matrixorigin · yangj1211 · Oct 28, 2024 · Oct 25, 2024
diff --git a/docs/MatrixOne/Develop/export-data/select-into-outfile.md b/docs/MatrixOne/Develop/export-data/select-into-outfile.md
@@ -7,15 +7,15 @@ MatrixOne 支持以下两种方式导出数据：
 
 本篇文档主要介绍如何使用 `SELECT INTO...OUTFILE` 导出数据。
 
-使用 `SELECT...INTO OUTFILE` 语法可以将表数据导出到主机上的文本文件中。
+使用 `SELECT...INTO OUTFILE` 语法可以将表数据导出到主机上的文本文件或者 stage 中。
 
 ## 语法结构
 
-`SELECT...INTO OUTFILE` 语法是 `SELECT` 语法和 `INTO OUTFILE filename` 的结合。默认输出格式与 `LOAD DATA` 命令相同。因此，以下语句是将名称为 **test** 的表导出到目录路径为 **/root/test** 的*. csv* 文件中。
+`SELECT...INTO OUTFILE` 语法是 `SELECT` 语法和 `INTO OUTFILE filename` 的结合。默认输出格式与 `LOAD DATA` 命令相同。
 
 ```
-mysql> SELECT * FROM TEST
-    -> INTO OUTFILE '/root/test.csv';
+mysql> SELECT * FROM <table_name>
+    -> INTO OUTFILE '<filepath>|<stage://stage_name>';
 ```
 
 你可以采用多种形式和选项更改输出格式，用于表示如何引用、分隔列和记录。
@@ -70,7 +70,11 @@ sudo docker run --name <name> --privileged -d -p 6001:6001 -v ${local_data_path}
     +------+-----------+------+
     ```
 
-2. 对于使用源代码或二进制文件的方式安装构建 MatrixOne，将表导出到本地目录，例如 *~/tmp/export_demo/export_datatable.txt*，命令示例如下：
+2. 数据导出
+
+   - 导出到本地
+
+   对于使用源代码或二进制文件的方式安装构建 MatrixOne，将表导出到本地目录，例如 *~/tmp/export_demo/export_datatable.txt*，命令示例如下：
 
     ```
     select * from user into outfile '~/tmp/export_demo/export_datatable.txt'
@@ -82,9 +86,29 @@ sudo docker run --name <name> --privileged -d -p 6001:6001 -v ${local_data_path}
     select * from user into outfile 'mo-data/export_datatable.txt';
     ```
 
-3. 到你本地 *export_datatable.txt* 文件下查看导出情况：
+    - 导出到 satge
 
+    ```sql
+    create stage stage_fs url = 'file:///Users/admin/test';
+    select * from user into outfile 'stage://stage_fs/user.csv';
+    ```
+
+3. 查看导出情况：
+
+    - 导出到本地
+
+    ```
+    (base) admin@192 test % cat export_datatable.txt 
+    id,user_name,sex
+    1,"weder","man"
+    2,"tom","man"
+    3,"wederTom","man"
     ```
+
+    - 导出到 stage
+
+    ```bash
+    (base) admin@192 test % cat user.csv 
     id,user_name,sex
     1,"weder","man"
     2,"tom","man"

diff --git a/docs/MatrixOne/Develop/import-data/bulk-load/1.1-load-s3.md b/docs/MatrixOne/Develop/import-data/bulk-load/1.1-load-s3.md
diff --git a/docs/MatrixOne/Develop/import-data/bulk-load/load-csv.md b/docs/MatrixOne/Develop/import-data/bulk-load/load-csv.md
@@ -7,37 +7,45 @@
 - 场景一：数据文件与 MatrixOne 服务器在同一台机器上：
 
 ```
-LOAD DATA
-INFILE 'file_name'
-INTO TABLE tbl_name
-[{FIELDS | COLUMNS}
-[TERMINATED BY 'string']
-[[OPTIONALLY] ENCLOSED BY 'char']
-]
-[LINES
-[STARTING BY 'string']
-[TERMINATED BY 'string']
-]
-[IGNORE number {LINES | ROWS}]
-[PARALLEL {'TRUE' | 'FALSE'}]
+> LOAD DATA 
+    INFILE '<file_name>|<stage://stage_name/filepath>'
+    INTO TABLE tbl_name
+    [CHARACTER SET charset_name]
+    [{FIELDS | COLUMNS}
+        [TERMINATED BY 'string']
+        [[OPTIONALLY] ENCLOSED BY 'char']
+        [ENCASPED BY 'char']
+    ]
+    [LINES
+        [STARTING BY 'string']
+        [TERMINATED BY 'string']
+    ]
+    [IGNORE number {LINES | ROWS}]
+    [SET column_name_1=nullif(column_name_1, expr1), column_name_2=nullif(column_name_2, expr2)...]
+    [PARALLEL {'TRUE' | 'FALSE'}]
+    [STRICT {'TRUE' | 'FALSE'}]
 ```
 
 - 场景二：数据文件与 MatrixOne 服务器在不同的机器上：
 
 ```
-LOAD DATA LOCAL
-INFILE 'file_name'
-INTO TABLE tbl_name
-[{FIELDS | COLUMNS}
-[TERMINATED BY 'string']
-[[OPTIONALLY] ENCLOSED BY 'char']
-]
-[LINES
-[STARTING BY 'string']
-[TERMINATED BY 'string']
-]
-[IGNORE number {LINES | ROWS}]
-[PARALLEL {'TRUE' | 'FALSE'}]
+> LOAD DATA LOCAL
+    INFILE '<file_name>|<stage://stage_name/filepath>'
+    INTO TABLE tbl_name
+    [CHARACTER SET charset_name]
+    [{FIELDS | COLUMNS}
+        [TERMINATED BY 'string']
+        [[OPTIONALLY] ENCLOSED BY 'char']
+        [ENCASPED BY 'char']
+    ]
+    [LINES
+        [STARTING BY 'string']
+        [TERMINATED BY 'string']
+    ]
+    [IGNORE number {LINES | ROWS}]
+    [SET column_name_1=nullif(column_name_1, expr1), column_name_2=nullif(column_name_2, expr2)...]
+    [PARALLEL {'TRUE' | 'FALSE'}]
+    [STRICT {'TRUE' | 'FALSE'}]
 ```
 
 ## 开始前准备

diff --git a/docs/MatrixOne/Reference/Data-Types/data-types.md b/docs/MatrixOne/Reference/Data-Types/data-types.md
@@ -573,4 +573,41 @@ mysql> select * from t1;
 | [1, 2, 3] | [4, 5] |
 +-----------+--------+
 1 row in set (0.00 sec)
+```
+
+## Datalink 数据类型
+
+|类型         | 解释                  |
+|------------|---------------------  |
+|datalink     | 用于存储指向文档 (例如 satge) 或文件链接的特殊数据类     |
+
+### 示例
+
+```sql
+drop table test01;
+create table test01 (col1 int, col2 datalink);
+create stage stage01 url='file:///Users/admin/case/';
+insert into test01 values (1, 'file:///Users/admin/case/t1.csv');
+insert into test01 values (2, 'file:///Users/admin/case/t1.csv?size=2');
+insert into test01 values (3, 'file:///Users/admin/case/t1.csv?offset=4');
+insert into test01 values (4, 'file:///Users/admin/case/t1.csv?offset=4&size=2');
+insert into test01 values (5, 'stage://stage01/t1.csv');
+insert into test01 values (6, 'stage://stage01/t1.csv?size=2');
+insert into test01 values (7, 'stage://stage01/t1.csv?offset=4');
+insert into test01 values (8, 'stage://stage01/t1.csv?offset=4&size=2');
+
+mysql> select * from test01;
++------+-------------------------------------------------+
+| col1 | col2                                            |
++------+-------------------------------------------------+
+|    1 | file:///Users/admin/case/t1.csv                 |
+|    2 | file:///Users/admin/case/t1.csv?size=2          |
+|    3 | file:///Users/admin/case/t1.csv?offset=4        |
+|    4 | file:///Users/admin/case/t1.csv?offset=4&size=2 |
+|    5 | stage://stage01/t1.csv                          |
+|    6 | stage://stage01/t1.csv?size=2                   |
+|    7 | stage://stage01/t1.csv?offset=4                 |
+|    8 | stage://stage01/t1.csv?offset=4&size=2          |
++------+-------------------------------------------------+
+8 rows in set (0.01 sec)
 ```
diff --git a/docs/MatrixOne/Reference/Data-Types/datalink-type.md b/docs/MatrixOne/Reference/Data-Types/datalink-type.md
@@ -0,0 +1,87 @@
+# DATALINK 类型
+
+`DATALINK` 类型用于存储指向文档 (例如 satge) 或文件链接的特殊数据类型。它的主要目的是在数据库中存储文档的链接地址，而不是存储文档本身。这种类型可以应用于各种场景，特别是在处理大规模文档管理时，提供对文档的快捷访问，而不需要将文档实际存储在数据库中。
+
+使用 `DATALINK` 数据类型可以：
+
+- 节省存储空间：文档实际存储在外部存储中（例如对象存储系统），而数据库只保存链接。
+- 方便的文档访问：通过存储链接，系统可以快速访问文档，无需额外的存储和处理。
+- 提高数据操作效率：避免了直接在数据库中处理大文件，提高了数据操作的速度和效率。
+
+## 插入 DATALINK 类型数据
+
+**语法结构**
+
+```
+INSERT INTO TABLE_NAME VALUES ('<file://<path>/<filename>>|<stage://<stage_name>/<path>/<file_name>>?<offset=xx>&<size=xxx>')
+```
+
+**参数释义**
+
+|  参数   | 说明 |
+|  ----  | ----  |
+| file   | 指向本地文件系统文件位置。|
+| stage  | 指向 stage 指向文件位置。|
+| offset | 非必填。偏移量，表明读的内容的起点。|
+| size   | 非必填。指定读取内容的大小，单位为子节。|
+
+## 读取 DATALINK 类型数据
+
+如果要读 `DATALINK` 指向文件链接的数据，可以使用 [load_file](../../Reference/Functions-and-Operators/Other/load_file.md) 函数。
+
+## 示例
+
+`/Users/admin/case` 下有文件 `t1.csv`
+
+```bash
+(base) admin@192 case % cat t1.csv 
+this is a test message
+```
+
+```sql
+drop table test01;
+create table test01 (col1 int, col2 datalink);
+create stage stage01 url='file:///Users/admin/case/';
+insert into test01 values (1, 'file:///Users/admin/case/t1.csv');
+insert into test01 values (2, 'file:///Users/admin/case/t1.csv?size=2');
+insert into test01 values (3, 'file:///Users/admin/case/t1.csv?offset=4');
+insert into test01 values (4, 'file:///Users/admin/case/t1.csv?offset=4&size=2');
+insert into test01 values (5, 'stage://stage01/t1.csv');
+insert into test01 values (6, 'stage://stage01/t1.csv?size=2');
+insert into test01 values (7, 'stage://stage01/t1.csv?offset=4');
+insert into test01 values (8, 'stage://stage01/t1.csv?offset=4&size=2');
+
+mysql> select * from test01;
++------+-------------------------------------------------+
+| col1 | col2                                            |
++------+-------------------------------------------------+
+|    1 | file:///Users/admin/case/t1.csv                 |
+|    2 | file:///Users/admin/case/t1.csv?size=2          |
+|    3 | file:///Users/admin/case/t1.csv?offset=4        |
+|    4 | file:///Users/admin/case/t1.csv?offset=4&size=2 |
+|    5 | stage://stage01/t1.csv                          |
+|    6 | stage://stage01/t1.csv?size=2                   |
+|    7 | stage://stage01/t1.csv?offset=4                 |
+|    8 | stage://stage01/t1.csv?offset=4&size=2          |
++------+-------------------------------------------------+
+8 rows in set (0.01 sec)
+
+mysql> select col1, load_file(col2) from test01;
++------+-------------------------+
+| col1 | load_file(col2)         |
++------+-------------------------+
+|    1 | this is a test message
+ |
+|    2 | th                      |
+|    3 |  is a test message
+     |
+|    4 |  i                      |
+|    5 | this is a test message
+ |
+|    6 | th                      |
+|    7 |  is a test message
+     |
+|    8 |  i                      |
++------+-------------------------+
+8 rows in set (0.01 sec)
+```
diff --git a/docs/MatrixOne/Reference/Functions-and-Operators/Other/load_file.md b/docs/MatrixOne/Reference/Functions-and-Operators/Other/load_file.md
@@ -0,0 +1,72 @@
+# **LOAD_FILE()**
+
+## **函数说明**
+
+`LOAD_FILE()` 函数用于读取 datalink 类型指向文件的内容。
+
+## **函数语法**
+
+```
+>LOAD_FILE(datalink_type_data) ;
+```
+
+## **参数释义**
+
+|  参数  | 说明 |
+|  ----  | ----  |
+| datalink_type_data | datalink 类型数据，可以使用[cast()](../../../Reference/Operators/operators/cast-functions-and-operators/cast/)函数进行转换|
+
+## 示例
+
+`/Users/admin/case` 下有文件 `t1.csv`
+
+```bash
+(base) admin@192 case % cat t1.csv 
+this is a test message
+```
+
+```sql
+create table t1 (col1 int, col2 datalink);
+create stage stage1 url='file:///Users/admin/case/';
+insert into t1 values (1, 'file:///Users/admin/case/t1.csv');
+insert into t1 values (2, 'stage://stage1//t1.csv');
+
+mysql> select * from t1;
++------+---------------------------------+
+| col1 | col2                            |
++------+---------------------------------+
+|    1 | file:///Users/admin/case/t1.csv |
+|    2 | stage://stage1//t1.csv          |
++------+---------------------------------+
+2 rows in set (0.00 sec)
+
+mysql> select col1, load_file(col2) from t1;
++------+-------------------------+
+| col1 | load_file(col2)         |
++------+-------------------------+
+|    1 | this is a test message
+ |
+|    2 | this is a test message
+ |
++------+-------------------------+
+2 rows in set (0.01 sec)
+
+
+mysql> select load_file(cast('file:///Users/admin/case/t1.csv' as datalink));
++--------------------------------------------------------------+
+| load_file(cast(file:///Users/admin/case/t1.csv as datalink)) |
++--------------------------------------------------------------+
+| this is a test message
+                                      |
++--------------------------------------------------------------+
+1 row in set (0.00 sec)
+
+mysql> select load_file(cast('stage://stage1//t1.csv' as datalink));
++-----------------------------------------------------+
+| load_file(cast(stage://stage1//t1.csv as datalink)) |
++-----------------------------------------------------+
+| this is a test message
+                             |
++-----------------------------------------------------+
+1 row in set (0.00 sec)
+```
diff --git a/docs/MatrixOne/Reference/Functions-and-Operators/Other/save_file.md b/docs/MatrixOne/Reference/Functions-and-Operators/Other/save_file.md
@@ -0,0 +1,49 @@
+# **SAVE_FILE()**
+
+## **函数说明**
+
+`SAVE_FILE()` 函数用于向 datalink 指向文件中写入内容，指行返回写入内容字节长度。
+
+## **函数语法**
+
+```
+>SAVE_FILE(datalink_type_data,content) ;
+```
+
+## **参数释义**
+
+|  参数  | 说明 |
+|  ----  | ----  |
+| datalink_type_data | datalink 类型数据，可以使用[cast()](../../../Reference/Operators/operators/cast-functions-and-operators/cast/)函数进行转换|
+| content | 需要写入 datalink 指向文件的内容|
+
+## 示例
+
+```
+drop stage if exists tab1;
+create stage stage01 url='file:///Users/admin/case/';
+mysql> select save_file(cast('stage://stage01/test.csv' as datalink), 'this is a test message');
++-------------------------------------------------------------------------------+
+| save_file(cast(stage://stage01/test.csv as datalink), this is a test message) |
++-------------------------------------------------------------------------------+
+|                                                                            22 |
++-------------------------------------------------------------------------------+
+1 row in set (0.00 sec)
+
+mysql> select save_file(cast('file:///Users/admin/case/test1.csv' as datalink), 'this is another test message');
++-----------------------------------------------------------------------------------------------+
+| save_file(cast(file:///Users/admin/case/test1.csv as datalink), this is another test message) |
++-----------------------------------------------------------------------------------------------+
+|                                                                                            28 |
++-----------------------------------------------------------------------------------------------+
+1 row in set (0.01 sec)
+
+```
+
+```bash
+(base) admin@192 case % cat test.csv
+this is a test message
+
+(base) admin@192 case % cat test1.csv
+this is another test message
+```