Sqlite3 for backends

[Andrin von Rechenberg]

Hi there

Is there any plan to support the sqlite3 module in python in GAE?

It would be very useful if someone would want to build an SQL like backend.

-Andrin

[Ernesto Oltra]

Hi Andrin,

The remote system is not writable, in case googlers add the module to python you could read the Sqlite3 files, but you couldn't write new data to them.

Ernesto

[Bart Thate]

Oi !

Agree completely, in fact really forgot about it ;]

I thought it was on the roadmap, not sure though .. maybe 2.7 supports sqlite as the python module is already builtin ?

Hmm just checked the roadmap and only 2.7 support is on it, lets hope they dont exclude "import sqlite3" from the deal.

Bart

programming schizofrenic -  http://tinyurl.com/bart-thate

--
You received this message because you are subscribed to the Google Groups "Google App Engine" group.
To post to this group, send email to google-a...@googlegroups.com.
To unsubscribe from this group, send email to google-appengi...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.

[Andrin von Rechenberg]

The idea would be to use sqlite in memory ( sqlite.connect(":memory:") ) so it would be ultra fast.

That's why I was talking about using it in backends... :)

-Andrin

[Ernesto Oltra]

I've heard another legal use here in this group: Read a complex database that changes infrequently but can benefit from JOINs.

Also sqlite3 is in C, so it's an extra effort to clean all this write-to-disc funcs in the code and test it. Post an issue and I'll start it, but it is a lot of work and probably won'y be short term.

[Gerald Tan]

You could use orientdb which is a pure java nosql db that can use a memory store, but how would the frontend instance communicate synchronously with the sqlite backend instance though?

[Ronoaldo José de Lana Pereira]

I guess that a B4 backend wold be a nice use case for this, right? You may implement an "in-memory database backend" that can be synced with the datastore. You can let this backend private and then implement a query handler in the backend that returns the data in a specific format (json, yaml, xml) that you can consume with urlfetch in the frontend.

I'm not sure how you will project this to scale as the data grows, but this may be a good choice for a moderate dataset.

- Ronoaldo