Hive：使用正则表达式分割字符串 [英] Hive: split string using regex

问题描述

我有一串 :: 分隔符的单词。我如何使用 Hive UDF regexp_extract（）来从字符串中提取单词？

解决方案

  regexp_extract（'2foa1fa :: 12hjk'，' ^（\\w。*）\\：{2}（\\w。*）$'，1）as word1 
  

OUTPUT： 2foa1fa

  regexp_extract（'2foa1fa :: 12hjk'，'^（\\w。*）\\：{2}（\\w。*）$'，2）as word2 
  

输出： 12hjk

• ^锚定到字符串的开始处
  
• \\w寻找单词字符，而。*表示任意数字时间


• \\：{2}在连续（这是您的::分隔符）中查找两个：
  
• $ anchors字符串到字符串结尾
  
• 指定regexp_extract中的第三个参数提取索引的（模式）
  
  
  

  现在是绝对的你可以使用一个分割函数来创建一个数组，然后用这个数组来查询阵列位置以及。看起来像这样：
  
  $ p $ 从
中选择my_array [2]（select split（'2foa1fa :: 12hjk '，'\\ ::'）as my_array from my_table）b;


  OUTPUT： 12hjk

  
  

  I have a string of words that are :: delimited. How can I use the Hive UDF regexp_extract() to extract words from the string?

  解决方案

  regexp_extract('2foa1fa::12hjk','^(\\w.*)\\:{2}(\\w.*)$',1) as word1
  

  OUTPUT: 2foa1fa

  regexp_extract('2foa1fa::12hjk','^(\\w.*)\\:{2}(\\w.*)$',2) as word2
  

  OUTPUT: 12hjk

  • ^ anchors to the beginning of the string
  • The \\w looks for a word character and .* means any number of times
  • The \\:{2} looks for two : in a row (this is your :: delimiter)
  • $ anchors the string to the end of the string
  • specifying the third parameter in regexp_extract extracts the indexed (pattern)

  Now just stick your column name in the place of the string literal and you should be good to go.

  You can also use a split function creating an array and then query by the array location as well. Which will look something like this:

  select my_array[2] from 
      (select split('2foa1fa::12hjk','\\::') as my_array from my_table) b;
  

  OUTPUT: 12hjk

  这篇关于Hive：使用正则表达式分割字符串的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！