1. Rob: database.e update for v2.5
- Posted by Greg Haberek <ghaberek at gmail.com> Oct 01, 2004
- 529 views
I don't like database.e crashing my Win32Lib apps. Its very messy, and I'm sure quite alarming for a user who doesn't expect a crash and a DOS box to popup. My biggest issue was with db_rename_table() crashing when a table didn't exist. I've updated it and turned it into a function. Please consider this for Euphoria Database v2.5. I'd like if database.e returns an error for all issues, rather than crashing with fatal().
global function db_rename_table(sequence name, sequence new_name) -- rename an existing table -- written by Jordah Ferguson -- updated by Greg Haberek atom table,table_ptr table = table_find(name) if table = -1 then return DB_OPEN_FAIL end if if table_find(new_name) != -1 then return DB_EXISTS_ALREADY end if safe_seek(table) db_free(get4()) table_ptr = db_allocate(length(new_name)+1) putn(new_name & 0) safe_seek(table) put4(table_ptr) return DB_OK end function
2. Re: Rob: database.e update for v2.5
- Posted by Hayden McKay <hmck1 at dodo.com.au> Oct 01, 2004
- 488 views
I somewhat agree with your idea to have database.e commands return a booleen value from a function. But I must point out that if your program is crashing because db_rename_table() tries to rename a table that does'nt exist then the problem is with the programes code, not database.e
3. Re: Rob: database.e update for v2.5
- Posted by Derek Parnell <ddparnell at bigpond.com> Oct 01, 2004
- 550 views
Hayden McKay wrote: > > I somewhat agree with your idea to have database.e commands return a booleen > value from a function. But I must point out that if your program is crashing > because db_rename_table() tries to rename a table that does'nt exist then > the problem is with the programes code, not database.e Exactly! That's why the program should handle the error and not database.e -- Derek Parnell Melbourne, Australia