DanielCarter-stack opened a new pull request, #11796:
URL: https://github.com/apache/seatunnel/pull/11796
## Summary
This PR improves five connector docs (English only) that have not been
touched by any open or recently merged (last 7 days) PR.
### Files Changed
- `docs/en/connectors/sink/Doris.md`
- `docs/en/connectors/sink/IoTDBv2.md`
- `docs/en/connectors/sink/Paimon.md`
- `docs/en/connectors/sink/SensorsData.md`
- `docs/en/connectors/sink/StarRocks.md`
### Doris Sink
- Repaired broken code fences: the `Use JSON format` and `Use CSV format`
examples were rendered as plain text instead of HOCON, and the closing `###
Case-Sensitive Configuration` header was nested inside the CSV block. Both
examples now use proper ` ```hocon ` fences and the case-sensitivity section
sits at the correct nesting level.
- Rewrote the `doris.config` option description to point at the actual
Stream Load keys (`format`, `read_json_by_line`, `column_separator`,
`row_delimiter`) instead of the vague "support operations such as insert,
delete, and update" wording.
- Replaced the misleading "undocumented `column_mapping`" reference in the
case-sensitivity FAQ with an accurate note about `case_sensitive = true|false`
and using a `Rename` transform to align upstream field names.
- Added two new examples aligned with the Doris e2e fixtures:
- `Custom SQL Pre-processing` covering `data_save_mode =
"CUSTOM_PROCESSING"` with a `custom_sql` that runs on the target cluster before
the sync task.
- `CDC With Schema Change` showing MySQL-CDC streaming into Doris with
`schema-changes.enabled = true`.
### StarRocks Sink
- Repaired missing ` ```hocon ` fences on the `Use JSON format to import
data`, `Use CSV format to import data`, and `Use save_mode function` examples
(all three were rendered as plain code blocks, breaking the docs site syntax
highlighting).
- Rewrote the `What is labelPrefix used for in StarRocks Sink?` FAQ entry.
The previous version opened with a self-referential sentence ("The current
StarRocks Sink page does not list exactly-once as a supported connector
capability.") which read like an editorial leftover; the new version explains
what `labelPrefix` does and explicitly references the Key Features matrix for
the exactly-once question.
- Added two examples drawn from the e2e fixtures:
- `CDC With Schema Change` showing MySQL-CDC streaming into StarRocks with
`schema-changes.enabled = true` and a custom `save_mode_create_template`.
- `Timer Flush With MySQL-CDC` showing `sink.flush.interval = 500`
together with `batch_max_rows = 100000` and `batch_max_bytes = 104857600`.
### Paimon Sink
- Fixed the description line which mixed full-width Chinese punctuation
(`cdc mode 、auto create table`) into the English description.
- Restructured the `## Changelog` section (which was actually
changelog-producer documentation but used a confusing H2 name) into a proper
`## Changelog Producer` section with English punctuation and a clearly labeled
four-mode list.
- Normalized the `## Schema Evolution` section: replaced `Cdc Ingestion`
with `CDC ingestion`, replaced Chinese full-width punctuation with ASCII, and
converted the bullet list into consistent Markdown form.
- Normalized the `## Filesystems` section: replaced full-width Chinese
punctuation, expanded the S3 config key list to a comma-separated form, and
dropped the duplicate "Besides, the warehouse should start with `s3a://`"
sentence.
- Added `Truncate target table on startup with Hive catalog` example showing
`schema_save_mode = "RECREATE_SCHEMA"` together with `data_save_mode =
"DROP_DATA"` for full-refresh batch jobs into a Hive-catalogued Paimon table.
### IoTDBv2 Sink
- Added `Case 4: Combined TAG, ATTRIBUTE, and FIELD columns` to the
IoTDB-table section. The existing Cases 1–3 cover the table dialect in pieces
(only required fields, then TAG+ATTRIBUTE, then FIELD-only); this fourth case
wires all three column categories together with both `key_measurement_fields`
and `key_tag_fields` / `key_attribute_fields` set, with the matching `IoTDB>
SELECT` and `IoTDB> DESC` output blocks.
### SensorsData Sink
- Fixed `Profile Property Updates` example which used `record_type =
profile` — not a valid value. The valid `record_type` values per the connector
source are `users`, `events`, `items`, and `details`; the example now uses
`record_type = users` (the correct user-profile-update case) and adds a brief
explanation paragraph.
- Added two missing examples drawn from the e2e fixtures:
- `User Details (Detail Records)` covering `record_type = "details"` with
`detail_id_column`, `distinct_id_column`, and `identity_fields`.
- `Dynamic Event Names With Multiple Identities` showing `event_name =
"${c_event}"` together with two `$identity_*` mappings and `skip_error_record =
true`.
### Out of Scope
- Version compatibility tables and `support version` markers are
intentionally not modified per the working agreement.
- No Chinese (zh) docs are touched in this batch — they remain blocked by
other open PRs or recent merges and will be addressed in a later batch.
--
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]
