Hello, Bruce. Thanks for an answer. Are there any thoughts to insert check into PostgreSQL code or we (developers) must take into account that UDT with leading underscore may exist?
You wrote: BM> I have applied the following documentaiton patch that states a leading BM> underscore "shouldn't" be used, rather than prohibited. BM> --------------------------------------------------------------------------- BM> Pavel Golub wrote: >> >> The following bug has been logged online: >> >> Bug reference: 2852 >> Logged by: Pavel Golub >> Email address: [EMAIL PROTECTED] >> PostgreSQL version: 8.x >> Operating system: Windows XP >> Description: User-defined type name begins with the underscore >> character (_) can be created >> Details: >> >> Hello, pgsql-bugs. >> >> Documentation says: >> "User-defined type names cannot begin with the >> underscore character (_) and can only be 62 >> characters long (or in general NAMEDATALEN - 2, >> rather than the NAMEDATALEN - 1 characters >> allowed for other names). Type names beginning >> with underscore are reserved for >> internally-created array type names. " >> >> However, such SQL may be executed: >> >> CREATE TYPE _my AS (id int4, id2 int4); >> >> And then server treats it as an array type. Thus next SQL will be executed >> too: >> >> CREATE TABLE my_table( >> my_arr my[] >> ); >> >> Checked on PostgreSQL versions (Windows XP): >> 8.0.6 >> 8.1.0 >> 8.2.0 >> >> ---------------------------(end of broadcast)--------------------------- >> TIP 4: Have you searched our list archives? >> >> http://archives.postgresql.org -- With best wishes, Pavel mailto:[EMAIL PROTECTED] ---------------------------(end of broadcast)--------------------------- TIP 2: Don't 'kill -9' the postmaster
