Count specific pattern in URLs bigquery sql

Question

I got a table which contains URLs and some other columns, for example dates. The URLs contain IDs, separated by different values. What the IDs have in common is that they contai

Answer 1

Below is for BigQuery Standard SQL

I'd like to construct a query that counts the amount of ID's in the URL

#standardSQL
SELECT date, 
  (
    SELECT COUNT(1)  
    FROM UNNEST(REGEXP_EXTRACT_ALL(url, r'[^[:punct:]]+')) part 
    WHERE NOT REGEXP_CONTAINS(part, r'[^\d]')
  ) IDs
FROM `project.dataset.table`

If to apply to sample data from your question - the output is

Row date        IDs 
1   01-01-1999  3        
2   01-02-1999  4        
3   01-02-1999  3        
4   01-01-1999  5        
5   01-01-1999  1        
6   01-01-1999  1        

Secondly, I'd like to group the "amounts" by date

#standardSQL
SELECT date, IDs, COUNT(1) combinations FROM (
  SELECT date, 
    (
      SELECT COUNT(1)  
      FROM UNNEST(REGEXP_EXTRACT_ALL(url, r'[^[:punct:]]+')) part 
      WHERE NOT REGEXP_CONTAINS(part, r'[^\d]')
    ) IDs
  FROM `project.dataset.table`
)
GROUP BY date, IDs   

If to apply to sample data from your question - the output is

Row date        IDs combinations     
1   01-01-1999  3   1    
2   01-02-1999  4   1    
3   01-02-1999  3   1    
4   01-01-1999  5   1    
5   01-01-1999  1   2