dc-l3-vienna-overpass-historical

Rerun

Installing collected packages: osm2geojson, overpass
  WARNING: The script osm2geojson is installed in '/home/agent/.local/bin' which is not on PATH.
  Consider adding this directory to PATH or, if you prefer to suppress this warning, use --no-warn-script-location.

Successfully installed osm2geojson-0.3.2 overpass-0.8.2

cat << 'PYEOF' > /work/solve.py
#!/usr/bin/env python3
"""
Fetch Vienna district boundaries (current + 2014-01-01) from Overpass,
compute changes, and output vienna_boundary_changes.geojson.
"""

import json
import sys
import re
import os
from collections import defaultdict

import geopandas as gpd
import pandas as pd
from shapely.geometry import shape, MultiPolygon, Polygon
from shapely.ops import unary_union
from shapely import wkt

# ---------------------------------------------------------------------------
# 1. Query Overpass for current and 2014-01-01 Vienna district (Bezirk) relations
# ---------------------------------------------------------------------------
# Vienna districts are relations with boundary=administrative, admin_level=9 or 10,
# name: "Wien" + district number + name.
# We'll use [out:json][timeout:300] and fetch via relation search.

OVERPASS_URL = "https://overpass-api.de/api/interpreter"

def build_overpass_query(bbox=None, date_prefix=""):
    """
    Build Overpass QL to get all admin_level=9 or 10 boundary relations
    in Vienna (rough bounding box). If date_prefix is not empty, we use
    [date:"2014-01-01T00:00:00Z"].
    """
    date_filter = ""
    if date_prefix:
        date_filter = f'[date:"{date_prefix}"]'

    # Bounding box covering Vienna (approx)
    # Vienna: lat ~48.1-48.35, lon ~16.2-16.6
    # We use a slightly larger box
    bbox_str = "48.05,16.15,48.40,16.70"
    
    query = f"""
[out:json][timeout:300]{date_filter};
(
  relation["boundary"="administrative"]["admin_level"="9"]({bbox_str});
  relation["boundary"="administrative"]["admin_level"="10"]({bbox_str});
);
out body geom;
"""
    return query

def fetch_overpass(query):
    """Fetch data from Overpass API."""
    import urllib.request
    import urllib.parse
    
    data = urllib.parse.urlencode({'data': query}).encode()
    req = urllib.request.Request(OVERPASS_URL, data=data,
                                 headers={'User-Agent': 'GIS-Analysis/1.0'})
    try:
        with urllib.request.urlopen(req, timeout=300) as resp:
            return json.loads(resp.read())
    except Exception as e:
        print(f"WARNING: Overpass error: {e}", file=sys.stderr)
        return None

def normalise_name(name):
    """
    Normalise Vienna district names to a canonical form.
    Examples:
      "Wien 1., Innere Stadt" -> "wien-1-innere-stadt"
      "Wien 1., Innere Stadt (1. Bezirk)" -> "wien-1-innere-stadt"
      "1., Innere Stadt" -> "wien-1-innere-stadt"
      "Wien 1. Bezirk, Innere Stadt" -> "wien-1-innere-stadt"
      "Innere Stadt" -> "wien-1-innere-stadt" (if we can infer number)
    Strategy: extract district number, then normalise the name.
    """
    if not name:
        return None
    
    # Extract district number
    num = None
    num_match = re.search(r'(\d+)\.', name)
    if num_match:
        num = int(num_match.group(1))
    
    # Also try "XX. Bezirk" or "XX.Bezirk"
    if num is None:
        num_match = re.search(r'(\d+)\s*\.?\s*Bezirk', name)
        if num_match:
            num = int(num_match.group(1))
    
    # Remove parenthetical remarks, "Wien" prefix, "Bezirk" suffix, etc.
    clean = name
    # Remove parentheses content
    clean = re.sub(r'\([^)]*\)', '', clean)
    # Remove leading "Wien" (with optional space/comma/dot)
    clean = re.sub(r'^Wien[\s,.]*', '', clean)
    # Remove "Bezirk" (with optional preceding number)
    clean = re.sub(r'\d*\s*\.?\s*Bezirk\s*', '', clean)
    # Remove leading numbers with dots
    clean = re.sub(r'^\d+\s*\.\s*', '', clean)
    # Clean whitespace
    clean = clean.strip()
    
    if num is not None:
        return f"wien-{num:02d}-{clean.lower().replace(' ', '-')}"
    elif clean:
        return f"wien-{clean.lower().replace(' ', '-')}"
    return name.lower().replace(' ', '-')

def extract_polygons_from_overpass(data):
    """Extract district polygons from Overpass response JSON."""
    features = []
    if not data or 'elements' not in data:
        return features
    
    for elem in data['elements']:
        if elem.get('type') != 'relation':
            continue
        
        tags = elem.get('tags', {})
        name = tags.get('name', '')
        if not name:
            continue
        
        # Check if it's a Vienna district (admin_level 9 or 10)
        admin_level = tags.get('admin_level', '')
        if admin_level not in ('9', '10'):
            continue
        
        # Build geometry from members
        members = elem.get('members', [])
        
        # Collect all way geometries from the relation
        # For simplicity, we build the geometry from the 'geometry' field if available,
        # or from the ways' node coordinates
        polys = []
        for member in members:
            if member.get('type') == 'way' and 'geometry' in member:
                geom_type = member.get('type')
                coords = [(p['lon'], p['lat']) for p in member['geometry']]
                if len(coords) >= 4:  # At least 4 points for a ring
                    polys.append(Polygon(coords))
            elif member.get('type') == 'way' and 'nodes' in member:
                # Geometry might be nested - skip, we rely on geometry field
                pass
        
        # Also check if the element itself has a geometry
        if 'geometry' in elem:
            coords = [(p['lon'], p['lat']) for p in elem['geometry']]
            # This might be a closed way geometry for the whole relation? 
            # Actually for relations, geometry is not directly on the element
            pass
        
        if polys:
            # Try to form a single polygon or multipolygon
            if len(polys) == 1:
                geom = polys[0]
            else:
                geom = unary_union(polys)
                if geom.geom_type == 'GeometryCollection':
                    # Filter to polygons
                    geom = unary_union([g for g in geom.geoms if g.geom_type in ('Polygon', 'MultiPolygon')])
                if geom.geom_type not in ('Polygon', 'MultiPolygon'):
                    continue
            
            normalised = normalise_name(name)
            features.append({
                'geometry': geom,
                'name': name,
                'normalised': normalised,
                'admin_level': admin_level
            })
    
    return features

# ---------------------------------------------------------------------------
# 2. Better approach: Use the osmium/overpass approach with raw data
# We'll use a simpler Overpass query that gets the relation's way geometries
# properly
# ---------------------------------------------------------------------------

def fetch_vienna_districts_alt(date_prefix=""):
    """
    More robust query: get all relations with name matching Vienna districts,
    then get their members with complete geometry.
    """
    date_filter = ""
    if date_prefix:
        date_filter = f'[date:"{date_prefix}"]'
    
    # bbox of Vienna
    bbox = "48.05,16.15,48.40,16.70"
    
    query = f"""
[out:json][timeout:300]{date_filter};
(
  relation["admin_level"="9"]({bbox});
  relation["admin_level"="10"]({bbox});
);
out body;
>;
out geom;
"""
    return fetch_overpass(query)

def process_vienna_data(data):
    """
    Process Overpass response to extract district boundaries.
    Returns list of (normalised_name, geometry, raw_name, admin_level)
    """
    if not data or 'elements' not in data:
        return []
    
    # Collect all ways with their geometries
    ways = {}
    for elem in data['elements']:
        if elem.get('type') == 'way':
            if 'geometry' in elem:
                coords = [(p['lon'], p['lat']) for p in elem['geometry']]
                if len(coords) >= 2:
                    ways[elem['id']] = coords
    
    # Collect relations that are districts
    districts = []
    
    for elem in data['elements']:
        if elem.get('type') != 'relation':
            continue
        
        tags = elem.get('tags', {})
        name = tags.get('name', '')
        admin_level = tags.get('admin_level', '')
        
        if not name or admin_level not in ('9', '10'):
            continue
        
        # Check if it's a Vienna district - name should contain "Wien" or be a known district
        if 'wien' not in name.lower():
            # Also check for just a number like "1., Innere Stadt"
            if not re.match(r'^\d+\.', name):
                continue
        
        # Build polygon from member ways
        members = elem.get('members', [])
        outer_rings = []
        inner_rings = []
        
        for member in members:
            if member.get('type') != 'way':
                continue
            way_id = member.get('ref')
            if way_id not in ways:
                continue
            
            coords = ways[way_id]
            role = member.get('role', '')
            
            if len(coords) >= 4:  # need at least 4 points to form polygon
                poly = Polygon(coords)
                if role == 'outer':
                    outer_rings.append(poly)
                elif role == 'inner':
                    inner_rings.append(poly)
                else:
                    # Default to outer if no role specified
                    outer_rings.append(poly)
        
        if not outer_rings:
            continue
        
        # Build the geometry
        if len(outer_rings) == 1:
            geom = Polygon(outer_rings[0].exterior.coords)
            # Add holes
            for inner in inner_rings:
                try:
                    geom = Polygon(geom.exterior.coords, [inner.exterior.coords])
                except:
                    pass
        else:
            # MultiPolygon from multiple outer rings
            polygons = []
            for outer in outer_rings:
                rings = [outer.exterior.coords]
                # Associate inner rings with this outer if they're inside
                remaining_inner = []
                for inner in inner_rings:
                    if outer.contains(inner.representative_point()):
                        rings.append(inner.exterior.coords)
                    else:
                        remaining_inner.append(inner)
                inner_rings = remaining_inner
                try:
                    polygons.append(Polygon(rings[0], rings[1:]))
                except:
                    polygons.append(Polygon(rings[0]))
            if polygons:
                geom = MultiPolygon(polygons) if len(polygons) > 1 else polygons[0]
            else:
                continue
        
        normalised = normalise_name(name)
        if normalised:
            districts.append({
                'normalised': normalised,
                'geometry': geom,
                'raw_name': name,
                'admin_level': admin_level
            })
    
    return districts

# ---------------------------------------------------------------------------
# 3. Main processing
# ---------------------------------------------------------------------------

def main():
    print("Fetching current Vienna districts from Overpass...", file=sys.stderr)
    current_data = fetch_vienna_districts_alt(date_prefix="")
    
    print("Fetching 2014-01-01 Vienna districts from Overpass...", file=sys.stderr)
    old_data = fetch_vienna_districts_alt(date_prefix="2014-01-01T00:00:00Z")
    
    if not current_data or not old_data:
        print("ERROR: Failed to fetch data from Overpass. Check connection.", file=sys.stderr)
        sys.exit(1)
    
    current_districts = process_vienna_data(current_data)
    old_districts = process_vienna_data(old_data)
    
    print(f"  Current districts: {len(current_districts)}", file=sys.stderr)
    print(f"  2014 districts: {len(old_districts)}", file=sys.stderr)
    
    if not current_districts or not old_districts:
        print("ERROR: No districts found.", file=sys.stderr)
        sys.exit(1)
    
    for d in current_districts:
        print(f"  Current: {d['raw_name']:40s} -> {d['normalised']}", file=sys.stderr)
    for d in old_districts:
        print(f"  2014:    {d['raw_name']:40s} -> {d['normalised']}", file=sys.stderr)
    
    # Match districts by normalised name
    current_by_name = {d['normalised']: d for d in current_districts}
    old_by_name = {d['normalised']: d for d in old_districts}
    
    common_names = set(current_by_name.keys()) & set(old_by_name.keys())
    only_current_names = set(current_by_name.keys()) - set(old_by_name.keys())
    only_old_names = set(old_by_name.keys()) - set(current_by_name.keys())
    
    print(f"  Common districts: {len(common_names)}", file=sys.stderr)
    print(f"  Only current: {only_current_names}", file=sys.stderr)
    print(f"  Only 2014: {only_old_names}", file=sys.stderr)
    
    if not common_names:
        print("ERROR: No matching districts between snapshots!", file=sys.stderr)
        sys.exit(1)
    
    # Compute changes for each district
    change_features = []
    
    # We'll work with all districts' geometries
    # For each district, compute:
    #   added = current_geom - old_geom (parts in current not in 2014)
    #   removed = old_geom - current_geom (parts in 2014 not in current)
    #   unchanged = current_geom ∩ old_geom
    
    for name in common_names:
        cur = current_by_name[name]
        old = old_by_name[name]
        
        cur_geom = cur['geometry']
        old_geom = old['geometry']
        
        # Ensure both are valid
        if not cur_geom.is_valid:
            cur_geom = cur_geom.buffer(0)
        if not old_geom.is_valid:
            old_geom = old_geom.buffer(0)
        
        # Added: in current but not in old
        added = cur_geom.difference(old_geom)
        # Removed: in old but not in current
        removed = old_geom.difference(cur_geom)
        # Unchanged: intersection
        unchanged = cur_geom.intersection(old_geom)
        
        # Add features for each non-empty result
        for geom, ctype in [(added, 'added_since_2014'), (removed, 'removed_since_2014'), (unchanged, 'unchanged')]:
            if geom.is_empty:
                continue
            if geom.geom_type == 'GeometryCollection':
                # Extract only polygons
                polys = [g for g in geom.geoms if g.geom_type in ('Polygon', 'MultiPolygon')]
                if not polys:
                    continue
                geom = unary_union(polys)
            if geom.geom_type not in ('Polygon', 'MultiPolygon'):
                continue
            if geom.area < 1e-10:  # very small, skip
                continue
            
            change_features.append({
                'geometry': geom,
                'district_name': name,
                'change_type': ctype
            })
    
    # For districts only in one snapshot, their entire geometry is added/removed
    for name in only_current_names:
        cur = current_by_name[name]
        geom = cur['geometry']
        if not geom.is_valid:
            geom = geom.buffer(0)
        if geom.area < 1e-10:
            continue
        change_features.append({
            'geometry': geom,
            'district_name': name,
            'change_type': 'added_since_2014'
        })
    
    for name in only_old_names:
        old = old_by_name[name]
        geom = old['geometry']
        if not geom.is_valid:
            geom = geom.buffer(0)
        if geom.area < 1e-10:
            continue
        change_features.append({
            'geometry': geom,
            'district_name': name,
            'change_type': 'removed_since_2014'
        })
    
    print(f"  Total change fragments: {len(change_features)}", file=sys.stderr)
    
    if not change_features:
        print("No changes detected.", file=sys.stderr)
        # Still output a file with an empty feature collection
        result = {"type": "FeatureCollection", "features": []}
        with open('/work/vienna_boundary_changes.geojson', 'w') as f:
            json.dump(result, f)
        print("Written empty vienna_boundary_changes.geojson", file=sys.stderr)
        return
    
    # Now determine touches_changed for each district
    # For each district, check if any of its fragments (or the district's full geometry)
    # borders any area that was added or removed.
    
    # Build a GeoDataFrame of change features
    gdf_changes = gpd.GeoDataFrame(change_features, crs='EPSG:4326')
    
    # Collect all "changed" (added or removed) geometries
    changed_mask = gdf_changes['change_type'].isin(['added_since_2014', 'removed_since_2014'])
    unchanged_mask = gdf_changes['change_type'] == 'unchanged'
    
    # For each district, check if its boundaries touch any changed area
    # We need to check if the district's geometry (current or old) touches any changed area
    
    # Build a mapping of district_name -> touches_changed
    touches_changed_map = {}
    
    # Get all unique district names
    all_district_names = set(f['district_name'] for f in change_features)
    
    # Build changed geometry union
    changed_geoms = gdf_changes.loc[changed_mask, 'geometry'].tolist()
    if changed_geoms:
        changed_union = unary_union(changed_geoms)
    else:
        changed_union = None
    
    # For each district, check if its boundary touches any changed area
    for dname in all_district_names:
        touches = False
        # Get this district's fragments
        district_frags = gdf_changes[gdf_changes['district_name'] == dname]
        
        for _, row in district_frags.iterrows():
            frag_geom = row['geometry']
            row_type = row['change_type']
            
            # Check if this fragment touches any changed area
            if changed_union is not None and not changed_union.is_empty:
                if frag_geom.touches(changed_union):
                    touches = True
                    break
                # Also check if they intersect at boundaries (buffer by 1mm)
                if frag_geom.buffer(1e-6).intersects(changed_union):
                    touches = True
                    break
        
        touches_changed_map[dname] = touches
    
    # Build final features
    features_out = []
    for _, row in gdf_changes.iterrows():
        geom = row['geometry']
        if geom.geom_type == 'Polygon':
            geom = MultiPolygon([geom])
        
        feat = {
            'type': 'Feature',
            'geometry': json.loads(gpd.GeoSeries([geom]).to_json()),
            'properties': {
                'change_type': row['change_type'],
                'district_name': row['district_name'],
                'touches_changed': touches_changed_map.get(row['district_name'], False)
            }
        }
        # The GeoJSON geometry might be nested, fix it
        features_out.append(feat)
    
    # Fix the geometry serialization
    result = {
        'type': 'FeatureCollection',
        'features': []
    }
    
    for i, (_, row) in enumerate(gdf_changes.iterrows()):
        geom = row['geometry']
        if geom.geom_type == 'Polygon':
            geom = MultiPolygon([geom])
        
        # Use shapely directly for serialization
        from shapely.geometry import mapping
        result['features'].append({
            'type': 'Feature',
            'geometry': mapping(geom),
            'properties': {
                'change_type': row['change_type'],
                'district_name': row['district_name'],
                'touches_changed': touches_changed_map.get(row['district_name'], False)
            }
        })
    
    with open('/work/vienna_boundary_changes.geojson', 'w') as f:
        json.dump(result, f, indent=2)
    
    print(f"Written /work/vienna_boundary_changes.geojson with {len(result['features'])} features", file=sys.stderr)
    
    # Summary
    print("\nSummary:", file=sys.stderr)
    for ctype in ['added_since_2014', 'removed_since_2014', 'unchanged']:
        count = sum(1 for f in result['features'] if f['properties']['change_type'] == ctype)
        print(f"  {ctype}: {count} features", file=sys.stderr)
    
    print("\nDistricts touching changes:", file=sys.stderr)
    for dname in sorted(all_district_names):
        tc = touches_changed_map[dname]
        print(f"  {dname}: touches_changed={tc}", file=sys.stderr)


if __name__ == '__main__':
    main()
PYEOF
python3 /work/solve.py

Fetching current Vienna districts from Overpass...
Fetching 2014-01-01 Vienna districts from Overpass...
  Current districts: 3
  2014 districts: 1
  Current: Katastralgemeinde Sulz im Wienerwald     -> wien-katastralgemeinde-sulz-im-wienerwald
  Current: Katastralgemeinde Wienerherberg          -> wien-katastralgemeinde-wienerherberg
  Current: Katastralgemeinde Wiener Neudorf         -> wien-katastralgemeinde-wiener-neudorf
  2014:    Gerasdorf bei Wien                       -> wien-gerasdorf-bei-wien
  Common districts: 0
  Only current: {'wien-katastralgemeinde-wienerherberg', 'wien-katastralgemeinde-sulz-im-wienerwald', 'wien-katastralgemeinde-wiener-neudorf'}
  Only 2014: {'wien-gerasdorf-bei-wien'}
ERROR: No matching districts between snapshots!

cat << 'PYEOF' > /work/debug_query.py
#!/usr/bin/env python3
"""Debug query to find proper tags for Vienna districts."""
import json, urllib.request, urllib.parse, sys

OVERPASS_URL = "https://overpass-api.de/api/interpreter"

# First, let's find Vienna's city boundary and its sub-areas
# Vienna (city) has osm_id 109166 (relation)
query = """
[out:json][timeout:300];
(
  area["name"="Wien"]["admin_level"="4"];
  area["name"="Wien"]["admin_level"="5"];
);
out body;
>; 
out skel qt;
"""
data = urllib.parse.urlencode({'data': query}).encode()
req = urllib.request.Request(OVERPASS_URL, data=data, headers={'User-Agent': 'GIS/1.0'})
with urllib.request.urlopen(req, timeout=120) as resp:
    result = json.loads(resp.read())

print("Elements from area query:")
for e in result.get('elements', []):
    print(f"  type={e.get('type')} id={e.get('id')} tags={e.get('tags', {})}")

# Let's try a different approach - get Vienna city relation
query2 = """
[out:json][timeout:300];
(
  rel["name:en"="Vienna"]["admin_level"="5"];
  rel["name"="Wien"]["admin_level"="5"];
  rel["name"="Wien"]["admin_level"="4"];
);
out body;
"""
data2 = urllib.parse.urlencode({'data': query2}).encode()
req2 = urllib.request.Request(OVERPASS_URL, data=data2, headers={'User-Agent': 'GIS/1.0'})
with urllib.request.urlopen(req2, timeout=120) as resp:
    result2 = json.loads(resp.read())

print("\nVienna city relations:")
for e in result2.get('elements', []):
    print(f"  type={e.get('type')} id={e.get('id')} tags={json.dumps(e.get('tags', {}), indent=4)}")
    print(f"  members count: {len(e.get('members', []))}")
PYEOF
python3 /work/debug_query.py

Traceback (most recent call last):
  File "/work/debug_query.py", line 40, in <module>
    with urllib.request.urlopen(req2, timeout=120) as resp:
         ~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.14/urllib/request.py", line 187, in urlopen
    return opener.open(url, data, timeout)
           ~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.14/urllib/request.py", line 493, in open
    response = meth(req, response)
  File "/usr/local/lib/python3.14/urllib/request.py", line 602, in http_response
    response = self.parent.error(
        'http', request, response, code, msg, hdrs)
  File "/usr/local/lib/python3.14/urllib/request.py", line 531, in error
    return self._call_chain(*args)
           ~~~~~~~~~~~~~~~~^^^^^^^
  File "/usr/local/lib/python3.14/urllib/request.py", line 464, in _call_chain
    result = func(*args)
  File "/usr/local/lib/python3.14/urllib/request.py", line 611, in http_error_default
    raise HTTPError(req.full_url, code, msg, hdrs, fp)
urllib.error.HTTPError: HTTP Error 429: Too Many Requests

cat << 'PYEOF' > /work/fetch.py
#!/usr/bin/env python3
"""Fetch Vienna districts from Overpass, save to GeoJSON files."""
import json, sys, time

try:
    import overpass
    api = overpass.API(timeout=300)
    
    # Vienna bounding box
    bbox = "48.05,16.15,48.40,16.70"
    
    # Query for current Vienna districts (Bezirke)
    # In Vienna, districts are admin_level=10 (Gemeindebezirk)
    # The city of Vienna itself is admin_level=5 or 4 (Statutarstadt)
    query_current = f"""
    area["name"="Wien"]["admin_level"="4"]->.wien;
    (
      rel(area.wien)["admin_level"="10"]["boundary"="administrative"];
      rel(area.wien)["admin_level"="9"]["boundary"="administrative"];
    );
    out body geom;
    """
    
    print("Querying current districts...", file=sys.stderr)
    current_data = api.get(query_current, verbosity='body geom')
    with open('/work/current_raw.json', 'w') as f:
        json.dump(current_data, f)
    print(f"Saved current_raw.json", file=sys.stderr)
    
    # Wait a bit to avoid rate limiting
    time.sleep(2)
    
    # Query for 2014-01-01 districts
    query_old = f"""
    [date:"2014-01-01T00:00:00Z"]
    area["name"="Wien"]["admin_level"="4"]->.wien;
    (
      rel(area.wien)["admin_level"="10"]["boundary"="administrative"];
      rel(area.wien)["admin_level"="9"]["boundary"="administrative"];
    );
    out body geom;
    """
    
    print("Querying 2014-01-01 districts...", file=sys.stderr)
    old_data = api.get(query_old, verbosity='body geom')
    with open('/work/old_raw.json', 'w') as f:
        json.dump(old_data, f)
    print(f"Saved old_raw.json", file=sys.stderr)
    
except Exception as e:
    print(f"Error: {e}", file=sys.stderr)
    # Check if files exist already
    import os
    for fn in ['/work/current_raw.json', '/work/old_raw.json']:
        if os.path.exists(fn):
            print(f"{fn} exists ({os.path.getsize(fn)} bytes)", file=sys.stderr)
        else:
            print(f"{fn} does not exist", file=sys.stderr)
    sys.exit(1)
PYEOF
python3 /work/fetch.py

/work/fetch.py:7: DeprecationWarning: overpass.API is deprecated and will be removed in overpass 1.0; use overpass.OverpassClient in 1.0 instead.
  api = overpass.API(timeout=300)
Querying current districts...
/work/fetch.py:25: DeprecationWarning: API.get(verbosity=...) is deprecated and will be replaced by OverpassClient.query(..., out=...) in overpass 1.0.
  current_data = api.get(query_current, verbosity='body geom')
Saved current_raw.json
Querying 2014-01-01 districts...
/work/fetch.py:45: DeprecationWarning: API.get(verbosity=...) is deprecated and will be replaced by OverpassClient.query(..., out=...) in overpass 1.0.
  old_data = api.get(query_old, verbosity='body geom')
Error: [out:json];
    [date:"2014-01-01T00:00:00Z"]
    area["name"="Wien"]["admin_level"="4"]->.wien;
    (
      rel(area.wien)["admin_level"="10"]["boundary"="administrative"];
      rel(area.wien)["admin_level"="9"]["boundary"="administrative"];
    );
    out body geom;out body geom;
/work/current_raw.json exists (1688261 bytes)
/work/old_raw.json does not exist

cat << 'PYEOF' > /work/fetch2.py
#!/usr/bin/env python3
"""Fetch Vienna districts from Overpass using raw HTTP requests."""
import json, urllib.request, urllib.parse, sys, time

OVERPASS = "https://overpass-api.de/api/interpreter"

def run_query(query, label=""):
    data = urllib.parse.urlencode({'data': query}).encode()
    req = urllib.request.Request(OVERPASS, data=data, 
                                 headers={'User-Agent': 'GIS-Analysis/1.0'})
    try:
        with urllib.request.urlopen(req, timeout=300) as resp:
            return json.loads(resp.read())
    except urllib.error.HTTPError as e:
        print(f"HTTP Error {e.code} for {label}: {e.read().decode()[:200]}", file=sys.stderr)
        return None
    except Exception as e:
        print(f"Error for {label}: {e}", file=sys.stderr)
        return None

# Query for current Vienna districts (admin_level 10 = Gemeindebezirk)
# We find the Vienna area first, then filter by admin_level 9 or 10
query_current = """
[out:json][timeout:300];
area["name"="Wien"]["admin_level"="4"]->.a;
(
  relation(area.a)["admin_level"="10"]["boundary"="administrative"];
  relation(area.a)["admin_level"="9"]["boundary"="administrative"];
);
out body;
>;
out geom;
"""

print("Fetching current data...", file=sys.stderr)
current = run_query(query_current, "current")
if current:
    with open('/work/current_raw.json', 'w') as f:
        json.dump(current, f)
    print(f"Saved current_raw.json ({len(current.get('elements',[]))} elements)", file=sys.stderr)
else:
    print("Failed to fetch current data", file=sys.stderr)

time.sleep(3)

query_old = """
[out:json][timeout:300][date:"2014-01-01T00:00:00Z"];
area["name"="Wien"]["admin_level"="4"]->.a;
(
  relation(area.a)["admin_level"="10"]["boundary"="administrative"];
  relation(area.a)["admin_level"="9"]["boundary"="administrative"];
);
out body;
>;
out geom;
"""

print("Fetching 2014-01-01 data...", file=sys.stderr)
old = run_query(query_old, "2014")
if old:
    with open('/work/old_raw.json', 'w') as f:
        json.dump(old, f)
    print(f"Saved old_raw.json ({len(old.get('elements',[]))} elements)", file=sys.stderr)
else:
    print("Failed to fetch old data", file=sys.stderr)
PYEOF
python3 /work/fetch2.py

Fetching current data...
Saved current_raw.json (12713 elements)
Fetching 2014-01-01 data...
Saved old_raw.json (5413 elements)

cat << 'PYEOF' > /work/inspect_data.py
import json, sys

with open('/work/current_raw.json') as f:
    cur = json.load(f)

with open('/work/old_raw.json') as f:
    old = json.load(f)

def list_relations(data, label):
    print(f"\n=== {label} relations ===")
    for e in data.get('elements', []):
        if e.get('type') == 'relation':
            tags = e.get('tags', {})
            name = tags.get('name', '')
            admin = tags.get('admin_level', '')
            boundary = tags.get('boundary', '')
            if admin in ('9', '10') and boundary == 'administrative':
                print(f"  id={e['id']} admin={admin} name='{name}'")
                # Also print other relevant tags
                for k,v in sorted(tags.items()):
                    if k not in ('name', 'admin_level', 'boundary', 'type'):
                        print(f"    {k}={v}")

list_relations(cur, "CURRENT")
list_relations(old, "OLD (2014)")

print("\n\n=== Checking for name differences ===")
cur_names = set()
for e in cur.get('elements', []):
    if e.get('type') == 'relation':
        tags = e.get('tags', {})
        if tags.get('admin_level', '') in ('9', '10') and tags.get('boundary') == 'administrative':
            cur_names.add(tags.get('name', ''))

old_names = set()
for e in old.get('elements', []):
    if e.get('type') == 'relation':
        tags = e.get('tags', {})
        if tags.get('admin_level', '') in ('9', '10') and tags.get('boundary') == 'administrative':
            old_names.add(tags.get('name', ''))

print("Current names:")
for n in sorted(cur_names):
    print(f"  '{n}'")
print("\nOld names (2014):")
for n in sorted(old_names):
    print(f"  '{n}'")
PYEOF
python3 /work/inspect_data.py

=== CURRENT relations ===
  id=1744661 admin=10 name='Katastralgemeinde Mauer'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Mauer
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1806
    wikidata=Q1910986
    wikipedia=de:Mauer (Wien)
  id=1990590 admin=9 name='Alsergrund'
    name:de=Alsergrund
    name:ru=Альзергрунд
    name:uk=Альзерґрунд
    population=39322
    ref=9
    ref:at:gkz=90901
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Alsergrund
    wikidata=Q257780
    wikipedia=de:Alsergrund
  id=1990591 admin=9 name='Hietzing'
    admin_title=Bezirk
    designation=Bezirk
    name:ru=Хитцинг
    name:uk=Гітцинґ
    official_status=Bezirk
    ref=13
    ref:at:gkz=91301
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Hietzing
    wikidata=Q636910
    wikipedia=de:Hietzing
  id=1990592 admin=9 name='Innere Stadt'
    name:uk=Внутрішнє місто
    population=17135
    ref=1
    ref:at:gkz=90101
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Innere Stadt
    wikidata=Q267329
    wikipedia=de:Innere Stadt (Wien)
  id=1990593 admin=9 name='Josefstadt'
    name:de=Josefstadt
    name:ru=Йозефштадт
    name:uk=Йозефштадт
    old_name=Josephstadt
    population=23389
    ref=8
    ref:at:gkz=90801
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Josefstadt
    wikidata=Q256778
    wikipedia=de:Josefstadt (Wien)
  id=1990594 admin=9 name='Leopoldstadt'
    name:ru=Леопольдштадт
    name:uk=Леопольдштадт
    ref=2
    ref:at:gkz=90201
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Leopoldstadt
    wikidata=Q259120
    wikipedia=de:Leopoldstadt
  id=1990595 admin=9 name='Mariahilf'
    admin_title=Bezirk
    coat_of_arms=File:Wien - Bezirk Mariahilf, Wappen.svg
    designation=Bezirk
    name:de=Mariahilf
    name:etymology:wikidata=Q1533009
    name:ru=Мариахильф
    name:uk=Маріагільф
    official_status=Bezirk
    population=28747
    ref=6
    ref:at:gkz=90601
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Mariahilf
    wikidata=Q654812
    wikipedia=de:Mariahilf
  id=1990596 admin=9 name='Meidling'
    admin_title=Bezirk
    designation=Bezirk
    name:ru=Майдлинг
    name:uk=Майдлінг
    official_status=Bezirk
    ref=12
    ref:at:gkz=91201
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Meidling
    wikidata=Q697557
    wikipedia=de:Meidling
  id=1990597 admin=9 name='Neubau'
    name:de=Neubau
    name:ru=Нойбау
    name:uk=Нойбау
    population=28823
    ref=7
    ref:at:gkz=90701
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Neubau
    wikidata=Q664835
    wikipedia=de:Neubau (Wien)
  id=1990598 admin=9 name='Penzing'
    name:ru=Пенцинг
    name:uk=Пенцінг
    population=89184
    ref=14
    ref:at:gkz=91401
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Penzing
    wikidata=Q690335
    wikipedia=de:Penzing (Wien)
  id=1990599 admin=9 name='Rudolfsheim-Fünfhaus'
    name:ru=Рудольфсхайм-Фюнфхаус
    name:uk=Рудольфсгайм-Фюнфгаус
    population=69078
    ref=15
    ref:at:gkz=91501
    ref:at:okz=17237
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Rudolfsheim-Fünfhaus
    wikidata=Q653289
    wikipedia=de:Rudolfsheim-Fünfhaus
  id=1990600 admin=9 name='Währing'
    name:en=Währing
    name:ru=Вэринг
    name:uk=Верінґ
    population=46622
    ref=18
    ref:at:gkz=91801
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Währing
    wikidata=Q690417
    wikipedia=de:Währing
  id=1991416 admin=9 name='Landstraße'
    admin_title=Bezirk
    designation=Bezirk
    name:ru=Ландштрассе
    name:uk=Ландштрассе
    official_status=Bezirk
    population=87953
    ref=3
    ref:at:gkz=90301
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Landstraße
    wikidata=Q519552
    wikipedia=de:Landstraße (Wien)
    wikipedia:ru=Ландштрассе
  id=1991433 admin=9 name='Brigittenau'
    name:ru=Бригиттенау
    name:uk=Бріґіттенау
    population=85380
    ref=20
    ref:at:gkz=92001
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Brigittenau
    wikidata=Q265655
    wikipedia=de:Brigittenau
  id=1991434 admin=9 name='Donaustadt'
    name:ru=Донауштадт
    population=144289
    ref=22
    ref:at:gkz=92201
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Donaustadt
    wikidata=Q258333
    wikipedia=de:Donaustadt
  id=1991435 admin=9 name='Döbling'
    name:ru=Дёблинг
    name:uk=Дьоблінґ
    population=66487
    ref=19
    ref:at:gkz=91901
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Döbling
    wikidata=Q267360
    wikipedia=de:Döbling
  id=1991436 admin=9 name='Favoriten'
    admin_title=Bezirk
    designation=Bezirk
    name:ru=Фаворитен
    name:uk=Фаворітен
    official_status=Bezirk
    population=189543
    ref=10
    ref:at:gkz=91001
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Favoriten
    wikidata=Q697548
    wikipedia=de:Favoriten
  id=1991437 admin=9 name='Floridsdorf'
    name:ru=Флоридсдорф
    population=151740
    ref=21
    ref:at:gkz=92101
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Floridsdorf
    wikidata=Q264645
    wikipedia=de:Floridsdorf
  id=1991438 admin=9 name='Hernals'
    name:en=Hernals
    name:ru=Хернальс
    name:uk=Гернальс
    population=50866
    ref=17
    ref:at:gkz=91701
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Hernals
    wikidata=Q262808
    wikipedia=de:Hernals
  id=1991439 admin=9 name='Liesing'
    admin_title=Bezirk
    designation=Bezirk
    name:ru=Лизинг
    name:uk=Лізінґ
    official_status=Bezirk
    population=96728
    ref=23
    ref:at:gkz=92301
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Liesing
    wikidata=Q278900
    wikipedia=de:Liesing (Wien)
  id=1991440 admin=9 name='Margareten'
    admin_title=Bezirk
    designation=Bezirk
    name:de=Margareten
    name:ru=Маргаретен
    name:uk=Марґаретен
    official_status=Bezirk
    population=51434
    ref=5
    ref:at:gkz=90501
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Margareten
    wikidata=Q252731
    wikipedia=de:Margareten
  id=1991441 admin=9 name='Ottakring'
    name:ru=Оттакринг
    name:uk=Оттакрінг
    population=100520
    ref=16
    ref:at:gkz=91601
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Ottakring
    wikidata=Q265945
    wikipedia=de:Ottakring
  id=1991442 admin=9 name='Simmering'
    admin_title=Bezirk
    designation=Bezirk
    name:ru=Зиммеринг
    official_status=Bezirk
    population=95162
    ref=11
    ref:at:gkz=91101
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Simmering
    wikidata=Q279028
    wikipedia=de:Simmering
  id=1991443 admin=9 name='Wieden'
    admin_title=Bezirk
    designation=Bezirk
    name:de=Wieden
    name:ru=Виден
    name:uk=Віден
    official_status=Bezirk
    population=29356
    ref=4
    ref:at:gkz=90401
    street_names:wikipedia=de:Liste der Straßennamen von Wien/Wieden
    wikidata=Q268001
    wikipedia=de:Wieden (Wien)
  id=5234381 admin=10 name='Katastralgemeinde Inzersdorf Stadt'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Inzersdorf Stadt
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1102
    wikidata=Q1671846
    wikipedia=de:Inzersdorf Stadt
  id=5241410 admin=10 name='Katastralgemeinde Favoriten'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Favoriten
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1101
    wikidata=Q1398933
    wikipedia=de:Favoriten (Wiener Bezirksteil)
  id=5241797 admin=10 name='Katastralgemeinde Oberlaa Stadt'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Oberlaa Stadt
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1105
    related:wikipedia=de:Oberlaa
    wikidata=Q27832649
  id=5241911 admin=10 name='Katastralgemeinde Rothneusiedl'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Rothneusiedl
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1106
    wikidata=Q1586529
    wikipedia=de:Rothneusiedl
  id=5241912 admin=10 name='Katastralgemeinde Oberlaa Land'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Oberlaa Land
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1104
    related:wikipedia=de:Oberlaa
    wikidata=Q27832622
  id=5241913 admin=10 name='Katastralgemeinde Unterlaa'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Unterlaa
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1108
    wikidata=Q2497878
    wikipedia=de:Unterlaa
  id=5243103 admin=10 name='Katastralgemeinde Inzersdorf'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Inzersdorf
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1803
    wikidata=Q1671849
    wikipedia=de:Inzersdorf (Wien)
  id=5247528 admin=10 name='Katastralgemeinde Siebenhirten'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Siebenhirten
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1808
    wikidata=Q2282351
    wikipedia=de:Siebenhirten (Wien)
  id=5247701 admin=10 name='Katastralgemeinde Erlaa'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Erlaa
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1802
    wikidata=Q1355197
    wikipedia=de:Erlaa
  id=5247849 admin=10 name='Katastralgemeinde Atzgersdorf'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Atzgersdorf
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1801
    wikidata=Q758336
    wikipedia=de:Atzgersdorf
  id=5248033 admin=10 name='Wieden'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    long_name=Katastralgemeinde Wieden
    official_name=Wieden
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01011
    wikidata=Q27833480
  id=5248034 admin=10 name='Katastralgemeinde Margarethen'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Margarethen
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01008
    wikidata=Q27833470
  id=5248153 admin=10 name='Katastralgemeinde Rodaun'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Rodaun
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1807
    wikidata=Q2160909
    wikipedia=de:Rodaun
  id=5248175 admin=10 name='Katastralgemeinde Liesing'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Liesing
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1805
    wikidata=Q1824343
    wikipedia=de:Liesing (Wiener Bezirksteil)
  id=5248176 admin=10 name='Katastralgemeinde Kalksburg'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Kalksburg
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1804
    wikidata=Q1678869
    wikipedia=de:Kalksburg
  id=5264565 admin=10 name='Katastralgemeinde Altmannsdorf'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Altmannsdorf
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1301
    wikidata=Q442551
    wikipedia=de:Altmannsdorf (Wien)
  id=5264583 admin=10 name='Katastralgemeinde Innere Stadt'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Innere Stadt
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01004
    wikidata=Q27834577
  id=5264614 admin=10 name='Katastralgemeinde Rosenberg'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Rosenberg
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1211
    wikidata=Q1503013
    wikipedia=de:Rosenberg (Wien)
  id=5280044 admin=10 name='Katastralgemeinde Albern'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Albern
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1109
    wikidata=Q2637387
    wikipedia=de:Albern
  id=5301360 admin=10 name='Katastralgemeinde Hetzendorf'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Hetzendorf
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1304
    wikidata=Q1616329
    wikipedia=de:Hetzendorf (Wien)
  id=5301402 admin=10 name='Katastralgemeinde Schönbrunn'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Schönbrunn
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01212
    wikidata=Q27837208
    wikipedia=de:Schönbrunn (Wien)
  id=5301600 admin=10 name='Katastralgemeinde Meidling'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Meidling
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1305
    wikidata=Q1917668
    wikipedia=de:Meidling (Wiener Bezirksteil)
  id=5301617 admin=10 name='Katastralgemeinde Gaudenzdorf'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Gaudenzdorf
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1303
    wikidata=Q1495934
    wikipedia=de:Gaudenzdorf
  id=5832478 admin=10 name='Katastralgemeinde Auhof'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Auhof
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01201
    wikidata=Q771042
    wikipedia=de:Auhof (Wien)
  id=5832617 admin=10 name='Katastralgemeinde Weidlingau'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Weidlingau
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01216
    wikidata=Q27835519
  id=5832830 admin=10 name='Katastralgemeinde Speising'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Speising
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1213
    wikidata=Q1718608
    wikipedia=de:Speising
  id=5832897 admin=10 name='Katastralgemeinde Lainz'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Lainz
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1207
    wikidata=Q1261352
    wikipedia=de:Lainz
  id=5833856 admin=10 name='Katastralgemeinde Hietzing'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Hietzing
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:a
[... truncated ...]ralgemeinde Gersthof'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Gersthof
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1501
    wikidata=Q1515094
    wikipedia=de:Gersthof
  id=5881429 admin=10 name='Katastralgemeinde Weinhaus'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Weinhaus
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1515
    wikidata=Q2556149
    wikipedia=de:Weinhaus (Wien)
  id=5881930 admin=10 name='Katastralgemeinde Währing'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Währing
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1514
    wikidata=Q1749983
    wikipedia=de:Währing (Wiener Bezirksteil)
  id=5886110 admin=10 name='Katastralgemeinde Neuwaldegg'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Neuwaldegg
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01404
    wikidata=Q1981987
    wikipedia=de:Neuwaldegg
  id=5891253 admin=10 name='Katastralgemeinde Dornbach'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Dornbach
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1401
    wikidata=Q13400194
    wikipedia=de:Dornbach (Wien)
  id=5891265 admin=10 name='Katastralgemeinde Hernals'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Hernals
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1402
    wikidata=Q1613671
    wikipedia=de:Hernals (Wiener Bezirksteil)
  id=5892045 admin=10 name='Katastralgemeinde Hadersdorf'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Hadersdorf
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01204
    wikidata=Q27835330
  id=5897544 admin=10 name='Katastralgemeinde Ottakring'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Ottakring
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1405
    wikidata=Q1496772
    wikipedia=de:Ottakring (Wiener Bezirksteil)
  id=5897594 admin=10 name='Katastralgemeinde Neulerchenfeld'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Neulerchenfeld
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1403
    wikidata=Q1980858
    wikipedia=de:Neulerchenfeld
  id=5899601 admin=10 name='Katastralgemeinde Sechshaus'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Sechshaus
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1307
    wikidata=Q2263631
    wikipedia=de:Sechshaus
  id=5899641 admin=10 name='Katastralgemeinde Rudolfsheim'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Rudolfsheim
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1306
    wikidata=Q2174345
    wikipedia=de:Rudolfsheim
  id=5899642 admin=10 name='Katastralgemeinde Fünfhaus'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Fünfhaus
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1302
    wikidata=Q1481965
    wikipedia=de:Fünfhaus
  id=5899920 admin=10 name='Katastralgemeinde Penzing'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Penzing
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1210
    wikidata=Q2069566
    wikipedia=de:Penzing (Wiener Bezirksteil)
  id=5900008 admin=10 name='Katastralgemeinde Breitensee'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Breitensee
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1202
    wikidata=Q906418
    wikipedia=de:Breitensee (Wien)
  id=5900027 admin=10 name='Katastralgemeinde Unterbaumgarten'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Unterbaumgarten
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01214
    wikidata=Q27835286
  id=5900074 admin=10 name='Katastralgemeinde Oberbaumgarten'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Oberbaumgarten
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=1208
    wikidata=Q27835307
  id=5900077 admin=10 name='Katastralgemeinde Hütteldorf'
    admin_title=Katastralgemeinde
    designation=Katastralgemeinde
    official_name=Hütteldorf
    official_status=Katastralgemeinde
    ref:at:gkz=90001
    ref:at:kgnr=01206
    wikidata=Q1423660
    wikipedia=de:Hütteldorf

=== OLD (2014) relations ===
  id=1643642 admin=9 name='Gerasdorf bei Wien'
  id=1744661 admin=10 name='KG Mauer'
  id=1990590 admin=9 name='Alsergrund'
    name:ru=Альзергрунд
    postal_code=1090
    ref=9
    ref:at:gkz=90901
    wikipedia:de=Alsergrund
    wikipedia:en=Alsergrund
    wikipedia:ru=Альзергрунд
    wikipedia:street_names:de=Liste_der_Straßennamen_von_Wien/Alsergrund
  id=1990591 admin=9 name='Hietzing'
    name:ru=Хитцинг
    postal_code=1130
    ref=13
    ref:at:gkz=91301
    wikipedia:de=Hietzing
    wikipedia:en=Hietzing
    wikipedia:ru=Хитцинг
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Hietzing
  id=1990592 admin=9 name='Innere Stadt'
    postal_code=1010
    ref=1
    ref:at:gkz=90101
    wikipedia:de=Innere Stadt (Wien)
    wikipedia:en=Innere Stadt
    wikipedia:ru=Внутренний город (Вена)
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Innere Stadt
  id=1990593 admin=9 name='Josefstadt'
    name:ru=Йозефштадт
    postal_code=1080
    ref=8
    ref:at:gkz=90801
    wikipedia:de=Josefstadt (Wien)
    wikipedia:en=Josefstadt
    wikipedia:ru=Йозефштадт
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Josefstadt
  id=1990594 admin=9 name='Leopoldstadt'
    name:ru=Леопольдштадт
    postal_code=1020
    ref=2
    ref:at:gkz=90201
    wikipedia:de=Leopoldstadt
    wikipedia:en=Leopoldstadt
    wikipedia:ru=Леопольдштадт
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Leopoldstadt
  id=1990595 admin=9 name='Mariahilf'
    name:ru=Мариахильф
    postal_code=1060
    ref=6
    ref:at:gkz=90601
    wikipedia:de=Mariahilf
    wikipedia:en=Mariahilf
    wikipedia:ru=Мариахильф
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Mariahilf
  id=1990596 admin=9 name='Meidling'
    name:ru=Майдлинг
    postal_code=1120
    ref=12
    ref:at:gkz=91201
    wikipedia:de=Meidling
    wikipedia:en=Meidling
    wikipedia:ru=Майдлинг
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Meidling
  id=1990597 admin=9 name='Neubau'
    name:ru=Нойбау
    postal_code=1070
    ref=7
    ref:at:gkz=90701
    wikipedia:de=Neubau (Wien)
    wikipedia:en=Neubau
    wikipedia:ru=Нойбау
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Neubau
  id=1990598 admin=9 name='Penzing'
    name:ru=Пенцинг
    postal_code=1140
    ref=14
    ref:at:gkz=91401
    wikipedia:de=Penzing (Wien)
    wikipedia:en=Penzing (Vienna)
    wikipedia:ru=Пенцинг_(Вена)
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Penzing
  id=1990599 admin=9 name='Rudolfsheim-Fünfhaus'
    is_in=Wien,Östereich,Europe
    population=69078
    postal_code=1150
    ref=15
    ref:at:gkz=91501
    ref:at:okz=17237
    wikipedia:de=Rudolfsheim-Fünfhaus
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Rudolfsheim-Fünfhaus
  id=1990600 admin=9 name='Währing'
    name:en=Währing
    name:ru=Вэринг
    postal_code=1180
    ref=18
    ref:at:gkz=91801
    wikipedia:de=Währing
    wikipedia:en=Währing
    wikipedia:ru=Вэринг
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Währing
  id=1991416 admin=9 name='Landstraße'
    name:ru=Ландштрассе
    postal_code=1030
    ref=3
    ref:at:gkz=90301
    wikipedia:de=Landstraße (Wien)
    wikipedia:en=Landstraße
    wikipedia:ru=Ландштрассе
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Landstraße
  id=1991433 admin=9 name='Brigittenau'
    name:ru=Бригиттенау
    postal_code=1200
    ref=20
    ref:at:gkz=92001
    wikipedia:de=Brigittenau
    wikipedia:en=Brigittenau
    wikipedia:ru=Бригиттенау
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Brigittenau
  id=1991434 admin=9 name='Donaustadt'
    is_in=Wien,Östereich,Europe
    name:ru=Донауштадт
    population=144289
    postal_code=1220
    ref=22
    ref:at:gkz=92201
    wikipedia:de=Donaustadt
    wikipedia:en=Donaustadt
    wikipedia:ru=Донауштадт
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Donaustadt
  id=1991435 admin=9 name='Döbling'
    is_in=Wien,Östereich,Europe
    name:ru=Дёблинг
    population=66487
    postal_code=1190
    ref=19
    ref:at:gkz=91901
    wikipedia:de=Döbling
    wikipedia:en=Döbling
    wikipedia:ru=Дёблинг
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Döbling
  id=1991436 admin=9 name='Favoriten'
    name:ru=Фаворитен
    postal_code=1100
    ref=10
    ref:at:gkz=91001
    wikipedia=de:Favoriten
    wikipedia:de=Favoriten
    wikipedia:en=Favoriten
    wikipedia:ru=Фаворитен
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Favoriten
  id=1991437 admin=9 name='Floridsdorf'
    name:ru=Флоридсдорф
    postal_code=1210
    ref=21
    ref:at:gkz=92101
    wikipedia:de=Floridsdorf
    wikipedia:en=Floridsdorf
    wikipedia:ru=Флоридсдорф
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Floridsdorf
  id=1991438 admin=9 name='Hernals'
    name:en=Hernals
    name:ru=Хернальс
    postal_code=1170
    ref=17
    ref:at:gkz=91701
    wikipedia:de=Hernals
    wikipedia:en=Hernals
    wikipedia:ru=Хернальс
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Hernals
  id=1991439 admin=9 name='Liesing'
    name:ru=Лизинг
    postal_code=1230
    ref=23
    ref:at:gkz=92301
    wikipedia:de=Liesing (Wien)
    wikipedia:en=Liesing
    wikipedia:ru=Лизинг_(Вена)
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Liesing
  id=1991440 admin=9 name='Margareten'
    is_in=Wien,Österreich,Europe
    name:ru=Маргаретен
    population=51434
    postal_code=1050
    ref=5
    ref:at:gkz=90501
    wikipedia:de=Margareten
    wikipedia:en=Margareten
    wikipedia:ru=Маргаретен
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Margareten
  id=1991441 admin=9 name='Ottakring'
    postal_code=1160
    ref=16
    ref:at:gkz=91601
    wikipedia:de=Ottakring
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Ottakring
  id=1991442 admin=9 name='Simmering'
    name:ru=Зиммеринг
    postal_code=1110
    ref=11
    ref:at:gkz=91101
    wikipedia:de=Simmering
    wikipedia:en=Simmering (Vienna)
    wikipedia:ru=Зиммеринг
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Simmering
  id=1991443 admin=9 name='Wieden'
    name:ru=Виден
    postal_code=1040
    ref=4
    ref:at:gkz=90401
    wikipedia:de=Wieden (Wien)
    wikipedia:en=Wieden
    wikipedia:ru=Виден (Вена)
    wikipedia:street_names:de=Liste der Straßennamen von Wien/Wieden


=== Checking for name differences ===
Current names:
  'Alsergrund'
  'Brigittenau'
  'Donaustadt'
  'Döbling'
  'Favoriten'
  'Floridsdorf'
  'Hernals'
  'Hietzing'
  'Innere Stadt'
  'Josefstadt'
  'Katastralgemeinde Albern'
  'Katastralgemeinde Alsergrund'
  'Katastralgemeinde Altmannsdorf'
  'Katastralgemeinde Aspern'
  'Katastralgemeinde Atzgersdorf'
  'Katastralgemeinde Auhof'
  'Katastralgemeinde Breitenlee'
  'Katastralgemeinde Breitensee'
  'Katastralgemeinde Brigittenau'
  'Katastralgemeinde Donaufeld'
  'Katastralgemeinde Dornbach'
  'Katastralgemeinde Erlaa'
  'Katastralgemeinde Eßling'
  'Katastralgemeinde Favoriten'
  'Katastralgemeinde Floridsdorf'
  'Katastralgemeinde Fünfhaus'
  'Katastralgemeinde Gaudenzdorf'
  'Katastralgemeinde Gersthof'
  'Katastralgemeinde Grinzing'
  'Katastralgemeinde Großjedlersdorf I'
  'Katastralgemeinde Großjedlersdorf II'
  'Katastralgemeinde Hacking'
  'Katastralgemeinde Hadersdorf'
  'Katastralgemeinde Heiligenstadt'
  'Katastralgemeinde Hernals'
  'Katastralgemeinde Hetzendorf'
  'Katastralgemeinde Hietzing'
  'Katastralgemeinde Hirschstetten'
  'Katastralgemeinde Hütteldorf'
  'Katastralgemeinde Innere Stadt'
  'Katastralgemeinde Inzersdorf'
  'Katastralgemeinde Inzersdorf Stadt'
  'Katastralgemeinde Jedlesee'
  'Katastralgemeinde Josefsdorf'
  'Katastralgemeinde Josefstadt'
  'Katastralgemeinde Kagran'
  'Katastralgemeinde Kahlenbergerdorf'
  'Katastralgemeinde Kaiserebersdorf'
  'Katastralgemeinde Kaiserebersdorf Herrschaft'
  'Katastralgemeinde Kaisermühlen'
  'Katastralgemeinde Kalksburg'
  'Katastralgemeinde Lainz'
  'Katastralgemeinde Landjägermeisteramt'
  'Katastralgemeinde Landstraße'
  'Katastralgemeinde Leopoldau'
  'Katastralgemeinde Leopoldstadt'
  'Katastralgemeinde Liesing'
  'Katastralgemeinde Margarethen'
  'Katastralgemeinde Mariahilf'
  'Katastralgemeinde Mauer'
  'Katastralgemeinde Meidling'
  'Katastralgemeinde Neubau'
  'Katastralgemeinde Neulerchenfeld'
  'Katastralgemeinde Neustift am Walde'
  'Katastralgemeinde Neuwaldegg'
  'Katastralgemeinde Nußdorf'
  'Katastralgemeinde Ober St. Veit'
  'Katastralgemeinde Oberbaumgarten'
  'Katastralgemeinde Oberdöbling'
  'Katastralgemeinde Oberlaa Land'
  'Katastralgemeinde Oberlaa Stadt'
  'Katastralgemeinde Obersievering'
  'Katastralgemeinde Ottakring'
  'Katastralgemeinde Penzing'
  'Katastralgemeinde Pötzleinsdorf'
  'Katastralgemeinde Rodaun'
  'Katastralgemeinde Rosenberg'
  'Katastralgemeinde Rothneusiedl'
  'Katastralgemeinde Rudolfsheim'
  'Katastralgemeinde Salmannsdorf'
  'Katastralgemeinde Schwarze Lackenau'
  'Katastralgemeinde Schönbrunn'
  'Katastralgemeinde Sechshaus'
  'Katastralgemeinde Siebenhirten'
  'Katastralgemeinde Simmering'
  'Katastralgemeinde Speising'
  'Katastralgemeinde Stadlau'
  'Katastralgemeinde Stammersdorf'
  'Katastralgemeinde Strebersdorf'
  'Katastralgemeinde Süssenbrunn'
  'Katastralgemeinde Unter St. Veit'
  'Katastralgemeinde Unterbaumgarten'
  'Katastralgemeinde Unterdöbling'
  'Katastralgemeinde Unterlaa'
  'Katastralgemeinde Untersievering'
  'Katastralgemeinde Weidlingau'
  'Katastralgemeinde Weinhaus'
  'Katastralgemeinde Währing'
  'Landstraße'
  'Leopoldstadt'
  'Liesing'
  'Margareten'
  'Mariahilf'
  'Meidling'
  'Neubau'
  'Ottakring'
  'Penzing'
  'Rudolfsheim-Fünfhaus'
  'Simmering'
  'Wieden'
  'Währing'

Old names (2014):
  'Alsergrund'
  'Brigittenau'
  'Donaustadt'
  'Döbling'
  'Favoriten'
  'Floridsdorf'
  'Gerasdorf bei Wien'
  'Hernals'
  'Hietzing'
  'Innere Stadt'
  'Josefstadt'
  'KG Mauer'
  'Landstraße'
  'Leopoldstadt'
  'Liesing'
  'Margareten'
  'Mariahilf'
  'Meidling'
  'Neubau'
  'Ottakring'
  'Penzing'
  'Rudolfsheim-Fünfhaus'
  'Simmering'
  'Wieden'
  'Währing'

cat << 'PYEOF' > /work/solve2.py
#!/usr/bin/env python3
"""
Fetch Vienna district polygons from Overpass data (both current and 2014-01-01),
compute boundary changes, and output vienna_boundary_changes.geojson.
Focuses on admin_level=9 (Bezirke).
"""

import json, sys, os, re
from collections import defaultdict

from shapely.geometry import Polygon, MultiPolygon, shape, mapping
from shapely.ops import unary_union
from shapely.validation import make_valid
import geopandas as gpd
import pandas as pd

# ---------------------------------------------------------------------------
# 1. Load raw Overpass data
# ---------------------------------------------------------------------------

with open('/work/current_raw.json') as f:
    cur_raw = json.load(f)

with open('/work/old_raw.json') as f:
    old_raw = json.load(f)

# ---------------------------------------------------------------------------
# 2. Build a dict of way_id -> list of coord tuples
# ---------------------------------------------------------------------------

def build_way_map(elements):
    """From a list of elements, extract all ways with their geometry."""
    ways = {}
    for el in elements:
        if el.get('type') == 'way' and 'geometry' in el:
            coords = [(p['lon'], p['lat']) for p in el['geometry']]
            if len(coords) >= 2:
                ways[el['id']] = coords
    return ways

cur_ways = build_way_map(cur_raw.get('elements', []))
old_ways = build_way_map(old_raw.get('elements', []))

print(f"Current ways: {len(cur_ways)}", file=sys.stderr)
print(f"Old ways: {len(old_ways)}", file=sys.stderr)

# ---------------------------------------------------------------------------
# 3. Extract district (admin_level=9) geometries from relations
# ---------------------------------------------------------------------------

def extract_districts(elements, ways, label):
    """Extract admin_level=9 boundary relations."""
    districts = {}
    
    # First pass: find relations
    for el in elements:
        if el.get('type') != 'relation':
            continue
        
        tags = el.get('tags', {})
        admin = tags.get('admin_level', '')
        
        # Only care about admin_level=9 districts (Bezirke)
        if admin != '9':
            continue
        
        name = tags.get('name', '')
        if not name:
            continue
        
        # Build geometry from member ways
        members = el.get('members', [])
        outer_rings = []
        inner_rings = []
        
        for m in members:
            if m.get('type') != 'way':
                continue
            wid = m.get('ref')
            if wid not in ways:
                continue
            coords = ways[wid]
            role = m.get('role', '')
            
            if len(coords) >= 4:
                try:
                    poly = Polygon(coords)
                    if not poly.is_valid:
                        poly = poly.buffer(0)
                    if poly.area > 0:
                        if role == 'inner':
                            inner_rings.append(poly)
                        else:
                            outer_rings.append(poly)
                except Exception as e:
                    print(f"  Error building polygon for way {wid}: {e}", file=sys.stderr)
        
        if not outer_rings:
            print(f"  No outer rings for {name}", file=sys.stderr)
            continue
        
        # Build the geometry
        if len(outer_rings) == 1:
            base_poly = outer_rings[0]
            # Add holes from inner rings
            holes = []
            for inner in inner_rings:
                # Check if inner is actually inside base
                if base_poly.contains(inner.representative_point()):
                    try:
                        holes.append(inner.exterior.coords)
                    except:
                        holes.append(list(inner.exterior.coords))
            if holes:
                base_poly = Polygon(base_poly.exterior.coords, holes)
            geom = base_poly
        else:
            # Multiple outer rings -> MultiPolygon
            polys = []
            for outer in outer_rings:
                # Collect inners that are inside this outer
                my_holes = []
                remaining = []
                for inner in inner_rings:
                    if outer.contains(inner.representative_point()):
                        my_holes.append(list(inner.exterior.coords))
                    else:
                        remaining.append(inner)
                inner_rings = remaining
                try:
                    if my_holes:
                        polys.append(Polygon(list(outer.exterior.coords), my_holes))
                    else:
                        polys.append(Polygon(list(outer.exterior.coords)))
                except:
                    polys.append(outer)
            if len(polys) == 1:
                geom = polys[0]
            else:
                geom = MultiPolygon(polys)
        
        if not geom.is_valid:
            geom = geom.buffer(0)
        
        if geom.area < 1:
            print(f"  Tiny geometry for {name}, skipping", file=sys.stderr)
            continue
        
        # Store with name as-is
        districts[name] = geom
    
    print(f"  {label}: extracted {len(districts)} districts", file=sys.stderr)
    for n in sorted(districts.keys()):
        print(f"    {n}", file=sys.stderr)
    return districts

current_districts = extract_districts(cur_raw.get('elements', []), cur_ways, "Current")
old_districts = extract_districts(old_raw.get('elements', []), old_ways, "Old (2014)")

# ---------------------------------------------------------------------------
# 4. Match districts by name
# ---------------------------------------------------------------------------
# The names are the same between snapshots for the Bezirke
common_names = set(current_districts.keys()) & set(old_districts.keys())
only_current = set(current_districts.keys()) - set(old_districts.keys())
only_old = set(old_districts.keys()) - set(current_districts.keys())

print(f"\nCommon: {len(common_names)}", file=sys.stderr)
print(f"Only current: {only_current}", file=sys.stderr)
print(f"Only old: {only_old}", file=sys.stderr)

if not common_names:
    print("ERROR: No matching district names!", file=sys.stderr)
    sys.exit(1)

# ---------------------------------------------------------------------------
# 5. Normalise names
# ---------------------------------------------------------------------------

def normalise(name):
    """Normalise Vienna district name to canonical form.
    E.g. 'Innere Stadt' -> 'wien-01-innere-stadt'
    We use the ref tag from current districts... but we don't have it easily.
    Let's build a mapping.
    """
    # Known mapping of district name -> number
    known = {
        'Innere Stadt': 1, 'Leopoldstadt': 2, 'Landstraße': 3, 'Wieden': 4,
        'Margareten': 5, 'Mariahilf': 6, 'Neubau': 7, 'Josefstadt': 8,
        'Alsergrund': 9, 'Favoriten': 10, 'Simmering': 11, 'Meidling': 12,
        'Hietzing': 13, 'Penzing': 14, 'Rudolfsheim-Fünfhaus': 15,
        'Ottakring': 16, 'Hernals': 17, 'Währing': 18, 'Döbling': 19,
        'Brigittenau': 20, 'Floridsdorf': 21, 'Donaustadt': 22, 'Liesing': 23
    }
    num = known.get(name)
    if num:
        return f"wien-{num:02d}-{name.lower().replace(' ', '-')}"
    return f"wien-{name.lower().replace(' ', '-')}"

# ---------------------------------------------------------------------------
# 6. Compute boundary changes for each district
# ---------------------------------------------------------------------------

change_features = []  # list of {geometry, district_name, change_type}

for name in common_names:
    cur_geom = current_districts[name]
    old_geom = old_districts[name]
    
    n_name = normalise(name)
    
    # Added: in current but not in old
    added = cur_geom.difference(old_geom)
    # Removed: in old but not in current
    removed = old_geom.difference(cur_geom)
    # Unchanged: intersection
    unchanged = cur_geom.intersection(old_geom)
    
    for geom, ctype in [
        (added, 'added_since_2014'),
        (removed, 'removed_since_2014'),
        (unchanged, 'unchanged')
    ]:
        if geom.is_empty:
            continue
        if geom.geom_type == 'GeometryCollection':
            polys = [g for g in geom.geoms if g.geom_type in ('Polygon', 'MultiPolygon')]
            if not polys:
                continue
            geom = unary_union(polys)
        if geom.geom_type not in ('Polygon', 'MultiPolygon'):
            continue
        if geom.area < 1e-6:
            continue
        if not geom.is_valid:
            geom = geom.buffer(0)
        
        change_features.append({
            'geometry': geom,
            'district_name': n_name,
            'change_type': ctype
        })

# Districts only in one snapshot
for name in only_current:
    geom = current_districts[name]
    if not geom.is_valid:
        geom = geom.buffer(0)
    if geom.area >= 1e-6:
        change_features.append({
            'geometry': geom,
            'district_name': normalise(name),
            'change_type': 'added_since_2014'
        })
for name in only_old:
    geom = old_districts[name]
    if not geom.is_valid:
        geom = geom.buffer(0)
    if geom.area >= 1e-6:
        change_features.append({
            'geometry': geom,
            'district_name': normalise(name),
            'change_type': 'removed_since_2014'
        })

print(f"\nTotal change fragments: {len(change_features)}", file=sys.stderr)

if not change_features:
    # Write empty GeoJSON
    result = {"type": "FeatureCollection", "features": []}
    with open('/work/vienna_boundary_changes.geojson', 'w') as f:
        json.dump(result, f, indent=2)
    print("No changes found. Written empty file.", file=sys.stderr)
    sys.exit(0)

# ---------------------------------------------------------------------------
# 7. Determine touches_changed for each district
# ---------------------------------------------------------------------------

# Build a union of all "changed" geometries (added + removed)
changed_geoms = [
    f['geometry'] for f in change_features
    if f['change_type'] in ('added_since_2014', 'removed_since_2014')
]

if changed_geoms:
    changed_union = unary_union(changed_geoms)
else:
    changed_union = None

# For each district name, check if any of its fragments touch changed areas
all_district_names = set(f['district_name'] for f in change_features)
touches_map = {}

for dname in all_district_names:
    touches = False
    district_frags = [f for f in change_features if f['district_name'] == dname]
    
    for f in district_frags:
        frag = f['geometry']
        if changed_union is not None and not changed_union.is_empty:
            # Check if they touch (share boundary)
            if frag.touches(changed_union):
                touches = True
                break
            # Also check if they intersect at boundary with tiny buffer
            if frag.buffer(1e-8).intersects(changed_union) and not frag.intersects(changed_union):
                touches = True
                break
            # More directly: check if the boundary touches
            boundary = frag.boundary
            if boundary.intersects(changed_union):
                touches = True
                break
    
    touches_map[dname] = touches

# ---------------------------------------------------------------------------
# 8. Write output GeoJSON
# ---------------------------------------------------------------------------

features_out = []
for f in change_features:
    geom = f['geometry']
    if geom.geom_type == 'Polygon':
        geom = MultiPolygon([geom])
    
    features_out.append({
        'type': 'Feature',
        'geometry': mapping(geom),
        'properties': {
            'change_type': f['change_type'],
            'district_name': f['district_name'],
            'touches_changed': touches_map.get(f['district_name'], False)
        }
    })

result = {
    'type': 'FeatureCollection',
    'features': features_out
}

with open('/work/vienna_boundary_changes.geojson', 'w') as f:
    json.dump(result, f, indent=2)

# Summary
print(f"\nWritten /work/vienna_boundary_changes.geojson", file=sys.stderr)
print(f"  Total features: {len(features_out)}", file=sys.stderr)
for ctype in ['added_since_2014', 'removed_since_2014', 'unchanged']:
    cnt = sum(1 for f in features_out if f['properties']['change_type'] == ctype)
    print(f"  {ctype}: {cnt}", file=sys.stderr)

print("\nDistricts:", file=sys.stderr)
for dname in sorted(all_district_names):
    tc = touches_map[dname]
    print(f"  {dname}: touches_changed={tc}", file=sys.stderr)

print("\nDone.", file=sys.stderr)
PYEOF
python3 /work/solve2.py

Current ways: 695
Old ways: 364
Traceback (most recent call last):
  File "/work/solve2.py", line 157, in <module>
    current_districts = extract_districts(cur_raw.get('elements', []), cur_ways, "Current")
  File "/work/solve2.py", line 140, in extract_districts
    geom = MultiPolygon(polys)
  File "/usr/local/lib/python3.14/site-packages/shapely/geometry/multipolygon.py", line 70, in __new__
    raise ValueError("Sequences of multi-polygons are not valid arguments")
ValueError: Sequences of multi-polygons are not valid arguments

Current ways: 695
Old ways: 364
  Tiny geometry for Alsergrund, skipping
  Tiny geometry for Hietzing, skipping
  Tiny geometry for Innere Stadt, skipping
  Tiny geometry for Josefstadt, skipping
  Tiny geometry for Leopoldstadt, skipping
  Tiny geometry for Mariahilf, skipping
  Tiny geometry for Meidling, skipping
  Tiny geometry for Neubau, skipping
  Tiny geometry for Penzing, skipping
  Tiny geometry for Rudolfsheim-Fünfhaus, skipping
  Tiny geometry for Währing, skipping
  Tiny geometry for Landstraße, skipping
  Tiny geometry for Brigittenau, skipping
  Tiny geometry for Donaustadt, skipping
  Tiny geometry for Döbling, skipping
  Tiny geometry for Favoriten, skipping
  Tiny geometry for Floridsdorf, skipping
  Tiny geometry for Hernals, skipping
  Tiny geometry for Liesing, skipping
  Tiny geometry for Margareten, skipping
  Tiny geometry for Ottakring, skipping
  Tiny geometry for Simmering, skipping
  Tiny geometry for Wieden, skipping
  Current: extracted 0 districts
  Tiny geometry for Gerasdorf bei Wien, skipping
  Tiny geometry for Alsergrund, skipping
  Tiny geometry for Hietzing, skipping
  Tiny geometry for Innere Stadt, skipping
  Tiny geometry for Josefstadt, skipping
  Tiny geometry for Leopoldstadt, skipping
  Tiny geometry for Mariahilf, skipping
  Tiny geometry for Meidling, skipping
  Tiny geometry for Neubau, skipping
  Tiny geometry for Penzing, skipping
  Tiny geometry for Rudolfsheim-Fünfhaus, skipping
  Tiny geometry for Währing, skipping
  Tiny geometry for Landstraße, skipping
  Tiny geometry for Brigittenau, skipping
  Tiny geometry for Donaustadt, skipping
  Tiny geometry for Döbling, skipping
  Tiny geometry for Favoriten, skipping
  Tiny geometry for Floridsdorf, skipping
  Tiny geometry for Hernals, skipping
  Tiny geometry for Liesing, skipping
  Tiny geometry for Margareten, skipping
  Tiny geometry for Ottakring, skipping
  Tiny geometry for Simmering, skipping
  Tiny geometry for Wieden, skipping
  Old (2014): extracted 0 districts

Common: 0
Only current: set()
Only old: set()
ERROR: No matching district names!

cat << 'PYEOF' > /work/debug2.py
import json, sys

with open('/work/current_raw.json') as f:
    data = json.load(f)

# Find a district relation and check its structure
for el in data['elements']:
    if el.get('type') == 'relation':
        tags = el.get('tags', {})
        if tags.get('admin_level') == '9':
            name = tags.get('name', '')
            members = el.get('members', [])
            print(f"Relation: {name}")
            print(f"  Members count: {len(members)}")
            # Check first few members
            way_refs = [m['ref'] for m in members if m.get('type') == 'way']
            print(f"  Way refs (first 5): {way_refs[:5]}")
            
            # Check if these ways exist in the elements
            way_dict = {}
            for e2 in data['elements']:
                if e2.get('type') == 'way':
                    way_dict[e2['id']] = e2
            
            for wid in way_refs[:5]:
                if wid in way_dict:
                    w = way_dict[wid]
                    has_geom = 'geometry' in w
                    if has_geom:
                        geom_len = len(w['geometry'])
                        print(f"  Way {wid}: has geometry ({geom_len} points), first pt: {w['geometry'][0]}")
                    else:
                        print(f"  Way {wid}: NO geometry, has nodes: {len(w.get('nodes', []))}")
                else:
                    print(f"  Way {wid}: NOT FOUND in elements!")
            
            # Check the relation's own geometry
            if 'geometry' in el:
                print(f"  Relation has direct geometry: {len(el['geometry'])} points")
            else:
                print(f"  Relation has NO direct geometry")
            
            if len(way_refs) > 0:
                break

# Also check the structure more broadly
print("\n\n--- Element type counts ---")
types = {}
for el in data['elements']:
    t = el.get('type')
    types[t] = types.get(t, 0) + 1
for t, c in types.items():
    print(f"  {t}: {c}")

# Count ways with geometry
ways_with_geom = sum(1 for el in data['elements'] if el.get('type') == 'way' and 'geometry' in el)
print(f"\nWays with geometry: {ways_with_geom}")
PYEOF
python3 /work/debug2.py

Relation: Alsergrund
  Members count: 9
  Way refs (first 5): [147468285, 8046654, 147396089, 147463455, 147463456]
  Way 147468285: has geometry (32 points), first pt: {'lat': 48.2151672, 'lon': 16.3408703}
  Way 8046654: has geometry (28 points), first pt: {'lat': 48.2151672, 'lon': 16.3408703}
  Way 147396089: has geometry (33 points), first pt: {'lat': 48.2184891, 'lon': 16.370149}
  Way 147463455: has geometry (15 points), first pt: {'lat': 48.2251075, 'lon': 16.3674575}
  Way 147463456: has geometry (22 points), first pt: {'lat': 48.236309, 'lon': 16.3616214}
  Relation has NO direct geometry


--- Element type counts ---
  relation: 112
  node: 11906
  way: 695

Ways with geometry: 695

cat << 'PYEOF' > /work/solve3.py
#!/usr/bin/env python3
"""
Compute boundary changes between current and 2014-01-01 Vienna districts.
Assembles boundary ways into proper polygon rings.
"""

import json, sys, os, re
from collections import defaultdict

from shapely.geometry import Polygon, MultiPolygon, LineString, shape, mapping, Point
from shapely.ops import unary_union, polygonize
from shapely.validation import make_valid
import geopandas as gpd
import pandas as pd

# ---------------------------------------------------------------------------
# 1. Load raw Overpass data
# ---------------------------------------------------------------------------

with open('/work/current_raw.json') as f:
    cur_raw = json.load(f)

with open('/work/old_raw.json') as f:
    old_raw = json.load(f)

# ---------------------------------------------------------------------------
# 2. Build way map
# ---------------------------------------------------------------------------

def build_way_map(elements):
    """From a list of elements, extract all ways with their geometry as LineStrings."""
    ways = {}
    for el in elements:
        if el.get('type') == 'way' and 'geometry' in el:
            coords = [(p['lon'], p['lat']) for p in el['geometry']]
            if len(coords) >= 2:
                ways[el['id']] = LineString(coords)
    return ways

cur_ways = build_way_map(cur_raw.get('elements', []))
old_ways = build_way_map(old_raw.get('elements', []))

print(f"Current ways: {len(cur_ways)}", file=sys.stderr)
print(f"Old ways: {len(old_ways)}", file=sys.stderr)

# ---------------------------------------------------------------------------
# 3. Assemble polygon from ordered boundary ways
# ---------------------------------------------------------------------------

def extract_district_geometry(relation, ways):
    """
    Given a relation element and a way map, reconstruct the polygon(s).
    Ways have roles 'outer' and 'inner'. We assemble them into rings.
    """
    members = relation.get('members', [])
    
    outer_ways = []
    inner_ways = []
    
    for m in members:
        if m.get('type') != 'way':
            continue
        wid = m.get('ref')
        if wid not in ways:
            continue
        role = m.get('role', '')
        if role == 'inner':
            inner_ways.append(ways[wid])
        else:
            outer_ways.append(ways[wid])
    
    # We need to assemble the ways into closed rings.
    # Since Overpass returns ways in order for most relations, we can try to
    # connect them end-to-end. Simple approach: polygonize the linework.
    
    all_rings = []
    result_polys = []
    
    # Try polygonize for outer
    if outer_ways:
        # Merge outer ways
        merged_outer = unary_union(outer_ways)
        # Polygonize
        polygonized = list(polygonize(outer_ways))
        if polygonized:
            result_polys.extend(polygonized)
        elif merged_outer.geom_type == 'LineString' and merged_outer.is_ring:
            result_polys.append(Polygon(merged_outer))
        else:
            # Try to join them manually
            # If merged is a MultiLineString that forms rings
            if merged_outer.geom_type == 'MultiLineString':
                for line in merged_outer.geoms:
                    if line.is_ring:
                        result_polys.append(Polygon(line))
            # Last resort: buffer and extract
            if not result_polys:
                buf = merged_outer.buffer(0.000001)
                if buf.geom_type == 'Polygon' and buf.area > 0:
                    result_polys.append(buf)
    
    # Add holes from inner ways
    if inner_ways and result_polys:
        merged_inner = unary_union(inner_ways)
        inner_polys = list(polygonize(inner_ways))
        if not inner_polys and merged_inner.geom_type in ('LineString', 'MultiLineString'):
            # Try to make holes
            if merged_inner.geom_type == 'LineString' and merged_inner.is_ring:
                inner_polys = [Polygon(merged_inner)]
            elif merged_inner.geom_type == 'MultiLineString':
                for line in merged_inner.geoms:
                    if line.is_ring:
                        inner_polys.append(Polygon(line))
        
        # Add holes to the outer polygon
        for inner_poly in inner_polys:
            for i, outer_poly in enumerate(result_polys):
                if outer_poly.contains(inner_poly.representative_point()):
                    try:
                        new_poly = Polygon(outer_poly.exterior.coords, [inner_poly.exterior.coords])
                        result_polys[i] = new_poly
                    except:
                        pass
    
    if not result_polys:
        return None
    
    # Filter valid polygons
    valid = []
    for p in result_polys:
        if p.is_valid and p.area > 1e-10:
            valid.append(p)
    
    if not valid:
        return None
    
    if len(valid) == 1:
        return valid[0]
    else:
        return MultiPolygon(valid)


def extract_districts(elements, ways, label):
    """Extract admin_level=9 boundary relations with proper geometries."""
    districts = {}
    
    for el in elements:
        if el.get('type') != 'relation':
            continue
        
        tags = el.get('tags', {})
        admin = tags.get('admin_level', '')
        
        if admin != '9':
            continue
        
        name = tags.get('name', '')
        if not name:
            continue
        
        geom = extract_district_geometry(el, ways)
        if geom is None:
            print(f"  Failed to build geometry for {name}", file=sys.stderr)
            continue
        
        if not geom.is_valid:
            geom = geom.buffer(0)
        
        districts[name] = geom
        print(f"  OK: {name} (area={geom.area:.6f} deg²)", file=sys.stderr)
    
    print(f"  {label}: extracted {len(districts)} districts", file=sys.stderr)
    return districts

current_districts = extract_districts(cur_raw.get('elements', []), cur_ways, "Current")
old_districts = extract_districts(old_raw.get('elements', []), old_ways, "Old (2014)")

# ---------------------------------------------------------------------------
# 4. Match districts by name
# ---------------------------------------------------------------------------

common_names = set(current_districts.keys()) & set(old_districts.keys())
only_current = set(current_districts.keys()) - set(old_districts.keys())
only_old = set(old_districts.keys()) - set(current_districts.keys())

print(f"\nCommon: {len(common_names)}", file=sys.stderr)
print(f"Only current: {only_current}", file=sys.stderr)
print(f"Only old: {only_old}", file=sys.stderr)

if not common_names:
    print("ERROR: No matching district names!", file=sys.stderr)
    sys.exit(1)

# ---------------------------------------------------------------------------
# 5. Normalise names
# ---------------------------------------------------------------------------

def normalise(name):
    known = {
        'Innere Stadt': 1, 'Leopoldstadt': 2, 'Landstraße': 3, 'Wieden': 4,
        'Margareten': 5, 'Mariahilf': 6, 'Neubau': 7, 'Josefstadt': 8,
        'Alsergrund': 9, 'Favoriten': 10, 'Simmering': 11, 'Meidling': 12,
        'Hietzing': 13, 'Penzing': 14, 'Rudolfsheim-Fünfhaus': 15,
        'Ottakring': 16, 'Hernals': 17, 'Währing': 18, 'Döbling': 19,
        'Brigittenau': 20, 'Floridsdorf': 21, 'Donaustadt': 22, 'Liesing': 23
    }
    num = known.get(name)
    if num:
        return f"wien-{num:02d}-{name.lower().replace(' ', '-')}"
    return f"wien-{name.lower().replace(' ', '-')}"

# ---------------------------------------------------------------------------
# 6. Compute boundary changes
# ---------------------------------------------------------------------------

change_features = []

for name in common_names:
    cur_geom = current_districts[name]
    old_geom = old_districts[name]
    n_name = normalise(name)
    
    added = cur_geom.difference(old_geom)
    removed = old_geom.difference(cur_geom)
    unchanged = cur_geom.intersection(old_geom)
    
    for geom, ctype in [
        (added, 'added_since_2014'),
        (removed, 'removed_since_2014'),
        (unchanged, 'unchanged')
    ]:
        if geom.is_empty:
            continue
        if geom.geom_type == 'GeometryCollection':
            polys = [g for g in geom.geoms if g.geom_type in ('Polygon', 'MultiPolygon')]
            if not polys:
                continue
            geom = unary_union(polys)
        if geom.geom_type not in ('Polygon', 'MultiPolygon'):
            continue
        if geom.area < 1e-10:
            continue
        if not geom.is_valid:
            geom = geom.buffer(0)
        
        change_features.append({
            'geometry': geom,
            'district_name': n_name,
            'change_type': ctype
        })

for name in only_current:
    geom = current_districts[name]
    if not geom.is_valid:
        geom = geom.buffer(0)
    if geom.area >= 1e-10:
        change_features.append({
            'geometry': geom,
            'district_name': normalise(name),
            'change_type': 'added_since_2014'
        })

for name in only_old:
    geom = old_districts[name]
    if not geom.is_valid:
        geom = geom.buffer(0)
    if geom.area >= 1e-10:
        change_features.append({
            'geometry': geom,
            'district_name': normalise(name),
            'change_type': 'removed_since_2014'
        })

print(f"\nTotal change fragments: {len(change_features)}", file=sys.stderr)

if not change_features:
    result = {"type": "FeatureCollection", "features": []}
    with open('/work/vienna_boundary_changes.geojson', 'w') as f:
        json.dump(result, f, indent=2)
    print("No changes found.", file=sys.stderr)
    sys.exit(0)

# ---------------------------------------------------------------------------
# 7. Determine touches_changed
# ---------------------------------------------------------------------------

changed_geoms = [
    f['geometry'] for f in change_features
    if f['change_type'] in ('added_since_2014', 'removed_since_2014')
]

if changed_geoms:
    changed_union = unary_union(changed_geoms)
else:
    changed_union = None

all_district_names = set(f['district_name'] for f in change_features)
touches_map = {}

for dname in all_district_names:
    touches = False
    district_frags = [f for f in change_features if f['district_name'] == dname]
    
    for f in district_frags:
        frag = f['geometry']
        if changed_union is not None and not changed_union.is_empty:
            if frag.touches(changed_union):
                touches = True
                break
            # Check boundary intersection
            boundary = frag.boundary
            if boundary.intersects(changed_union):
                touches = True
                break
    
    touches_map[dname] = touches

# ---------------------------------------------------------------------------
# 8. Write output GeoJSON
# ---------------------------------------------------------------------------

features_out = []
for f in change_features:
    geom = f['geometry']
    if geom.geom_type == 'Polygon':
        geom = MultiPolygon([geom])
    
    features_out.append({
        'type': 'Feature',
        'geometry': mapping(geom),
        'properties': {
            'change_type': f['change_type'],
            'district_name': f['district_name'],
            'touches_changed': touches_map.get(f['district_name'], False)
        }
    })

result = {
    'type': 'FeatureCollection',
    'features': features_out
}

with open('/work/vienna_boundary_changes.geojson', 'w') as f:
    json.dump(result, f, indent=2)

print(f"\nWritten /work/vienna_boundary_changes.geojson", file=sys.stderr)
print(f"  Total features: {len(features_out)}", file=sys.stderr)
for ctype in ['added_since_2014', 'removed_since_2014', 'unchanged']:
    cnt = sum(1 for f in features_out if f['properties']['change_type'] == ctype)
    print(f"  {ctype}: {cnt}", file=sys.stderr)

print("\nDistricts:", file=sys.stderr)
for dname in sorted(all_district_names):
    tc = touches_map[dname]
    print(f"  {dname}: touches_changed={tc}", file=sys.stderr)

print("\nDone.", file=sys.stderr)
PYEOF
python3 /work/solve3.py

Current ways: 695
Old ways: 364
  OK: Alsergrund (area=0.000359 deg²)
  OK: Hietzing (area=0.004560 deg²)
  OK: Innere Stadt (area=0.000347 deg²)
  OK: Josefstadt (area=0.000132 deg²)
  OK: Leopoldstadt (area=0.002327 deg²)
  OK: Mariahilf (area=0.000176 deg²)
  OK: Meidling (area=0.000980 deg²)
  OK: Neubau (area=0.000195 deg²)
  OK: Penzing (area=0.004086 deg²)
  OK: Rudolfsheim-Fünfhaus (area=0.000474 deg²)
  OK: Währing (area=0.000768 deg²)
  OK: Landstraße (area=0.000896 deg²)
  OK: Brigittenau (area=0.000692 deg²)
  OK: Donaustadt (area=0.012381 deg²)
  OK: Döbling (area=0.003021 deg²)
  OK: Favoriten (area=0.003846 deg²)
  OK: Floridsdorf (area=0.005385 deg²)
  OK: Hernals (area=0.001380 deg²)
  OK: Liesing (area=0.003875 deg²)
  OK: Margareten (area=0.000243 deg²)
  OK: Ottakring (area=0.001049 deg²)
  OK: Simmering (area=0.002812 deg²)
  OK: Wieden (area=0.000215 deg²)
  Current: extracted 23 districts
  OK: Gerasdorf bei Wien (area=0.002136 deg²)
  OK: Alsergrund (area=0.000359 deg²)
  OK: Hietzing (area=0.004560 deg²)
  OK: Innere Stadt (area=0.000347 deg²)
  OK: Josefstadt (area=0.000132 deg²)
  OK: Leopoldstadt (area=0.002335 deg²)
  OK: Mariahilf (area=0.000176 deg²)
  OK: Meidling (area=0.000980 deg²)
  OK: Neubau (area=0.000194 deg²)
  OK: Penzing (area=0.004086 deg²)
  OK: Rudolfsheim-Fünfhaus (area=0.000474 deg²)
  OK: Währing (area=0.000768 deg²)
  OK: Landstraße (area=0.000895 deg²)
  OK: Brigittenau (area=0.000692 deg²)
  OK: Donaustadt (area=0.012374 deg²)
  OK: Döbling (area=0.003021 deg²)
  OK: Favoriten (area=0.003846 deg²)
  OK: Floridsdorf (area=0.005385 deg²)
  OK: Hernals (area=0.001379 deg²)
  OK: Liesing (area=0.003876 deg²)
  OK: Margareten (area=0.000243 deg²)
  OK: Ottakring (area=0.001050 deg²)
  OK: Simmering (area=0.002811 deg²)
  OK: Wieden (area=0.000215 deg²)
  Old (2014): extracted 24 districts

Common: 23
Only current: set()
Only old: {'Gerasdorf bei Wien'}

Total change fragments: 70

Written /work/vienna_boundary_changes.geojson
  Total features: 70
  added_since_2014: 23
  removed_since_2014: 24
  unchanged: 23

Districts:
  wien-01-innere-stadt: touches_changed=True
  wien-02-leopoldstadt: touches_changed=True
  wien-03-landstraße: touches_changed=True
  wien-04-wieden: touches_changed=True
  wien-05-margareten: touches_changed=True
  wien-06-mariahilf: touches_changed=True
  wien-07-neubau: touches_changed=True
  wien-08-josefstadt: touches_changed=True
  wien-09-alsergrund: touches_changed=True
  wien-10-favoriten: touches_changed=True
  wien-11-simmering: touches_changed=True
  wien-12-meidling: touches_changed=True
  wien-13-hietzing: touches_changed=True
  wien-14-penzing: touches_changed=True
  wien-15-rudolfsheim-fünfhaus: touches_changed=True
  wien-16-ottakring: touches_changed=True
  wien-17-hernals: touches_changed=True
  wien-18-währing: touches_changed=True
  wien-19-döbling: touches_changed=True
  wien-20-brigittenau: touches_changed=True
  wien-21-floridsdorf: touches_changed=True
  wien-22-donaustadt: touches_changed=True
  wien-23-liesing: touches_changed=True
  wien-gerasdorf-bei-wien: touches_changed=True

Done.

python3 -c "
import json
with open('/work/vienna_boundary_changes.geojson') as f:
    data = json.load(f)
print('Type:', data['type'])
print('Features:', len(data['features']))
for feat in data['features']:
    props = feat['properties']
    geom = feat['geometry']
    print(f'  {props[\"change_type\"]:20s} {props[\"district_name\"]:35s} touches={props[\"touches_changed\"]}  geom_type={geom[\"type\"]}')
"

Type: FeatureCollection
Features: 70
  added_since_2014     wien-21-floridsdorf                 touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-21-floridsdorf                 touches=True  geom_type=MultiPolygon
  unchanged            wien-21-floridsdorf                 touches=True  geom_type=MultiPolygon
  added_since_2014     wien-15-rudolfsheim-fünfhaus        touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-15-rudolfsheim-fünfhaus        touches=True  geom_type=MultiPolygon
  unchanged            wien-15-rudolfsheim-fünfhaus        touches=True  geom_type=MultiPolygon
  added_since_2014     wien-01-innere-stadt                touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-01-innere-stadt                touches=True  geom_type=MultiPolygon
  unchanged            wien-01-innere-stadt                touches=True  geom_type=MultiPolygon
  added_since_2014     wien-04-wieden                      touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-04-wieden                      touches=True  geom_type=MultiPolygon
  unchanged            wien-04-wieden                      touches=True  geom_type=MultiPolygon
  added_since_2014     wien-05-margareten                  touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-05-margareten                  touches=True  geom_type=MultiPolygon
  unchanged            wien-05-margareten                  touches=True  geom_type=MultiPolygon
  added_since_2014     wien-18-währing                     touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-18-währing                     touches=True  geom_type=MultiPolygon
  unchanged            wien-18-währing                     touches=True  geom_type=MultiPolygon
  added_since_2014     wien-13-hietzing                    touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-13-hietzing                    touches=True  geom_type=MultiPolygon
  unchanged            wien-13-hietzing                    touches=True  geom_type=MultiPolygon
  added_since_2014     wien-09-alsergrund                  touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-09-alsergrund                  touches=True  geom_type=MultiPolygon
  unchanged            wien-09-alsergrund                  touches=True  geom_type=MultiPolygon
  added_since_2014     wien-22-donaustadt                  touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-22-donaustadt                  touches=True  geom_type=MultiPolygon
  unchanged            wien-22-donaustadt                  touches=True  geom_type=MultiPolygon
  added_since_2014     wien-07-neubau                      touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-07-neubau                      touches=True  geom_type=MultiPolygon
  unchanged            wien-07-neubau                      touches=True  geom_type=MultiPolygon
  added_since_2014     wien-14-penzing                     touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-14-penzing                     touches=True  geom_type=MultiPolygon
  unchanged            wien-14-penzing                     touches=True  geom_type=MultiPolygon
  added_since_2014     wien-23-liesing                     touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-23-liesing                     touches=True  geom_type=MultiPolygon
  unchanged            wien-23-liesing                     touches=True  geom_type=MultiPolygon
  added_since_2014     wien-03-landstraße                  touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-03-landstraße                  touches=True  geom_type=MultiPolygon
  unchanged            wien-03-landstraße                  touches=True  geom_type=MultiPolygon
  added_since_2014     wien-20-brigittenau                 touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-20-brigittenau                 touches=True  geom_type=MultiPolygon
  unchanged            wien-20-brigittenau                 touches=True  geom_type=MultiPolygon
  added_since_2014     wien-11-simmering                   touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-11-simmering                   touches=True  geom_type=MultiPolygon
  unchanged            wien-11-simmering                   touches=True  geom_type=MultiPolygon
  added_since_2014     wien-08-josefstadt                  touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-08-josefstadt                  touches=True  geom_type=MultiPolygon
  unchanged            wien-08-josefstadt                  touches=True  geom_type=MultiPolygon
  added_since_2014     wien-19-döbling                     touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-19-döbling                     touches=True  geom_type=MultiPolygon
  unchanged            wien-19-döbling                     touches=True  geom_type=MultiPolygon
  added_since_2014     wien-17-hernals                     touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-17-hernals                     touches=True  geom_type=MultiPolygon
  unchanged            wien-17-hernals                     touches=True  geom_type=MultiPolygon
  added_since_2014     wien-10-favoriten                   touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-10-favoriten                   touches=True  geom_type=MultiPolygon
  unchanged            wien-10-favoriten                   touches=True  geom_type=MultiPolygon
  added_since_2014     wien-02-leopoldstadt                touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-02-leopoldstadt                touches=True  geom_type=MultiPolygon
  unchanged            wien-02-leopoldstadt                touches=True  geom_type=MultiPolygon
  added_since_2014     wien-12-meidling                    touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-12-meidling                    touches=True  geom_type=MultiPolygon
  unchanged            wien-12-meidling                    touches=True  geom_type=MultiPolygon
  added_since_2014     wien-06-mariahilf                   touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-06-mariahilf                   touches=True  geom_type=MultiPolygon
  unchanged            wien-06-mariahilf                   touches=True  geom_type=MultiPolygon
  added_since_2014     wien-16-ottakring                   touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-16-ottakring                   touches=True  geom_type=MultiPolygon
  unchanged            wien-16-ottakring                   touches=True  geom_type=MultiPolygon
  removed_since_2014   wien-gerasdorf-bei-wien             touches=True  geom_type=MultiPolygon

cat << 'PYEOF' > /work/check_areas.py
import json, sys
from shapely.geometry import shape

with open('/work/current_raw.json') as f:
    cur = json.load(f)
with open('/work/old_raw.json') as f:
    old = json.load(f)

# Let me get the raw coordinates directly
# For Alsergrund (current), check the ways
for el in cur['elements']:
    if el.get('type') == 'relation' and el.get('tags',{}).get('name') == 'Alsergrund':
        members = el.get('members', [])
        for m in members[:3]:
            if m.get('type') == 'way':
                print(f"Current way {m['ref']} role={m.get('role')}")
                # Find this way
                for e2 in cur['elements']:
                    if e2.get('type') == 'way' and e2['id'] == m['ref']:
                        geom = e2.get('geometry', [])
                        print(f"  First 2 points: {geom[:2]}")
                        print(f"  Last 2 points: {geom[-2:]}")
                        break
        break

print("\n--- For 2014 ---")
for el in old['elements']:
    if el.get('type') == 'relation' and el.get('tags',{}).get('name') == 'Alsergrund':
        members = el.get('members', [])
        for m in members[:3]:
            if m.get('type') == 'way':
                print(f"Old way {m['ref']} role={m.get('role')}")
                for e2 in old['elements']:
                    if e2.get('type') == 'way' and e2['id'] == m['ref']:
                        geom = e2.get('geometry', [])
                        print(f"  First 2 points: {geom[:2]}")
                        print(f"  Last 2 points: {geom[-2:]}")
                        break
        break
PYEOF
python3 /work/check_areas.py

Current way 147468285 role=outer
  First 2 points: [{'lat': 48.2151672, 'lon': 16.3408703}, {'lat': 48.2152918, 'lon': 16.3409357}]
  Last 2 points: [{'lat': 48.2171389, 'lon': 16.3419295}, {'lat': 48.2172473, 'lon': 16.3419883}]
Current way 8046654 role=outer
  First 2 points: [{'lat': 48.2151672, 'lon': 16.3408703}, {'lat': 48.2152357, 'lon': 16.3433909}]
  Last 2 points: [{'lat': 48.2146161, 'lon': 16.3558154}, {'lat': 48.2143408, 'lon': 16.3565102}]
Current way 147396089 role=outer
  First 2 points: [{'lat': 48.2184891, 'lon': 16.370149}, {'lat': 48.2184449, 'lon': 16.370089}]
  Last 2 points: [{'lat': 48.214294, 'lon': 16.3568662}, {'lat': 48.2143408, 'lon': 16.3565102}]

--- For 2014 ---
Old way 147468285 role=outer
  First 2 points: [{'lat': 48.2151672, 'lon': 16.3408703}, {'lat': 48.2152918, 'lon': 16.3409357}]
  Last 2 points: [{'lat': 48.2171354, 'lon': 16.3419378}, {'lat': 48.2172453, 'lon': 16.3419972}]
Old way 8046654 role=outer
  First 2 points: [{'lat': 48.2151672, 'lon': 16.3408703}, {'lat': 48.2152357, 'lon': 16.3433909}]
  Last 2 points: [{'lat': 48.2154172, 'lon': 16.3460301}, {'lat': 48.2154152, 'lon': 16.3462104}]
Old way 104962880 role=outer
  First 2 points: [{'lat': 48.2151074, 'lon': 16.3495419}, {'lat': 48.2151523, 'lon': 16.3491487}]
  Last 2 points: [{'lat': 48.2154018, 'lon': 16.3463431}, {'lat': 48.2154152, 'lon': 16.3462104}]

Current ways: 695
Old ways: 364
  OK: Alsergrund (area=0.00035913 deg²)
  OK: Hietzing (area=0.00456027 deg²)
  OK: Innere Stadt (area=0.00034707 deg²)
  OK: Josefstadt (area=0.00013189 deg²)
  OK: Leopoldstadt (area=0.00232716 deg²)
  OK: Mariahilf (area=0.00017618 deg²)
  OK: Meidling (area=0.00097967 deg²)
  OK: Neubau (area=0.00019452 deg²)
  OK: Penzing (area=0.00408591 deg²)
  OK: Rudolfsheim-Fünfhaus (area=0.00047390 deg²)
  OK: Währing (area=0.00076838 deg²)
  OK: Landstraße (area=0.00089591 deg²)
  OK: Brigittenau (area=0.00069161 deg²)
  OK: Donaustadt (area=0.01238082 deg²)
  OK: Döbling (area=0.00302130 deg²)
  OK: Favoriten (area=0.00384630 deg²)
  OK: Floridsdorf (area=0.00538470 deg²)
  OK: Hernals (area=0.00137956 deg²)
  OK: Liesing (area=0.00387493 deg²)
  OK: Margareten (area=0.00024329 deg²)
  OK: Ottakring (area=0.00104948 deg²)
  OK: Simmering (area=0.00281169 deg²)
  OK: Wieden (area=0.00021484 deg²)
  Current: extracted 23 districts
  OK: Gerasdorf bei Wien (area=0.00213568 deg²)
  OK: Alsergrund (area=0.00035903 deg²)
  OK: Hietzing (area=0.00456027 deg²)
  OK: Innere Stadt (area=0.00034722 deg²)
  OK: Josefstadt (area=0.00013207 deg²)
  OK: Leopoldstadt (area=0.00233471 deg²)
  OK: Mariahilf (area=0.00017579 deg²)
  OK: Meidling (area=0.00097973 deg²)
  OK: Neubau (area=0.00019409 deg²)
  OK: Penzing (area=0.00408590 deg²)
  OK: Rudolfsheim-Fünfhaus (area=0.00047445 deg²)
  OK: Währing (area=0.00076839 deg²)
  OK: Landstraße (area=0.00089523 deg²)
  OK: Brigittenau (area=0.00069151 deg²)
  OK: Donaustadt (area=0.01237388 deg²)
  OK: Döbling (area=0.00302121 deg²)
  OK: Favoriten (area=0.00384632 deg²)
  OK: Floridsdorf (area=0.00538470 deg²)
  OK: Hernals (area=0.00137908 deg²)
  OK: Liesing (area=0.00387577 deg²)
  OK: Margareten (area=0.00024324 deg²)
  OK: Ottakring (area=0.00104998 deg²)
  OK: Simmering (area=0.00281086 deg²)
  OK: Wieden (area=0.00021481 deg²)
  Old (2014): extracted 24 districts

Common: 23
Only current: set()
Only old: {'Gerasdorf bei Wien'}

Total change fragments: 70

Written /work/vienna_boundary_changes.geojson
  Total features: 70
  added_since_2014: 23
  removed_since_2014: 24
  unchanged: 23

Districts:
  wien-01-innere-stadt: touches_changed=True
  wien-02-leopoldstadt: touches_changed=True
  wien-03-landstraße: touches_changed=True
  wien-04-wieden: touches_changed=True
  wien-05-margareten: touches_changed=True
  wien-06-mariahilf: touches_changed=True
  wien-07-neubau: touches_changed=True
  wien-08-josefstadt: touches_changed=True
  wien-09-alsergrund: touches_changed=True
  wien-10-favoriten: touches_changed=True
  wien-11-simmering: touches_changed=True
  wien-12-meidling: touches_changed=True
  wien-13-hietzing: touches_changed=True
  wien-14-penzing: touches_changed=True
  wien-15-rudolfsheim-fünfhaus: touches_changed=True
  wien-16-ottakring: touches_changed=True
  wien-17-hernals: touches_changed=True
  wien-18-währing: touches_changed=True
  wien-19-döbling: touches_changed=True
  wien-20-brigittenau: touches_changed=True
  wien-21-floridsdorf: touches_changed=True
  wien-22-donaustadt: touches_changed=True
  wien-23-liesing: touches_changed=True
  wien-gerasdorf-bei-wien: touches_changed=True

Done.

python3 -c "
import json
from shapely.geometry import shape

with open('/work/vienna_boundary_changes.geojson') as f:
    data = json.load(f)

for feat in data['features']:
    p = feat['properties']
    geom = shape(feat['geometry'])
    area_m2_approx = geom.area * (111320**2)  # rough conversion to sq meters
    print(f'{p[\"change_type\"]:20s} {p[\"district_name\"]:35s} area={geom.area:.10f} deg² (~{area_m2_approx:.1f} m²)')
"

added_since_2014     wien-08-josefstadt                  area=0.0000000165 deg² (~204.3 m²)
removed_since_2014   wien-08-josefstadt                  area=0.0000002046 deg² (~2535.2 m²)
unchanged            wien-08-josefstadt                  area=0.0001318690 deg² (~1634139.1 m²)
added_since_2014     wien-15-rudolfsheim-fünfhaus        area=0.0000005993 deg² (~7426.8 m²)
removed_since_2014   wien-15-rudolfsheim-fünfhaus        area=0.0000011551 deg² (~14313.9 m²)
unchanged            wien-15-rudolfsheim-fünfhaus        area=0.0004732979 deg² (~5865175.4 m²)
added_since_2014     wien-04-wieden                      area=0.0000001682 deg² (~2084.8 m²)
removed_since_2014   wien-04-wieden                      area=0.0000001432 deg² (~1774.8 m²)
unchanged            wien-04-wieden                      area=0.0002146688 deg² (~2660206.6 m²)
added_since_2014     wien-20-brigittenau                 area=0.0000002889 deg² (~3579.9 m²)
removed_since_2014   wien-20-brigittenau                 area=0.0000001869 deg² (~2316.2 m²)
unchanged            wien-20-brigittenau                 area=0.0006913255 deg² (~8567003.7 m²)
added_since_2014     wien-09-alsergrund                  area=0.0000002837 deg² (~3515.8 m²)
removed_since_2014   wien-09-alsergrund                  area=0.0000001798 deg² (~2227.9 m²)
unchanged            wien-09-alsergrund                  area=0.0003588467 deg² (~4446879.5 m²)
added_since_2014     wien-18-währing                     area=0.0000005988 deg² (~7420.9 m²)
removed_since_2014   wien-18-währing                     area=0.0000006098 deg² (~7556.6 m²)
unchanged            wien-18-währing                     area=0.0007677835 deg² (~9514482.6 m²)
added_since_2014     wien-22-donaustadt                  area=0.0000074600 deg² (~92445.9 m²)
removed_since_2014   wien-22-donaustadt                  area=0.0000005141 deg² (~6370.4 m²)
unchanged            wien-22-donaustadt                  area=0.0123733617 deg² (~153332459.7 m²)
added_since_2014     wien-07-neubau                      area=0.0000005658 deg² (~7011.7 m²)
removed_since_2014   wien-07-neubau                      area=0.0000001305 deg² (~1617.6 m²)
unchanged            wien-07-neubau                      area=0.0001939567 deg² (~2403539.5 m²)
added_since_2014     wien-19-döbling                     area=0.0000003974 deg² (~4924.1 m²)
removed_since_2014   wien-19-döbling                     area=0.0000003100 deg² (~3841.3 m²)
unchanged            wien-19-döbling                     area=0.0030209022 deg² (~37435450.7 m²)
added_since_2014     wien-10-favoriten                   area=0.0000017919 deg² (~22205.5 m²)
removed_since_2014   wien-10-favoriten                   area=0.0000018115 deg² (~22449.0 m²)
unchanged            wien-10-favoriten                   area=0.0038445076 deg² (~47641685.3 m²)
added_since_2014     wien-17-hernals                     area=0.0000014370 deg² (~17807.2 m²)
removed_since_2014   wien-17-hernals                     area=0.0000009507 deg² (~11780.8 m²)
unchanged            wien-17-hernals                     area=0.0013781266 deg² (~17077940.7 m²)
added_since_2014     wien-05-margareten                  area=0.0000000783 deg² (~969.7 m²)
removed_since_2014   wien-05-margareten                  area=0.0000000278 deg² (~343.9 m²)
unchanged            wien-05-margareten                  area=0.0002432144 deg² (~3013947.6 m²)
added_since_2014     wien-06-mariahilf                   area=0.0000004291 deg² (~5317.0 m²)
removed_since_2014   wien-06-mariahilf                   area=0.0000000377 deg² (~466.8 m²)
unchanged            wien-06-mariahilf                   area=0.0001757549 deg² (~2177980.3 m²)
added_since_2014     wien-13-hietzing                    area=0.0000003476 deg² (~4308.0 m²)
removed_since_2014   wien-13-hietzing                    area=0.0000003514 deg² (~4354.9 m²)
unchanged            wien-13-hietzing                    area=0.0045599174 deg² (~56507146.0 m²)
added_since_2014     wien-14-penzing                     area=0.0000013448 deg² (~16664.4 m²)
removed_since_2014   wien-14-penzing                     area=0.0000013301 deg² (~16482.9 m²)
unchanged            wien-14-penzing                     area=0.0040845677 deg² (~50616544.8 m²)
added_since_2014     wien-11-simmering                   area=0.0000010153 deg² (~12581.5 m²)
removed_since_2014   wien-11-simmering                   area=0.0000001879 deg² (~2328.2 m²)
unchanged            wien-11-simmering                   area=0.0028106740 deg² (~34830273.0 m²)
added_since_2014     wien-02-leopoldstadt                area=0.0000003926 deg² (~4864.9 m²)
removed_since_2014   wien-02-leopoldstadt                area=0.0000079394 deg² (~98386.3 m²)
unchanged            wien-02-leopoldstadt                area=0.0023267692 deg² (~28833654.8 m²)
added_since_2014     wien-12-meidling                    area=0.0000004019 deg² (~4980.3 m²)
removed_since_2014   wien-12-meidling                    area=0.0000004635 deg² (~5744.1 m²)
unchanged            wien-12-meidling                    area=0.0009792693 deg² (~12135245.0 m²)
added_since_2014     wien-01-innere-stadt                area=0.0000000896 deg² (~1109.8 m²)
removed_since_2014   wien-01-innere-stadt                area=0.0000002384 deg² (~2954.7 m²)
unchanged            wien-01-innere-stadt                area=0.0003469770 deg² (~4299788.1 m²)
added_since_2014     wien-03-landstraße                  area=0.0000009005 deg² (~11158.8 m²)
removed_since_2014   wien-03-landstraße                  area=0.0000002286 deg² (~2832.9 m²)
unchanged            wien-03-landstraße                  area=0.0008950056 deg² (~11091037.0 m²)
added_since_2014     wien-21-floridsdorf                 area=0.0000002849 deg² (~3530.4 m²)
removed_since_2014   wien-21-floridsdorf                 area=0.0000002916 deg² (~3613.7 m²)
unchanged            wien-21-floridsdorf                 area=0.0053844131 deg² (~66724413.9 m²)
added_since_2014     wien-23-liesing                     area=0.0000011885 deg² (~14727.9 m²)
removed_since_2014   wien-23-liesing                     area=0.0000020249 deg² (~25092.4 m²)
unchanged            wien-23-liesing                     area=0.0038737405 deg² (~48003943.9 m²)
added_since_2014     wien-16-ottakring                   area=0.0000006142 deg² (~7611.7 m²)
removed_since_2014   wien-16-ottakring                   area=0.0000011111 deg² (~13768.5 m²)
unchanged            wien-16-ottakring                   area=0.0010488677 deg² (~12997718.0 m²)
removed_since_2014   wien-gerasdorf-bei-wien             area=0.0021356800 deg² (~26465650.2 m²)

python3 -c "
from shapely.geometry import shape
from shapely.ops import unary_union
import json

with open('/work/vienna_boundary_changes.geojson') as f:
    data = json.load(f)

# Count total area of added and removed
added_total = 0
removed_total = 0
for feat in data['features']:
    p = feat['properties']
    geom = shape(feat['geometry'])
    if p['change_type'] == 'added_since_2014':
        added_total += geom.area
    elif p['change_type'] == 'removed_since_2014':
        removed_total += geom.area

print(f'Total added area: {added_total:.10f} deg² ({added_total*111320*111320:.0f} m²)')
print(f'Total removed area: {removed_total:.10f} deg² ({removed_total*111320*111320:.0f} m²)')

# Check: do added and removed sum to near zero (net balance)?
# Total current area minus total old area should equal added - removed
total_cur = 0
total_old = 0
for feat in data['features']:
    p = feat['properties']
    geom = shape(feat['geometry'])
    if p['change_type'] == 'unchanged':
        total_cur += geom.area
        total_old += geom.area
    elif p['change_type'] == 'added_since_2014':
        total_cur += geom.area
    elif p['change_type'] == 'removed_since_2014':
        total_old += geom.area

print(f'Total current area: {total_cur:.8f} deg²')
print(f'Total old area: {total_old:.8f} deg²')
print(f'Difference (cur - old): {total_cur - total_old:.10f} deg²')
print(f'Net (added - removed): {added_total - removed_total:.10f} deg²')
"

Total added area: 0.0000206947 deg² (256452 m²)
Total removed area: 0.0021561085 deg² (26718803 m²)
Total current area: 0.05019851 deg²
Total old area: 0.05233393 deg²
Difference (cur - old): -0.0021354138 deg²
Net (added - removed): -0.0021354138 deg²

python3 -c "
from shapely.geometry import shape, mapping
import json

# Let's examine a specific district change to see if it's real
with open('/work/current_raw.json') as f:
    cur = json.load(f)
with open('/work/old_raw.json') as f:
    old = json.load(f)

# Check Donaustadt (district 22) - largest change
# Look at the relation contents
for el in cur['elements']:
    if el.get('type') == 'relation' and el.get('tags',{}).get('name') == 'Donaustadt':
        members = el.get('members', [])
        way_ids = [m['ref'] for m in members if m.get('type') == 'way']
        print(f'Current Donaustadt: {len(members)} members, {len(way_ids)} ways')
        print(f'  Way IDs: {sorted(way_ids)}')
        break

for el in old['elements']:
    if el.get('type') == 'relation' and el.get('tags',{}).get('name') == 'Donaustadt':
        members = el.get('members', [])
        way_ids = [m['ref'] for m in members if m.get('type') == 'way']
        print(f'Old Donaustadt: {len(members)} members, {len(way_ids)} ways')
        print(f'  Way IDs: {sorted(way_ids)}')
        break
"

Current Donaustadt: 66 members, 66 ways
  Way IDs: [8030102, 10594407, 30583087, 32488076, 32493565, 32499166, 32506491, 32556105, 33118978, 46627169, 47089864, 59178617, 75435700, 101381543, 125133462, 147463463, 147463465, 147463466, 148189440, 148189441, 148189442, 148189448, 148189449, 148189451, 148189452, 148189457, 148189458, 148189460, 148611510, 171912825, 171979503, 171979504, 171979731, 171980200, 171980201, 263625782, 351930181, 390609705, 390633881, 390638777, 390641760, 390641763, 390724677, 390724681, 390724723, 390754903, 390754905, 390754906, 390858104, 657311974, 657311975, 680969149, 769052960, 769438299, 769438301, 769438302, 796318652, 796318655, 796318656, 796318657, 796318658, 796318659, 917103355, 917103356, 1285228021, 1390987692]
Old Donaustadt: 38 members, 38 ways
  Way IDs: [8030102, 10594407, 29183431, 30583087, 32488076, 32493565, 32499166, 32506491, 32556105, 33118978, 46627169, 47089864, 59178617, 75435700, 101381543, 125133462, 147463463, 147463465, 147463466, 148189440, 148189441, 148189442, 148189448, 148189449, 148189451, 148189452, 148189457, 148189458, 148189460, 148189461, 148611510, 171912825, 171979503, 171979504, 171979731, 171980200, 171980201, 193191812]

python3 -c "
import json
# Validate the GeoJSON
with open('/work/vienna_boundary_changes.geojson') as f:
    data = json.load(f)

assert data['type'] == 'FeatureCollection'
print(f'FeatureCollection with {len(data[\"features\"])} features')
for i, feat in enumerate(data['features']):
    assert feat['type'] == 'Feature'
    assert feat['geometry']['type'] == 'MultiPolygon'
    props = feat['properties']
    assert props['change_type'] in ('added_since_2014', 'removed_since_2014', 'unchanged')
    assert isinstance(props['district_name'], str)
    assert isinstance(props['touches_changed'], bool)
    # Check that indices for coordinates are floats
    for poly in feat['geometry']['coordinates']:
        for ring in poly:
            for pt in ring:
                assert len(pt) == 2
                assert isinstance(pt[0], (int, float))
                assert isinstance(pt[1], (int, float))

print('All validations passed!')
print(f'File size: {len(json.dumps(data))} chars')
"