ENH: add dtype parameter for append() - #33165

PR summary

This PR is continuing this closed PR #22429. Solving issue #22427 and #25640.

Before

import numpy as np

a = np.array([1, 2], dtype=np.float32)
print(a.dtype) # prints float32
b = np.append(a, 42)
print(b.dtype) # prints float64

After adding dtype argument

a = np.array([1, 2], dtype=np.float32)
print(a.dtype) # prints float32
b = np.append(a, 42, dtpye=np.float32)
print(b.dtype) # prints float32

I'm interested working on this PR because I did solved array type issue previously (#32555) and want to keep working on open source project. This will help me and other users have exact type for their array when using append().

I'm aware there are lot of discussion in other PR about append() method. My current implementation just continuing the closed unmerged PR. I'm open with new suggestion how the new append() should be written for future release (even totally different approach).

AI Disclosure

I use Opencode to understand the repository and create initial draft. I rewrite the code and test it manually.