LEMDO 2025-05-12 to 2025-05-16
to : Martin Holmes
Minutes: 1025
Over the weekend, clearing space on Jenkins, I cleared the workspace of the
LEMDO job, and then the build broke because it was downloading fresh materials
from the TEI (Stylesheets, schemas etc.). Diagnosing on Monday, the break turned out to be caused by
the new TEI <constraintDecl>
element, which is being added
automatically into ODD files by the Stylesheets, but which wasn’t part of our
schema. Adding it to our ODD fixed that. Then I received a set of instructions
from JJ on changes to the PDF formatting, and turned those into a GH issue with
a collection of sub-issues, partly to test out the new sub-issue feature. Then
started working through the issues, and got three cleared by the end of the day.
Tuesday I continued working on the set of issues, now up to seventeen, and got all but a handful cleared.
On Wednesday, I wrestled the remaining issues into submission, so we now have working PDFs as far as I know, with only one related ticket on how to encode and render bylines, which needs JJ’s and NH’s input.
On Thursday, I re-engaged with issues #235 and #327: these are requesting the movement of annotation and collation markers so that they come after any closing punctuation rather than before it. This is a very tricky problem, and after determining that the bug in issue #327 is not fixable using the approach piloted in response to the original request in issue #235, I started from scratch with a different approach. Instead of trying to move the punctuation, I decided to try to move the markers instead. This works better, although the XPath that walks the preceding:: axis is necessarily slow.
On Friday, with SS’s help confirmed that this is all working now, and then cleared a couple more tickets; the PDF generation is all complete as far as I’m concerned, as is the DRE post-processing for punctuation position. Debugged a couple of build breaks too.