string.Formatter does not support key/attribute access on unnumbered fields

[tbeadle]

BPO	27307
Nosy	@ericvsmith, @tekknolagi
PRs	gh-71494: string.Formatter unnumbered key/attributes #21767
Files	0001-Issue-27307-Support-index-attribute-access-for-unnum.patch

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = None
created_at = <Date 2016-06-13.15:32:42.624>
labels = ['3.7', '3.8', 'type-bug', 'library', '3.9']
title = 'string.Formatter does not support key/attribute access on unnumbered fields'
updated_at = <Date 2020-08-07.15:32:47.479>
user = 'https://bugs.python.org/tbeadle'

bugs.python.org fields:

activity = <Date 2020-08-07.15:32:47.479>
actor = 'python-dev'
assignee = 'none'
closed = False
closed_date = None
closer = None
components = ['Library (Lib)']
creation = <Date 2016-06-13.15:32:42.624>
creator = 'tbeadle'
dependencies = []
files = ['43376']
hgrepos = []
issue_num = 27307
keywords = ['patch']
message_count = 5.0
messages = ['268452', '364613', '364623', '364630', '364642']
nosy_count = 4.0
nosy_names = ['eric.smith', 'python-dev', 'tbeadle', 'tekknolagi']
pr_nums = ['21767']
priority = 'normal'
resolution = None
stage = 'patch review'
status = 'open'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue27307'
versions = ['Python 3.7', 'Python 3.8', 'Python 3.9']

[tbeadle]

Support for unnumbered fields in string.Formatter.format was added in http://bugs.python.org/issue13598, however, it does not support accessing an index or attribute of an unnumbered field like str.format does. Instead, it raises an unhelpful "KeyError: ''":

In [1]: import string

In [2]: fmt = string.Formatter()

In [3]: fmt.format('{[0]}', ['a', 'b'])
\---------------------------------------------------------------------------
```pytb
KeyError                                  Traceback (most recent call last)
<ipython-input-3-a923004fe8e8> in <module>()
----> 1 fmt.format('{[0]}', ['a', 'b'])
```


/usr/lib64/python2.7/string.pyc in format(*args, \*\*kwargs)
    557                 raise TypeError("format() missing 1 required positional "
    558                                 "argument: 'format_string'")
--\> 559         return self.vformat(format_string, args, kwargs)
    560 
    561     def vformat(self, format_string, args, kwargs):

/usr/lib64/python2.7/string.pyc in vformat(self, format_string, args, kwargs)
    561     def vformat(self, format_string, args, kwargs):
    562         used_args = set()
--\> 563         result = self.\_vformat(format_string, args, kwargs, used_args, 2)
    564         self.check_unused_args(used_args, args, kwargs)
    565         return result

/usr/lib64/python2.7/string.pyc in \_vformat(self, format_string, args, kwargs, used_args, recursion_depth)
    583                 # given the field_name, find the object it references
    584                 #  and the argument it came from
--\> 585                 obj, arg_used = self.get_field(field_name, args, kwargs)
    586                 used_args.add(arg_used)
    587 

/usr/lib64/python2.7/string.pyc in get_field(self, field_name, args, kwargs)
    644         first, rest = field_name.\_formatter_field_name_split()
    645 
--\> 646         obj = self.get_value(first, args, kwargs)
    647 
    648         # loop through the rest of the field_name, doing

/usr/lib64/python2.7/string.pyc in get_value(self, key, args, kwargs)
    603             return args[key]
    604         else:
--\> 605             return kwargs[key]
    606 
    607 

KeyError: ''

The attached patch adds this functionality.

aronancher asked in http://bugs.python.org/issue13598#msg218114 if the original patch was going to make it in to python 2.7. Perhaps that could get a look?

[ericvsmith]

It would be good if someone could convert this to a pull request and beef up the tests.

[tekknolagi]

I'll take a look at the patch and see if this solves my problem. If it does, I'll update my PR with tests.

[tekknolagi]

Looks like the patch solves my problem, so I am going to update my PR sometime today.

[tekknolagi]

Okay, well it doesn't provide the desired behavior of raising the error when switching back and forth between manual and auto numbering, so I am looking into that.