From 7fffd08d798cefd9c62d0897fa5f60015a21da3b Mon Sep 17 00:00:00 2001
From: Yanick Fratantonio <yanickf@google.com>
Date: Tue, 21 Jan 2025 16:54:53 +0000
Subject: [PATCH] gh workflow: add poc to show different onnxruntime results on
 windows

---
 .../onnxruntime-windows-diff-poc.yml          | 45 +++++++++++++++++++
 python/scripts/extract_features.py            | 39 ++++++++++++++++
 2 files changed, 84 insertions(+)
 create mode 100644 .github/workflows/onnxruntime-windows-diff-poc.yml
 create mode 100644 python/scripts/extract_features.py

diff --git a/.github/workflows/onnxruntime-windows-diff-poc.yml b/.github/workflows/onnxruntime-windows-diff-poc.yml
new file mode 100644
index 00000000..86da7679
--- /dev/null
+++ b/.github/workflows/onnxruntime-windows-diff-poc.yml
@@ -0,0 +1,45 @@
+name: POC for debugging windows onnxruntime diff
+
+on:
+  workflow_dispatch:
+  pull_request:
+
+permissions:
+  contents: read
+
+jobs:
+  run-inference:
+    strategy:
+      matrix:
+        python-version: ["3.12"]
+        os: [ "ubuntu-latest", "macos-latest", "windows-latest" ]
+    runs-on: ${{ matrix.os }}
+    steps:
+      - uses: actions/checkout@eef61447b9ff4aafe5dcd4e0bbf5d482be7e7871 # pin@v4
+      - name: Setup Python
+        uses: actions/setup-python@0b93645e9fea7318ecaed2b359559ac225c90a2b # pin@v5
+        with:
+          python-version: '${{ matrix.python-version }}'
+      - name: Install magika with pip
+        run: python3 -m pip install magika==0.6.0rc3
+      - run: python3 -c "import magika, pathlib; m = magika.Magika(); res = m.identify_path(pathlib.Path('tests_data/basic/python/code.py')); print(res)"
+      - run: python3 ./python/scripts/extract_features.py
+
+      # do hexdump
+      - if: matrix.os == 'windows-latest'
+        shell: pwsh
+        run: |
+          Format-Hex tests_data/basic/python/code.py
+      - if: matrix.os == 'ubuntu-latest'
+        run: hexdump -C tests_data/basic/python/code.py
+
+      # get the hash
+      - if: matrix.os == 'windows-latest'
+        shell: pwsh
+        run: |
+          Get-FileHash tests_data/basic/python/code.py
+      - if: matrix.os == 'ubuntu-latest'
+        run: sha256sum tests_data/basic/python/code.py
+      - if: matrix.os == 'macos-latest'
+        run: shasum -a256 tests_data/basic/python/code.py
+
diff --git a/python/scripts/extract_features.py b/python/scripts/extract_features.py
new file mode 100644
index 00000000..f506b221
--- /dev/null
+++ b/python/scripts/extract_features.py
@@ -0,0 +1,39 @@
+import hashlib
+from pathlib import Path
+
+import magika
+
+test_path = Path(__file__).parent.parent.parent / "tests_data/basic/python/code.py"
+
+m = magika.Magika()
+
+fs = m._extract_features_from_path(
+    test_path,
+    beg_size=1024,
+    mid_size=0,
+    end_size=1024,
+    padding_token=256,
+    block_size=4096,
+    use_inputs_at_offsets=False,
+)
+
+
+def serialize(fs):
+    return hashlib.sha256(str(fs.beg + fs.end).encode("ascii")).hexdigest()
+
+
+print(fs)
+print(serialize(fs))
+
+f = open(test_path, "rb")
+content_bytes = f.read()
+print(f"content bytes len: {len(content_bytes)}")
+f.close()
+
+content_bytes_2 = test_path.read_bytes()
+print(f"content bytes 2 len: {len(content_bytes_2)}")
+
+print(f"file size: {test_path.stat().st_size}")
+
+content_bytes_ints = list(map(int, content_bytes[:64]))
+print(content_bytes_ints)