gws.plugin.alkis.data¶
ALKIS data: reading, indexing, searching and exporting.
This package holds the data layer of the ALKIS plugin. It reads ALKIS source
tables from a PostgreSQL database, builds a compact search index from them in
a separate schema, runs Flurstueck (parcel) and address searches against that
index, and exports search results to files. The alkis action and the
gws alkis CLI commands are built on top of it.
Submodules:
types: the data model. ALKIS entities (Flurstueck,Lage,Gebaeude,Buchungsblatt,Person,Partand so on), their records,EnumPaircode/text values, query objects (FlurstueckQuery,AdresseQueryand their options), theIndexStatusand theReaderinterface for source data.norbit: aReaderfor source tables written by the norBIT ALKIS import.gid: the GeoInfoDok schema used by the reader and the indexer,geo_info_dok.gid6orgeo_info_dok.gid7depending onGWS_ALKIS_GID_VERSION(default7).indexer: builds the index. It reads all relevant ALKIS object types through aReader, links them together in memory and writes the result into the index tables.index: the index node. It defines the index tables, reports the index status, finds Flurstuecke and addresses and loads them with their related data. It also holds helpers to serialize objects to JSON, to normalize search strings and to flatten a Flurstueck into rows for export. The index node is created by thealkisaction, which passes the database,crs,schema,excludeGemarkungandgemarkungFilteroptions. IfgemarkungFilteris set, searches and the street list are restricted to these Gemarkung codes.exporter: a configurable exporter that writes Flurstuecke to CSV or GeoJSON, using models whose field names are flat keys (seeindex.flatten_fs). If no models are configured, a default model with basic Flurstueck fields is used.geo_info_dok: Python modules generated from the GeoInfoDok schema (gid6,gid7) and the generator that creates them.
Design:
Every ALKIS object is represented by an entity with a list of records
(recs), one per version in the object’s life span, sorted by start date.
A record is historic when its life span has ended; an entity is historic when
all of its records are. Code list values are stored as EnumPair objects
holding both the code and the display text.
The indexer works in phases: places (Land, Kreis, Gemeinde, Gemarkung and so
on), Buchung (Buchungsblatt, Buchungsstelle, Namensnummer, Person,
Anschrift), Lage with Gebaeude, Flurstueck data, Parts (Nutzung, Festlegung,
Bewertung, computed by intersecting the source geometries with the
Flurstuecke) and finally the flat search tables. Each object is stored as a
JSONB document; the flat index* tables hold the searchable columns and
refer to Flurstuecke by their uid. Index tables are named
alkis_<VERSION>_<table> and live in the configured index schema.
A search selects matching Flurstueck uids from the flat tables, then loads the JSONB documents and attaches related data depending on the requested display themes, dropping historic objects unless history display is requested.
Example:
from gws.plugin.alkis.data import index, indexer, types as dt
# build the index from the source schema (ix is an ``index.Object``)
indexer.run(ix, 'public', with_force=True)
# find a Flurstueck with its address and buildings
q = dt.FlurstueckQuery(
gemarkungCode='1234',
flurnummer='5',
zaehler='12',
options=dt.FlurstueckQueryOptions(
displayThemes=[dt.DisplayTheme.lage, dt.DisplayTheme.gebaeude],
),
)
for fs in ix.find_flurstueck(q):
print(fs.fsnummer, [la.recs[-1].strasse for la in fs.lageList])
Example:
actions+ {
type "alkis"
dbUid "DB_ALKIS"
dataSchema "public"
indexSchema "gws8"
crs 25832
ui { useExport true }
exporters+ {
type "csv"
title "Basisdaten (CSV)"
models+ {
fields [
{ type "text" name "fs_flurstueckskennzeichen" title "Kennzeichen" }
{ type "float" name "fs_recs_amtlicheFlaeche" title "Flaeche" }
]
}
}
}
Source code: gws.plugin.alkis.data