DanielCarter-stack opened a new pull request, #11756:
URL: https://github.com/apache/seatunnel/pull/11756
### Summary
Docs-only pass that improves the `Console`, `DB2`, `Snowflake`,
`FakeSource`, and `TiDB-CDC` connector pages in both `docs/en` and `docs/zh`.
No source, config, SPI, or runtime code is touched.
### Why these five
All five were last touched well outside the seven-day "do not double-touch"
window used by the connector-docs batches (verified across merged commits and
open PRs from `DanielCarter-stack`).
### What changed
**DB2** (`docs/en/connectors/{source,sink}/DB2.md` + `docs/zh/...`)
- Corrected the `partition_num` Default cell from the misleading `"job
parallelism"` to the actual hardcoded default of `10` (mirrors
`JdbcSourceTableConfig.DEFAULT_PARTITION_NUMBER`).
- Corrected `partition_lower_bound`/`partition_upper_bound` types from
`BigDecimal` to `String` to match
`JdbcSourceOptions.PARTITION_LOWER_BOUND`/`PARTITION_UPPER_BOUND`.
- Added a description column to the options table for every row.
- Converted example code blocks to `hocon` fences and removed the misplaced
`partition_column` tip from the sink page (it is a source-only concept).
- Refreshed the `Key Features` block: source is **not** exactly-once; sink
**is** exactly-once (XA) and supports CDC via primary-key merge SQL;
batch/stream/multi-table/timer-flush capabilities are now marked explicitly.
**Snowflake** (`docs/en/connectors/{source,sink}/Snowflake.md` +
`docs/zh/...`)
- Same partition-related fixes as DB2.
- Documented Snowflake-specific JDBC URL parameters such as
`GEOGRAPHY_OUTPUT_FORMAT` (both in the source options description and in a
Notes callout) and pointed readers to the Snowflake Geospatial Data Types
reference.
- Tidied the page structure (`Support Those Engines` heading, consistent
table layout, hocon fences).
**FakeSource** (`docs/en/connectors/source/FakeSource.md` + `docs/zh/...`)
- Renamed the engines section from `Supported Engines` to `Support Those
Engines` for consistency with the rest of the connector docs.
- Standardised the Required column casing to `No`/`Yes`.
- Added a `Key Features` callout describing the per-subtask split
enumeration model (driven by `split.num`/`split.read-interval`).
**TiDB-CDC** (`docs/en/connectors/source/TiDB-CDC.md` + `docs/zh/...`)
- Removed the misleading `Spark` entry from the engines block; TiDB-CDC only
supports `SeaTunnel Zeta` and `Flink`. The previous doc listed `Spark` as a
supported engine even though no Spark source is registered.
- Expanded the options table with prose descriptions for `pd-addresses`,
`startup.mode`, and the `tikv.grpc.*` / `tikv.batch_*_concurrency` knobs, plus
reorganised the source/sink example into proper `hocon` fences.
**Console** (`docs/en/connectors/sink/Console.md` + `docs/zh/...`)
- Added a `Key Features` callout about schema-evolution handling
(`ADD_COLUMN`, `DROP_COLUMN`, `RENAME_COLUMN`, `UPDATE_COLUMN`) that the sink
already implements via `SupportSchemaEvolutionSink`.
### Backward compatibility
Pure documentation change. No config defaults, option contracts, or runtime
behavior are altered. No changelog entries need updating.
### Checklist
- [x] Docs-only change
- [x] `docs/en` and `docs/zh` updated together where the connector has a
Chinese mirror
- [x] No source files modified
- [x] No config option names, types, or defaults changed
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
