Merge pull request #2 from pontikos-lab/fix_py_dcm_version

Fix py dcm version in metadata.json

.github/workflows/release.yml

@@ -1,17 +1,39 @@
-name: Test and Release
+name: Test and Release Workflow
 
 on:
+  pull_request:
+    branches:
+      - "*" # Run tests on all branches for PRs
   push:
     branches:
-      - main
+      - main # Run tests and potentially release on pushes to main
     paths:
       - "**/*.py"
       - pyproject.toml
       - poetry.lock
 
 jobs:
-  test-and-release:
-    name: Test and Release
+  run-tests:
+    name: Run Tests
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+      - name: Install dependencies
+        run: |
+          pip install poetry
+          poetry install
+      - name: Run tests
+        run: poetry run pytest -k 'not test_main_mapping_example_dir_relative and not test_main_mapping_example_dir and not test_process_dcm'
+
+  release:
+    name: Release
+    needs: run-tests
+    if: github.event_name == 'push' && github.ref == 'refs/heads/main'
     runs-on: ubuntu-latest
     permissions:
       contents: write
@@ -20,20 +42,14 @@ jobs:
         uses: actions/checkout@v4
         with:
           fetch-depth: 0
-
       - name: Set up Python
         uses: actions/setup-python@v5
         with:
           python-version: "3.11"
-
       - name: Install dependencies
         run: |
           pip install poetry commitizen
           poetry install
-
-      - name: Run tests
-        run: poetry run pytest -k 'not test_main_mapping_example_dir_relative and not test_main_mapping_example_dir and not test_process_dcm'
-
       - name: Check if version bump is needed
         id: check
         continue-on-error: true
@@ -42,6 +58,7 @@ jobs:
           echo "bump=$(cz bump --dry-run 2>&1 | grep -q 'bump:' && echo 'yes' || echo 'no')" >> $GITHUB_OUTPUT
 
       - name: Set up Git user
+        if: steps.check.outputs.bump == 'yes'
         run: |
           git config --global user.email "github-actions[bot]@users.noreply.github.com"
           git config --global user.name "github-actions[bot]"
@@ -54,6 +71,7 @@ jobs:
           echo "version=$(git describe --tags --abbrev=0)" >> $GITHUB_OUTPUT
 
       - name: Build project
+        if: steps.check.outputs.bump == 'yes'
         run: poetry build
 
       - name: Generate Changelog

.pre-commit-config.yaml

@@ -26,18 +26,18 @@ repos:
       - id: check-merge-conflict
       - id: check-docstring-first
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.6.5
+    rev: v0.6.8
     hooks:
       - id: ruff
         args: ["--fix"]
         files: ^hooks
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.6.5
+    rev: v0.6.8
     hooks:
       - id: ruff-format
         files: ^hooks
   - repo: https://github.com/commitizen-tools/commitizen
-    rev: v3.29.0
+    rev: v3.29.1
     hooks:
       - id: commitizen
         stages:

.talismanrc

@@ -1,10 +1,4 @@
 fileignoreconfig:
   - filename: poetry.lock
-    checksum: 1c2b16ed82e9ee932d71256bd8951b7bb6615a0bed7b0aae452ba97e5fcbe769
-  - filename: tests/test_utils.py
-    checksum: a674113efa3a5e2dba900599efe3bdbe16ce88dca297ec57752ad20afb399908
-  - filename: tests/test_main.py
-    checksum: 59a23ed618d9f4f2b0e357444be748cd00233bcf7428a8afde1b3be89258a103
-  - filename: process_dcm/utils.py
-    checksum: c7e686f07cf036a427a8dcb3719affc72d5de9357b48625c688aab01614f57f3
+    checksum: 0ad40dc799769c4734b1f5791c32eb5657f6a337fd70492d16b4b8878bdcbf9f
 version: ""

.vscode/settings.json

@@ -39,5 +39,6 @@
       "source.organizeImports": "explicit"
     },
     "editor.formatOnType": true
-  }
+  },
+  "conventionalCommits.scopes": ["deps"]
 }

process_dcm/main.py

@@ -105,7 +105,7 @@ def main(
             fg=typer.colors.BRIGHT_YELLOW,
         )
 
-    len_sf, base_dir, subfolders = find_dicom_folders_with_base(input_dir)
+    len_sf, base_dir, subfolders = find_dicom_folders_with_base(os.path.abspath(input_dir))
     output_dirs = []
 
     if os.path.isabs(output_dir):
@@ -121,7 +121,7 @@ def main(
             output_dirs = [os.path.join(x, output_dir) for x in subfolders]
         else:
             output_dir = os.path.abspath(output_dir)
-            output_dirs = [x.replace(base_dir, output_dir) for x in subfolders]
+            output_dirs = [x.replace(os.path.abspath(base_dir), output_dir) for x in subfolders]
 
     tasks = list(zip(subfolders, output_dirs))
 

process_dcm/utils.py

@@ -21,6 +21,7 @@
 from pydicom.dataset import FileDataset
 from pydicom.filereader import dcmread
 
+from process_dcm import __version__
 from process_dcm.const import RESERVED_CSV, ImageModality
 
 warnings.filterwarnings("ignore", category=UserWarning, message="A value of type *")
@@ -153,7 +154,7 @@ def process_dcm_meta(
     metadata["series"] = {}
     metadata["images"]["images"] = []
     metadata["parser_version"] = [1, 5, 2]
-    metadata["py_dcm_version"] = [0, 1, 0]
+    metadata["py_dcm_version"] = list(map(int, __version__.split(".")))
 
     keep_gender = "g" in keep
     keep_names = "n" in keep
@@ -449,18 +450,22 @@ def find_dicom_folders_with_base(root_folder: str) -> tuple[int, str, list[str]]
     return len_ins, base_dir, natsorted(folders)
 
 
-def get_md5(file_path: Path | str | list[str]) -> str:
-    """Calculate the MD5 checksum of a file or list of files."""
+def get_md5(file_path: Path | str | list[str], minus: int = 0) -> str:
+    """Calculate the MD5 checksum of a file or list of files, optionally suppressing lines from the bottom."""
     md5_hash = hashlib.md5()
-    if isinstance(file_path, str) or isinstance(file_path, Path):
-        with open(file_path, "rb") as f:
-            while chunk := f.read(4096):
-                md5_hash.update(chunk)
+
+    def process_file(file: Path | str) -> None:
+        with open(file, "rb") as f:
+            lines = f.readlines()
+            for line in lines[:-minus] if minus > 0 else lines:
+                md5_hash.update(line)
+
+    if isinstance(file_path, str | Path):
+        process_file(file_path)
     elif isinstance(file_path, list):
         for fname in file_path:
-            with open(fname, "rb") as f:
-                while chunk := f.read(4096):
-                    md5_hash.update(chunk)
+            process_file(fname)
+
     return md5_hash.hexdigest()
 
 

pyproject.toml

@@ -21,7 +21,7 @@ classifiers = [
 python = "^3.10"
 opencv-python-headless = "^4.10.0.84"
 typer = "^0.12.5"
-pydicom = "^3.0.0"
+pydicom = "^3.0.1"
 natsort = "^8.4.0"
 pillow = "^10.4.0"
 python-gdcm = "^3.0.24.1"
@@ -30,21 +30,21 @@ python-gdcm = "^3.0.24.1"
 process-dcm = "process_dcm.main:cli"
 
 [tool.poetry.group.dev.dependencies]
-boto3-stubs = { extras = ["lambda", "s3"], version = "1.35.19" }
+boto3-stubs = { extras = ["lambda", "s3"], version = "1.35.29" }
 ipdb = "0.13.13"
 ipython = "8.27.0"
 jupyterlab = "4.2.5"
 mypy = "1.11.2"
-mypy-boto3-lambda = "1.35.3"
-mypy-boto3-s3 = "1.35.16"
+mypy-boto3-lambda = "1.35.28"
+mypy-boto3-s3 = "1.35.22"
 pandas-stubs = "2.2.2.240909"
 pdbpp = "0.10.3"
 pip = "24.2"
 pre-commit = "3.8.0"
 pytest-cov = "5.0.0"
-pytest-env = "1.1.4"
+pytest-env = "1.1.5"
 pytest-xdist = "3.6.1"
-ruff = "0.6.5"
+ruff = "0.6.8"
 typed-argument-parser = "1.10.1"
 wheel = "0.44.0"
 pytest-mock = "^3.14.0"

tests/conftest.py

@@ -13,6 +13,8 @@
 from process_dcm import __version__ as version
 from process_dcm.const import ImageModality
 
+bottom = 11
+
 
 def pytest_report_header():
     return f">>>\tVersion: {version}\n"
@@ -46,7 +48,8 @@ def del_file_paths(file_paths: list[str]) -> None:
             os.rmdir(path)
 
 
-def create_directory_structure(base_path, structure):
+def create_directory_structure(base_path: Path, structure: dict) -> None:
+    """Creates a directory structure recursively."""
     for item, content in structure.items():
         path = base_path / item
         if isinstance(content, dict):

tests/test_main.py

@@ -8,7 +8,7 @@
 from process_dcm.const import RESERVED_CSV
 from process_dcm.main import app, process_task
 from process_dcm.utils import get_md5
-from tests.conftest import remove_ansi_codes
+from tests.conftest import bottom, remove_ansi_codes
 
 
 def test_main_defaults(runner):
@@ -59,9 +59,9 @@ def test_cli_without_args(runner):
 @pytest.mark.parametrize(
     "md5, meta, keep",
     [
-        (["837808d746aef8e2dd08defbdbc70818"], "0a9a930806f2784aa4e60d47b3bad6ed", "pndg"),
-        (["7a355bb7e0c95155d1541c7fe0941c5e"], "fd6c5a84aca6499b0ea8b99d4e25dc92", "pnDg"),
-        (["2319181ecfc33d35b01dcec65ab2c568"], "35fe295648681e3521da8dddaed63705", ""),
+        (["5ba37cc43233db423394cf98c81d5fbc"], "27e4fa04ad730718b2509af36743c995", "pndg"),
+        (["5ba37cc43233db423394cf98c81d5fbc"], "3a15fdd18a67b3d4ce7be6164f58f073", "pnDg"),
+        (["5ba37cc43233db423394cf98c81d5fbc"], "ba5973bc8dd8e15aa6bef95bcd248fbf", ""),
     ],
 )
 def test_main(md5, meta, keep, janitor, runner):
@@ -83,9 +83,10 @@ def test_main(md5, meta, keep, janitor, runner):
         ]
         result = runner.invoke(app, args)
         assert result.exit_code == 0
-        of = sorted(glob(f"{output_dir}/**/*"))
-        assert len(of) == 51
-        assert get_md5(output_dir / "example-dcms/metadata.json") == meta
+        tof = sorted(glob(f"{output_dir}/**/*"))
+        of = [x for x in tof if "metadata.json" not in x]
+        assert len(tof) == 51
+        assert get_md5(output_dir / "example-dcms/metadata.json", bottom) == meta
         assert get_md5(of) in md5
 
 
@@ -94,10 +95,11 @@ def test_main_dummy(janitor, runner):
     args = ["tests/dummy_ex", "-o", "dummy_dir", "-k", "p"]
     result = runner.invoke(app, args)
     assert result.exit_code == 0
-    of = sorted(glob("dummy_dir/**/*"))
-    assert len(of) == 2
-    assert get_md5("dummy_dir/dummy_ex/metadata.json") == "1cabdb14492a0e62d80cfc0a3fe304e9"
-    assert get_md5(of) in "b19bbfca59584915295f67e9259880d7"
+    tof = sorted(glob("dummy_dir/**/*"))
+    of = [x for x in tof if "metadata.json" not in x]
+    assert len(tof) == 2
+    assert get_md5("dummy_dir/dummy_ex/metadata.json", bottom) == "0fbf0de5556e77447925ff3bfaf8168e"
+    assert get_md5(of) in ["377b5a17c284226518ccef9bddff25af"]
 
 
 def test_main_mapping(janitor, runner):
@@ -121,10 +123,11 @@ def test_main_mapping(janitor, runner):
         result = runner.invoke(app, args)
         assert result.exit_code == 0
         assert "WARN: '--relative' x 'absolute --output_dir'" in result.output
-        of = sorted(glob(f"{output_dir}/**/*"))
-        assert len(of) == 51
-        assert get_md5(output_dir / "example-dcms/metadata.json") == "261826ad2e067e9adb7143bb6c053dbc"
-        assert get_md5(of) in "6ff8e2fe69c5fbe86f81f44f74496cab"
+        tof = sorted(glob(f"{output_dir}/**/*"))
+        of = [x for x in tof if "metadata.json" not in x]
+        assert len(tof) == 51
+        assert get_md5(output_dir / "example-dcms/metadata.json", bottom) == "450e2e40d321a24219c1c9ec15b2c80e"
+        assert get_md5(of) in ["5ba37cc43233db423394cf98c81d5fbc"]
 
 
 def test_main_abort(runner):
@@ -151,8 +154,8 @@ def test_main_mapping_example_dir(janitor, runner):
         assert result.exit_code == 0
         of = sorted(glob(f"{output_dir}/**/**/*"))
         assert len(of) == 262
-        assert get_md5(output_dir / "012345/20180724_L/metadata.json") == "1b46961177c80daf69e7dea7379fcc31"
-        assert get_md5(output_dir / "3517807670/20180926_R/metadata.json") == "bbf5c47f9fb28f46b4cc1bf08c311593"
+        assert get_md5(output_dir / "012345/20180724_L/metadata.json", bottom) == "93fff12758d6c0f9098e7fd5e8c8304e"
+        assert get_md5(output_dir / "3517807670/20180926_R/metadata.json", bottom) == "b9ff35a765db6b1eaeac4253c93a6044"
 
 
 # skip this test for CI
@@ -170,8 +173,8 @@ def test_main_mapping_example_dir_relative(janitor, runner):
     janitor.append(path1)
     janitor.append(path2)
     assert len(of) == 262
-    assert get_md5(path1 / "20180724_L/dummy/metadata.json") == "1b46961177c80daf69e7dea7379fcc31"
-    assert get_md5(path2 / "20180926_R/dummy/metadata.json") == "bbf5c47f9fb28f46b4cc1bf08c311593"
+    assert get_md5(path1 / "20180724_L/dummy/metadata.json", bottom) == "93fff12758d6c0f9098e7fd5e8c8304e"
+    assert get_md5(path2 / "20180926_R/dummy/metadata.json", bottom) == "b9ff35a765db6b1eaeac4253c93a6044"
 
 
 def test_process_task():