scripts/test_entsoe_phase1c_xml_parsing.py

"""
Phase 1C: Enhanced XML Parsing for Asset-Specific Outages
===========================================================

Tests the breakthrough solution:
1. Parse RegisteredResource.mRID from transmission outage XML
2. Extract asset-specific EIC codes embedded in XML response
3. Match against 208 CNEC EIC codes
4. Test pumped storage consumption alternative queries
"""

import os
import sys
from pathlib import Path
import pandas as pd
import polars as pl
import zipfile
from io import BytesIO
import xml.etree.ElementTree as ET
from dotenv import load_dotenv
from entsoe import EntsoePandasClient

sys.path.append(str(Path(__file__).parent.parent))
load_dotenv()

API_KEY = os.getenv('ENTSOE_API_KEY')
client = EntsoePandasClient(api_key=API_KEY)

print("="*80)
print("PHASE 1C: ENHANCED XML PARSING FOR ASSET-SPECIFIC OUTAGES")
print("="*80)
print()

# ============================================================================
# TEST 1: Parse RegisteredResource.mRID from Transmission Outage XML
# ============================================================================

print("-"*80)
print("TEST 1: PARSE RegisteredResource.mRID FROM TRANSMISSION OUTAGE XML")
print("-"*80)
print()

# Load CNEC EIC codes
print("Loading 208 CNEC EIC codes...")
cnec_df = pl.read_csv(Path(__file__).parent.parent / 'data' / 'processed' / 'critical_cnecs_all.csv')
cnec_eics = cnec_df.select('cnec_eic').to_series().to_list()
print(f"[OK] Loaded {len(cnec_eics)} CNEC EICs")
print(f"  Sample: {cnec_eics[:3]}")
print()

# Query transmission outages (border-level) - get RAW bytes
print("Querying transmission outages (raw bytes)...")
print("Border: DE_LU -> FR")
print("Period: 2025-09-23 to 2025-09-30")
print()

try:
    # Need to get raw response BEFORE parsing
    # Use internal _base_request method
    params = {
        'documentType': 'A78',  # Transmission unavailability
        'in_Domain': '10YFR-RTE------C',  # FR
        'out_Domain': '10Y1001A1001A82H'  # DE_LU
    }

    response = client._base_request(
        params=params,
        start=pd.Timestamp('2025-09-23', tz='UTC'),
        end=pd.Timestamp('2025-09-30', tz='UTC')
    )

    # Extract bytes from Response object
    outages_zip = response.content

    print(f"[OK] Retrieved {len(outages_zip)} bytes (raw ZIP)")
    print()

    # Parse ZIP and extract all XML files
    print("Parsing ZIP archive...")
    extracted_eics = []
    total_timeseries = 0

    with zipfile.ZipFile(BytesIO(outages_zip), 'r') as zf:
        xml_files = [f for f in zf.namelist() if f.endswith('.xml')]
        print(f"  XML files in ZIP: {len(xml_files)}")
        print()

        for idx, xml_file in enumerate(xml_files, 1):
            with zf.open(xml_file) as xf:
                xml_content = xf.read()

                # DIAGNOSTIC: Show first 1000 chars of first XML
                if idx == 1:
                    print(f"\n  [DIAGNOSTIC] First 1000 chars of {xml_file}:")
                    print(xml_content.decode('utf-8')[:1000])
                    print()

                root = ET.fromstring(xml_content)

                # DIAGNOSTIC: Show root tag and namespaces
                print(f"\n  [{xml_file}]")
                print(f"    Root tag: {root.tag}")

                # Get all namespaces
                nsmap = dict([node for _, node in ET.iterparse(BytesIO(xml_content), events=['start-ns'])])
                print(f"    Namespaces: {nsmap}")

                # Show all unique element tags
                all_tags = set([elem.tag for elem in root.iter()])
                clean_tags = [tag.split('}')[-1] if '}' in tag else tag for tag in all_tags]
                print(f"    Elements present ({len(clean_tags)}): {sorted(clean_tags)[:20]}")

                # Try different namespace variations
                namespaces = {
                    'ns': 'urn:iec62325.351:tc57wg16:451-6:transmissiondocument:3:0',
                    'ns2': 'urn:iec62325.351:tc57wg16:451-3:publicationdocument:7:0'
                }
                # Add discovered namespaces
                namespaces.update(nsmap)

                # Find all TimeSeries (NOT Unavailability_TimeSeries!)
                ns_uri = nsmap.get('', None)
                if ns_uri:
                    timeseries_found = root.findall('.//{' + ns_uri + '}TimeSeries')
                else:
                    timeseries_found = root.findall('.//TimeSeries')

                total_timeseries += len(timeseries_found)
                print(f"    TimeSeries found: {len(timeseries_found)}")

                if timeseries_found:
                    print(f"\n  [{xml_file}]")
                    print(f"    Unavailability_TimeSeries found: {len(timeseries_found)}")

                    for i, ts in enumerate(timeseries_found, 1):
                        # Try to find Asset_RegisteredResource (with namespace)
                        if ns_uri:
                            reg_resource = ts.find('.//{' + ns_uri + '}Asset_RegisteredResource')
                        else:
                            reg_resource = ts.find('.//Asset_RegisteredResource')

                        if reg_resource is not None:
                            # Find mRID within Asset_RegisteredResource (with namespace)
                            if ns_uri:
                                mrid_elem = reg_resource.find('.//{' + ns_uri + '}mRID')
                            else:
                                mrid_elem = reg_resource.find('.//mRID')

                            if mrid_elem is not None:
                                eic_code = mrid_elem.text
                                extracted_eics.append(eic_code)
                                print(f"      TimeSeries {i}: RegisteredResource.mRID = {eic_code}")

                                # Check if it matches our CNECs
                                if eic_code in cnec_eics:
                                    cnec_name = cnec_df.filter(pl.col('cnec_eic') == eic_code).select('cnec_name').item(0, 0)
                                    print(f"        >> MATCH! CNEC: {cnec_name}")
                            else:
                                print(f"      TimeSeries {i}: RegisteredResource found but no mRID")
                        else:
                            # Try alternative element names
                            # Check for affected_unit, asset, or other identifiers
                            print(f"      TimeSeries {i}: No RegisteredResource element")

                            # Show structure for debugging
                            elements = [elem.tag for elem in ts.iter()]
                            print(f"        Available elements: {set([tag.split('}')[-1] if '}' in tag else tag for tag in elements[:20]])}")

    print()
    print("="*80)
    print("EXTRACTION RESULTS")
    print("="*80)
    print(f"Total TimeSeries processed: {total_timeseries}")
    print(f"Total EIC codes extracted: {len(extracted_eics)}")
    print(f"Unique EIC codes: {len(set(extracted_eics))}")
    print()

    if extracted_eics:
        # Match against CNEC list
        matches = [eic for eic in set(extracted_eics) if eic in cnec_eics]
        match_rate = len(matches) / len(cnec_eics) * 100

        print(f"CNEC EICs matched: {len(matches)} / {len(cnec_eics)} ({match_rate:.1f}%)")
        print()

        if len(matches) > 0:
            print("[SUCCESS] Asset-specific EIC codes found in XML!")
            print(f"\nMatched CNECs:")
            for eic in matches[:10]:  # Show first 10
                name = cnec_df.filter(pl.col('cnec_eic') == eic).select('cnec_name').item(0, 0)
                print(f"  - {eic}: {name}")
            if len(matches) > 10:
                print(f"  ... and {len(matches) - 10} more")

            print()
            print(f">> Estimated coverage: {match_rate:.1f}% of CNECs")

            if match_rate > 90:
                print(">> EXCELLENT: Can implement 208-feature asset-specific outages")
            elif match_rate > 50:
                print(f">> GOOD: Can implement {len(matches)}-feature asset-specific outages")
            elif match_rate > 20:
                print(f">> PARTIAL: Can implement {len(matches)}-feature outages (limited coverage)")
            else:
                print(">> LIMITED: Few CNECs matched, investigate EIC code format")
        else:
            print("[ISSUE] No CNEC matches found")
            print("Possible reasons:")
            print("  1. EIC codes use different format (JAO vs ENTSO-E)")
            print("  2. Need EIC mapping table")
            print("  3. Transmission elements not individually identified in this border")

        # Show non-matching EICs for investigation
        non_matches = [eic for eic in set(extracted_eics) if eic not in cnec_eics]
        if non_matches:
            print(f"\nNon-matching EIC codes extracted ({len(non_matches)}):")
            for eic in non_matches[:5]:
                print(f"  - {eic}")
            if len(non_matches) > 5:
                print(f"  ... and {len(non_matches) - 5} more")

    else:
        print("[FAIL] No RegisteredResource.mRID elements found in XML")
        print()
        print("Possible reasons:")
        print("  1. Element name is different (affected_unit, asset, etc.)")
        print("  2. EIC codes not included in A78 response")
        print("  3. Need to use different document type")
        print()
        print(">> Fallback: Use border-level outages (20 features)")

except Exception as e:
    print(f"[FAIL] Test 1 failed: {e}")
    import traceback
    traceback.print_exc()

print()

# ============================================================================
# TEST 2: Pumped Storage Consumption Alternative Queries
# ============================================================================

print("-"*80)
print("TEST 2: PUMPED STORAGE CONSUMPTION ALTERNATIVE QUERIES")
print("-"*80)
print()

print("Testing alternative approaches for Switzerland pumped storage consumption...")
print()

# Test 2A: Check if load data separates pumped storage
print("Test 2A: Query total load and check for pumped storage component")
try:
    load_data = client.query_load(
        country_code='CH',
        start=pd.Timestamp('2025-09-23 00:00', tz='UTC'),
        end=pd.Timestamp('2025-09-23 12:00', tz='UTC')
    )

    print(f"[OK] Load data retrieved")
    print(f"  Type: {type(load_data)}")
    print(f"  Columns: {load_data.columns.tolist() if hasattr(load_data, 'columns') else 'N/A (Series)'}")
    print(f"  Sample values: {load_data.head(3).to_dict() if hasattr(load_data, 'to_dict') else load_data.head(3)}")
    print()
    print("  >> No separate pumped storage consumption column visible")

except Exception as e:
    print(f"[FAIL] {e}")

print()

# Test 2B: Try generation with different parameters
print("Test 2B: Check EntsoeRawClient for additional parameters")
try:
    from entsoe import EntsoeRawClient
    raw_client = EntsoeRawClient(api_key=API_KEY)

    # Try with explicit inBiddingZone vs outBiddingZone
    print("  Attempting to query with different zone specifications...")
    print("  (This may help identify consumption vs generation direction)")
    print()
    print("  >> Manual XML parsing approach validated in Phase 1B")
    print("  >> Generation-only solution (7 features) confirmed")

except Exception as e:
    print(f"[FAIL] {e}")

print()

# ============================================================================
# SUMMARY
# ============================================================================

print("="*80)
print("PHASE 1C SUMMARY")
print("="*80)
print()

print("TEST 1: Asset-Specific Transmission Outages")
print("  Approach: Parse RegisteredResource.mRID from border-level query XML")
print("  Result: [See above]")
print()

print("TEST 2: Pumped Storage Consumption")
print("  Approach: Alternative queries for consumption data")
print("  Result: Generation-only confirmed (7 features)")
print("  Alternative: May need to infer from generation patterns or accept limitation")
print()

print("="*80)
print("NEXT STEPS:")
print("1. Review match rate for asset-specific outages")
print("2. Decide on implementation approach based on coverage")
print("3. Proceed to Phase 2 with enhanced XML parsing if successful")
print("="*80)