目前来说，在 3.x 里最推荐的分享方式是文本的表格，而不是 CSV（可读性差）/ dot 文件（缺失信息）/ 截图（。。。。）
- 获得执行计划
从 console 中，先执行:profile <输出文件名>.log再执行 PROFILE 或者 EXPLAIN 开头的查询，就会把可以直接贴过来的执行计划导出到进入 console 所在的目录。
:profile profile.log
PROFILE MATCH xxxx RETURN yyy
- 贴到论坛
可以用这个代码的格式包裹起来
也可以作为附件上传 profile.log
比如：
(root@nebula) [basketballplayer]> profile GO FROM 'player100' OVER follow WHERE rank(edge) < 100 YIELD src(edge), dst(edge), rank(edge)
+-------------+-------------+------------+
| src(EDGE) | dst(EDGE) | rank(EDGE) |
+-------------+-------------+------------+
| "player100" | "player101" | 0 |
| "player100" | "player125" | 0 |
+-------------+-------------+------------+
Got 2 rows (time spent 4.356ms/37.68475ms)
Execution Plan (optimize time 264 us)
-----+-----------+--------------+--------------------------------------+------------------------------
| id | name | dependencies | profiling data | operator info |
-----+-----------+--------------+--------------------------------------+------------------------------
| 5 | Project | 4 | { | outputVar: { |
| | | | "execTime": "29(us)", | "colNames": [ |
| | | | "rows": 2, | "src(EDGE)", |
| | | | "totalTime": "33(us)", | "dst(EDGE)", |
| | | | "version": 0 | "rank(EDGE)" |
| | | | } | ], |
| | | | | "type": "DATASET", |
| | | | | "name": "__Project_5" |
| | | | | } |
| | | | | inputVar: __Filter_4 |