've done some prepatory work on such a thing which is I think merged
already - but I think its an isomorphic problem to be able to interpret
an abstract description which is capable of describing anything our data
can; and getting that description from a populated branch/repository. 

> We want it to be easy for people to make the right decision about
> which tests should be applied to readonly formats and when the
> conversion to readonly is done.  It might be easy to accidentally omit
> to really exercise the readonly format by continuing to use the
> writable one.

I wonder if we can tastefully have the mutating tests assert that the
repo was mutated (after the setup phase).

> In some ways it would be nice to have a different base class for these
> tests, so that they can be stored in the same file as the other tests
> for that logical area, but also clearly distinct.

Thats another way of separating them, but it has the same failure mode -
put the method on the wrong subclass and it won't run on readonly
repositories.

-Rob

-- 
GPG key available at: <http://www.robertcollins.net/keys.txt>.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20070622/32a67bbc/attachment-0001.pgp