Skip to content

Apply normalization consistently in VLenBytes #570

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open

Apply normalization consistently in VLenBytes #570

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
d3a8fc0
Apply normalization consistently in VLenBytes
folded Aug 28, 2024
9efe051
Alternative: skip bytes to string conversion and memcpy if zero length.
folded Aug 29, 2024
19cf050
Update vlen.pyx
folded Aug 29, 2024
3b224f1
Merge branch 'main' into patch-1
folded Sep 2, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 4 additions & 1 deletion numcodecs/vlen.pyx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -250,7 +250,10 @@ class VLenBytes(Codec):
l = lengths[i]
store_le32(<uint8_t*>data, l)
data += 4
encv = PyBytes_AS_STRING(values[i])
b = values[i]
if b is None or b == 0: # treat these as missing value, normalize
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Like on line 231? It seems like l = 0 in this case, there is nothing to copy, and these operations are wasted. I might be wrong...

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If you don't pass a bytes object, you will hit this assert when calling PyBytes_AS_STRING: https://github.com/python/cpython/blob/40fff90ae3d46843bb9d27c6a53ef61c861a3bb4/Include/cpython/bytesobject.h#L21

it would be equivalent to skip the PyBytes_AS_STRING and memcpy entirely if l = 0. Shall I update the PR to do that instead?

b = b''
encv = PyBytes_AS_STRING(b)
memcpy(data, encv, l)
data += l

Expand Down