New submission from Łukasz Langa luk...@langa.pl:
SUMMARY
===
This is a refactoring of configparser raised from discussions on #6517 and
#9876. It solves both issues and makes the code quite elegant. Best part is,
it's fully backwards compatible [1].
In the patch I basically did:
a) move the interpolation functionality out of the parsers
b) deprecate RCP and CP in favor of SCP
c) refactor the resulting code to get rid of duplicates
This means that actually despite the generalization and introduction of
deprecation warnings, configparser lost 22 lines in the process. The
implementation is complete and unit tested.
WHAT'S LEFT TO DO
=
1. Obviously the new `interpolation` __init__ argument needs docs. This
includes the docstrings on *Interpolation classes that describe how the
algorithms work.
2. `ExtendedInterpolation` class implementing the buildout syntax, that would
satisfy the request from #9876.
3. Once we have more interpolation implementations new test cases will be added
to test_cfgparser.py.
Will do all three when I have any opinions on the current direction.
REVIEW HINTS
I need some thumbs up or advice on the design decisions made.
1. `interpolation` now takes a class, __init__ instantiates an object on its
own. Do tell if taking an object upfront would be a better approach.
2. *Interpolation classes have methods which take the parser as an argument. Do
tell if making it an object attribute would be a better approach.
All other points of interest during a review naturally apply here too.
[1] Assuming _names are private and can be moved around when needed.
--
assignee: lukasz.langa
components: Library (Lib)
files: configparser_interpolation.diff
keywords: needs review, patch
messages: 122078
nosy: eric.araujo, fdrake, georg.brandl, lukasz.langa, rhettinger
priority: normal
severity: normal
stage: patch review
status: open
title: Modular interpolation in configparser
type: feature request
versions: Python 3.2
Added file: http://bugs.python.org/file19767/configparser_interpolation.diff
___
Python tracker rep...@bugs.python.org
http://bugs.python.org/issue10499
___
___
Python-bugs-list mailing list
Unsubscribe:
http://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com
