create github actions for managing prs and commits

.github/get_latest_index.py

@@ -0,0 +1,95 @@
+import os
+import re
+import requests
+import sys
+import uuid
+from google.cloud import bigquery
+
+class IDCIndexManager:
+    def __init__(self, project_id):
+        print("Initializing IDCIndexManager...")
+        self.project_id = project_id
+        self.client = bigquery.Client(project=project_id)
+
+    def get_latest_idc_release_version(self, view_id):
+        print("Getting latest IDC release version...")
+        view = self.client.get_table(view_id)
+        latest_idc_release_version=int(re.search(r"idc_v(\d+)", view.view_query).group(1))
+        return latest_idc_release_version
+
+    def extract_index_version(self, file_path):
+        print(f"Extracting index version from {file_path}...")
+        with open(file_path, "r") as file:
+            for line in file:
+                if "def get_idc_version(self):" in line:
+                    return int(re.findall(r"v(\d+)", next(file))[0])
+
+    def update_index_version(self, file_path, latest_idc_release_version):
+        print(f"Updating index version in {file_path}...")
+        with open(file_path, "r") as file:
+            lines = file.readlines()
+        with open(file_path, "w") as file:
+            for i in range(len(lines)):
+                if "def get_idc_version(self):" in lines[i]:
+                    lines[i + 1] = re.sub(
+                        r"v(\d+)", f"v{latest_idc_release_version}", lines[i + 1]
+                    )
+                file.write(lines[i])
+
+    def update_sql_queries_folder(
+        self, dir, current_index_version, latest_idc_release_version
+    ):
+        print(f"Updating SQL queries from {dir}...")
+        for file_name in os.listdir(dir):
+            if file_name.endswith(".sql"):
+                file_path = os.path.join(dir, file_name)
+                with open(file_path, "r") as file:
+                    sql_query = file.read()
+                modified_sql_query = sql_query.replace(
+                    f"idc_v{current_index_version}", f"idc_v{latest_idc_release_version}"
+                )
+                with open(file_path, "w") as file:
+                    file.write(modified_sql_query)
+                return modified_sql_query
+
+    def execute_sql_query(self, file_path):
+        print(f"Executing SQL query from {file_path}...")
+        with open(file_path, "r") as file:
+            sql_query = file.read()
+        df = self.client.query(sql_query).to_dataframe()
+        csv_file_name = f"{os.path.basename(file_path).split('.')[0]}.csv.zip"
+        return df, csv_file_name
+
+    def create_csv_zip_from_df(self, df, csv_file_name):
+        print(f"Creating CSV zip file {csv_file_name}...")
+        df.to_csv(csv_file_name, compression={'method': 'zip'}, escapechar="\\")
+
+    def run_queries_folder(self, dir):
+        print(f"Running queries from {dir}...")
+        for file_name in os.listdir(dir):
+            if file_name.endswith(".sql"):
+                file_path = os.path.join(dir, file_name)
+                df, csv_file_name = self.execute_sql_query(file_path)
+                self.create_csv_zip_from_df(df, csv_file_name)
+
+    def set_multiline_output(self, name, value):
+        print(f"Setting multiline output {name}...")
+        with open(os.environ["GITHUB_OUTPUT"], "a") as fh:
+            delimiter = uuid.uuid1()
+            print(f"{name}<<{delimiter}", file=fh)
+            print(value, file=fh)
+            print(delimiter, file=fh)
+
+    def run(self):
+        print("Running IDCIndexManager...")
+        latest_idc_release_version = self.get_latest_idc_release_version("bigquery-public-data.idc_current.dicom_all_view")
+        print(f"Latest IDC release version: {latest_idc_release_version}")
+        current_index_version = self.extract_index_version("idc_index/index.py")
+        print(f"Current index version: {current_index_version}")
+        self.set_multiline_output("current_index_version", int(current_index_version))
+        self.set_multiline_output("latest_idc_release_version", int(latest_idc_release_version))
+
+
+if __name__ == "__main__":
+    manager = IDCIndexManager("gcp-project-id")
+    manager.run()

.github/workflows/commit-pr-release-manager.yml

@@ -0,0 +1,146 @@
+name: commit-pr-idc-index-release-manager
+on:
+  workflow_dispatch:
+  push: 
+  pull_request:    
+
+jobs:
+  update_idc_index:
+    runs-on: ubuntu-latest
+    environment:
+      name: pypi
+      url: https://pypi.org/p/idc-index
+    permissions:
+      id-token: write
+      contents: write
+      pull-requests: read 
+
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v4
+      with:
+        fetch-depth: 0
+
+    - name: Set up Python
+      uses: actions/setup-python@v4
+      with:
+        python-version: 3.11
+
+    - name: Install dependencies
+      run: |
+        pip install requests==2.31.0 pandas==2.1.1 google-cloud-bigquery==3.12.0 \
+        pyarrow==13.0.0 db-dtypes==1.1.1 PyGithub==2.1.1 flake8==6.1.0 \
+        duckdb==0.9.2
+      shell: bash
+
+    - name: Authorize Google Cloud
+      uses: google-github-actions/auth@v1
+      with:
+        credentials_json: ${{ secrets.SERVICE_ACCOUNT_KEY }}
+        create_credentials_file: true
+        export_environment_variables: true    
+
+    - name: Check if queries folder changed in the latest commit  or pull request 
+      uses: dorny/paths-filter@v2
+      id: pr_proposed_changes
+      with:       
+        filters: |
+          queries:
+            - 'queries/**'
+
+    - name: If queries are modified, run them with bigquery
+      id: initialize_idc_manager_class
+      if: steps.pr_proposed_changes.outputs.queries == 'true' 
+      shell: python
+      run: |
+        import sys
+        import os
+        sys.path.append(".github")
+
+        from get_latest_index import IDCIndexManager
+
+        project_id = os.environ['GCP_PROJECT_ID']
+        manager = IDCIndexManager(project_id)
+
+        manager.run_queries_folder("queries/")
+      env: 
+        GCP_PROJECT_ID: ${{ secrets.GCP_PROJECT_ID }} 
+
+    - name: If queries are not modified download latest index from github
+      if: steps.pr_proposed_changes.outputs.queries == 'false' || startsWith(github.ref, 'refs/tags/v')
+      run:
+        wget -q https://github.com/ImagingDataCommons/idc-index/releases/download/latest/idc_index.csv.zip
+      shell: bash
+
+    - name: Setup testing
+      run: |
+        wget -q "https://github.com/peak/s5cmd/releases/download/v2.2.2/s5cmd_2.2.2_Linux-64bit.tar.gz"\
+        && tar -xvzf "s5cmd_2.2.2_Linux-64bit.tar.gz"\
+        && rm "s5cmd_2.2.2_Linux-64bit.tar.gz"\
+        && mv s5cmd /usr/local/bin/s5cmd
+      shell: bash       
+
+    - name: If queries are modified by a pull request, change latest_index_url to use locally available csv
+      if: steps.pr_proposed_changes.outputs.queries == 'true' && github.event_name == 'pull_request'
+      run: |
+        import os
+        from pathlib import Path
+        home_dir = str(Path.home())
+        with open('idc_index/index.py', 'r') as file:
+            filedata = file.read()
+            filedata = filedata.replace('https://github.com/ImagingDataCommons/idc-index/releases/download/latest/idc_index.csv.zip', os.path.join(home_dir, 'work/idc-index/idc-index/idc_index.csv.zip'))    
+        with open('idc_index/index.py', 'w') as file:
+          file.write(filedata)
+      shell: python
+
+    - name: Test package
+      run: |
+        python -m unittest -vv tests/idcindex.py
+      shell: bash
+
+    - name: Create Tagged Release
+      id: create_tagged_release
+      if: (startsWith(github.ref, 'refs/tags/v')) && (github.event_name != 'pull_request')
+      #if: (startsWith(github.ref, 'refs/tags/v') && github.event_name != 'pull_request' && github.ref == 'refs/heads/main')      
+      uses: ncipollo/release-action@v1
+      with:
+        prerelease: false
+        body: 'Versioned idc-index'
+        artifacts: "*.zip"
+
+    - name: Create latest Release
+      if: (github.event_name != 'pull_request' && github.ref == 'refs/heads/main')
+      #if: (github.event_name != 'pull_request' && github.ref == 'refs/heads/main')
+      uses: crowbarmaster/GH-Automatic-Releases@latest
+      with:
+        repo_token: "${{ secrets.GITHUB_TOKEN }}"
+        automatic_release_tag: "latest"
+        #generate_notes: false
+        body: "Latest idc-index"
+        prerelease: true
+        title: "Latest idc-index"
+        files: |
+            *.zip
+
+    - name: Get version
+      uses: mtkennerly/dunamai-action@v1
+      #if: (startsWith(github.ref, 'refs/tags/v') && github.event_name != 'pull_request' && github.ref == 'refs/heads/main')
+      with:
+        env-var: set_pypi_idc_index_version
+        args: --style semver
+
+    - name: Echo soon to be released pypi version
+      #if: (startsWith(github.ref, 'refs/tags/v') && github.event_name != 'pull_request' && github.ref == 'refs/heads/main')
+      run: |
+        echo $set_pypi_idc_index_version
+
+    - name: Build a source tarball
+      #if: (startsWith(github.ref, 'refs/tags/v') && github.event_name != 'pull_request' && github.ref == 'refs/heads/main')
+      run: python setup.py sdist 
+
+    - name: Publish distribution to PyPI
+      #if: (startsWith(github.ref, 'refs/tags/v') && github.event_name != 'pull_request' && github.ref == 'refs/heads/main')
+      if: startsWith(github.ref, 'refs/tags/v')
+      uses: pypa/gh-action-pypi-publish@release/v1
+      with:
+        repository-url: https://test.pypi.org/legacy/      

.github/workflows/get-latest-index.yml

@@ -0,0 +1,83 @@
+name: idc-index release manager
+on:
+  workflow_dispatch: 
+  schedule:
+    - cron: 0 12 */1 * *
+
+jobs:
+  update_idc_index:
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout code
+      uses: actions/checkout@v4
+
+    - name: Set up Python
+      uses: actions/setup-python@v4
+      with:
+        python-version: 3.11
+
+    - name: Install dependencies
+      run: pip install requests==2.31.0 pandas==2.1.1 google-cloud-bigquery==3.12.0 pyarrow==13.0.0 db-dtypes==1.1.1 PyGithub==2.1.1
+
+    - name: Authorize Google Cloud
+      uses: google-github-actions/auth@v1
+      with:
+        credentials_json: ${{ secrets.SERVICE_ACCOUNT_KEY }}
+        create_credentials_file: true
+        export_environment_variables: true 
+
+    - name: Run script to get the latest idc index
+      id: initialize_idc_manager_class
+      shell: python
+      run: |
+        import sys
+        import os
+        sys.path.append(".github")
+
+        from get_latest_index import IDCIndexManager
+
+        project_id = os.environ['GCP_PROJECT_ID']
+        manager = IDCIndexManager(project_id)
+
+        current_index_version = manager.extract_index_version("idc_index/index.py")
+        latest_idc_release_version = manager.get_latest_idc_release_version("bigquery-public-data.idc_current.dicom_all_view")
+
+        if current_index_version < latest_idc_release_version:
+            manager.update_index_version("idc_index/index.py", latest_idc_release_version)
+            manager.update_sql_queries_folder("queries/", current_index_version, latest_idc_release_version)
+            manager.run_queries_folder("queries/")
+
+        manager.set_multiline_output("current_index_version", int(current_index_version))
+        manager.set_multiline_output("latest_idc_release_version", int(latest_idc_release_version))
+      env: 
+        GCP_PROJECT_ID: ${{ secrets.GCP_PROJECT_ID }} 
+
+    - name: Create latest Release
+      id: create_release
+      if: ${{ steps.initialize_idc_manager_class.outputs.current_index_version  != steps.initialize_idc_manager_class.outputs.latest_idc_release_version }}
+      uses: ncipollo/release-action@v1
+      with:
+        tag: latest
+        prerelease: true
+        makeLatest: true
+        allowUpdates: true
+        body: 'Latest idc-index'
+        artifacts: "*.zip"
+
+    - name: Create Pull Request
+      uses: peter-evans/create-pull-request@v5
+      with:
+        title: Update to v${{ steps.initialize_idc_manager_class.outputs.latest_idc_release_version }}
+        body:  Update sql queries and/or index.py to v${{ steps.initialize_idc_manager_class.outputs.latest_idc_release_version }}
+        base:  main
+        branch: update-sql-queries-and-or-index
+        add-paths: |
+          queries/*.sql
+          idc_index/index.py
+      env:
+        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+
+
+

idc_index/index.py

@@ -8,13 +8,14 @@
 import zipfile
 import duckdb
 
+latest_index_url= 'https://github.com/vkt1414/idc-index/releases/download/latest/idc_index.csv.zip'
 
 class IDCClient:
     def __init__(self):
         current_dir = os.path.dirname(os.path.abspath(__file__))
         file_path = os.path.join(current_dir, 'idc_index.csv.zip')
         if not os.path.exists(file_path):
-            self.index=pd.read_csv('https://github.com/ImagingDataCommons/idc-index/releases/download/latest/idc_index.csv.zip', dtype=str, encoding='utf-8')
+            self.index=pd.read_csv(latest_index_url, dtype=str, encoding='utf-8')
         else:
             self.index = pd.read_csv(file_path, dtype=str, encoding='utf-8')
         self.index = self.index.astype(str).replace('nan', '')

setup.py

@@ -70,8 +70,8 @@ def run(self):
             except Exception as e:
                 logging.error('Failed to download s5cmd:', e)
 setup(
-    name='idc_index',
-    version=package_version,
+    name='test-index-vamsi',
+    version=os.environ['set_pypi_idc_index_version'],
     packages=find_packages(),
     include_package_data=True,
     install_requires=['pandas', 'requests', 'duckdb'],