Enhance compliance checking with centralized output gathering

- Add reusable gather_test_outputs.sh script for QA pipelines
- Run comprehensive test suite (CMIP6 + CMIP7 examples)
- Categorize outputs by CMIP version (cmip6/cmip7/other)
- Generate manifest and summary files for traceability
- Update workflow to use centralized gathering approach
- Upload both test outputs and compliance reports as artifacts
- Document reusable approach for other QA tools (ESGF-QA, etc.)
- Focus on CMIP7 examples per user request

Agent-Logs-Url: https://github.com/ilaflott/fremorizer/sessions/80a31504-3038-4dea-9af9-6f5ceb1e3641

Co-authored-by: ilaflott <6273252+ilaflott@users.noreply.github.com>

Author: Claude

.github/scripts/gather_test_outputs.sh

@@ -0,0 +1,246 @@
+#!/bin/bash
+# Script to gather all NetCDF test outputs into a centralized location
+# This script is designed to be reusable by other QA pipelines (IOOS compliance-checker, ESGF-QA, etc.)
+#
+# Usage: gather_test_outputs.sh <output_directory>
+#
+# The script:
+# 1. Searches for all NetCDF files generated by tests
+# 2. Copies them to a centralized directory with preserved structure
+# 3. Creates a manifest file listing all gathered files
+# 4. Categorizes files by CMIP version (CMIP6 vs CMIP7)
+
+set -e
+
+# Default output directory if not specified
+OUTPUT_DIR="${1:-/tmp/qa_test_outputs}"
+
+echo "=========================================="
+echo "Gathering Test Outputs for QA Validation"
+echo "=========================================="
+echo "Output directory: ${OUTPUT_DIR}"
+echo ""
+
+# Create the centralized output directory
+mkdir -p "${OUTPUT_DIR}"
+mkdir -p "${OUTPUT_DIR}/cmip6"
+mkdir -p "${OUTPUT_DIR}/cmip7"
+mkdir -p "${OUTPUT_DIR}/other"
+
+# Initialize counters
+total_files=0
+cmip6_files=0
+cmip7_files=0
+other_files=0
+
+# Create manifest file
+MANIFEST="${OUTPUT_DIR}/manifest.txt"
+echo "# Test Output Manifest" > "${MANIFEST}"
+echo "# Generated: $(date -u +%Y-%m-%dT%H:%M:%SZ)" >> "${MANIFEST}"
+echo "# " >> "${MANIFEST}"
+
+# Find all NetCDF files in test output directories
+# Primary locations:
+# - fremorizer/tests/test_files/outdir (from test_cmor_run_subtool.py)
+# - fremorizer/tests/test_files/outdir_ppan_only (from test_cmor_run_subtool_further_examples.py)
+
+echo "Searching for NetCDF files in test directories..."
+echo ""
+
+# Function to categorize and copy files
+gather_files() {
+    local search_dir=$1
+    local label=$2
+
+    if [ ! -d "${search_dir}" ]; then
+        echo "Warning: Directory ${search_dir} not found, skipping"
+        return
+    fi
+
+    echo "Gathering from: ${search_dir} (${label})"
+
+    # Find all .nc files
+    while IFS= read -r -d '' ncfile; do
+        if [ -f "${ncfile}" ]; then
+            # Determine CMIP version from file path or content
+            cmip_version="other"
+            if [[ "${ncfile}" == *"CMIP6"* ]]; then
+                cmip_version="cmip6"
+                ((cmip6_files++))
+            elif [[ "${ncfile}" == *"CMIP7"* ]]; then
+                cmip_version="cmip7"
+                ((cmip7_files++))
+            else
+                # Try to detect from file attributes
+                if command -v ncdump &> /dev/null; then
+                    if ncdump -h "${ncfile}" 2>/dev/null | grep -q 'mip_era = "CMIP7"'; then
+                        cmip_version="cmip7"
+                        ((cmip7_files++))
+                    elif ncdump -h "${ncfile}" 2>/dev/null | grep -q 'mip_era = "CMIP6"'; then
+                        cmip_version="cmip6"
+                        ((cmip6_files++))
+                    else
+                        ((other_files++))
+                    fi
+                else
+                    ((other_files++))
+                fi
+            fi
+
+            # Create unique filename to avoid collisions
+            basename=$(basename "${ncfile}")
+            # Add a hash of the full path to ensure uniqueness
+            path_hash=$(echo "${ncfile}" | md5sum | cut -c1-8)
+            unique_name="${path_hash}_${basename}"
+
+            # Copy to categorized directory
+            dest_file="${OUTPUT_DIR}/${cmip_version}/${unique_name}"
+            cp "${ncfile}" "${dest_file}"
+
+            # Add to manifest
+            echo "${ncfile} -> ${cmip_version}/${unique_name}" >> "${MANIFEST}"
+
+            ((total_files++))
+
+            echo "  ✓ ${ncfile}"
+            echo "    → ${cmip_version}/${unique_name}"
+        fi
+    done < <(find "${search_dir}" -name "*.nc" -type f -print0 2>/dev/null)
+}
+
+# Gather files from different test output directories
+gather_files "fremorizer/tests/test_files/outdir" "CMIP6 basic tests"
+gather_files "fremorizer/tests/test_files/outdir_ppan_only" "Extended test examples"
+
+# Add summary to manifest
+echo "" >> "${MANIFEST}"
+echo "# Summary" >> "${MANIFEST}"
+echo "# Total files: ${total_files}" >> "${MANIFEST}"
+echo "# CMIP6 files: ${cmip6_files}" >> "${MANIFEST}"
+echo "# CMIP7 files: ${cmip7_files}" >> "${MANIFEST}"
+echo "# Other files: ${other_files}" >> "${MANIFEST}"
+
+echo ""
+echo "=========================================="
+echo "Gathering Complete"
+echo "=========================================="
+echo "Total files gathered: ${total_files}"
+echo "  - CMIP6: ${cmip6_files}"
+echo "  - CMIP7: ${cmip7_files}"
+echo "  - Other: ${other_files}"
+echo ""
+echo "Output directory: ${OUTPUT_DIR}"
+echo "Manifest file: ${MANIFEST}"
+echo ""
+
+# Create a summary file for easy reference
+SUMMARY="${OUTPUT_DIR}/SUMMARY.md"
+cat > "${SUMMARY}" << EOF
+# Test Output Summary
+
+**Generated:** $(date -u +%Y-%m-%dT%H:%M:%SZ)
+
+## Statistics
+
+- **Total NetCDF files:** ${total_files}
+- **CMIP6 files:** ${cmip6_files} (in \`cmip6/\` directory)
+- **CMIP7 files:** ${cmip7_files} (in \`cmip7/\` directory)
+- **Other files:** ${other_files} (in \`other/\` directory)
+
+## Directory Structure
+
+\`\`\`
+${OUTPUT_DIR}/
+├── cmip6/          # CMIP6-specific outputs
+├── cmip7/          # CMIP7-specific outputs
+├── other/          # Unclassified outputs
+├── manifest.txt    # Detailed file mapping
+└── SUMMARY.md      # This file
+\`\`\`
+
+## Files by Category
+
+### CMIP6 Files (${cmip6_files})
+EOF
+
+# List CMIP6 files
+if [ ${cmip6_files} -gt 0 ]; then
+    find "${OUTPUT_DIR}/cmip6" -name "*.nc" -type f | sort | while read -r f; do
+        echo "- \`$(basename "$f")\`" >> "${SUMMARY}"
+    done
+else
+    echo "- None" >> "${SUMMARY}"
+fi
+
+cat >> "${SUMMARY}" << EOF
+
+### CMIP7 Files (${cmip7_files})
+EOF
+
+# List CMIP7 files
+if [ ${cmip7_files} -gt 0 ]; then
+    find "${OUTPUT_DIR}/cmip7" -name "*.nc" -type f | sort | while read -r f; do
+        echo "- \`$(basename "$f")\`" >> "${SUMMARY}"
+    done
+else
+    echo "- None" >> "${SUMMARY}"
+fi
+
+cat >> "${SUMMARY}" << EOF
+
+### Other Files (${other_files})
+EOF
+
+# List other files
+if [ ${other_files} -gt 0 ]; then
+    find "${OUTPUT_DIR}/other" -name "*.nc" -type f | sort | while read -r f; do
+        echo "- \`$(basename "$f")\`" >> "${SUMMARY}"
+    done
+else
+    echo "- None" >> "${SUMMARY}"
+fi
+
+cat >> "${SUMMARY}" << EOF
+
+## Usage for QA Pipelines
+
+This centralized output directory can be used by various QA tools:
+
+### IOOS Compliance Checker
+\`\`\`bash
+for ncfile in ${OUTPUT_DIR}/*/*.nc; do
+    compliance-checker --test=cf --criteria=normal "\${ncfile}"
+done
+\`\`\`
+
+### ESGF-QA
+\`\`\`bash
+for ncfile in ${OUTPUT_DIR}/*/*.nc; do
+    esgf-qa --project CMIP6 "\${ncfile}"
+done
+\`\`\`
+
+### Custom Validation
+\`\`\`bash
+# Process all files
+find ${OUTPUT_DIR} -name "*.nc" -type f -exec your-validator {} \;
+
+# Process only CMIP7 files
+find ${OUTPUT_DIR}/cmip7 -name "*.nc" -type f -exec your-validator {} \;
+\`\`\`
+
+## See Also
+
+- \`manifest.txt\` - Full mapping of source files to gathered files
+- Individual QA pipeline documentation for specific usage instructions
+EOF
+
+cat "${SUMMARY}"
+
+# Exit successfully if we found any files
+if [ ${total_files} -gt 0 ]; then
+    exit 0
+else
+    echo "::warning::No NetCDF files were found in test output directories"
+    exit 1
+fi