Problem with NumPy documentation (suggest fails)
oscargus opened this issue · 1 comments
Brief description
(base) lnx00073 [/local/data1/matplotlib/doc] (mlabdocumentation)> sphobjinv su -u https://numpy.org/doc/stable/ fft
No inventory at provided URL.
Attempting "https://numpy.org/doc/stable/objects.inv" ...
Remote inventory found.
No results found.
but
(base) lnx00073 [/local/data1/matplotlib/doc] (mlabdocumentation)> sphobjinv su -u https://numpy.org/doc/stable/ numpy | grep fft
No inventory at provided URL.
Attempting "https://numpy.org/doc/stable/objects.inv" ...
Remote inventory found.
y
:py:module:`numpy.fft`
:py:function:`numpy.fft.fft`
:py:function:`numpy.fft.fft2`
:py:function:`numpy.fft.fftfreq`
:py:function:`numpy.fft.fftn`
:py:function:`numpy.fft.fftshift`
:py:function:`numpy.fft.hfft`
:py:function:`numpy.fft.ifft`
:py:function:`numpy.fft.ifft2`
:py:function:`numpy.fft.ifftn`
:py:function:`numpy.fft.ifftshift`
:py:function:`numpy.fft.ihfft`
:py:function:`numpy.fft.irfft`
:py:function:`numpy.fft.irfft2`
:py:function:`numpy.fft.irfftn`
:py:function:`numpy.fft.rfft`
:py:function:`numpy.fft.rfft2`
:py:function:`numpy.fft.rfftfreq`
:py:function:`numpy.fft.rfftn`
Expected behavior
Some hits.
Actual behavior
No hits.
To reproduce
See above.
Attachments
System information
- Device:
- OS: CentOS 7
Python environment
Python
Python 3.8.13 | packaged by conda-forge | (default, Mar 25 2022, 06:04:10)
[GCC 10.3.0]
Libraries
Additional information
sphobjinv v2.2.2
I opened numpy/numpy#21294 before greping. As stated there, it seems like sphinx cannot find it either, but still investigating if I was just messing up.
sphobjinv suggest
is showing no results because all of the objects in the inventory yield a match score lower than the default threshold of '75'.
If you lower that threshold, results do appear.
>sphobjinv su -u https://numpy.org/doc/stable/ fft -t 60
No inventory at provided URL.
Attempting "https://numpy.org/doc/stable/objects.inv" ...
Remote inventory found.
Display all 39 results (Y/N)? y
:py:module:`numpy.fft`
:py:function:`numpy.fft.fft`
:py:function:`numpy.fft.fft2`
:py:function:`numpy.fft.fftfreq`
:py:function:`numpy.fft.fftn`
:py:function:`numpy.fft.fftshift`
:py:function:`numpy.fft.hfft`
:py:function:`numpy.fft.ifft`
:py:function:`numpy.fft.ifft2`
:py:function:`numpy.fft.ifftn`
:py:function:`numpy.fft.ifftshift`
:py:function:`numpy.fft.ihfft`
:py:function:`numpy.fft.irfft`
:py:function:`numpy.fft.irfft2`
:py:function:`numpy.fft.irfftn`
:py:function:`numpy.fft.rfft`
:py:function:`numpy.fft.rfft2`
:py:function:`numpy.fft.rfftfreq`
:py:function:`numpy.fft.rfftn`
:std:doc:`reference/generated/numpy.fft.fft`
:std:doc:`reference/generated/numpy.fft.fft2`
:std:doc:`reference/generated/numpy.fft.fftfreq`
:std:doc:`reference/generated/numpy.fft.fftn`
:std:doc:`reference/generated/numpy.fft.fftshift`
:std:doc:`reference/generated/numpy.fft.hfft`
:std:doc:`reference/generated/numpy.fft.ifft`
:std:doc:`reference/generated/numpy.fft.ifft2`
:std:doc:`reference/generated/numpy.fft.ifftn`
:std:doc:`reference/generated/numpy.fft.ifftshift`
:std:doc:`reference/generated/numpy.fft.ihfft`
:std:doc:`reference/generated/numpy.fft.irfft`
:std:doc:`reference/generated/numpy.fft.irfft2`
:std:doc:`reference/generated/numpy.fft.irfftn`
:std:doc:`reference/generated/numpy.fft.rfft`
:std:doc:`reference/generated/numpy.fft.rfft2`
:std:doc:`reference/generated/numpy.fft.rfftfreq`
:std:doc:`reference/generated/numpy.fft.rfftn`
:std:doc:`reference/routines.fft`
:std:label:`routines.fft`
I have #145 in mind as a better default (or at least alternative) suggest
behavior, that would avoid this sort of thing.
If nothing else, though, in the short term I should really improve the message emitted by the CLI. "No results found." is not accurate. It should really provide a short report like:
12373 results found.
0 results found with match score exceeding threshold of 75.
I'll create an issue for that UI change shortly.