[misc] rejstrik trestu PO

data/misc/README.md

@@ -0,0 +1,5 @@
+# Misc
+
+Náhodné další datasety, které jsou a) malé, b) jednotabulkové
+
+- [Rejstřík trestů právnických osob](https://eservice-po.rejtr.justice.cz/public/odsouzeni;jsessionid=E925E6A2467577AB9730F7F9BA96D3E9.pocluster1?0)

data/misc/main.py

@@ -0,0 +1,19 @@
+import csv
+import io
+import os
+from urllib.request import urlopen
+
+URL_ODSOUZENE_PO = "https://eservice-po.rejtr.justice.cz/public/odsouzeni_csv"
+
+
+def main(outdir: str, partial: bool = False):
+    with urlopen(URL_ODSOUZENE_PO) as r, open(
+        os.path.join(outdir, "odsouzene_po.csv"), "w", encoding="utf-8"
+    ) as fw:
+        cw = csv.writer(fw, lineterminator="\n")
+        for row in csv.reader(io.TextIOWrapper(r, encoding="utf-8")):
+            cw.writerow(row)
+
+
+if __name__ == "__main__":
+    main(".")

data/misc/schema.py

@@ -0,0 +1,27 @@
+from sqlalchemy import Column, MetaData, Table
+from sqlalchemy.sql.sqltypes import Integer, Text
+
+meta = MetaData()
+
+
+schema = [
+    Table(
+        "odsouzene_po",
+        meta,
+        Column("ico", Integer, nullable=True, index=True),
+        Column("obchodni_jmeno", Text, nullable=False),
+        Column("sidlo", Text, nullable=True),
+        Column("stat", Text, nullable=False),
+        Column("odsouzeni", Text, nullable=False),
+    )
+]
+
+
+if __name__ == "__main__":
+    from sqlalchemy import create_engine
+    from sqlalchemy.schema import CreateTable
+
+    engine = create_engine("sqlite:///:memory:")
+    for table in schema:
+        print(f"-- {table.name} as created in SQLite")
+        print(CreateTable(table).compile(engine))