CVS hierarchy and file name convention

Hi,

I had an action item to find someone to describe BenToWeb CVS hierarchy and 
file name convention.


The hierarchy is the following:

* metadata

* testfiles
   + resources
     - applets
     - audio
     - css
     - dtds
     - flash
     - images
     - scripts
     - video

So we have metatdata and test samples in two separate directories.
The 'metadata' directory contains the TCDL files.
The 'testfiles' directory contains the test samples, but any supporting 
files (images, CSS, JavaScript go into a separate directory under the 
'resources' directory.

We should also consider if we want this at the top level or under a 
directory per technology,
i.e. an 'XHTMLsamples' directory containing the above hierarchy, an 
'SVGsamples' directory (if we get that far), etc. Do we want a separate 
structure for XHTML samples with and without JavaScript? (We don't do that 
in BenToWeb.)


For the file name convention we used to have something complex but in the 
new version of our test suite, we adopted something simpler.

First the old, complex convention:
technology_wcag2_date_x.y_lz_scz_nnn, where the sections separated by 
underscores have the following meaning:
- technology: the technology for which the test suite is developed, so 
xhtml1, xhtml2, xforms1, etc.,
- wcag2: a constant referring to WCAG 2.0,
- date: the date on which the WCAG 2.0 draft was published, in yyyymmdd 
format,
- x.y: the number of WCAG 2.0 guideline, e.g. 1.1, 1.2, ...,
- lz: the WCAG 2.0 level: l1 | l2 | l3,
- scz: the number of the WCAG 2.0 success criterion, e.g. sc1, sc2, ...,
- nnn: the number of the test case (for this success criterion, not the 
number in the complete test suite), e.g. 001, 002, ...

We used this convention for both the TCDL file names, the test file names 
and the IDs used in the TCDL files (ID attribute on root element).
The ID for a XHTML test case for guideline 1.1 level 1 success criterion 1 
may look like this: xhtml1_wcag2_20060427_1.1_l1_sc1_001 (when referring to 
the 27 April 2006 working draft).


Now the current convention:
sca.b.c_lz_nnn, where the sections separated by underscores have the 
following meaning:
- sca.b.c: the number of the success criterion,
- lz: the level of the success criterion,
- nnn: the number of the test case.

We use this convention for both the TCDL file names, the test file names 
and the IDs used in the TCDL files (ID attribute on root element).
The ID for a XHTML test case for success criterion 1.1.1 (level 1) may look 
like this: sc1.1.1_l1_001 (so the ID contains no reference to the date of 
the working draft, but those references are in the rules section anyway).
With WCAG's current numbering system, it is also possible to work without a 
level indicator in the file names, but the level indicator enables quick 
and easy filtering by level.

Also, using the same naming convention for both TCDL and the test file 
names makes it easy to see which test files go with which TCDL file without 
looking into the TCDL file.

Finally, if more than one test file is necessary for the same test case, we 
add an additional _001, _002, etc to the test file names (before the file 
name extension).


I think the second convention will be more popular, but please correct me 
if I am wrong ;-)


Best regards,

Christophe


-- 
Christophe Strobbe
K.U.Leuven - Departement of Electrical Engineering - Research Group on 
Document Architectures
Kasteelpark Arenberg 10 - 3001 Leuven-Heverlee - BELGIUM
tel: +32 16 32 85 51
http://www.docarch.be/ 


Disclaimer: http://www.kuleuven.be/cwis/email_disclaimer.htm

Received on Tuesday, 29 August 2006 13:41:11 UTC