{"id":4800,"date":"2020-12-08T11:36:54","date_gmt":"2020-12-08T03:36:54","guid":{"rendered":"https:\/\/nj.transwarp.cn:8180\/?p=4800"},"modified":"2020-12-08T15:36:23","modified_gmt":"2020-12-08T07:36:23","slug":"udtf%e6%a0%b7%e4%be%8b%e4%b9%8b%e6%8b%86%e5%88%86map","status":"publish","type":"post","link":"https:\/\/kbwp.transwarp.cn\/?p=4800","title":{"rendered":"UDTF\u6837\u4f8b\u4e4b\u62c6\u5206map\u7684key\/value"},"content":{"rendered":"<h3>\u539f\u7406\u63cf\u8ff0<\/h3>\n<hr \/>\n<p>UDTF(User-Defined Table-Generating Functions)  \u7528\u6765\u89e3\u51b3 \u8f93\u5165\u4e00\u884c\u8f93\u51fa\u591a\u884c(One-to-many maping) \u7684\u9700\u6c42\u3002<\/p>\n<p>UDTF\u6709\u4e24\u79cd\u4f7f\u7528\u65b9\u6cd5\uff0c\u4e00\u79cd\u76f4\u63a5\u653e\u5230 <strong>select<\/strong> \u540e\u9762\uff0c\u4e00\u79cd\u548c <strong>lateral view<\/strong> \u4e00\u8d77\u4f7f\u7528\u3002<\/p>\n<pre><code class=\"language-ruby\">1\uff1a\u76f4\u63a5select\u4e2d\u4f7f\u7528\uff1a<code>select explode_map(properties) as (col1,col2) from src;<\/code>\n\u4e0d\u53ef\u4ee5\u6dfb\u52a0\u5176\u4ed6\u5b57\u6bb5\u4f7f\u7528\uff1a<code>select a, explode_map(properties) as (col1,col2) from src<\/code>\n\u4e0d\u53ef\u4ee5\u5d4c\u5957\u8c03\u7528\uff1a<code>select explode_map(explode_map(properties)) from src<\/code>\n\u4e0d\u53ef\u4ee5\u548c<code>group by\/cluster by\/distribute by\/sort by<\/code>\u4e00\u8d77\u4f7f\u7528\uff1a<code>select explode_map(properties) as (col1,col2) from src group by col1, col2<\/code>\n2\uff1a\u548clateral view\u4e00\u8d77\u4f7f\u7528\uff1a<code>select src.id, mytable.col1, mytable.col2 from src lateral view explode_map(properties) mytable as col1, col2;<\/code>\n\u6b64\u65b9\u6cd5\u66f4\u4e3a\u65b9\u4fbf\u65e5\u5e38\u4f7f\u7528\u3002\u6267\u884c\u8fc7\u7a0b\u76f8\u5f53\u4e8e\u5355\u72ec\u6267\u884c\u4e86\u4e24\u6b21\u62bd\u53d6\uff0c\u7136\u540eunion\u5230\u4e00\u4e2a\u8868\u91cc\u3002<\/code><\/pre>\n<p>\u672c\u7bc7\u6587\u7ae0\u4e3b\u8981\u56f4\u7ed5UDTF\u51fd\u6570\u8fdb\u884c\u4ecb\u7ecd\uff0c\u5e38\u89c1\u7684UDTF\u51fd\u6570\uff0c\u6bd4\u5982<code>explode<\/code>,<code>json_tuple<\/code>,<code>stack<\/code>\u7b49\u7b49\u3002<\/p>\n<p><strong>\u6807\u51c6UDTF\u7684\u5b9e\u73b0\u6b65\u9aa4\u5982\u4e0b\uff1a<\/strong><\/p>\n<pre><code class=\"language-markdown\">- 1.\u7ee7\u627forg.apache.hadoop.hive.ql.udf.generic.GenericUDTF,\u5b9e\u73b0initialize, process, close\u4e09\u4e2a\u65b9\u6cd5\u3002\n- 2.UDTF\u9996\u5148\u4f1a\u8c03\u7528initialize\u65b9\u6cd5\uff0c\u6b64\u65b9\u6cd5\u8fd4\u56deUDTF\u7684\u8fd4\u56de\u884c\u7684\u4fe1\u606f\uff08\u8fd4\u56de\u4e2a\u6570\uff0c\u7c7b\u578b\uff09\u3002\n- 3.\u521d\u59cb\u5316\u5b8c\u6210\u540e\uff0c\u4f1a\u8c03\u7528process\u65b9\u6cd5,\u771f\u6b63\u7684\u5904\u7406\u8fc7\u7a0b\u5728process\u51fd\u6570\u4e2d\uff0c\u5728process\u4e2d\uff0c\u6bcf\u4e00\u6b21forward()\u8c03\u7528\u4ea7\u751f\u4e00\u884c\uff1b\u5982\u679c\u4ea7\u751f\u591a\u5217\u53ef\u4ee5\u5c06\u591a\u4e2a\u5217\u7684\u503c\u653e\u5728\u4e00\u4e2a\u6570\u7ec4\u4e2d\uff0c\u7136\u540e\u5c06\u8be5\u6570\u7ec4\u4f20\u5165\u5230forward()\u51fd\u6570\u3002\n- 4.\u6700\u540eclose()\u65b9\u6cd5\u8c03\u7528\uff0c\u5bf9\u9700\u8981\u6e05\u7406\u7684\u65b9\u6cd5\u8fdb\u884c\u6e05\u7406\u3002<\/code><\/pre>\n<p>\u4e0b\u9762\u6211\u4eec\u5f00\u59cb\u7f16\u5199\u81ea\u5df1\u7684UDTF\u3002\u7528\u6765\u5207\u5206&quot;key:value;key:value;&quot;\u8fd9\u79cd\u5b57\u7b26\u4e32\uff0c\u8fd4\u56de\u7ed3\u679c\u4e3akey, value\u4e24\u4e2a\u5b57\u6bb5\u3002<\/p>\n<h3>\u5b9e\u73b0\u6b65\u9aa4<\/h3>\n<hr \/>\n<h4>1. \u4ee3\u7801\u5b9e\u73b0<\/h4>\n<blockquote>\n<p>\u4ee3\u7801\u4e2d\u7684\u7c7b\uff0c\u53ef\u4ee5\u5230inceptor\u5bb9\u5668\u5185\/usr\/lib\/inceptor\/lib\/\u8def\u5f84\u4e0b\u83b7\u53d6\u5bf9\u5e94\u7684hive-exec-0.12.0-transwarp-X.X.X.jar<\/p>\n<\/blockquote>\n<pre><code class=\"language-java\">import java.util.ArrayList;\n\nimport org.apache.hadoop.hive.ql.exec.UDFArgumentException;\nimport org.apache.hadoop.hive.ql.metadata.HiveException;\nimport org.apache.hadoop.hive.ql.udf.generic.GenericUDTF;\nimport org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;\nimport org.apache.hadoop.hive.serde2.objectinspector.ObjectInspectorFactory;\nimport org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector;\nimport org.apache.hadoop.hive.serde2.objectinspector.primitive.PrimitiveObjectInspectorFactory;\n\n\/**\n * Demo UDTF class\n *\n * @author transwarp\n * @date 2017\/04\/19\n *\/\n\npublic class SplitKeyValue extends GenericUDTF {\n    public  String trim(String str){\n        return str.trim();\n    }\n    @Override\n    public void close() throws HiveException {\n    }\n    @Override\n    public StructObjectInspector initialize(ObjectInspector[] arg0) throws UDFArgumentException {\n        if (arg0.length != 1) {\n            \/\/throw new UDFArgumentLengthException(\"ExplodeMap takes only one argument\");\n        }\n        if (arg0[0].getCategory() != ObjectInspector.Category.PRIMITIVE) {\n            \/\/throw new UDFArgumentException(\"ExplodeMap takes string as a parameter\");\n        }\n        ArrayList<String> fieldNames = new ArrayList<String>();\n        ArrayList<ObjectInspector> fieldOIs = new ArrayList<ObjectInspector>();\n        fieldNames.add(\"col1\");\n        fieldOIs.add(PrimitiveObjectInspectorFactory.javaStringObjectInspector);\n        fieldNames.add(\"col2\");\n        fieldOIs.add(PrimitiveObjectInspectorFactory.javaStringObjectInspector);\n\n        return ObjectInspectorFactory.getStandardStructObjectInspector(fieldNames, fieldOIs);\n    }\n    @Override\n    public void process(Object[] arg0) throws HiveException {\n        String input = arg0[0].toString();\n        String[] test = input.split(\";\");\n        for (int i = 0; i < test.length; i++) {\n            try {\n                String[] result = test[i].split(\":\");\n                forward(result);\n            } catch (Exception e) {\n                continue;\n            }\n        }\n    }\n}<\/code><\/pre>\n<h4>2. \u6253\u5305\u6210jar\u5305<\/h4>\n<p>\u6b64\u5904\u5ffd\u7565\uff0c\u53ef\u81ea\u884c\u767e\u5ea6\u4e0d\u540c\u547d\u4ee4\u6216IDE\u5de5\u5177\u7684\u6253\u5305\u65b9\u5f0f\u3002<\/p>\n<p><font color=red>\u6ce8\u610f\uff1a\u6253\u5305\u65f6\uff0cjdk\u7248\u672c\u8981\u548cinceptor\u5185\u7684jdk\u7248\u672c\u4fdd\u6301\u4e00\u81f4\uff0c\u5426\u5219\u4f1a\u62a5java.lang.UnsupportedClassVersionError: InceptorUDAF : Unsupported major.minor version 52.0\u7684\u95ee\u9898<\/font><\/p>\n<h4>3. UDTF\u6301\u4e45\u5316<\/h4>\n<pre><code class=\"language-ruby\">**4.X\u7248\u672c\uff1a**\n1) udf.jar\u653e\u5230\/usr\/lib\/hive\/lib\u4e2d\uff08\u6bcf\u4e2aexecutor\u8282\u70b9\uff09;\n2) add jar  \/usr\/lib\/hive\/lib\/udf.jar;\n3) create permanent function func_name as 'package.classname';\n\n**5.X\u7248\u672c\u4ee5\u4e0a**\n1) \u628audf.jar\u653e\u5230\/usr\/lib\/inceptor\/lib\u76ee\u5f55\u4e0b\uff0cpush\u5230\u955c\u50cf\u4ed3\u5e93;\n2) \u91cd\u542finceptor\u6765\u83b7\u53d6\u6700\u65b0\u955c\u50cf;\n3) create permanent function func_name as 'package.classname';\n\n\u6ce8\u610f\uff1a\n1) udf\u65e0\u8bba\u662f\u4e34\u65f6\u8fd8\u662f\u6c38\u4e45\uff0c\u5982\u679c\u8981\u5220\u9664\u5e76\u91cd\u65b0\u521b\u5efa\u4f7f\u7528\u76f8\u540c\u7c7b\u6216\u8005jar\u7684\u8bdd\uff0c\u90fd\u8981\u91cd\u542finceptor server;\n2) \u5b98\u65b9\u4e0d\u63a8\u8350\u4f7f\u7528using jar\u7684\u65b9\u5f0f\u662f\u56e0\u4e3a\u6b64\u65b9\u6cd5\u5b58\u5728\u5f02\u5e38\u9690\u60a3\uff0c\u4f1a\u62a5\u627e\u4e0d\u5230jar\u5305\u7684\u9519\u8bef\u3002<\/code><\/pre>\n<p><strong>\u8fd9\u91cc\u4ee5TDH6.0.2\u7248\u672c\u73af\u5883\u4e3a\u4f8b\uff1a<\/strong><br \/>\n\u9996\u5148\u5c06SplitKeyValue.jar\u653e\u5230\/usr\/lib\/inceptor\/lib\u76ee\u5f55\u4e0b\uff0cpush\u5230\u955c\u50cf\u4ed3\u5e93\u4e2d\uff0c\u91cd\u542finceptor\u6765\u62c9\u53d6\u6700\u65b0\u955c\u50cf;<\/p>\n<pre><code class=\"language-sql\">--\u521b\u5efa\u6c38\u4e45\u51fd\u6570\n> CREATE PERMANENT FUNCTION explode_map AS 'SplitKeyValue';<\/code><\/pre>\n<pre><code class=\"language-sql\">--\u6d4b\u8bd5UDF\u51fd\u6570\uff0c\u5206\u522b\u4ee5\u5982\u4e0b5\u79cd\u573a\u666f\u8fdb\u884c\u6d4b\u8bd5\u3002\u7ecf\u6d4b\u8bd5\uff0cUDTF\u652f\u63012\u79cd\u4f7f\u7528\u65b9\u6cd5\uff1a\u4e00\u79cd\u76f4\u63a5\u653e\u5230select\u540e\u9762\uff081\uff09\uff0c\u4e00\u79cd\u548clateral view\u4e00\u8d77\u4f7f\u7528\uff085\uff09<\/code><\/pre>\n<pre><code class=\"language-sql\">--1\u3001\u76f4\u63a5select\u4e2d\u4f7f\u7528\nWITH tb AS \n(\nSELECT 1 AS id ,\"aaaaa:aa;bbbbb:bb\" AS map_col FROM system.dual UNION ALL \nSELECT 1 AS id ,\"ccccc:cc;ddddd:dd\" AS map_col FROM system.dual UNION ALL \nSELECT 2 AS id ,\"eeeee:ee;eeeee:ee\" AS map_col FROM system.dual\n)\nselect explode_map(map_col) as (col1,col2) from tb;\n\ncol1  |col2 |\n------|-----|\naaaaa |aa   |\nbbbbb |bb   |\nccccc |cc   |\nddddd |dd   |\neeeee |ee   |\neeeee |ee   |<\/code><\/pre>\n<pre><code class=\"language-sql\">--2\u3001\u4e0d\u53ef\u4ee5\u6dfb\u52a0\u5176\u4ed6\u5b57\u6bb5\u4f7f\u7528\uff1aselect a, explode_map(properties) as (col1,col2) from src\nWITH tb AS \n(\nSELECT 1 AS id ,\"aaaaa:aa;bbbbb:bb\" AS map_col FROM system.dual UNION ALL \nSELECT 1 AS id ,\"ccccc:cc;ddddd:dd\" AS map_col FROM system.dual UNION ALL \nSELECT 2 AS id ,\"eeeee:ee;eeeee:ee\" AS map_col FROM system.dual\n)\nselect id,explode_map(map_col) as (col1,col2) from tb;\n--SQL \u9519\u8bef [10088] [42000]: COMPILE FAILED: Semantic error: \n--[Error 10088] Line 7:40 AS clause has an invalid number of aliases. Error encountered near token 'col2'<\/code><\/pre>\n<pre><code class=\"language-sql\">--3\u3001\u4e0d\u53ef\u4ee5\u5d4c\u5957\u8c03\u7528\uff1aselect explode_map(explode_map(properties)) from src\nWITH tb AS \n(\nSELECT 1 AS id ,\"aaaaa:aa;bbbbb:bb\" AS map_col FROM system.dual UNION ALL \nSELECT 1 AS id ,\"ccccc:cc;ddddd:dd\" AS map_col FROM system.dual UNION ALL \nSELECT 2 AS id ,\"eeeee:ee;eeeee:ee\" AS map_col FROM system.dual\n)\nselect explode_map(explode_map(map_col)) from tb;\n--SQL \u9519\u8bef [10081] [42000]: COMPILE FAILED: Semantic error: \n--[Error 10081] UDTF's are not supported outside the SELECT clause, nor nested in expressions<\/code><\/pre>\n<pre><code class=\"language-sql\">--4\u3001\u4e0d\u53ef\u4ee5\u548cgroup by\/cluster by\/distribute by\/sort by\u4e00\u8d77\u4f7f\u7528\uff1a\n--select explode_map(properties) as (col1,col2) from src group by col1, col2\nWITH tb AS \n(\nSELECT 1 AS id ,\"aaaaa:aa;bbbbb:bb\" AS map_col FROM system.dual UNION ALL \nSELECT 1 AS id ,\"ccccc:cc;ddddd:dd\" AS map_col FROM system.dual UNION ALL \nSELECT 2 AS id ,\"eeeee:ee;eeeee:ee\" AS map_col FROM system.dual\n)\nselect id,explode_map(map_col) as (col1,col2) from tb GROUP BY col1,col2;\n--SQL \u9519\u8bef [10004] [42000]: COMPILE FAILED: Semantic error: \n--[Error 10004] Line 7:63 Invalid table alias or column reference: (possible column names are: id, map_col). \n--Error encountered near token 'col1'<\/code><\/pre>\n<pre><code class=\"language-sql\">--5\u3001\u548clateral view\u4e00\u8d77\u4f7f\u7528\uff1a\n--e.g. select src.id, mytable.col1, mytable.col2 from src lateral view explode_map(properties) mytable as col1, col2;\n--\u6b64\u65b9\u6cd5\u66f4\u4e3a\u65b9\u4fbf\u65e5\u5e38\u4f7f\u7528\u3002\u6267\u884c\u8fc7\u7a0b\u76f8\u5f53\u4e8e\u5355\u72ec\u6267\u884c\u4e86\u4e24\u6b21\u62bd\u53d6\uff0c\u7136\u540eunion\u5230\u4e00\u4e2a\u8868\u91cc\u3002\nWITH tb AS \n(\nSELECT 1 AS id ,\"aaaaa:aa;bbbbb:bb\" AS map_col FROM system.dual UNION ALL \nSELECT 1 AS id ,\"ccccc:cc;ddddd:dd\" AS map_col FROM system.dual UNION ALL \nSELECT 2 AS id ,\"eeeee:ee;eeeee:ee\" AS map_col FROM system.dual\n)\nselect id, mytable.col1, mytable.col2 from tb lateral view explode_map(map_col) mytable as col1, col2;\nid |col1  |col2 |\n---|------|-----|\n1  |aaaaa |aa   |\n1  |bbbbb |bb   |\n1  |ccccc |cc   |\n1  |ddddd |dd   |\n2  |eeeee |ee   |\n2  |eeeee |ee   |<\/code><\/pre>\n","protected":false},"excerpt":{"rendered":"<p>\u539f\u7406\u63cf\u8ff0 UDTF(User-Defined Table-Generating Functions) \u7528\u6765\u89e3\u51b3 ..<\/p>\n<div class=\"clear-fix\"><\/div>\n<p><a href=\"https:\/\/kbwp.transwarp.cn\/?p=4800\" title=\"read more...\">Read more<\/a><\/p>\n","protected":false},"author":12,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[38],"tags":[],"class_list":["post-4800","post","type-post","status-publish","format-standard","hentry","category-configuration"],"acf":[],"_links":{"self":[{"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=\/wp\/v2\/posts\/4800","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=\/wp\/v2\/users\/12"}],"replies":[{"embeddable":true,"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=4800"}],"version-history":[{"count":2,"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=\/wp\/v2\/posts\/4800\/revisions"}],"predecessor-version":[{"id":4801,"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=\/wp\/v2\/posts\/4800\/revisions\/4801"}],"wp:attachment":[{"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4800"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4800"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/kbwp.transwarp.cn\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4800"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}