Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.python > #95260 > unrolled thread

Re: Data integrity problem with sqlite3 - solved

Back to article view | Back to comp.lang.python

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: Data integrity problem with sqlite3 - solved "Frank Millman" <frank@chagford.com> - 2015-08-11 15:17 +0200

#95260 — Re: Data integrity problem with sqlite3 - solved

"Frank Millman"  wrote in message news:mqcmie$po9$1@ger.gmane.org...

> Hi all
>
> I have a 'data integrity' problem with sqlite3 that I have been battling 
> with for a while. I have not got to the bottom of it yet but I do have 
> some useful info, so I thought I would post it here in the hope that 
> someone with some knowledge of the internals of the python sqlite3 module 
> can throw some light on it.

Oops, I have just spotted my mistake.

There are times when I want to issue a SELECT statement with a lock, as it 
will be followed by an UPDATE and I do not want anything to change in 
between.

MS SQL Server allows you to add 'WITH (UPDLOCK)' to a SELECT statement, 
PostgreSQL allows you to add 'FOR UPDATE'.

I could not find an equivalent for sqlite3, but in my wisdom (this was some 
time ago) I decided that issuing a 'BEGIN IMMEDIATE' would do the trick.

I had not anticipated that this would generate an implied COMMIT first, but 
it makes sense, and this is what has bitten me. Now I must try to figure out 
a better solution.

Apologies for any wasted time.

Frank

[toc] | [standalone]

Back to top | Article view | comp.lang.python

csiph-web