rivet-cli 0.20.0

Rivet: PostgreSQL/MySQL/SQL Server/MongoDB → Parquet/CSV (local, S3, GCS, Azure). Crate name rivet-cli; binary rivet.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91

# Cross config × database coverage matrix — cross-cutting config options /
# behaviours, is each proven per engine? Same shape + drift-guard as the other
# matrices. Cell = { test } | { gap } | { na }.
#
# Many gaps here are honest holes (cloud destinations, codec parity, csv, tuning
# profile) on the non-PostgreSQL engines — PG has the deepest coverage, the others
# ride the shared matrix/differential harness for the happy path. The ratchet makes
# the holes visible and un-growable; fill them by writing the per-engine test.

engines: [postgres, mysql, mssql, mongo]

scenarios:
  # ── destinations ──────────────────────────────────────────────────────────
  - id: dest_local
    what: "destination: local — one/many parquet parts with all rows"
    postgres: { test: local_destination_produces_one_parquet_with_all_rows }
    mysql:    { test: mysql_full_mode_repeated_run_accumulates_manifest_entries }
    mssql:    { test: mssql_full_mode_repeated_run_accumulates_manifest_entries }
    mongo:    { test: mongo_batch_keyset_no_loss_or_dup }

  - id: dest_gcs
    what: "destination: gcs — fake-GCS round-trip, all rows"
    postgres: { test: gcs_fake_destination_produces_one_parquet_with_all_rows }
    mysql:    { test: stand_dest_gcs_mysql }
    mssql:    { test: stand_dest_gcs_mssql }
    mongo:    { test: stand_dest_gcs_mongo }

  - id: dest_s3
    what: "destination: s3 — MinIO round-trip, all rows"
    postgres: { test: s3_minio_destination_produces_one_parquet_with_all_rows }
    mysql:    { test: stand_dest_s3_mysql }
    mssql:    { test: stand_dest_s3_mssql }
    mongo:    { test: stand_dest_s3_mongo }

  # ── format / compression ──────────────────────────────────────────────────
  - id: format_parquet
    what: "format: parquet — the default, exercised by every matrix export"
    postgres: { test: pg_golden_full_type_matrix_schema_coverage }
    mysql:    { test: mysql_golden_full_type_matrix_schema_coverage }
    mssql:    { test: duckdb_validates_mssql_type_matrix_parquet }
    mongo:    { test: mongo_batch_keyset_no_loss_or_dup }

  - id: format_csv
    what: "format: csv — decimal-as-text, unicode/newline escaping"
    postgres: { test: postgres_type_matrix_csv_decimal_and_row_count }
    mysql:    { test: mysql_type_matrix_csv_decimal_text_preserved }
    mssql:    { test: stand_format_csv_mssql }
    mongo:    { test: stand_format_csv_mongo }

  - id: compression_codecs
    what: "compression: full codec matrix (zstd/snappy/gzip/none) reads back correctly"
    postgres: { test: duckdb_validates_postgres_parquet_compression_codecs }
    mysql:    { test: stand_compression_codecs_mysql }
    mssql:    { test: stand_compression_codecs_mssql }
    mongo:    { test: stand_compression_codecs_mongo }

  # ── enrichment / verification / tuning ────────────────────────────────────
  - id: meta_columns
    what: "meta_columns — _rivet_exported_at + _rivet_row_hash land in the output"
    postgres: { test: stand_meta_columns_postgres }
    mysql:    { na: "engine-agnostic post-read enrichment (src/enrich.rs); one e2e test covers every engine" }
    mssql:    { na: "engine-agnostic — see the postgres cell" }
    mongo:    { na: "engine-agnostic — see the postgres cell" }

  - id: reconcile
    what: "--reconcile — destination count matches source (exit 0), mismatch exits nonzero"
    postgres: { test: reconcile_all_match_exits_zero_with_pretty_output }
    mysql:    { test: mysql_reconcile_all_match_exits_zero_with_pretty_output }
    mssql:    { test: mssql_reconcile_all_match_exits_zero_with_pretty_output }
    mongo:    { test: mongo_run_reconcile_matches_source_count }

  - id: parallel_n
    what: "parallel: N — concurrent workers, no loss or dup"
    postgres: { test: dense_ties_pg_parallel_no_loss_no_dup }
    mysql:    { test: dense_ties_mysql_parallel_no_loss_no_dup }
    mssql:    { test: dense_ties_mssql_parallel_no_loss_no_dup }
    mongo:    { test: mongo_batch_parallel_integer_id_no_loss_or_dup }

  - id: environment_tuning_profile
    what: "source.environment → tuning profile default (local → fast)"
    postgres: { test: source_environment_local_runs_fast_profile_by_default }
    mysql:    { test: stand_environment_profile_mysql }
    mssql:    { test: stand_environment_profile_mssql }
    mongo:    { na: "the tuning-profile governor (Log Flush Waits back-pressure) is SQL-only; no Mongo analogue" }

  - id: cross_db_parity
    what: "same seed on two engines → equal row count + equal PK id-set + identical empty-export behaviour"
    postgres: { test: pg_and_mysql_full_exports_agree_on_row_count_and_id_set }
    mysql:    { test: pg_and_mysql_chunked_exports_agree_on_row_count_and_id_set }
    mssql:    { test: pg_and_mssql_full_exports_agree_on_row_count_and_id_set }
    mongo:    { na: "Mongo's JSON-blob model has no row-for-row SQL parity to compare against" }