Closed Gertrudius closed 1 year ago
Side thing, I corrected the text markup entries for the initial šu-ma on line 73 from omitted to superfluous and deleted its text_discourse entry.
Also, part of the issue here is that the discourse number "02eacbec-21a5-4388-b3e1-2a8622844c8e" should only be 15, but it is also erroneously assigned 0.6666 from line 80. This should be its own number entry.
SELECT * FROM text_epigraphy WHERE discourse_uuid = "02eacbec-21a5-4388-b3e1-2a8622844c8e";
EDIT: this is now fixed, and it was the reason that one of the "Debt record" paragraphs was pulled out of position.
I've gone ahead and reordered the obj_in_text
now that the issue with that number is fixed, and things look like they are correct now.
I noticed when building my PHP script for issue #7, that the logic will position many paragraphs with phrases and sentences as their parents, seemingly just in this text. Some of these could be addressed by issue #42, but several of those I looked at do resolve to a word or number whose uuid will connect to a
text_epigraphy
.object_on_tablet
value through itsdiscourse_uuid
. This would suggest that either itstext_discourse
.obj_in_text
value is incorrect, or maybe the retained child/grandchild etc. that is a number or word was not the first or last one under that paragraph that fell out of the OCHRE export, and so it places it in the middle of what should be its children. If the latter issue is the problem, and confined to this text, it might be easiest to address this manually, which will allow us to retain the given paragraphs.