Code cleanup, styling, and formatting #753
Conversation
|
Looks great @hsbadr ! Just 2 comments:
|
Sure, I'll automate it in GHA. Also, it seems that Window doesn't like that |
- gen-*.* - lantern.* - RcppExports.*
@dfalbel GHA needs more work than this PR. So, I'd go with the documentation and update the workflow later. Here's the customized Rscript -e 'styler::style_pkg(exclude_files = list.files("./R", pattern = "^gen-*.*|^RcppExports.*"))'
find . -type f \( -name 'DESCRIPTION' -o -name "*.R" \) ! -path "*/gen-*.*" ! -path "*/RcppExports.*" -exec sed -i -e 's/[ \t]*$//' {} \;
find . -type f \( -name '*.h' -o -name '*.hpp' -o -name '*.c' -o -name '*.cc' -o -name '*.cpp' -o -name '*.cxx' \) ! -path "*/gen-*.*" ! -path "*/lantern.*" ! -path "*/RcppExports.*" -exec sed -i -e 's/[ \t]*$//' {} \;
find . -type f \( -name '*.h' -o -name '*.hpp' -o -name '*.c' -o -name '*.cc' -o -name '*.cpp' -o -name '*.cxx' \) ! -path "*/gen-*.*" ! -path "*/lantern.*" ! -path "*/RcppExports.*" -exec clang-format -style=Google --verbose -i {} \;The above lines will reproduce this PR. Would you prefer to run them on pull request or if |
|
Oh wow, we had one, but it seems that we have unintentionally removed at some point. Can't find when though :( I think running in PR's is great! What do you think? Maybe we can also add this custom script to |
Should I recover the latest version in this PR?
Yes, I usually run it on PR... but that commits format/style changes if needed. We can also run it on demand by adding
Yeah, that makes sense and could be better that listing specific commands in |
|
@dfalbel This is ready. I'll take a cut at GHA workflow later. |
|
@hsbadr Thanks for getting the CONTRIBUTING guide back! I'm trying to run the
Maybe I'm missing something? |
It seems that you use Posix Sed (BSD/macOS) not GNU Sed (Linux). I run the script on Linux (same in GHA). |
Co-authored-by: Daniel Falbel <dfalbel@gmail.com>
@dfalbel Can you try now? I've replaced GNU |
There was a problem hiding this comment.
Thanks @hsbadr
Now I think formatting worked, but I don't completely reproduce the same changes in this PR. I get a large diff eg:
(Of course we can ignore the workflows change, but I would expect .R files and .cpp files to be identical)
❯ git diff --stat style
.github/CONTRIBUTING.md | 109 -------------------------------------------------------------------------------------------------------------
.github/workflows/lantern.yaml | 2 +-
.github/workflows/main.yaml | 38 ++++++++++++++++++--------------------
.github/workflows/website.yaml | 2 +-
NEWS.md | 2 --
R/autograd.R | 6 ------
R/device.R | 1 -
R/distributions-categorical.R | 1 -
R/distributions-chi2.R | 1 -
R/distributions-mixture_same_family.R | 1 -
R/distributions-multivariate_normal.R | 1 -
R/distributions-normal.R | 1 -
R/gen-method.R | 2 +-
R/gen-namespace.R | 4 ++--
R/generator.R | 10 ++++++----
R/indexing.R | 1 -
R/jit-compile.R | 1 -
R/linalg.R | 26 --------------------------
R/nn-activation.R | 25 -------------------------
R/nn-batchnorm.R | 3 ---
R/nn-conv.R | 5 -----
R/nn-distance.R | 1 -
R/nn-dropout.R | 3 ---
R/nn-init.R | 12 ------------
R/nn-linear.R | 3 ---
R/nn-loss.R | 12 ------------
R/nn-normalization.R | 2 --
R/nn-pooling.R | 19 -------------------
R/nn-rnn.R | 3 ---
R/nn-sparse.R | 1 -
R/nn-upsampling.R | 1 -
R/nn-utils-rnn.R | 3 ---
R/nn.R | 3 ---
R/nnf-activation.R | 2 --
R/optim.R | 1 -
R/quantization.R | 1 +
R/stack.R | 1 -
R/tensor.R | 10 ----------
R/trace.R | 4 ----
R/with-indices.R | 10 ----------
R/wrapers.R | 1 -
README.md | 94 +++++++++++++++++++++++++++++++++++++++++++++++++---------------------------------------------
inst/include/lantern/types.h | 10 +++++-----
inst/include/torch.h | 2 --
inst/include/torch_types.h | 2 --
lantern/include/lantern/types.h | 10 +++++-----
lantern/src/Allocator.cpp | 2 --
lantern/src/AllocatorCuda.cpp | 5 ++---
lantern/src/Autograd.cpp | 4 ++--
lantern/src/Backends.cpp | 4 ++--
lantern/src/Compile.cpp | 1 -
lantern/src/Contrib/SortVertices/cuda_utils.h | 1 -
lantern/src/Contrib/SortVertices/sort_vert.cpp | 2 --
lantern/src/Contrib/SortVertices/sort_vert_cpu.cpp | 1 -
lantern/src/Contrib/Sparsemax.cpp | 2 --
lantern/src/Delete.cpp | 3 ++-
lantern/src/Device.cpp | 3 ++-
lantern/src/Dimname.cpp | 3 ++-
lantern/src/Dtype.cpp | 3 ++-
lantern/src/Function.cpp | 1 -
lantern/src/Function.h | 3 ++-
lantern/src/Generator.cpp | 3 ++-
lantern/src/IValue.cpp | 1 -
lantern/src/Indexing.cpp | 3 ++-
lantern/src/JITTypes.cpp | 2 --
lantern/src/Layout.cpp | 3 ++-
lantern/src/MemoryFormat.cpp | 3 ++-
lantern/src/NNUtilsRnn.cpp | 3 ++-
lantern/src/QScheme.cpp | 3 ++-
lantern/src/Quantization.cpp | 3 ++-
lantern/src/Reduction.cpp | 3 ++-
lantern/src/Save.cpp | 2 +-
lantern/src/Scalar.cpp | 3 ++-
lantern/src/ScriptModule.cpp | 1 -
lantern/src/Stack.cpp | 1 -
lantern/src/Storage.cpp | 3 ++-
lantern/src/Tensor.cpp | 3 ++-
lantern/src/TensorList.cpp | 3 ++-
lantern/src/TensorOptions.cpp | 3 ++-
lantern/src/Threads.cpp | 1 -
lantern/src/Trace.cpp | 1 -
lantern/src/utils.cpp | 4 ++--
lantern/tests/init.cpp | 4 ++--
lantern/tests/main.cpp | 3 ++-
Maybe this was caused by a different version of styler or clang-format?
I have styler v1.6.2
clang-format version 8.0.0 (tags/google/stable/2019-01-18)
|
@dfalbel Note that this also fixes some GHA issues such as failed |
|
@dfalbel Any idea? |
|
Oh I see, we can't run win-gpu tests on forked repos. |
Fixes: ``` Warning: google-github-actions/setup-gcloud is pinned at HEAD. We strongly advise against pinning to "@master" as it may be unstable. Please update your GitHub Action YAML from: uses: 'google-github-actions/setup-gcloud@master' to: uses: 'google-github-actions/setup-gcloud@v0' ```
This reverts commit 226c176.
|
@dfalbel It seems that the GPU runs are stuck, after the I think it's ready to go anyway. |
Close #752.