regex - REGEXP_EXTRACT with URL in Hive

Question

I want to extract a word between '/bla-bla-bla/' and 'a12345' in the URL, which is "this-is-the-word" using regexp_extract in Hive.

INPUT: www.website.com/bla-bla-bla/this-is-the-word.a12345.anotherword.blabla

DESIRED OUTPUT: this-is-the-word

I've tried below, but none of them worked. What RegEx will achieve my desired output from this input?

regexp_extract(URL,'^.*[/]bla[-]bla[-]bla[/]([a-z]+)\\.(a([0-9]+))*$',1)
regexp_extract(URL,'^.*[/]bla-bla-bla[/]([a-z]*)[.]a([0-9]+)*$',1)

标签： regexhive

Answer 1

You may use

regexp_extract(URL,'^.*/bla-bla-bla/([^/.]+)\.a[0-9].*$', 1)

See this regex demo

It matches

• ^ - start of string
• .* - any 0+ chars other than line break chars, as many as possible
• /bla-bla-bla/ - a literal /bla-bla-bla/ substring
• ([^/.]+) - Group 1 (what you will get since the next argument is 1): 1 or more chars other than / and .
• \.a - a .a substring
• [0-9] - a digit
• .*$ - the rest of the string to its end.