RE: Update to test suite

Hi Leroy,

 

I think that the best way is how you are currently doing it.

Just like in locqualityissue6html.html of the Test Suite that corresponds to
example 83 of the draft.

/html/body[1]/p[1]/span[1]     locQualityIssueComment="should be thomas
cahill."        locQualityIssueProfileRef="http://example.org/qamovel/v1"
locQualityIssueSeverity="100"  locQualityIssueType="inconsistentEntities"

/html/body[1]/p[1]/span[1]/@data-mytool-qacode

/html/body[1]/p[1]/span[1]/@its-loc-quality-issue-comment

/html/body[1]/p[1]/span[1]/@its-loc-quality-issue-profile-ref

/html/body[1]/p[1]/span[1]/@its-loc-quality-issue-severity

/html/body[1]/p[1]/span[1]/@its-loc-quality-issue-type

 

With the attributes in alphabetical order, for me at least it's very clear.

 

By the way locqualityissue6htmloutput.txt I miss the rest of attributes of
/html/body[1]/p[1]/span[2]
I can only see /html/body[1]/p[1]/span[2]/@its-loc-quality-issue-severity

 

Finally sorry if I miss the confirmation (I don't find any email), but I'm
not sure if you added my user to the github so I can make commits. I would
like to help you correct these files and above all the provenance examples,
since It's a data category I'm interested in, and as I saw in the minutes
from today and also in the draft is moving forward.

 

Cheers,

 

Everyone,

 

I have added in loc quality and precis output and corrected the input files
as well as adding in its:param tests for this data category . One key
question i want to ask the  group is how to handle or best represent  inline
markup in the output files for the loc quality issue???? (See example 83
http://www.w3.org/TR/its20/#lqissue).

 

Thanks,

Leroy

On 20 November 2012 12:36, Leroy Finn <finnle@tcd.ie> wrote:

Fredrik, Everyone,

 

I have updated all the output files to reflect the error found by Fredrik
where its: was removed from the path section of the output files. Thanks for
noticing that Fredrik. The updated output is now on github more updates to
follow today.

 

Thanks,

Leroy

 

On 20 November 2012 10:24, Leroy Finn <finnle@tcd.ie> wrote:

Hey Fredrik,

 

Sorry i didn't notice that. It was a accident i didn't mean to remove those.
I will correct this now and repost the output ASAP.

 

Thanks for spotting this,

Leroy

 

On 20 November 2012 10:18, Fredrik Liden <fliden@enlaso.com> wrote:

Hi Leroy,

 

Thanks for the updates and working on all the test files. The new output
makes it a lot easier to compare the files.

 

I updated a few of the "Translate" files and posted them to a new okapi
branch off of your Master branch. You can compare the branches to
review/approve the changes.

wasn't sure if you prefer us to fork it instead to keep it even more
separate so let us know if this works for you and we'll continue with the
other categories.

 

 

Btw, I see the its: prefix has been removed from all its:translate="no" to
translate="no" which was agreed on.

However, it seems the its: namespaces from the left side data have been
removed as well (see @its:version and its:rules below). Since those are
representation of real xml entities it would make sense to keep those
namespaces intact. 1 Because implementations are most likely already
keeping track of namespaces in the parsed tree. 2. It would be possible to
have two elements/attributes conflicting if they have the same name but
different namespace. 3. To be consistent we might as well either remove all
namespaces or keep all of them.

 

Each alternative, removing only the its: namespaces, removing all
namespaces, or keep all namespaces may render a different line sort order so
it a fairly significant decision. 

Again this is only for the list of elements and attributes on the left hand
side. Not the ITS information then printed for each one. Thoughts, anyone?

 

Sample Input file

  <its:rules xmlns:its="http://www.w3.org/2005/11/its"

             xmlns:xlink="http://www.w3.org/1999/xlink"

             version="2.0"

             xlink:type="simple"

             xlink:href="translate10xmlrules.xml" />

 

Current okapi:

/doc/@its:version        translate="no"   (this line is just an example from
another file)

/doc/its:rules[1]               translate="no"

/doc/its:rules[1]/@version          translate="no"

/doc/its:rules[1]/@xlink:href     translate="no"

/doc/its:rules[1]/@xlink:type     translate="no"

 

Latest testsuite:

/doc/@version translate="no" (this line is just an example from another
file)

/doc/rules[1]     translate="no"

/doc/rules[1]/@version                translate="no"

/doc/rules[1]/@xlink:href           translate="no"

/doc/rules[1]/@xlink:type           translate="no"

 

Cheers,

Fredrik

 

From: Leroy Finn [mailto:finnle@tcd.ie] 
Sent: Monday, November 19, 2012 9:29 AM
To: Multilingual Web LT-TESTS Public
Subject: Update to test suite

 

Everyone,

 

I have added a major update to the test suite and further updates will come
tomorrow as well. The update today is as follows "corrected data category
files and added in new output format for all data categories except for
ruby,directionality,loc quality and precis and disambiguation". The missing
categories will be updated tomorrow. I will send out a mail in the morning
with more details of the changes that i have made which will also be visible
on the git hub page tomorrow.

 

Thanks,

Leroy