It looks like there is at least one entry in monuments_config which is not getting used in either fill_table_monuments_all.sql (or fill_table_wlpa_all.sql) thus making monuments_config even harder to deal with.
These datasets should be identified and moved out of monuments_config. Ideally we would also have a test which checks both this and (more importantly) that fill_table_monuments_all.sql (or fill_table_wlpa_all.sql) are not looking for a table which is not present in monuments_config.
| Subject | Repo | Branch | Lines +/- | |
|---|---|---|---|---|
| Ensure all monuments_config entries are used | labs/tools/heritage | master | +5 -454 |
Change 292533 had a related patch set uploaded (by Lokal Profil):
[DON'T MERGE] Ensure all monuments_config entries are used
monuments_pl-old_(pl), monuments_de-nrw_(de), monuments-old_ch_(it), monuments-old_ch_(de), monuments-old_ch_(fr), monuments_hk-hb_(en) were the ones I identified as unused.
Unless someone has a good reason for me not to do it I'll remove these from the configuration during next week.
I've already removed a commented out section of update_monuments_all which referred to a different database than heritage.
I just noticed that these unused config have some adverse effects. See https://fr.wikipedia.org/w/index.php?title=Projet:Monuments_historiques/Images_de_monuments_suisse_non_utilis%C3%A9es&action=history : ErfgoedBot updates the page with ch config, and then empties it with the ch-old config.
Ok that is all the ammunition I need. Feel free to merge the patch if it looks good.
Mentioned in SAL [2016-06-10T10:30:17Z] <JeanFred> Deployed latest from Git: d25eda5 (T136704)