1

Please login or register to vote for this query.

(click on this box to dismiss)

Q&A about the site for professional and independent game developers

declare @past int = 4;
declare @Today smalldatetime = getdate();
declare @After smalldatetime = DATEADD(week, @past * -1, @Today); 

with unit as (
  
  select top 1000000
    id,creationDate
  from users u
  where 
    u.CreationDate > @after
    and reputation > 100
  order by 
    creationDate desc

),

-- retrieve batch

main as (

  select 
    p.id postId,u.id userId,score,title,p.body,p.tags,
      try_convert(xml,p.body) tmp,
    p.creationDate postDate, 
    u.creationDate signupDate 
  from posts p 
  inner join unit u on 
    p.OwnerUserId = u.Id and p.postTypeId = 1
  where 
    p.CreationDate > @after

),

-- post process text nodes
-- better replace alternative:
-- https://stackoverflow.com/a/18635337/24561636

post as (

  SELECT postId,
    REPLACE(LOWER(REPLACE(TRANSLATE(T2.Loc.value('.','nvarchar(100)'), 
     '0123456789,.\/<>?;:"[]{}-=_+!@#$%^&*()`~', 
     '########################################'), 
     '#', '|')
     ),'| I ','I ') text
    
  FROM main
  CROSS APPLY tmp.nodes('/p/text()') AS T2(Loc)   
  
),

dist as (

  select top 1024
    value as phrase,
    count(distinct postId) questions 
  from post 
  cross apply string_split(text,'|',1)
  where 
    len(value) > 5 and trim(value) like '% %'
  group by value
  order by questions desc
)

select * from dist

view current version

edited apr 24

created apr 24 Dennis.Leeftink