Small fixes to make it work with Python 3 #158
Conversation
Note that two tests fail due to internal changes in Python 3:
- test_iterable_literals_eval: The order of keys in dict changed in Py3
- test_no_evaluation: ast.literal_eval('5 -2') gives 3 in Py3
|
|
|
|
| @@ -950,7 +950,7 @@ def process_doc(doc, | |||
| if do_check_order and symbols.redefined: | |||
| warning("Document is incompatible to --inorder processing.") | |||
| warning("The following properties were redefined after usage:") | |||
| for k, v in symbols.redefined.iteritems(): | |||
| for k, v in symbols.redefined.items(): | |||
There was a problem hiding this comment.
Although iteritems() is more efficient than items() in python2, I think for the sake of simplicity, we can switch to items().
| @@ -50,7 +50,7 @@ | |||
| this_dir_cwd = os.getcwd() | |||
| os.chdir(cur_dir) | |||
| # Remove this dir from path | |||
| sys.path = filter(lambda a: a not in [this_dir, this_dir_cwd], sys.path) | |||
| sys.path = [a for a in sys.path if a not in [this_dir, this_dir_cwd]] | |||
There was a problem hiding this comment.
This is definitely more pythonic ;-)
src/xacro/xmlutils.py
Outdated
| @@ -31,6 +31,7 @@ | |||
| # Maintainer: Morgan Quigley <morgan@osrfoundation.org> | |||
|
|
|||
| import xml | |||
There was a problem hiding this comment.
The "import xml" can be removed then.
Parsing of number and boolean literals was done with ast.literal_eval(). However, since python3 this function evaluates expressions like "5 -3" as the number result and not a string anymore. Hence resort to explicit type-casting to int, float, or boolean.
|
As soon as #161 is merged, I will add some more commits, cleanup and re-issue as a new PR. |
|
I believe all these changes (and more) were incorporated into #163, which just merged. Thanks @kartikmohta and @rhaschke for iterating on this! Cheers |
Note that two tests fail due to internal changes in Python 3: