Re: [cowiki-dev] Parser test document
"Daniel T. Gorski" <[email protected]>
| Newsgroups | gmane.comp.php.cowiki.devel |
|---|---|
| Message-ID | <20050404003428.GN28092@bantha> |
On 03 Apr 13:46, PAUL HANCHETT wrote: > Daniel T. Gorski wrote: > >I would prefer to have this test document in CVS with a name like > >TEST.IDIOM or TEST.COWIKI.IDIOM (for future use, I already mentioned this > >somewhere). The developers would have to use this document to check the > >parser and if the parser behaviour should have change, it should by > >discussed anyway first on the dev- list. This would lead to appropiate > >changes to the test document first. > > > >If its finished, the right place for it is CVS, not coWiki pages. > How about putting it into a new directory <coWiki_Root>/misc/test/parser > as ReferenceMarkup.txt? > > This document may actually end up being a series documents: > > * One document with only valid markup. > (Does it need to demonstrate all valid combinations?) I don't know if its possible to demonstrate all valid combinations and permutation, but its a "test" document, isn't it? An test should cover things that have to be tested :) This would be the best soultion. > For validation, probably it should also > have a parallel document representing the XML version in the data > store. This would make an automated check of the parser > possible. Potentially, this is a series of individual markup > checks, but still needs to check valid aggregations of markups. > * Disallowed combinations of markup and the resulting (expected) XML > store > > * Similarly, there should be examples of the XML store and the HTML > it should be translated to. > If these things were done, checks of the parser and display formatting > would could be automated... Sounds good for me :) regards dtg