On 27/11/2007, Michael Bayer <[EMAIL PROTECTED]> wrote: > > > On Nov 27, 2007, at 11:30 AM, Yuri Pimenov wrote: > > >> Quote it. > > oh right, duh, let me answer post-coffee next time.... > > SQLite is the reason for the truncation on the dot, and I got both > test cases (i.e. using sqlite and you dont expect the dot, or using > dot-separated labels and you do expect the dot) to be supported > including with SQLite in r3836. My main concern, which is slowing > down of ResultProxy, doesn't seem to be an issue here...if the dot > isnt present (the majority of cases), the truncation doesnt occur now > so that saves time in itself. > > > >> > > I've found 'use_labels' parameter in select(). Everything is good but > > it uses _ as separator between table and column. That makes things > > not so easy because it is common practice to use _ in table and column > > names. > > Why not to use . as separator? > > an underscore is pretty portable and doesnt require quoting, making it > easier to read and type. also underscore qualified labels are > definitely the most common practice i have seen, using ORM like > Hibernate and others... I have never seen a quoted "tablename.colname" > label being used before (hence it tripped me up). > > Also due to the sqlite issue, we necessarily have to perform a > truncation step in the result if a dot is present in the colname, so > if anything we gain speed by not standardizing that syntax.
You are right. I just mixed up labeling with table qualified column names. My actual problem is that i use crossproduct select queries and i want to distinguish between columns of different tables. > > > > > > > My other blue sky dream is about ORM and unicode column names > > (python3.0?). > > we support unicode column names and table names as well as the > databases themselves do....just go browse through the unit tests in > test/sql/unicode.py (there are others elsewhere). for ORM you need > to redefine the mapped properties using ascii names (i.e. in the > 'properties' dict) so you can access them. Python3.0 will use unicode identifiers. So there will be no need to do ascii -> non ascii mapping. -- Yuri Pimenov --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "sqlalchemy" group. To post to this group, send email to sqlalchemy@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sqlalchemy?hl=en -~----------~----~----~----~------~----~------~--~---
