Data Lake Analytics + OSS数据文件格式处理大全

0. 前言

Data Lake Analytics是Serverless化的云上交互式查询分析服务。用户可以使用标准的SQL语句，对存储在OSS、TableStore上的数据无需移动，直接进行查询分析。

目前该产品已经正式登陆阿里云，欢迎大家申请试用，体验更便捷的数据分析服务。
请参考https://help.aliyun.com/document_detail/70386.html 进行产品开通服务申请。

在上一篇教程中，我们介绍了如何分析CSV格式的TPC-H数据集。除了纯文本文件(例如，CSV，TSV等)，用户存储在OSS上的其他格式的数据文件，也可以使用Data Lake Analytics进行查询分析，包括ORC, PARQUET, JSON, RCFILE, AVRO甚至ESRI规范的地理JSON数据，还可以用正则表达式匹配的文件等。

本文详细介绍如何根据存储在OSS上的文件格式使用Data Lake Analytics （下文简称 DLA）进行分析。DLA内置了各种处理文件数据的SerDe（Serialize/Deserilize的简称,目的是用于序列化和反序列化）实现，用户无需自己编写程序，基本上能选用DLA中的一款或多款SerDe来匹配您OSS上的数据文件格式。如果还不能满足您特殊文件格式的处理需求，请联系我们，尽快为您实现。

1. 存储格式与SerDe

用户可以依据存储在OSS上的数据文件进行建表，通过STORED AS 指定数据文件的格式。
例如，

CREATE EXTERNAL TABLE nation (
    N_NATIONKEY INT, 
    N_NAME STRING, 
    N_REGIONKEY INT, 
    N_COMMENT STRING
) 
ROW FORMAT DELIMITED FIELDS TERMINATED BY '|' 
STORED AS TEXTFILE 
LOCATION 'oss://test-bucket-julian-1/tpch_100m/nation';

建表成功后可以使用SHOW CREATE TABLE语句查看原始建表语句。

mysql> show create table nation;
+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| Result                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                            |
+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| CREATE EXTERNAL TABLE `nation`(
  `n_nationkey` int,
  `n_name` string,
  `n_regionkey` int,
  `n_comment` string)
ROW FORMAT DELIMITED
    FIELDS TERMINATED BY '|'
STORED AS `TEXTFILE`
LOCATION
  'oss://test-bucket-julian-1/tpch_100m/nation'|
+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1 row in set (1.81 sec)

下表中列出了目前DLA已经支持的文件格式，当针对下列格式的文件建表时，可以直接使用STORED AS，DLA会选择合适的SERDE/INPUTFORMAT/OUTPUTFORMAT。

存储格式	描述
STORED AS TEXTFILE	数据文件的存储格式为纯文本文件。默认的文件类型。 文件中的每一行对应表中的一条记录。
STORED AS ORC	数据文件的存储格式为ORC。
STORED AS PARQUET	数据文件的存储格式为PARQUET。
STORED AS RCFILE	数据文件的存储格式为RCFILE。
STORED AS AVRO	数据文件的存储格式为AVRO。
STORED AS JSON	数据文件的存储格式为JSON (Esri ArcGIS的地理JSON数据文件除外)。

在指定了STORED AS 的同时，还可以根据具体文件的特点，指定SerDe (用于解析数据文件并映射到DLA表)，特殊的列分隔符等。
后面的部分会做进一步的讲解。

2. 示例

2.1 CSV文件

CSV文件，本质上还是纯文本文件，可以使用STORED AS TEXTFILE。
列与列之间以逗号分隔，可以通过ROW FORMAT DELIMITED FIELDS TERMINATED BY ',' 表示。

普通CSV文件

例如，数据文件oss://bucket-for-testing/oss/text/cities/city.csv的内容为

Beijing,China,010
ShangHai,China,021
Tianjin,China,022

建表语句可以为

CREATE EXTERNAL TABLE city (
    city STRING, 
    country STRING, 
    code INT
) 
ROW FORMAT DELIMITED FIELDS TERMINATED BY ',' 
STORED AS TEXTFILE 
LOCATION 'oss://bucket-for-testing/oss/text/cities';

使用OpenCSVSerde__处理引号__引用的字段

OpenCSVSerde在使用时需要注意以下几点：

1. 用户可以为行的字段指定字段分隔符、字段内容引用符号和转义字符，例如：WITH SERDEPROPERTIES ("separatorChar" = ",", "quoteChar" = "`", "escapeChar" = "" )；
2. 不支持字段内嵌入的行分割符；
3. 所有字段定义STRING类型；
4. 其他数据类型的处理，可以在SQL中使用函数进行转换。
   例如，

CREATE EXTERNAL TABLE test_csv_opencsvserde (
  id STRING,
  name STRING,
  location STRING,
  create_date STRING,
  create_timestamp STRING,
  longitude STRING,
  latitude STRING
) 
ROW FORMAT SERDE 'org.apache.hadoop.hive.serde2.OpenCSVSerde'
with serdeproperties(
'separatorChar'=',',
'quoteChar'='"',
'escapeChar'='\'
)
STORED AS TEXTFILE LOCATION 'oss://test-bucket-julian-1/test_csv_serde_1';

自定义分隔符

需要自定义列分隔符(FIELDS TERMINATED BY)，转义字符(ESCAPED BY)，行结束符(LINES TERMINATED BY)。
需要在建表语句中指定

ROW FORMAT DELIMITED
    FIELDS TERMINATED BY '	'
    ESCAPED BY '\'
    LINES TERMINATED BY '
'

忽略CSV文件中的HEADER

在csv文件中，有时会带有HEADER信息，需要在数据读取时忽略掉这些内容。这时需要在建表语句中定义skip.header.line.count。

例如，数据文件oss://my-bucket/datasets/tpch/nation_csv/nation_header.tbl的内容如下：

N_NATIONKEY|N_NAME|N_REGIONKEY|N_COMMENT
0|ALGERIA|0| haggle. carefully final deposits detect slyly agai|
1|ARGENTINA|1|al foxes promise slyly according to the regular accounts. bold requests alon|
2|BRAZIL|1|y alongside of the pending deposits. carefully special packages are about the ironic forges. slyly special |
3|CANADA|1|eas hang ironic, silent packages. slyly regular packages are furiously over the tithes. fluffily bold|
4|EGYPT|4|y above the carefully unusual theodolites. final dugouts are quickly across the furiously regular d|
5|ETHIOPIA|0|ven packages wake quickly. regu|

相应的建表语句为：

CREATE EXTERNAL TABLE nation_header (
    N_NATIONKEY INT, 
    N_NAME STRING, 
    N_REGIONKEY INT, 
    N_COMMENT STRING
) 
ROW FORMAT DELIMITED FIELDS TERMINATED BY '|' 
STORED AS TEXTFILE 
LOCATION 'oss://my-bucket/datasets/tpch/nation_csv/nation_header.tbl'
TBLPROPERTIES ("skip.header.line.count"="1");

skip.header.line.count的取值x和数据文件的实际行数n有如下关系:

• 当x<=0时，DLA在读取文件时，不会过滤掉任何信息，即全部读取；
• 当0
• 当x>=n时，DLA在读取文件时，会过滤掉所有的文件内容。

2.2 TSV文件

与CSV文件类似，TSV格式的文件也是纯文本文件，列与列之间的分隔符为Tab。

例如，数据文件oss://bucket-for-testing/oss/text/cities/city.tsv的内容为

Beijing    China    010
ShangHai    China    021
Tianjin    China    022

建表语句可以为

CREATE EXTERNAL TABLE city (
    city STRING, 
    country STRING, 
    code INT
) 
ROW FORMAT DELIMITED FIELDS TERMINATED BY '	' 
STORED AS TEXTFILE 
LOCATION 'oss://bucket-for-testing/oss/text/cities';

2.3 多字符数据字段分割符文件

假设您的数据字段的分隔符包含多个字符，可采用如下示例建表语句，其中每行的数据字段分割符为“||”，可以替换为您具体的分割符字符串。

ROW FORMAT SERDE 'org.apache.hadoop.hive.contrib.serde2.MultiDelimitSerDe'
with serdeproperties(
"field.delim"="||"
)

示例：

CREATE EXTERNAL TABLE test_csv_multidelimit (
  id STRING,
  name STRING,
  location STRING,
  create_date STRING,
  create_timestamp STRING,
  longitude STRING,
  latitude STRING
) 
ROW FORMAT SERDE 'org.apache.hadoop.hive.contrib.serde2.MultiDelimitSerDe'
with serdeproperties(
"field.delim"="||"
)
STORED AS TEXTFILE LOCATION 'oss://bucket-for-testing/oss/text/cities/';

2.4 JSON文件

DLA可以处理的JSON文件通常以纯文本的格式存储，数据文件的编码方式需要是UTF-8。
在JSON文件中，每行必须是一个完整的JSON对象。
例如，下面的文件格式是不被接受的

{"id": 123, "name": "jack", 
"c3": "2001-02-03 12:34:56"}
{"id": 456, "name": "rose", "c3": "1906-04-18 05:12:00"}
{"id": 789, "name": "tom", "c3": "2001-02-03 12:34:56"}
{"id": 234, "name": "alice", "c3": "1906-04-18 05:12:00"}

需要改写成：

{"id": 123, "name": "jack", "c3": "2001-02-03 12:34:56"}
{"id": 456, "name": "rose", "c3": "1906-04-18 05:12:00"}
{"id": 789, "name": "tom", "c3": "2001-02-03 12:34:56"}
{"id": 234, "name": "alice", "c3": "1906-04-18 05:12:00"}

不含嵌套的JSON数据

建表语句可以写

CREATE EXTERNAL TABLE t1 (id int, name string, c3 timestamp)
STORED AS JSON
LOCATION 'oss://path/to/t1/directory';

含有嵌套的JSON文件

使用struct和array结构定义嵌套的JSON数据。
例如，用户原始数据（注意：无论是否嵌套，一条完整的JSON数据都只能放在一行上，才能被Data Lake Analytics处理）：

{       "DocId": "Alibaba",         "User_1": {             "Id": 1234,             "Username": "bob1234",          "Name": "Bob",          "ShippingAddress": {                    "Address1": "969 Wenyi West St.",                     "Address2": null,                       "City": "Hangzhou",                      "Province": "Zhejiang"           },              "Orders": [{                            "ItemId": 6789,                                 "OrderDate": "11/11/2017"                       },                      {                               "ItemId": 4352,                                 "OrderDate": "12/12/2017"                       }               ]       } }

使用在线JSON格式化工具格式化后，数据内容如下：

{
    "DocId": "Alibaba", 
    "User_1": {
        "Id": 1234, 
        "Username": "bob1234", 
        "Name": "Bob", 
        "ShippingAddress": {
            "Address1": "969 Wenyi West St.", 
            "Address2": null, 
            "City": "Hangzhou", 
            "Province": "Zhejiang"
        }, 
        "Orders": [
            {
                "ItemId": 6789, 
                "OrderDate": "11/11/2017"
            }, 
            {
                "ItemId": 4352, 
                "OrderDate": "12/12/2017"
            }
        ]
    }
}

则建表语句可以写成如下（注意：LOCATION中指定的路径必须是JSON数据文件所在的目录，该目录下的所有JSON文件都能被识别为该表的数据）：

CREATE EXTERNAL TABLE json_table_1 (
    docid string,
    user_1 struct<
            id:INT,
            username:string,
            name:string,
            shippingaddress:struct<
                            address1:string,
                            address2:string,
                            city:string,
                            province:string
                            >,
            orders:array<
                    struct<
                        itemid:INT,
                        orderdate:string
                    >
            >
    >
)
STORED AS JSON
LOCATION 'oss://xxx/test/json/hcatalog_serde/table_1/';

对该表进行查询：

select * from json_table_1;

+---------+----------------------------------------------------------------------------------------------------------------+
| docid   | user_1                                                                                                         |
+---------+----------------------------------------------------------------------------------------------------------------+
| Alibaba | [1234, bob1234, Bob, [969 Wenyi West St., null, Hangzhou, Zhejiang], [[6789, 11/11/2017], [4352, 12/12/2017]]] |
+---------+----------------------------------------------------------------------------------------------------------------+

对于struct定义的嵌套结构，可以通过“.”进行层次对象引用，对于array定义的数组结构，可以通过“[数组下标]”（注意：数组下标从1开始）进行对象引用。

select DocId,
       User_1.Id,
       User_1.ShippingAddress.Address1,
       User_1.Orders[1].ItemId
from json_table_1
where User_1.Username = 'bob1234'
  and User_1.Orders[2].OrderDate = '12/12/2017';

+---------+------+--------------------+-------+
| DocId   | id   | address1           | _col3 |
+---------+------+--------------------+-------+
| Alibaba | 1234 | 969 Wenyi West St. |  6789 |
+---------+------+--------------------+-------+

使用JSON函数处理数据

例如，把“value_string”的嵌套JSON值作为字符串存储：

{"data_key":"com.taobao.vipserver.domains.meta.biz.alibaba.com","ts":1524550275112,"value_string":"{"appName":"","apps":[],"checksum":"50fa0540b430904ee78dff07c7350e1c","clusterMap":{"DEFAULT":{"defCkport":80,"defIPPort":80,"healthCheckTask":null,"healthChecker":{"checkCode":200,"curlHost":"","curlPath":"/status.taobao","type":"HTTP"},"name":"DEFAULT","nodegroup":"","sitegroup":"","submask":"0.0.0.0/0","syncConfig":{"appName":"trade-ma","nodegroup":"tradema","pubLevel":"publish","role":"","site":""},"useIPPort4Check":true}},"disabledSites":[],"enableArmoryUnit":false,"enableClientBeat":false,"enableHealthCheck":true,"enabled":true,"envAndSites":"","invalidThreshold":0.6,"ipDeleteTimeout":1800000,"lastModifiedMillis":1524550275107,"localSiteCall":true,"localSiteThreshold":0.8,"name":"biz.alibaba.com","nodegroup":"","owners":["junlan.zx","张三","李四","cui.yuanc"],"protectThreshold":0,"requireSameEnv":false,"resetWeight":false,"symmetricCallType":null,"symmetricType":"warehouse","tagName":"ipGroup","tenantId":"","tenants":[],"token":"1cf0ec0c771321bb4177182757a67fb0","useSpecifiedURL":false}"}

使用在线JSON格式化工具格式化后，数据内容如下：

{
    "data_key": "com.taobao.vipserver.domains.meta.biz.alibaba.com", 
    "ts": 1524550275112, 
    "value_string": "{"appName":"","apps":[],"checksum":"50fa0540b430904ee78dff07c7350e1c","clusterMap":{"DEFAULT":{"defCkport":80,"defIPPort":80,"healthCheckTask":null,"healthChecker":{"checkCode":200,"curlHost":"","curlPath":"/status.taobao","type":"HTTP"},"name":"DEFAULT","nodegroup":"","sitegroup":"","submask":"0.0.0.0/0","syncConfig":{"appName":"trade-ma","nodegroup":"tradema","pubLevel":"publish","role":"","site":""},"useIPPort4Check":true}},"disabledSites":[],"enableArmoryUnit":false,"enableClientBeat":false,"enableHealthCheck":true,"enabled":true,"envAndSites":"","invalidThreshold":0.6,"ipDeleteTimeout":1800000,"lastModifiedMillis":1524550275107,"localSiteCall":true,"localSiteThreshold":0.8,"name":"biz.alibaba.com","nodegroup":"","owners":["junlan.zx","张三","李四","cui.yuanc"],"protectThreshold":0,"requireSameEnv":false,"resetWeight":false,"symmetricCallType":null,"symmetricType":"warehouse","tagName":"ipGroup","tenantId":"","tenants":[],"token":"1cf0ec0c771321bb4177182757a67fb0","useSpecifiedURL":false}"
}

建表语句为

CREATE external TABLE json_table_2 (
   data_key string,
   ts bigint,
   value_string string
)
STORED AS JSON
LOCATION 'oss://xxx/test/json/hcatalog_serde/table_2/';

表建好后，可进行查询：

select * from json_table_2;

+---------------------------------------------------+---------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| data_key                                          | ts            | value_string                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                               |
+---------------------------------------------------+---------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| com.taobao.vipserver.domains.meta.biz.alibaba.com | 1524550275112 | {"appName":"","apps":[],"checksum":"50fa0540b430904ee78dff07c7350e1c","clusterMap":{"DEFAULT":{"defCkport":80,"defIPPort":80,"healthCheckTask":null,"healthChecker":{"checkCode":200,"curlHost":"","curlPath":"/status.taobao","type":"HTTP"},"name":"DEFAULT","nodegroup":"","sitegroup":"","submask":"0.0.0.0/0","syncConfig":{"appName":"trade-ma","nodegroup":"tradema","pubLevel":"publish","role":"","site":""},"useIPPort4Check":true}},"disabledSites":[],"enableArmoryUnit":false,"enableClientBeat":false,"enableHealthCheck":true,"enabled":true,"envAndSites":"","invalidThreshold":0.6,"ipDeleteTimeout":1800000,"lastModifiedMillis":1524550275107,"localSiteCall":true,"localSiteThreshold":0.8,"name":"biz.alibaba.com","nodegroup":"","owners":["junlan.zx","张三","李四","cui.yuanc"],"protectThreshold":0,"requireSameEnv":false,"resetWeight":false,"symmetricCallType":null,"symmetricType":"warehouse","tagName":"ipGroup","tenantId":"","tenants":[],"token":"1cf0ec0c771321bb4177182757a67fb0","useSpecifiedURL":false}       |
+---------------------------------------------------+---------------+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+

下面SQL示例json_parse，json_extract_scalar，json_extract等常用JSON函数的使用方式：

mysql> select json_extract_scalar(json_parse(value), '$.owners[1]') from json_table_2;

+--------+
| _col0  |
+--------+
| 张三    |
+--------+

mysql> select json_extract_scalar(json_obj.json_col, '$.DEFAULT.submask') 
from (
  select json_extract(json_parse(value), '$.clusterMap') as json_col from json_table_2
) json_obj
where json_extract_scalar(json_obj.json_col, '$.DEFAULT.healthChecker.curlPath') = '/status.taobao';

+-----------+
| _col0     |
+-----------+
| 0.0.0.0/0 |
+-----------+

mysql> with json_obj as (select json_extract(json_parse(value), '$.clusterMap') as json_col from json_table_2)
select json_extract_scalar(json_obj.json_col, '$.DEFAULT.submask')
from json_obj 
where json_extract_scalar(json_obj.json_col, '$.DEFAULT.healthChecker.curlPath') = '/status.taobao';

+-----------+
| _col0     |
+-----------+
| 0.0.0.0/0 |
+-----------+

2.5 ORC文件

Optimized Row Columnar(ORC)是Apache开源项目Hive支持的一种优化的列存储文件格式。与CSV文件相比，不仅可以节省存储空间，还可以得到更好的查询性能。

对于ORC文件，只需要在建表时指定 STORED AS ORC。
例如，

CREATE EXTERNAL TABLE orders_orc_date (
    O_ORDERKEY INT, 
    O_CUSTKEY INT, 
    O_ORDERSTATUS STRING, 
    O_TOTALPRICE DOUBLE, 
    O_ORDERDATE DATE, 
    O_ORDERPRIORITY STRING, 
    O_CLERK STRING, 
    O_SHIPPRIORITY INT, 
    O_COMMENT STRING
) 
STORED AS ORC 
LOCATION 'oss://bucket-for-testing/datasets/tpch/1x/orc_date/orders_orc';

2.6 PARQUET文件

Parquet是Apache开源项目Hadoop支持的一种列存储的文件格式。
使用DLA建表时，需要指定STORED AS PARQUET即可。
例如，

CREATE EXTERNAL TABLE orders_parquet_date (
    O_ORDERKEY INT, 
    O_CUSTKEY INT, 
    O_ORDERSTATUS STRING, 
    O_TOTALPRICE DOUBLE, 
    O_ORDERDATE DATE, 
    O_ORDERPRIORITY STRING, 
    O_CLERK STRING, 
    O_SHIPPRIORITY INT, 
    O_COMMENT STRING
) 
STORED AS PARQUET 
LOCATION 'oss://bucket-for-testing/datasets/tpch/1x/parquet_date/orders_parquet';

2.7 RCFILE文件

Record Columnar File (RCFile), 列存储文件，可以有效地将关系型表结构存储在分布式系统中，并且可以被高效地读取和处理。
DLA在建表时，需要指定STORED AS RCFILE。
例如，

CREATE EXTERNAL TABLE lineitem_rcfile_date (
    L_ORDERKEY INT, 
    L_PARTKEY INT, 
    L_SUPPKEY INT, 
    L_LINENUMBER INT, 
    L_QUANTITY DOUBLE, 
    L_EXTENDEDPRICE DOUBLE, 
    L_DISCOUNT DOUBLE, 
    L_TAX DOUBLE, 
    L_RETURNFLAG STRING, 
    L_LINESTATUS STRING, 
    L_SHIPDATE DATE, 
    L_COMMITDATE DATE, 
    L_RECEIPTDATE DATE, 
    L_SHIPINSTRUCT STRING, 
    L_SHIPMODE STRING, 
    L_COMMENT STRING
) 
STORED AS RCFILE
LOCATION 'oss://bucke-for-testing/datasets/tpch/1x/rcfile_date/lineitem_rcfile'

2.8 AVRO文件

DLA针对AVRO文件建表时，需要指定STORED AS AVRO，并且定义的字段需要符合AVRO文件的schema。

如果不确定可以通过使用Avro提供的工具，获得schema，并根据schema建表。
在Apache Avro官网下载avro-tools-.jar到本地，执行下面的命令获得Avro文件的schema：

java -jar avro-tools-1.8.2.jar getschema /path/to/your/doctors.avro
{
  "type" : "record",
  "name" : "doctors",
  "namespace" : "testing.hive.avro.serde",
  "fields" : [ {
    "name" : "number",
    "type" : "int",
    "doc" : "Order of playing the role"
  }, {
    "name" : "first_name",
    "type" : "string",
    "doc" : "first name of actor playing role"
  }, {
    "name" : "last_name",
    "type" : "string",
    "doc" : "last name of actor playing role"
  } ]
}

建表语句如下，其中fields中的name对应表中的列名，type需要参考本文档中的表格转成DLA支持的类型

CREATE EXTERNAL TABLE doctors(
number int,
first_name string,
last_name string)
STORED AS AVRO
LOCATION 'oss://mybucket-for-testing/directory/to/doctors';

大多数情况下，Avro的类型可以直接转换成DLA中对应的类型。如果该类型在DLA不支持，则会转换成接近的类型。具体请参照下表：

Avro类型	对应DLA类型
null	void
boolean	boolean
int	int
long	bigint
float	float
double	double
bytes	binary
string	string
record	struct
map	map
list	array
union	union
enum	string
fixed	binary

2.9 可以用正则表达式匹配的文件

通常此类型的文件是以纯文本格式存储在OSS上的，每一行代表表中的一条记录，并且每行可以用正则表达式匹配。
例如，Apache WebServer日志文件就是这种类型的文件。

某日志文件的内容为：

127.0.0.1 - frank [10/Oct/2000:13:55:36 -0700] "GET /apache_pb.gif HTTP/1.0" 200 2326
127.0.0.1 - - [26/May/2009:00:00:00 +0000] "GET /someurl/?track=Blabla(Main) HTTP/1.1" 200 5864 - "Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US) AppleWebKit/525.19 (KHTML, like Gecko) Chrome/1.0.154.65 Safari/525.19"

每行文件可以用下面的正则表达式表示，列之间使用空格分隔：

([^ ]*) ([^ ]*) ([^ ]*) (-|\[[^\]]*\]) ([^ "]*|"[^"]*") (-|[0-9]*) (-|[0-9]*)(?: ([^ "]*|"[^"]*") ([^ "]*|"[^"]*"))?

针对上面的文件格式，建表语句可以表示为：

CREATE EXTERNAL TABLE serde_regex(
  host STRING,
  identity STRING,
  userName STRING,
  time STRING,
  request STRING,
  status STRING,
  size INT,
  referer STRING,
  agent STRING)
ROW FORMAT SERDE 'org.apache.hadoop.hive.serde2.RegexSerDe'
WITH SERDEPROPERTIES (
  "input.regex" = "([^ ]*) ([^ ]*) ([^ ]*) (-|\[[^\]]*\]) ([^ "]*|"[^"]*") (-|[0-9]*) (-|[0-9]*)(?: ([^ "]*|"[^"]*") ([^ "]*|"[^"]*"))?"
)
STORED AS TEXTFILE
LOCATION 'oss://bucket-for-testing/datasets/serde/regex';

查询结果

mysql> select * from serde_regex;
+-----------+----------+-------+------------------------------+---------------------------------------------+--------+------+---------+--------------------------------------------------------------------------------------------------------------------------+
| host      | identity | userName | time                         | request                                     | status | size | referer | agent                                                                                                                    |
+-----------+----------+-------+------------------------------+---------------------------------------------+--------+------+---------+--------------------------------------------------------------------------------------------------------------------------+
| 127.0.0.1 | -        | frank | [10/Oct/2000:13:55:36 -0700] | "GET /apache_pb.gif HTTP/1.0"               | 200    | 2326 | NULL    | NULL                                                                                                                     |
| 127.0.0.1 | -        | -     | [26/May/2009:00:00:00 +0000] | "GET /someurl/?track=Blabla(Main) HTTP/1.1" | 200    | 5864 | -       | "Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US) AppleWebKit/525.19 (KHTML, like Gecko) Chrome/1.0.154.65 Safari/525.19" |
+-----------+----------+-------+------------------------------+---------------------------------------------+--------+------+---------+--------------------------------------------------------------------------------------------------------------------------+

2.10 Esri ArcGIS的地理JSON数据文件

DLA支持Esri ArcGIS的地理JSON数据文件的SerDe处理，关于这种地理JSON数据格式说明，可以参考：https://github.com/Esri/spatial-framework-for-hadoop/wiki/JSON-Formats

示例：

CREATE EXTERNAL TABLE IF NOT EXISTS california_counties
(
    Name string,
    BoundaryShape binary
)
ROW FORMAT SERDE 'com.esri.hadoop.hive.serde.JsonSerde'
STORED AS INPUTFORMAT 'com.esri.json.hadoop.EnclosedJsonInputFormat'
OUTPUTFORMAT 'org.apache.hadoop.hive.ql.io.HiveIgnoreKeyTextOutputFormat'
LOCATION 'oss://test_bucket/datasets/geospatial/california-counties/'

3. 总结

通过以上例子可以看出，DLA可以支持大部分开源存储格式的文件。对于同一份数据，使用不同的存储格式，在OSS中存储文件的大小，DLA的查询分析速度上会有较大的差别。推荐使用ORC格式进行文件的存储和查询。

为了获得更快的查询速度，DLA还在不断的优化中，后续也会支持更多的数据源，为用户带来更好的大数据分析体验。

 

原文链接
更多技术干货 请关注阿里云云栖社区微信号 ：yunqiinsight