avro_schema_serialize_deserialize_file_quickstart.py

1import avro.schema
2from avro.datafile import DataFileReader, DataFileWriter
3from avro.io import DatumReader, DatumWriter
4
5# 1. Define the schema
6schema_json = """
7{
8    "namespace": "example.avro",
9    "type": "record",
10    "name": "User",
11    "fields": [
12        {"name": "name", "type": "string"},
13        {"name": "favorite_number",  "type": ["int", "null"]},
14        {"name": "favorite_color", "type": ["string", "null"]}
15    ]
16}
17"""
18schema = avro.schema.parse(schema_json)
19
20# 2. Write data to an Avro file
21with open('users.avro', 'wb') as f:
22    writer = DataFileWriter(f, DatumWriter(), schema)
23    writer.append({"name": "Alyssa", "favorite_number": 256})
24    writer.append({"name": "Ben", "favorite_number": 7, "favorite_color": "red"})
25    writer.close()
26
27# 3. Read data back from the Avro file
28with open('users.avro', 'rb') as f:
29    reader = DataFileReader(f, DatumReader())
30    for user in reader:
31        print(user)
32    reader.close()