W3C home > Mailing lists > Public > public-xml-schema-testsuite@w3.org > June 2010

Re: Mystery files in the XML Schema test suite (.imp, .red, .inc, .scm, etc.)

From: C. M. Sperberg-McQueen <cmsmcq@blackmesatech.com>
Date: Sat, 19 Jun 2010 16:52:47 -0600
Cc: public-xml-schema-testsuite@w3.org
Message-Id: <36634D02-66F2-40F5-B1E8-615F7D59EC1C@blackmesatech.com>
To: "C. M. Sperberg-McQueen" <cmsmcq@blackmesatech.com>

On 19 Jun 2010, at 15:47 , C. M. Sperberg-McQueen wrote:

> In the distributed version of the XML Schema test suite,
> various subdirectories of msData have a variety of files
> with names like addB087.imp, addB094.red, etc.
>
> As far as I know, nothing in the test suite definition or
> in the relevant specs requires XML instance test documents
> to be given filenames ending in .xml, or schema documents
> to be given names ending in .xsd, but this caught my eye.
> None of these documents appear to be mentioned in any of the
> metadata files in msMeta, so I wonder:
>
>  - What are they?
>  - Are they needed for anything?  Or are they just
>    artefacts of the tools originally used to produce
>    the ms tests?

Closer examination suggests that the .imp, .red, and .inc
files are schema documents used in tests of import, redefine,
and include.  They aren't listed in the metadata perhaps
because if they were passed in as parameters it would defeat
the point of the test.

>  - Would it make any difference (for better or worse) if
>    the test suite strove for the state of having no documents
>    in the xxxData directories except documents referred to
>    by one or more entries in the metadata?

Further reflection shows that this is tightly bound up
with the question "Just what schema documents is an
implementation running the test suite expected to consult?"
The schema wg is currently working on clarifying that
question.

 From a cleanliness point of view, it would be nice to
have all, somewhere, an explicit list of all the documents
which form part of the test suite (preferably with information
about which tests they are part of), but in the current state
of the XSTS vocabulary that would be hard to manage.

Further information would still be welcome.

-- 
****************************************************************
* C. M. Sperberg-McQueen, Black Mesa Technologies LLC
* http://www.blackmesatech.com
* http://cmsmcq.com/mib
* http://balisage.net
****************************************************************
Received on Saturday, 19 June 2010 22:53:17 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Saturday, 19 June 2010 22:53:17 GMT