From 3617f9b6d9146f9f6955e604b209244ef9c2ee21 Mon Sep 17 00:00:00 2001 From: Philip Sargent Date: Thu, 24 Mar 2022 20:59:36 +0000 Subject: bugfix and docum --- parsers/logbooks.py | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) (limited to 'parsers/logbooks.py') diff --git a/parsers/logbooks.py b/parsers/logbooks.py index 4ebfe0a..9d1a8f0 100644 --- a/parsers/logbooks.py +++ b/parsers/logbooks.py @@ -32,20 +32,19 @@ todo=''' -- far too many uses of Django field dereferencing to get values, which is SLOW - - Loogbooks 1987, 1988, 1989 all crash on MySql - but not sqlite - with db constraint fail. Edit logbook to fix. - import/parse/re-export-as-html the 'artisanal-format' old logbooks so that we keep only a modern HTML05 format. Then we can retiure the old parsers and reduce the volume of code here substantially. -- edit LoadLogbooks() to use coroutines to speed up import substantially, - but perhaps we had better profile it first? - - rewrite to use generators rather than storing everything intermediate in lists - to reduce memory impact. - the object store will need additional functions to replicate the persontrip calculation and storage. For the moment we leave all that to be done in the django db + Concurrent synchronisation would be nice.. + +- DB lock currently prevents multiple threads for loading logbooks. But asyncio might work..? - We should ensure logbook.html is utf-8 and stop this crap: file_in = open(logbookfile,'rb') -- cgit v1.2.3