Re: [binarycloud-dev] QueryManager / Standards

[Alex B. <en...@tu...>]

>>> Wouldn't it be better just to define column and table names as
>>> case-insensitive and always uppercase/lowercase them in
>>> QueryManager before doing a MetaBase query?
> 
>> 2)
>> I don't like it because I think it's a hell of a lof more elegant to have
>> 'name sync' between entities, entity paths, and sql.
> 
> I'm with you. If we establish a standard like proposed, we must keep it
> throughout the system. It makes live lot easier.

Agreed.

I don't want to put in any code that necessarily enforces the standard, but
I want to document and adhere to it.

>>> be nice. E.g. "FurbeeId" is more readable than "furbeeid" or "FURBEEID".
>> The above is opinion, not doctrine... does anyone else have an opinion on
>> this one?
> Personally I like the furbeeId, userId, furbeeName syntax.  But with the
> compatibility issues and the complexity of the system in mind  it might be
> really wired in large applications. Its easier to track the path from a
> entity to the database if everything is in sync.

actually, I agree: I prefer MooId or mooId, etc. But I think if we have to
go with 'single case' the best is lowercase field_name

> Additionally I prefer do add the strtolower() thing in buffer creation ( as
> originally done with strotupper() ). So users have the freedom to name the
> database fields mixed case (FurbeeId) but the system gets them as all
> lowercase (furbeeid). But for the repository modules it should stay as
> proposed.

hmm. that is enforcing policy, but I think I'm willing to break my statement
above and say "yeah" :)

> Note to the query syntax: As I said before I like to write queris with SQL
> statements UPPERCASE. I'd love to add this to the standard 'caus it make the
> queries MUCH more readable.

yes, ditto.

> So from my point of view I think the following is very acceptable:
> 
> For modules to be distributed with binarycloud or available from the
> (upcoming) module repository, as well as for binarycloud core components
> you must:
> - Use field and table/entity names with less than 32 chars
> - Use lowercase for all database, field and table/entity names
> - Use CAPS for SQL command (SELECT, AS, FROM, DELETE...)
> - Verify that your SQL & Schemas work with Oracle, MySQL, and PostgreSQL.

deal.

anyone else have comments/etc?

I'm preparing a significant revision on the binarycloud coding standards...
this will make it in.

best,

_alex