Merge pull request #14 from wp-labs/feature/postgres

Add the configuration description document for the postgres sink

Author: seeker-zuo

docs-en/10-user/05-connectors/02-sinks/23-postgresql_sink.md

@@ -0,0 +1,68 @@
+# PostgreSQL Sink
+
+The PostgreSQL sink writes parsed records into PostgreSQL tables. It builds `INSERT IGNORE` statements from `columns`, which is suitable for idempotent retry scenarios. It only accepts Record data (raw input is not supported).
+
+## Connector Definition
+
+Use the built-in template (located at `connectors/sink.d/50-mysql.toml`; PostgreSQL uses the same configuration set as MySQL):
+
+```toml
+[[connectors]]
+id = "postgresql_sink"
+type = "postgresql"
+allow_override = ["endpoint", "username", "password", "database", "table", "columns", "batch_size"]
+
+[connectors.params]
+endpoint = "localhost:5432"
+username = "postgres"
+password = "123456"
+database = "wparse"
+table = "nginx_logs"
+columns = ["sip", "timestamp", "http/request", "status", "size", "referer", "http/agent", "wp_event_id"]
+batch_size = 1024
+```
+
+## Parameters
+
+| Parameter | Type | Description |
+|------|------|------|
+| `endpoint` | string | PostgreSQL endpoint (`host:port`, required) |
+| `username` | string | Username (optional, default `postgres`) |
+| `password` | string | Password (optional) |
+| `database` | string | Target database (required) |
+| `table` | string | Target table name (required) |
+| `columns` | array | Column list that defines write order (required) |
+| `batch_size` | int | Batch insert size (optional) |
+
+## Configuration Example
+
+### Basic Usage
+
+```toml
+version = "2.0"
+
+[sink_group]
+name = "all"
+oml = ["/*"]
+batch_timeout_ms=5000   # Auto flush when batch size is not reached within this time window
+parallel = 8
+
+[[sink_group.sinks]]
+name = "main"
+connect = "postgresql_sink"
+
+[sink_group.sinks.params]
+endpoint = "localhost:5432"
+username = "postgres"
+password = "123456"
+database = "wparse"
+table = "nginx_logs"
+columns = ["sip", "timestamp", "http/request", "status", "size", "referer", "http/agent", "wp_event_id"]
+batch_size = 1024
+```
+
+## Notes
+
+- Field names in `columns` must match OML output fields. Missing table columns are written as `NULL`.
+- You can override the connection string with `POSTGRESQL_URL` (format: `postgresql://user:pass@host:port/db`).
+- PostgreSQL uses the same configuration keys as MySQL. Only connector `type`, port, and connection details differ.

docs-en/10-user/05-connectors/02-sinks/README.md

@@ -11,6 +11,7 @@ This guide introduces how to configure and use various data outputs (Sinks) in t
 | `syslog` | Output to Syslog server (UDP/TCP) | [Syslog Sink Configuration](./13-syslog_sink.md) |
 | `tcp` | Output to TCP server | [TCP Sink Configuration](./15-tcp_sink.md) |
 | `kafka` | Output to Kafka | - |
+| `postgresql` | Output to PostgreSQL | [PostgreSQL Sink Configuration](./23-postgresql_sink.md) |
 | `doris` | Output to Doris | [Doris Sink Configuration](./17-doris.md) |
 | `elasticsearch` | Output to Elasticsearch | [Elasticsearch Sink Configuration](./20-elasticsearch_sink.md) |
 | `prometheus` | Prometheus metrics exposure | [Prometheus Sink Configuration](./14-prometheus_sink.md) |

docs-zh/10-user/05-connectors/02-sinks/23-postgresql_sink.md

@@ -0,0 +1,68 @@
+# PostgreSQL Sink
+
+PostgreSQL sink 用于将解析后的记录写入 PostgreSQL 表。它会根据 `columns` 生成 `INSERT IGNORE` 语句，适合幂等重试场景；仅接受 Record 数据（不支持 raw 输入）。
+
+## 连接器定义
+
+推荐使用仓库自带模板（位于 `connectors/sink.d/50-mysql.toml`，配置项与 MySQL 保持一致）：
+
+```toml
+[[connectors]]
+id = "postgresql_sink"
+type = "postgresql"
+allow_override = ["endpoint", "username", "password", "database", "table", "columns", "batch_size"]
+
+[connectors.params]
+endpoint = "localhost:5432"
+username = "postgres"
+password = "123456"
+database = "wparse"
+table = "nginx_logs"
+columns = ["sip", "timestamp", "http/request", "status", "size", "referer", "http/agent", "wp_event_id"]
+batch_size = 1024
+```
+
+## 可用参数
+
+| 参数 | 类型 | 说明 |
+|------|------|------|
+| `endpoint` | string | PostgreSQL 地址（`host:port`，必填） |
+| `username` | string | 用户名（可选，默认 `postgres`） |
+| `password` | string | 密码（可选） |
+| `database` | string | 目标数据库（必填） |
+| `table` | string | 目标表名（必填） |
+| `columns` | array | 列名列表，决定写入字段顺序（必填） |
+| `batch_size` | int | 批量写入条数（可选） |
+
+## 配置示例
+
+### 基础用法
+
+```toml
+version = "2.0"
+
+[sink_group]
+name = "all"
+oml = ["/*"]
+batch_timeout_ms=5000   # 当数据在设置的这个时间范围内不满足批量插入数量时自动插入
+parallel = 8
+
+[[sink_group.sinks]]
+name = "main"
+connect = "postgresql_sink"
+
+[sink_group.sinks.params]
+endpoint = "localhost:5432"
+username = "postgres"
+password = "123456"
+database = "wparse"
+table = "nginx_logs"
+columns = ["sip", "timestamp", "http/request", "status", "size", "referer", "http/agent", "wp_event_id"]
+batch_size = 1024
+```
+
+## 注意事项
+
+- `columns` 中的字段名需与 OML 输出字段一致；缺失的表字段会以 `NULL` 写入。
+- 可通过环境变量 `POSTGRESQL_URL` 覆盖连接串（格式：`postgresql://user:pass@host:port/db`）。
+- PostgreSQL 的配置项与 MySQL 完全一致，仅连接器 `type`、端口与连接信息不同。

docs-zh/10-user/05-connectors/02-sinks/README.md

@@ -12,6 +12,7 @@
 | `tcp` | 输出到 TCP 服务端 | [TCP Sink 配置](./15-tcp_sink.md) |
 | `kafka` | 输出到 Kafka | [Kafka Sink 配置](./18-kafka_sink.md) |
 | `mysql` | 输出到 MySQL | [MySQL Sink 配置](./19-mysql_sink.md) |
+| `postgresql` | 输出到 PostgreSQL | [PostgreSQL Sink 配置](./23-postgresql_sink.md) |
 | `doris` | 输出到 Doris | [Doris Sink 配置](./17-doris.md) |
 | `elasticsearch` | 输出到 Elasticsearch | [Elasticsearch Sink 配置](./20-elasticsearch_sink.md) |
 | `prometheus` | Prometheus 指标暴露 | [Prometheus Sink 配置](./14-prometheus_sink.md) |