Meeting minutes
<john> Steven around yet?
Previous Actions
Action: Tom: to raise ticket re: space definitions in Ixml grammar
<trackbot> Sorry, but no Tracker is associated with this channel.
Steven: only action remaining from last meeting is mine re: spaces
review of conformance section and conformance language
<Steven> https://
review of conformance section and conformance language
<Steven> https://
<Steven> That's the updated version. Please check
Action: group: validate new conformance changes at URL above
<trackbot> Sorry, but no Tracker is associated with this channel.
Grammar allows element content in attributes
<Steven> https://
<Steven> Steven: I have corrected this
grammar allows elements in attributes
(see link above)
Steven: I've fixed this issue.
Michael: I can confirm!
<cmsmcq> Issues at https://
github issues
Steven: I hadn't seen these before
Michael: Looks like issue #1 (s/may not/must not/) is now resolved.
Michael: on issue#2 the problem is that the whitespace is declared at the end of a rule definition, rather than inbetween them.
Michael: this makes the XML less than graceful: the comments get parsed within rule elements, not inbetween them
Steven: good suggestion
Action: Steven: to implement Michael's suggestion
<trackbot> Sorry, but no Tracker is associated with this channel.
<Steven> For rules separated by spaces rather than terminated by them
<Steven> https://
Positive Test Case Coverage
https://
Michael: The 'dogfood' test for converting the IXML grammar to XML covers 2/3 of the cases
Steven: we could do with this list!
Action: Michael: Transcribe list of cases not covered by dogfood test, and add to github
<trackbot> Sorry, but no Tracker is associated with this channel.
<cmsmcq> test case catalog straw proposal: https://
Michael: I have this test suite, but I don't know how to use github to create pages for them.
Michael: You can check this out and see the basic HTML, such as https://
Michael: inputs and outputs can be either embedded in the test catalogue, or linked to in an external file.
Michael: can we use this as a group test specification structure?
Michael: it would be good to have other users of the test cases, along with any feedback for what it can't do
Tom: I will value this in the future
Tom: I can also help automate publishing as web pages
Action: Tom: automate Michael's test suite publishing
<trackbot> Sorry, but no Tracker is associated with this channel.
Michael: I am interested in having a non-xml syntax
John: that is interesting - lots of quoting problems?
Also John: I am too busy on the railway :)
Steven: I foresee a paper in converting XML to non-XML representations
Tom: absolutely
John: this will be a good use case
All: let's revisit when we have more working implementations
<Steven> s/there's a apaper/I forsee a paper/
<cmsmcq> s/forsee/foresee/
<Tom> s/there's a apaper/I foresee a paper/
<Tom> s/apaper/paper
Encouraging implementation
(end of https://
Michael: reporting on status of implementation in meetings would be a good way of encouraging progress on papers
Michael: can we go around the table now?
Steven: OK, do you want to start
Michael: Aparecium (XQuery implementation) works according to (my interpretation of) older versions of the grammar; In order to support refactoring, I have been working through the code adding regression tests, which is currently in progress.
Michael: I have also been writing a set of tools for working with grammars using XSLT.
Tom: I have a minimally working XSLT parser
Tom: time is a little limited: I know that there are improvements to be made on performance and stability, some of which is planned, some of which is not.
Aleksei: I've started on a SAXON based XML reader; works with streams. Should be available in .NET
Aleksei: I've started with the XML version of the grammar.
John: just to note I ran Tomos' JayParser, and managed to get it working for simple grammars. Happy to act as an independent tester!
Michael: A question for Aleksei: Is all I would need to run your implementation is to install mono?
Aleksei: .net 5 or 6 are available for linux/mac os x - installing the core is enough to get it working.
Steven: I have my ABC version; the next task is to convert it to C (which may be easier said than done!)
John: how large is your abc implementation out of interest?
Steven: around 1800 lines (666 of abc)
Implicit String Generation
https://
Steven: I think we decided on this last week - to introduce the '+' mark
Action: Steven: add '+' to grammar
<trackbot> Sorry, but no Tracker is associated with this channel.
next meeting
Tom: I'm on holiday on the 10th
Steven: How about the 17th?
All: agree to meet on 17th August.
<Steven> rrswagent, make minutes