On Fri, 19 Oct 2018 at 12:16, Shane Curcuru <[email protected]> wrote: > > sebb wrote on 10/19/18 5:51 AM: > > I think it would be useful for the Secretary workbench to be able to > > reparse a message if necessary. > > > > For example, there is a message (201810/6d84281261/) which has two > > attachments, but one of them is not shown in the YAML summary. When I > > tried parsing the message on my system, it correctly detected them > > both. Either one of the attachments was dropped from the YAML file, or > > it was never added in the first place. It would be useful to know > > which is the case. > > > > I think it would just be a question of re-running the deliver.rb > > script with the appropriate file. > > > > WDYT? > > Yes, *as long as* it's bulletproof and crystal-clear to a user when this > is happening, to prevent unintentionally either adding another record or > overwriting an existing record that you didn't mean to. Not sure if > that's debug / only show what would happen / don't actually write data > flags or what. > > Maybe my concern isn't important, but I want to make sure that in the > future a new Secretary helper couldn't mistakenly use this to re-process > all of last month's emails or something, which could make quite a mess > if the script mistakenly re-updated a bunch of data.
I was thinking of a script to reprocess a single email by providing its month and hash. The user would have to know these. Ideally the script would run the existing deliver script; however it does not return the hash or the new YAML entry so it would not be easy to detect what changed. One way round this might be to replicate the code in a new script which does not automatically perform the update unless certain conditions are met, e.g.. no change to hash or month and no change to the mail source copy - i.e. only update the YAML hash. However this means keeping two sources in step, so perhaps the deliver script could take a parameter which makes it perform the checks.> -- > > - Shane > Whimsy PMC > The Apache Software Foundation
