KVP and data that contains a forward slash

Derek Atkins warlord at MIT.EDU
Tue Feb 9 11:57:08 EST 2016

John Ralls <jralls at ceridwen.us> writes:

>> On Feb 7, 2016, at 4:03 AM, Robert Fewell <14ubobit at gmail.com> wrote:
>> Hi,
>> While looking at the Bayesian kvp entries and looking to convert the way
>> they are stored based on the guid as opposed to the full account path I
>> have come across a difference between 2.6.11 and master.
>> What I have done is create two basic new files, one from 2.6.11 and one
>> from master with the account separator of / and used the CSV transaction
>> importer to import the enclosed file once with Bayesian turned on and once
>> with it of. If you look at the resulting kvp trees, in 2.6.11 you can have
>> a / in a slot but master treats it as an other level, you can see this more
>> easily in the storage of the full account path. You will also notice with
>> Bayesian turned off, the entries stored under "desc" appear to have over
>> written each other so you have only the last entry.
>> I can get around this by replacing any / with a \ for the saved Bayesian
>> token or description and the full account path would be a guid anyway.
>> So the question is should the new kvp procedures allow a / in a slot or kvp
>> path.
> Thanks for catching that. We absolutely don't want to create deep
> keys, performance will suffer in SQL. Please file a bug so that I
> don't forget.

Don't we still have a single function that parses a full key-path from a
single string?  How would it know how to part something like:

  /foo/bar/Assets/Bank  into  [foo] [bar] [Assets/Bank] ??

> Regards,
> John Ralls

       Derek Atkins, SB '93 MIT EE, SM '95 MIT Media Laboratory
       Member, MIT Student Information Processing Board  (SIPB)
       URL: http://web.mit.edu/warlord/    PP-ASEL-IA     N1NWH
       warlord at MIT.EDU                        PGP key available

More information about the gnucash-devel mailing list