On Tue, Jun 28, 2011 at 11:09:52AM -0500, dennis jenkins wrote:
> Any suggestions on how to name tables when table names contain both
> multi-word nouns and mutli-table "many-many" mappings?
>
> Example: Suppose that I have a table called "foo" and another table called
> "barBiz" (or "bar_biz" if you prefer). Further, both of these tables have a
> "serial" primary key. Now I want to create a third table that represents a
> many-to-many relationship between "foo" and "barBiz". So far I have been
> keeping compound-noun table names in camel case, but mapping tables separate
> the base table names with underscores. Thus the table name would be
> "foo_barBiz".
>
> However, I find the above distasteful, for many of the reasons that Merlin
> and others have outlined. Yet naming the table "foo_bar_biz" seems
> ambiguous to me, as does using just lower-case "foo_barbiz" / "barbiz".
> These examples are contrived. The real table names are normal English words
> with subjective meaning.
>
> I'd like in ask the pgsql community for suggestions on how they name tables.
GNUmed uses
lnk_foo2bar_biz
or
lnk_bar_biz2foo
depending on what makes more sense.
Karsten
--
GPG key ID E4071346 @ gpg-keyserver.de
E167 67FD A291 2BEA 73BD 4537 78B9 A9F9 E407 1346