Data Import Tests¶

View source code

There are 16 tests focused on CSV import (documents, vertices), delimiter handling, type inference nuances, NULL/empty values, stats, error cases, batch commits, Unicode, and a small performance smoke.

Quick Start¶

CSV Import (documents)¶

import arcadedb_embedded as arcadedb

with arcadedb.create_database("./mydb") as db:
    db.command("sql", "CREATE DOCUMENT TYPE Person")

    stats = arcadedb.import_csv(db, "data.csv", type_name="Person")
    print(stats)

CSV Import (vertices)¶

with arcadedb.create_database("./mydb") as db:
    db.command("sql", "CREATE VERTEX TYPE Product")

    stats = arcadedb.import_csv(
        db,
        "products.csv",
        type_name="Product",
        import_type="vertices",
        typeIdProperty="id",
    )

Test Cases¶

1) CSV as documents¶

Creates a Person document type and imports three rows; asserts stats documents == 3, errors == 0, and verifies properties for Alice/New York.

2) CSV as vertices¶

Creates a Product vertex type and imports three rows as vertices using typeIdProperty="id"; asserts vertices == 3, documents == 0, and checks name/category values.

3) Custom delimiter (TSV)¶

csv_content = """name|age|city
Alice|30|NYC"""

stats = arcadedb.import_csv(db, "data.tsv", type_name="Item", delimiter="\t")

4) CSV type inference¶

csv_content = """name,age,active,score,notes
Alice,30,true,98.5,
Bob,25,false,87.3,Some text"""

stats = arcadedb.import_csv(db, "data.csv", type_name="Person")

result = db.query("sql", "SELECT FROM Person WHERE name = 'Alice'")
alice = list(result)[0]

assert isinstance(record.get("count"), int)
assert isinstance(record.get("price"), float)
# Booleans are strings with the Java importer
assert isinstance(record.get("active"), str)

Type inference observations from tests:

Numeric strings map to int/float
Empty strings may be None or empty string, depending on importer/schema
Boolean strings are imported as strings (e.g., "true", "false")

Import Options¶

Common Options¶

stats = arcadedb.import_csv(
    db,
    file_path,
    type_name="MyType",
    commitEvery=1000,
)

CSV Options¶

stats = arcadedb.import_csv(
    db,
    "data.csv",
    type_name="Person",
    import_type="documents",  # "documents" or "vertices" (edges not covered in tests)
    delimiter=",",            # Field delimiter
    commitEvery=1000
)

Note: JSON/JSONL import via IMPORT DATABASE is not exercised in this test file.

Import Statistics¶

Importer returns statistics:

stats = arcadedb.import_csv(db, "data.csv", type_name="Person")

print(stats)
# {
#     'documents': 3,
#     'vertices': 0,
#     'edges': 0,
#     'errors': 0,
#     'duration_ms': 123
# }

Running These Tests¶

# Run all import tests
pytest tests/test_importer.py -v

# Run with output
pytest tests/test_importer.py -v -s

Best Practices¶

✅ Use appropriate batch size¶

# Large files: increase batch size
arcadedb.import_csv(db, "huge_file.csv", type_name="Data", commitEvery=100)

# Small files: default is fine
arcadedb.import_csv(db, "small_file.csv", type_name="Data")

✅ Create types before importing¶

# Define schema first for better performance
db.command("sql", "CREATE DOCUMENT TYPE Person")
arcadedb.import_csv(db, "people.csv", type_name="Person")

✅ Handle import errors¶

try:
    stats = arcadedb.import_csv(db, "data.csv", type_name="Person")
    print(f"Imported {stats['records_imported']} records")
except Exception as e:
    print(f"Import failed: {e}")
    # Handle error, rollback, etc.