revamp whole EA flow. Server created EAs when we do certain jobs (transform, delete_files, restore_files), then instead of faking amendments in the jscript, get job creation to return EA from job ORM, then check is now generic for end of any amendment job, and when it finishes, use that to clear our amendments in document, and redraw through normal UI code. No smarts in client, all driven by state from server, and if we reload a page mid jobs, it has required state, and because an amendment job is still progressing, it runs check code again

This commit is contained in:
2025-10-25 18:21:28 +11:00
parent d3ae9b788f
commit 9bdd9d5b78
6 changed files with 236 additions and 171 deletions

9
TODO
View File

@@ -1,12 +1,3 @@
* new viewing model (get ids of query on first load, then paginate only inside that known list)
- BUT, when we finish a delete, what do I do with pageList / entryList???
- start by showing them as deleted (via amend)
- then on success, remove the ids from the *List arrays in js -- but do
this via repagination, invalidate page cache fully, then getPage(currentPage)
(e.g. assume 1, 2, 3 ... 40 in eList). delete 23, 24,
then reset lists to remove 23 and 24, pageList would then
get reset to page with: 21,22,25,26 ... 30, 31
? get rid of style and just use class -- think this should work, so change in
templates/files.html for throbber, etc. and dont set style as much in view_support.js