As a further response, regarding that Martina said "insert ... new row and get
back the id of this last inserted row", I believe that this likely reflects a
poor design. Wherever possible, every field of a row to insert including its
identifiers should be known BEFORE inserting the row, and that way, you don't
need to ask afterwards what the id is, you already know because it is the value
you told it to use when doing the insert. -- Darren Duncan
On 2016-06-27 11:58 PM, Hick Gunter wrote:
Do not use SQLite for concurrent access over a network connection. Locking
semantics are broken for most network filesystems, so you will have corruption
issues that are no fault of SQLite. If your application requires concurrent
network access, you should be using either a network client/server DBMS or
programming your own Client/Server pair with the server process calling the
SQLite API to modify the db file held on the servers' local filesystem.
SQLite uses file level locking as documented on the SQLite main page.
A cursor (CURrent Set Of Records) is a concept designed for reading data, not writing. SQLite does not
support navigation other than retrieving the result set in the order specified in the select statement. No
"paging" backwards. It is possible to simulate a cursor, but the "simple" solutions may
be imperformant for large result sets (e.g. producing half the result set to display a small number of
records on a "middle page").
I assume you are looking for "transactions". SQLite supports both explicit transactions
(BEGIN ... COMMIT/ROLLBACK) and "automatic" Transactions (each statement is ist own
transaction). If you are just inserting/modfying a set of rows (this also includes a single row)
within a single table, automatic transactions are sufficient. If you require multiple changes to
multiple rows in one or more tables to complete together or not at all, then you need an explicit
transaction.
Gunter
-----Ursprüngliche Nachricht-----
Von: sqlite-users-boun...@mailinglists.sqlite.org
[mailto:sqlite-users-boun...@mailinglists.sqlite.org] Im Auftrag von Martina
Pasini
Gesendet: Montag, 27. Juni 2016 20:51
An: sqlite-users@mailinglists.sqlite.org
Betreff: [sqlite] How to use "cursors" in c#
Hello to Everyone!
I am sorry to bother you, but I am new to DB and SQLite and I have a doubt.
I am programming in c# and I have to insert into a .sqlite file a new row and get back
the id of this last inserted row. This file "unfortunately"
might be located on an internal network and other users might try to add other
rows in the meantime.
I have not studied yet how to manage locking etc (will I need it only when
editing a row or also when adding a new one?)
For this case, when I add a new row, could it be sufficient to use a cursor or
something similar? Could you give me some advice/links I should read to
understand how to manage this process?
Thank you very much for you kind attention!
Best Regards,
Martina
_______________________________________________
sqlite-users mailing list
sqlite-users@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
