vesoft-inc · cooper-lzy · Sep 7, 2022 · Sep 1, 2022 · Sep 1, 2022 · Sep 1, 2022
diff --git a/docs-2.0/3.ngql-guide/8.clauses-and-options/unwind.md b/docs-2.0/3.ngql-guide/8.clauses-and-options/unwind.md
@@ -4,70 +4,199 @@
 
 `UNWIND`可以作为单独语句或语句中的子句使用。
 
-## 语法
+## openCypher 兼容性
 
-```ngql
+nGQL 语句中兼容 openCypher 的`UNWIND`。
+
+## UNWIND 语句
+
+```ngql 
 UNWIND <list> AS <alias> <RETURN clause>;
 ```
 
-## 拆分列表
-
-```ngql
-nebula> UNWIND [1,2,3] AS n RETURN n;
-+---+
-| n |
-+---+
-| 1 |
-| 2 |
-| 3 |
-+---+
-```
+## 示例
+
+- 拆分列表。
+
+  ```ngql
+  nebula> UNWIND [1,2,3] AS n RETURN n;
+  +---+
+  | n |
+  +---+
+  | 1 |
+  | 2 |
+  | 3 |
+  +---+
+  ```
+
+## UNWIND 子句
+
+### 语法
+
+- nGQL 语句中使用`UNWIND`子句。
+
+  !!! note
+
+        在 nGQL 原生语句中使用`UNWIND`子句时，需要用在管道符`|`之后，并使用`$-`作为变量前缀。如果`UNWIND`后使用语句或子句，需要使用管道符`|`并且使用`$-`作为变量前缀。
+
+  ```ngql
+  <statement> | UNWIND $-.<var> AS <alias> <|> <clause>;
+  ```
+
+
+- openCypher 语句中使用`UNWIND`子句。
+
+  ```
+  <statement> UNWIND <list> AS <alias> <RETURN clause>；
+  ```
+
+
+### 示例
+
+- 在`UNWIND`语句中使用`WITH DISTINCT`可以将列表中的重复项忽略，返回去重后的结果。
+
+  !!! note
+
+      原生 nGQL 语句不支持`WITH DISTINCT`。  
+
+  1. 拆分列表`[1,1,2,2,3,3]`。
+  2. 删除重复行。
+  3. 排序行。
+  4. 将行转换为列表。
+
+  ```ngql
+  nebula> WITH [1,1,2,2,3,3] AS n \
+          UNWIND n AS r \
+          WITH DISTINCT r AS r \
+          ORDER BY r \
+          RETURN collect(r);
+  +------------+
+  | collect(r) |
+  +------------+
+  | [1, 2, 3]  |
+  +------------+
+  ```
+
+- `MATCH`语句中使用`UNWIND`。
+
+  1. 将匹配路径上的顶点输出到列表中。
+  2. 拆分列表。
+  3. 删除重复行。
+  4. 将行转换为列表。
+
+  ```ngql
+  nebula> MATCH p=(v:player{name:"Tim Duncan"})--(v2) \
+          WITH nodes(p) AS n \
+          UNWIND n AS r \
+          WITH DISTINCT r AS r \
+          RETURN collect(r);
+  +----------------------------------------------------------------------------------------------------------------------+
+  | collect(r)                                                                                                           |
+  +----------------------------------------------------------------------------------------------------------------------+
+  | [("player100" :player{age: 42, name: "Tim Duncan"}), ("player101" :player{age: 36, name: "Tony Parker"}),            |
+  |("team204" :team{name: "Spurs"}), ("player102" :player{age: 33, name: "LaMarcus Aldridge"}),                          |
+  |("player125" :player{age: 41, name: "Manu Ginobili"}), ("player104" :player{age: 32, name: "Marco Belinelli"}),       |
+  |("player144" :player{age: 47, name: "Shaquile O'Neal"}), ("player105" :player{age: 31, name: "Danny Green"}),         |
+  |("player113" :player{age: 29, name: "Dejounte Murray"}), ("player107" :player{age: 32, name: "Aron Baynes"}),         |
+  |("player109" :player{age: 34, name: "Tiago Splitter"}), ("player108" :player{age: 36, name: "Boris Diaw"})]           |  
+  +----------------------------------------------------------------------------------------------------------------------+
+  ```
+
+- `GO`语句中使用`UNWIND`。
+
+  ```ngql 
+  // 在点列表中查询点关联的边。
+  nebula> YIELD ['player101', 'player100'] AS a | UNWIND $-.a AS  b | GO FROM $-.b OVER follow YIELD edge AS e;
+  +----------------------------------------------------+
+  | e                                                  |
+  +----------------------------------------------------+
+  | [:follow "player101"->"player100" @0 {degree: 95}] |
+  | [:follow "player101"->"player102" @0 {degree: 90}] |
+  | [:follow "player101"->"player125" @0 {degree: 95}] |
+  | [:follow "player100"->"player101" @0 {degree: 95}] |
+  | [:follow "player100"->"player125" @0 {degree: 95}] |
+  +----------------------------------------------------+
+  ```
+
+- `LOOKUP`语句中使用`UNWIND`。
+
+  ```ngql
+  // 查询年龄大于 46 岁球员的所有属性，去掉重复属性，并将结果转换为行。
+  nebula> LOOKUP ON player \
+          WHERE player.age > 46 \
+          YIELD DISTINCT keys(vertex) as p | UNWIND $-.p as a | YIELD $-.a AS a;
+  +--------+
+  | a      |
+  +--------+
+  | "age"  |
+  | "name" |
+  +--------+
+  ```
+
+- `FETCH`语句中使用`UNWIND`。
+
+  ```ngql
+  //查询 player101 点的所有 Tag，并将结果转换为行。
+  nebula> CREATE TAG hero(like string, height int);
+          INSERT VERTEX hero(like, height) VALUES "player101":("deep", 182);
+          FETCH PROP ON * "player101" \
+          YIELD tags(vertex) as t | UNWIND $-.t as a | YIELD $-.a AS a;
+  +----------+
+  | a        |
+  +----------+
+  | "hero"   |
+  | "player" |
+  +----------+
+  ```
+
+- `GET SUBGRAPH`语句中使用`UNWIND`。
+
+  ```ngql
+  // 查询从点 player100 开始、0~2 跳、serve 类型的出边和入边的子图，并将结果转换为行。
+  GET SUBGRAPH 2 STEPS FROM "player100" BOTH serve YIELD edges as e | UNWIND $-.e as a | YIELD $-.a AS a;
+  +----------------------------------------------+
+  | a                                            |
+  +----------------------------------------------+
+  | [:serve "player100"->"team204" @0 {}]        |
+  | [:serve "player101"->"team204" @0 {}]        |
+  | [:serve "player102"->"team204" @0 {}]        |
+  | [:serve "player103"->"team204" @0 {}]        |
+  | [:serve "player105"->"team204" @0 {}]        |
+  | [:serve "player106"->"team204" @0 {}]        |
+  | [:serve "player107"->"team204" @0 {}]        |
+  | [:serve "player108"->"team204" @0 {}]        |
+  | [:serve "player109"->"team204" @0 {}]        |
+  | [:serve "player110"->"team204" @0 {}]        |
+  | [:serve "player111"->"team204" @0 {}]        |
+  | [:serve "player112"->"team204" @0 {}]        |
+  | [:serve "player113"->"team204" @0 {}]        |
+  | [:serve "player114"->"team204" @0 {}]        |
+  | [:serve "player125"->"team204" @0 {}]        |
+  | [:serve "player138"->"team204" @0 {}]        |
+  | [:serve "player104"->"team204" @20132015 {}] |
+  | [:serve "player104"->"team204" @20182019 {}] |
+  +----------------------------------------------+
+  ```
+
+- `FIND PATH`语句中使用`UNWIND`。
+
+  ```
+  // 找出 player101 到 team204 延 serve 类型边的最短路径上的所有点，并将结果转换为行。
+  FIND SHORTEST PATH FROM "player101" TO "team204" OVER serve YIELD path as p | YIELD nodes($-.p) AS nodes | UNWIND $-.nodes AS a | YIELD $-.a AS a;
+  +---------------+
+  | a             |
+  +---------------+
+  | ("player101") |
+  | ("team204")   |
+  +---------------+
+  ```
+
+
+
+
 
-## 返回去重列表
 
-在`UNWIND`语句中使用`WITH DISTINCT`可以将列表中的重复项忽略，返回去重后的结果。
 
-### 示例 1
 
-1. 拆分列表`[1,1,2,2,3,3]`。
-2. 删除重复行。
-3. 排序行。
-4. 将行转换为列表。
 
-```ngql
-nebula> WITH [1,1,2,2,3,3] AS n \
-        UNWIND n AS r \
-        WITH DISTINCT r AS r \
-        ORDER BY r \
-        RETURN collect(r);
-+------------+
-| collect(r) |
-+------------+
-| [1, 2, 3]  |
-+------------+
-```
 
-### 示例 2
-
-1. 将匹配路径上的顶点输出到列表中。
-2. 拆分列表。
-3. 删除重复行。
-4. 将行转换为列表。
-
-```ngql
-nebula> MATCH p=(v:player{name:"Tim Duncan"})--(v2) \
-        WITH nodes(p) AS n \
-        UNWIND n AS r \
-        WITH DISTINCT r AS r \
-        RETURN collect(r);
-+----------------------------------------------------------------------------------------------------------------------+
-| collect(r)                                                                                                           |
-+----------------------------------------------------------------------------------------------------------------------+
-| [("player100" :player{age: 42, name: "Tim Duncan"}), ("player101" :player{age: 36, name: "Tony Parker"}),            |
-|("team204" :team{name: "Spurs"}), ("player102" :player{age: 33, name: "LaMarcus Aldridge"}),                          |
-|("player125" :player{age: 41, name: "Manu Ginobili"}), ("player104" :player{age: 32, name: "Marco Belinelli"}),       |
-|("player144" :player{age: 47, name: "Shaquile O'Neal"}), ("player105" :player{age: 31, name: "Danny Green"}),         |
-|("player113" :player{age: 29, name: "Dejounte Murray"}), ("player107" :player{age: 32, name: "Aron Baynes"}),         |
-|("player109" :player{age: 34, name: "Tiago Splitter"}), ("player108" :player{age: 36, name: "Boris Diaw"})]           |  
-+----------------------------------------------------------------------------------------------------------------------+
-```