abhin4v/ringo
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Changes dimension table population query for coalesce changes.

Browse Source 
- Replaces "IS NOT DISTICNT FROM" with "="
- Adds not-all-columns-null check for full refresh.
This commit is contained in:
Abhinav Sarkar 2015-12-29 16:22:58 +05:30
parent aeed16570e
commit d758dc3310
1 changed files with 14 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
src/Ringo/Generator.hs
View File

@ -104,21 +104,23 @@ dimensionTablePopulateSQL popMode fact dimTableName = do
, let col = fromJust . findColumn cName $ tableColumns factTable ]
baseSelectC = "SELECT DISTINCT\n" <> joinColumnNames selectCols
<> "\nFROM " <> factTableName fact
insertC selectC = "INSERT INTO " <> dimTableName
baseWhereC = "(\n"
<> Text.intercalate "\nOR " [ c <> " IS NOT NULL" | (_, c) <- colMapping ]
<> "\n)"
insertC selectC whereCs =
"INSERT INTO " <> dimTableName
<> " (\n" <> joinColumnNames (map fst colMapping) <> "\n) "
<> "SELECT x.* FROM (\n" <> selectC <> ") x"
<> "SELECT x.* FROM (\n"
<> selectC <> "\nWHERE " <> Text.intercalate " AND\n" whereCs
<> ") x"
timeCol = head [ cName | DimTime cName <- factColumns fact ]
return $ case popMode of
FullPopulation -> insertC baseSelectC
FullPopulation -> insertC baseSelectC [baseWhereC]
IncrementalPopulation ->
insertC (baseSelectC <> "\nWHERE "
<> timeCol <> " > ? AND " <> timeCol <> " <= ?"
<> " AND (\n"
<> Text.intercalate "\nOR " [ c <> " IS NOT NULL" | (_, c) <- colMapping ]
<> "\n)")
insertC baseSelectC [baseWhereC, timeCol <> " > ?", timeCol <> " <= ?"]
<> "\nLEFT JOIN " <> dimTableName <> " ON\n"
<> Text.intercalate " \nAND "
[ fullColName dimTableName c1 <> " IS NOT DISTINCT FROM " <> fullColName "x" c2 -- TODO can be replaced with = ?
[ fullColName dimTableName c1 <> " = " <> fullColName "x" c2
| (c1, c2) <- colMapping ]
<> "\nWHERE " <> Text.intercalate " \nAND "
[ fullColName dimTableName c <> " IS NULL" | (c, _) <- colMapping ]
@ -237,10 +239,7 @@ factTablePopulateSQL popMode fact = do
$ fullColName factSourceTableName dimFKIdColName
else "SELECT " <> dimIdColName <> " FROM " <> tableName <> "\nWHERE "
<> Text.intercalate "\n AND " dimLookupWhereClauses
insertSQL' = if factSourceTableName == fTableName
then insertSQL
else coalesceFKId insertSQL -- TODO always coalesce
in (dimFKIdColName, insertSQL', True)
in (dimFKIdColName, coalesceFKId insertSQL, True)
colMap = [ (cName, (sql, groupByColPrefix <> cName), addToGroupBy)
| (cName, sql, addToGroupBy) <- factColMap ++ dimColMap ]