Re: [jats-list] Can an Editor Write Schematron?

["Gareth Oakes goakes@xxxxxxx" <jats-list-service@xxxxxxxxxxxxxxxxxxxxxx>]

Hi Wendell,

Thanks that is great input and plenty to think over. You made me also think
about an XSLT project we did years ago where we embedded requirement IDs in a
structured way (XSLTDoc from memory?). That really helped with traceability of
functionality to requirements and also gave a neat way to track code and test
coverage.

Most likely webre way off track for this list so Ibll hold up there.

// Gareth Oakes
// VP Content Technologies, GPSL
// www.gpsl.co

From: "Wendell Piez wapiez@xxxxxxxxxxxxxxx"
<jats-list-service@xxxxxxxxxxxxxxxxxxxxxx>
Reply to: "jats-list@xxxxxxxxxxxxxxxxxxxxxx"
<jats-list@xxxxxxxxxxxxxxxxxxxxxx>
Date: Saturday, 13 February 2021 at 04:32
To: "jats-list@xxxxxxxxxxxxxxxxxxxxxx" <jats-list@xxxxxxxxxxxxxxxxxxxxxx>
Subject: Re: [jats-list] Can an Editor Write Schematron?

Gareth,

This is totally doable, but you end up wanting an enhanced Schematron, that
carries code numbers per rule (possibly structured in some way) for binding
points, plus also documentation. There are actually two sets of information
that need to be represented on both sides (Schematron source and spreadsheet
in this case): the Schematron behavior (including the messaging it emits); and
its rationale and "flagging" (such as those controlled IDs). The latter can be
represented in the Schematron in an extension namespace. Then the Schematron
has a fairly rich data set you can externalize in that spreadsheet or in other
ways for the consumption of people who need to see and validate the "intent"
of the logic, without its implementation.

I've heard tell of this approach working well producing HTML reports (for both
rule set management, and validation reports that link to them), not worrying
about spreadsheets.

In the ideal scenario, of course, all this is also aligned with test sets (see
the JATS-Con articles by Sasha Schwartzman and Vince Lizzi, at least), and
subject matter experts can sometimes validate test sets even when they can't
write the XPath. So you get an extra vector: SME confirms that both
rationales, and test examples, are correct; this gives the developer something
external to test the XPath against. When it works, an XPath-literate third
pair of eyes looks at it all together (and makes improvements), eh voila.

In a somewhat similar approach, the documentation/metadata per rule is not
embedded in the Schematron, but rather the Schematron itself is generated by
machine from a language optimized to provide for support of certain kinds of
rules ... another topic.

Cheers, Wendell




I have heard tell of this working well, but with HTML reports rather than
spreadsheets.

On Thu, Feb 11, 2021 at 6:02 PM Gareth Oakes
goakes@xxxxxxx<mailto:goakes@xxxxxxx>
<jats-list-service@xxxxxxxxxxxxxxxxxxxxxx<mailto:jats-list-service@xxxxxxxxxx
errytech.com>> wrote:
Codifying the logic as English statements in a spreadsheet seems like a good
approach. I wonder if you can "sync" this with the Schematron by generating
English language statements from the Schematron asserts/reports? That way once
the programmer interprets the latest spreadsheet they can regenerate the
spreadsheet from the Schematron file, ready for the next time an editor wants
to update the spreadsheet. This gives a handy point of reference for the
current logic as well as a clue to the type of language (and thought process)
the editor should be using.

// Gareth Oakes
// VP Content Technologies, GPSL
// www.gpsl.co<http://www.gpsl.co>

On 12/2/21, 08:55, "Debbie Lapeyre
dalapeyre@xxxxxxxxxxxxxxxx<mailto:dalapeyre@xxxxxxxxxxxxxxxx>"
<jats-list-service@xxxxxxxxxxxxxxxxxxxxxx<mailto:jats-list-service@xxxxxxxxxx
errytech.com>> wrote:

    Flip answer: Yes if they know XPath.

    More useful answer:

    1) They are more likely to RUN Schematron, as set up
    by programmers, than to write it. Whole test suites
    can be set up for them to run single-click.

    2) But, they can also SPECIFY Schematron (in a natural language such as
    English) IN A SPREADSHEET that tells the programmers what to check and
    what error message the Editors want back.

    The editors are in charge of the requirements document (the spreadsheet).
    The programmers are in charge of interpreting those requirements in
    Schematron and of updating the spreadsheet to say which module(s)
    check each requirement.

    Some basic spreadsheet logic of giving each request a unique
    number, error severity (such as editorial error, XML error | human must
    look at this one, caution, fatal, make up whatever categories you
    think are useful), status, QA or management signoff flag, whatever
    you need is useful.

    This is a very handy system that works well for many. Most
    Editors know spreadsheets. There are no pointy brackets in sight.
    It has the additional psychological advantage that the Editors
    feel that the programmers are working for them, finding the
    errors Editors want fixed, not imposing rules from above.

    --dal


    > On Feb 11, 2021, at 5:30 PM, Gareth Oakes
goakes@xxxxxxx<mailto:goakes@xxxxxxx>
<jats-list-service@xxxxxxxxxxxxxxxxxxxxxx<mailto:jats-list-service@xxxxxxxxxx
errytech.com>> wrote:
    >
    > Sorry to jump in but that is an interesting thought. I suspect your
average PE would get stuck on the fact that you need to know XPath and be
quite familiar with the JATS schema in order to create good Schematron
assertions or reports. I wonder if there is a UI solution that would help in
this case though? I'm not aware of one but am interested to hear from others.
It would sure save the trouble of having an XML person come in to define and
maintain those basic editorial rules.
    >
    > I think the pre-requisites for Schematron are experience with XML, good
knowledge of XPath, and working knowledge of the schema you are writing rules
against.
    >
    > // Gareth Oakes
    > // VP Content Technologies, GPSL
    > // www.gpsl.co<http://www.gpsl.co>
    >
    > On 12/2/21, 06:44, "Charles O'Connor
coconnor@xxxxxxxxxxxx<mailto:coconnor@xxxxxxxxxxxx>"
<jats-list-service@xxxxxxxxxxxxxxxxxxxxxx<mailto:jats-list-service@xxxxxxxxxx
errytech.com>> wrote:
    >
    >    Hi Liam,
    >
    >
    >
    >    When the Schematron course comes together, please ping me off list.
    >
    >
    >
    >    I did a webinar yesterday on our (Aries) XML-through production
workflow, which includes the ability to configure a task to run Schematron.
Question from a participant was, "This Schematron thing sounds cool, but is it
something the average production editor could write?"
    >
    >
    >
    >    My answer was that a somewhat ambitious one certainly could. There
may also be interest among others at Aries.
    >
    >
    >
    >    Also, let me know whether there are prerequisites beyond a basic
knowledge of what XML is and how it works.
    >
    >
    >
    >    Thanks,
    >
    >    Charles
    >
    >
    >
    >    -----Original Message-----
    >
    >    From: Liam R. E. Quin
liam@xxxxxxxxxxxxxxxx<mailto:liam@xxxxxxxxxxxxxxxx>
<jats-list-service@xxxxxxxxxxxxxxxxxxxxxx<mailto:jats-list-service@xxxxxxxxxx
errytech.com>>
    >
    >    Sent: Wednesday, February 10, 2021 2:55 PM
    >
    >    To: jats-list
<jats-list@xxxxxxxxxxxxxxxxxxxxxx<mailto:jats-list@xxxxxxxxxxxxxxxxxxxxxx>>
    >
    >    Subject: [jats-list] [ANN] XSLT 3 training - dates for February,
March, April 2021
    >
    >
    >
    >    *** External email: use caution ***
    >
    >
    >
    >
    >
    >
    >
    >    I'll be running XSLT 3 training on
    >
    >    * 23rd, 24th, 25th of February,
    >
    >      (likely in CET times, e.g. 09:30-15:00 depending on numbers of
    >
    >    people)
    >
    >
    >
    >    * 23rd, 24th, 25t of March,
    >
    >    * 27th, 28th, 28th of April
    >
    >
    >
    >    The courses are live video (not pre-recorded); classes are limited to
eight people at a time. Shoes optional :)
    >
    >
    >
    >    There's a course overview at
    >
    >    https://www.delightfulcomputing.com/
    >
    >
    >
    >    In addition, i know you've been missing XML Prague, so a Schematron
course is in the works.
    >
    >
    >
    >    The single most important thing to me in teaching is empowerment:
    >
    >    helping people to see how to find their own answers, and to go out
into the world and solve new problems.
    >
    >
    >
    >    Liam
    >
    >
    >
    >    PS: there's some flexibility in the dates, and the time of day can be
adjusted to meet your needs, depending on who is participating.
    >
    >
    >
    >
    >
    >    --
    >
    >    Liam Quin, https://www.delightfulcomputing.com/
    >
    >    Available for XML/Document/Information Architecture/XSLT/
XSL/XQuery/Web/Text Processing/A11Y training, work & consulting.
    >
    >    Barefoot Web-slave, antique illustrations:
http://www.fromoldbooks.org
    >
    >
    >
    >
    >
    >
    >
    >

    ================================================================
    Deborah A Lapeyre
mailto:dalapeyre@xxxxxxxxxxxxxxxx<mailto:dalapeyre@xxxxxxxxxxxxxxxx>
    Mulberry Technologies, Inc.      http://www.mulberrytech.com
    17 West Jefferson Street         Phone: 301-315-9631 (USA)
    Suite 207                        Fax:   301-315-8385
    Rockville, MD 20850
    ----------------------------------------------------------------
    Mulberry Technologies: Consultancy for XML, XSLT, and Schematron
    ================================================================




--
...Wendell Piez... ...wendell -at- nist -dot- gov...
...wendellpiez.com... ...pellucidliterature.org... ...pausepress.org...
...github.com/wendellpiez.<http://github.com/wendellpiez.>..
...gitlab.coko.foundation/wendell...
JATS-List info and archive<http://www.mulberrytech.com/JATS/JATS-List/>
EasyUnsubscribe<http://lists.mulberrytech.com/unsub/jats-list/2708257> (by
email<>)