Take the first one, please...

Captain

Consider this very highly anonymized pseudocode. It's a summary of the code I actually read through. I wrote the summary because the real Java has a ton of irrelevant nonsense going on in it, like setting irrelevant variables...

while (rs5.next) # imagine this code is spread out over a 5000 line function... with broken indentation...
  if(prevId != null && prevId = rs5.getID_PRSN)
    if(rs5.getCD_ROLEX != RP)
      if(rs5.getCD_ROLEX not in m_parentRole) # imagine there are 5 of these conditionals, one for each of CD_ROLE1, CD_ROLE2, etc.
        m_parentRole += rs5.getCD_ROLEX # imagine this is a nicely formatted concat, with commas, etc
    m_casePart.setTemplateField(m_prevCnt, "p_rolesAdl", m_parentRole) # Note that "p_rolesAdl" is the wrong field ANYWAY
    continue; # Java semantics...

  m_parentRole = ""
  m_role = ""

  if (rs5.getCD_ROLEX not in [RP, AV, OC])
    m_role += rs5.getCD_ROLEX

  m_parentRole = m_role
  m_casePart.setTemplateField(m_parentCounter, "p_rolesPrn", m_role) # finally the right field

  m_prevCnt = m_parentCounter
  m_prevId = rs5.getID_PRSN

What do we think this does guys? I'll give you a hint: the increasingly poorly described Haskell oneliner looks like:

    filter (\x -> not( x `elem` ["AV", "RP", "OC"]))
  . fmap (^.CD_ROLEX)
  . join 
  . fmap (take 1)
  . fmap (sortBy ( \a b -> compare (a ^.CD_ROLEX, b^.CD_ROLEX)))
  . groupBy (\a b -> a ^. ID_PRSN == b^.ID_PRSN)
  $ rs5

Oh, and the Java only works if the result set comes in with a specific sort order, which isn't even set in the SQL query.

dkf

@Captain said in Take the first one, please...:

Oh, and the Java only works if the result set comes in with a specific sort order, which isn't even set in the SQL query.

I believe it is possible to set a default sort order in the DDL, but often people don't (and it isn't good practice).

boomzilla

@Captain said in Take the first one, please...:

What do we think this does guys? I'll give you a hint: the increasingly poorly described Haskell oneliner looks like:

A hint in the category of endohints?

error

What do we think this does guys? I'll give you a hint: the increasingly poorly described Haskell oneliner looks like:
  . fmap (sortBy ( \a b -> compare (a ^.CD_ROLEX, b^.CD_ROLEX)))
  . groupBy (\a b -> a ^. ID_PRSN == b^.ID_PRSN)

Detects fake CD_ROLEXes and groups dealers into ID_PRISON?

Captain

@error Other way around. You group a list into a list of lists (where each sublist has the same prison id), and then sort each sublist on the fake rolex key.