[Bug]: invalid formatting of complex with sign='+'/' ' and without format type

[skirpichev]

Describe the bug

Python 3.12.8 (Tue Jun 23 15:26:01 UTC 2026)
[Graal, Oracle GraalVM, Java 25.0.3 (amd64)] on 'linux'
Type "help", "copyright", "credits" or "license" for more information.
>>> format(1j, '+.1')
'1j'
>>> format(1j, ' .2')
'1j'

c.f.

Python 3.12.13 (tags/v3.12.13:3bb231a6a5d, May 14 2026, 11:06:30) [GCC 14.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> format(1j, '+.1')
'+1j'
>>> format(1j, ' .2')
' 1j'

Docs says:

If the presentation type is missing, the result will match the output of str() (complex numbers with a non-zero real part are also surrounded by parentheses), possibly altered by other format modifiers.

Edit:
This affects also thousands separators:

Python 3.12.8 (Tue Jun 23 15:26:01 UTC 2026)
[Graal, Oracle GraalVM, Java 25.0.3 (amd64)] on 'linux'
Type "help", "copyright", "credits" or "license" for more information.
>>> format(1000j, ',.4')
'1000j'

vs

Python 3.12.13 (tags/v3.12.13:3bb231a6a5d, May 14 2026, 11:06:30) [GCC 14.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> format(1000j, ',.4')
'1,000j'

Operating system

Linux

CPU architecture

x86_64

GraalPy version

25.1.3

JDK version

No response

Context configuration

No response

Steps to reproduce

See description

Expected behavior

Match CPython

Stack trace

Additional context

No response

[saisharan0103]

The repro details point to the no-presentation-type path for complex formatting dropping modifiers that CPython still applies: +.1 should preserve +, .2 should preserve the leading space, and ,.4 should apply grouping.

For a PR, I would start by checking the complex formatter branch used when the format type is omitted, then add focused CPython parity cases for sign and thousands separators.