Hi!
One year ago Michael Widenius posted this:
....
We have on the other hand played with the idea to at some point
implement SQL commands of type:
ISAM INSERT table_name (...)
ISAM UPDATE table_name SET ... WHERE key_name=(key_part1,key_part2,...);
ISAM READ table_name FIRST
ISAM READ table_name NEXT
ISAM READ table_name NEXT LIMIT 10 #This would read the next 10 rows
ISAM READ table_name PREV
ISAM READ table_name LAST
ISAM READ table_name WHERE key_name=(key_part1,key_part2,...);
Most of the above would be trivial to parse but could give you direct
access to the lover ISAM/MyISAM levels with fill cacheing.
....
Is this still only an idea? I'm asking because I want to migrate a lot of
applications from a mainframe to Unix which all use ISAM. I could simply use
native ISAM on Unix too of course but I want an SQL-interface too. Since its
not possible to access ISAM tables directly while the same tables are under
control of mysqld I have to use the SQL-parser.
Besides the performance issue there is another problem with SQL:
The ISAM-API has a different meaning of a cursor than in SQL. I can set this
cursor to the first or last row or to any row by supplying a primary key.
Then I can read forward and backward row by row and this I can't simply
emulate
with SQL. Or at least this is very difficult and slow to accomplish with
SQL.
Summary: Is there a lower level API for accessing ISAM tables (BDB, ISAM,
MyISAM)
which is usable in parallel to or in combination with the
SQL-parser?
Bye,
Volker
---------------------------------------------------------------------
Before posting, please check:
http://www.mysql.com/manual.php (the manual)
http://lists.mysql.com/ (the list archive)
To request this thread, e-mail <[EMAIL PROTECTED]>
To unsubscribe, e-mail <[EMAIL PROTECTED]>
Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php
