From 012e186c2407483c516107fa25945a0d24c480bb Mon Sep 17 00:00:00 2001
From: IanCa <ianrcallanan@gmail.com>
Date: Thu, 11 Jan 2024 17:39:49 -0600
Subject: [PATCH] Add an automatic validation check to schemas when changed

---
 .github/workflows/check_schemas.yaml | 45 ++++++++++++++++++++++++++++
 tests/validate_schemas.py            | 43 ++++++++++++++++++++++++++
 2 files changed, 88 insertions(+)
 create mode 100644 .github/workflows/check_schemas.yaml
 create mode 100644 tests/validate_schemas.py

diff --git a/.github/workflows/check_schemas.yaml b/.github/workflows/check_schemas.yaml
new file mode 100644
index 00000000..226855e9
--- /dev/null
+++ b/.github/workflows/check_schemas.yaml
@@ -0,0 +1,45 @@
+name: Process Files
+
+on:
+  push:
+    branches: [ * ]
+  pull_request:
+    branches: [ * ]
+
+jobs:
+  process-changed-files:
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v4
+      with:
+        fetch-depth: 0  # OR "2" -> To retrieve the preceding commit.
+
+    - name: Set up Python
+      uses: actions/setup-python@v5
+      with:
+        python-version: '3.10'
+
+    - name: Get changed files
+      id: changed-files
+      uses: tj-actions/changed-files@v41
+      # To compare changes between the current commit and the last pushed remote commit set `since_last_remote_commit: true`. e.g
+      # with:
+      #   since_last_remote_commit: true
+
+    - name: List all changed files
+      env:
+        ALL_CHANGED_FILES: ${{ steps.changed-files.outputs.all_changed_files }}
+      run: |
+        for file in "$ALL_CHANGED_FILES"; do
+          echo "$file was changed"
+        done
+
+    - name: Install dependencies
+      run: pip install git+https://github.com/hed-standard/hed-python.git@develop
+
+    - name: Run command on file list
+      env:
+        ALL_CHANGED_FILES: ${{ steps.changed-files.outputs.all_changed_files }}
+      run: |
+        python tests/validate_schemas.py $ALL_CHANGED_FILES
diff --git a/tests/validate_schemas.py b/tests/validate_schemas.py
new file mode 100644
index 00000000..7800aebf
--- /dev/null
+++ b/tests/validate_schemas.py
@@ -0,0 +1,43 @@
+import sys
+from hed.schema import load_schema, from_string
+from hed.errors import get_printable_issue_string
+
+def main():
+    validation_issues = []
+    saving_failures = []
+    for file_path in sys.argv[1:]:  # sys.argv[1:] contains all the arguments passed to the script
+        # print(file_path)
+        if file_path.endswith(".xml") or file_path.endswith(".mediawiki"):
+            base_schema = load_schema(file_path)
+            issues = base_schema.check_compliance()
+            if issues:
+                validation_issues.extend(issues)
+                print(get_printable_issue_string(issues, title=file_path))
+
+            mediawiki_string = base_schema.get_as_mediawiki_string()
+            reloaded_schema = from_string(mediawiki_string, schema_format=".mediawiki")
+
+            if reloaded_schema != base_schema:
+                error_text = f"Failed to reload {file_path} as mediawiki.  There is either a problem with the source file, or the saving/loading code."
+                saving_failures.append(error_text)
+                print(error_text)
+
+            xml_string = base_schema.get_as_xml_string()
+            reloaded_schema = from_string(xml_string, schema_format=".xml")
+
+            if reloaded_schema != base_schema:
+                error_text = f"Failed to reload {file_path} as xml.  There is either a problem with the source file, or the saving/loading code."
+                saving_failures.append(error_text)
+                print(error_text)
+
+    # if saving_failures:
+    #     for issue in saving_failures:
+    #         print(issue)
+
+    if validation_issues or saving_failures:
+        return 1
+    return 0
+
+
+if __name__ == "__main__":
+    exit(main())