Add support for lazy attributes via a decorator

[nthiery]

The lazy-attribute-...-nt patch on the sage-combinat patch servers is a first implementation of lazy attributes (see the doc in the patch).
Comments and alternative implementations welcome! See the current caveats in the doc

Cheers,
Nicolas

Component: misc

Issue created by migration from https://trac.sagemath.org/ticket/4371

[nthiery]

Patch

[nthiery]

comment:1

Attachment: lazy_attributes-4371-nt.patch.gz

In case I did not make myself clear: this patch is not ready for insertion in sage.
It's more of a request for comments!

[sagetrac-mvngu]

comment:2

Replying to @nthiery:

In case I did not make myself clear: this patch is not ready for insertion in sage.
It's more of a request for comments!

Here are some possible fixes for improving the documentation of the patch lazy_attributes-4371-nt.patch:

-easilly override a given attribute; you don't need to call the
+easily override a given attribute; you don't need to call the

-the internal dictionnary of the object:
+the internal dictionary of the object:

Otherwise, the doc looks OK to me.

[nthiery]

comment:3

Fixes applied. Thanks. Lazy attributes now work well with new style classes:

http://sage.math.washington.edu:2144/file/03b4130fb25d/lazy_attributes-4371-nt.patch

[sagetrac-mvngu]

comment:4

Replying to @nthiery:

Fixes applied. Thanks. Lazy attributes now work well with new style classes:

http://sage.math.washington.edu:2144/file/03b4130fb25d/lazy_attributes-4371-nt.patch

For your new patch at the above URL, here's a fix to improve your documentation:

-Invoking Descriptors in the python reference manual).
+Invoking Descriptors in the Python reference manual).

[sagetrac-mabshoff]

comment:5

Nicolas,

should this patch be reviewed?

Cheers,

Michael

[nthiery]

comment:6

Comments on it are more than welcome (in particular for the hasattr feature).
No need to waste time on a complete final review though.

Thanks,
Nicolas