Merge pull request #25 from TidierOrg/fix-athena-join-syntax

fixes join syntax for athena
TidierOrg · Jun 5, 2024 · abfa4b7 · abfa4b7
2 parents c0b8718 + 4978017
commit abfa4b7
Show file tree

Hide file tree

Showing 4 changed files with 23 additions and 54 deletions.
diff --git a/NEWS.md b/NEWS.md
@@ -1,5 +1,8 @@
 # TidierDB.jl updates
 
+## v0.1.5 - 2024-06-05
+- Adjusts Athena backend join syntax to match all other backends
+
 ## v0.1.4 - 2024-05-14
 - Adds Google Big Query support
 - use `connect` with GBQ JSON credentials and project id establish connection

diff --git a/docs/examples/UserGuide/athena.jl b/docs/examples/UserGuide/athena.jl
@@ -1,4 +1,4 @@
-# To use the Athena AWS backend with TidierDB, set up and a few minimal but key syntax differences are covered here. 
+# To use the Athena AWS backend with TidierDB, set up and a small syntax difference are covered here. 
 
 # ## Connecting 
 # Connection is established through AWS.jl as shwon below.
@@ -62,37 +62,4 @@
 #    2 │     8  15.1
 # ``` 
 
-# ## Joining Syntax 
-# Since running queries requires athena_params to be passed, and all of the joins pull in the new table metadata with a query, when performing joins in Athena, the final argument of the join must be the Athena Parameters. This syntax difference will hopefully be resolved in the future.
-# ```julia
-# query = @chain from_query(mtcars) begin
-#     @group_by cyl
-#     @summarize begin
-#         across(mpg, (mean, minimum, maximum))
-#         num_cars = n()
-#         end
-#     @mutate begin
-#         efficiency = case_when(
-#             mean_mpg >= 25, "High",
-#             mean_mpg >= 15, "Moderate",
-#             "Low" )
-#        end
-# end;
-
-# @chain from_query(query) begin
-#    @full_join(demodb.mtcars, cyl, cyl, athena_params)
-#    @group_by(efficiency)
-#    @summarize(avg_hp = mean(hp))
-#    @collect
-# end
-# ```
-# ```
-# 2×2 DataFrame
-#  Row │ efficiency  avg_hp   
-#      │ String      Float64  
-# ─────┼──────────────────────
-#    1 │ High         82.6364
-#    2 │ Moderate    180.238
-# ```
-
 # I would like to acknowledge the work of Manu Francis and this [blog post](https://medium.com/@manuedavakandam/beginners-guide-to-aws-athena-with-julia-a0192f7f4b4a), which helped guide this process  
diff --git a/src/TidierDB.jl b/src/TidierDB.jl
@@ -131,7 +131,7 @@ function finalize_query(sqlquery::SQLQuery)
      "FROM )" => ")" ,  "SELECT SELECT " => "SELECT ", "SELECT  SELECT " => "SELECT ", "DISTINCT SELECT " => "DISTINCT ", 
      "SELECT SELECT SELECT " => "SELECT ", "PARTITION BY GROUP BY" => "PARTITION BY", "GROUP BY GROUP BY" => "GROUP BY", "HAVING HAVING" => "HAVING", )
 
-    if current_sql_mode[] == :postgres || current_sql_mode[] == :duckdb || current_sql_mode[] == :mysql || current_sql_mode[] == :mssql || current_sql_mode[] == :clickhouse
+    if current_sql_mode[] == :postgres || current_sql_mode[] == :duckdb || current_sql_mode[] == :mysql || current_sql_mode[] == :mssql || current_sql_mode[] == :clickhouse || current_sql_mode[] == :athena || current_sql_mode[] == :gbq
         complete_query = replace(complete_query, "\"" => "'", "==" => "=")
     end
 

diff --git a/src/joins_sq.jl b/src/joins_sq.jl
@@ -12,7 +12,7 @@ end
 """
 $docstring_left_join
 """
-macro left_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=nothing)
+macro left_join(sqlquery, join_table, lhs_column, rhs_column)
     # Convert column references to string
     lhs_col_str = string(lhs_column)
     rhs_col_str = string(rhs_column)
@@ -43,10 +43,10 @@ macro left_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=noth
                 sq.from *= join_clause
             end
 
-            if $(esc(athena_params)) === nothing
+            if current_sql_mode[] != :athena
                 new_metadata = get_table_metadata(sq.db, string($(esc(join_table))))
             else
-                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), $(esc(athena_params)))
+                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), sq.athena_params)
             end
             sq.metadata = vcat(sq.metadata, new_metadata)
         else
@@ -59,7 +59,7 @@ end
 """
 $docstring_right_join
 """
-macro right_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=nothing)
+macro right_join(sqlquery, join_table, lhs_column, rhs_column)
     # Convert column references to string
     lhs_col_str = string(lhs_column)
     rhs_col_str = string(rhs_column)
@@ -90,10 +90,10 @@ macro right_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=not
                 sq.from *= join_clause
             end
 
-            if $(esc(athena_params)) === nothing
+            if current_sql_mode[] != :athena
                 new_metadata = get_table_metadata(sq.db, string($(esc(join_table))))
             else
-                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), $(esc(athena_params)))
+                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), sq.athena_params)
             end
             sq.metadata = vcat(sq.metadata, new_metadata)
         else
@@ -107,7 +107,7 @@ end
 """
 $docstring_inner_join
 """
-macro inner_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=nothing)
+macro inner_join(sqlquery, join_table, lhs_column, rhs_column)
     # Convert column references to string
     lhs_col_str = string(lhs_column)
     rhs_col_str = string(rhs_column)
@@ -138,10 +138,10 @@ macro inner_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=not
                 sq.from *= join_clause
             end
 
-            if $(esc(athena_params)) === nothing
+            if current_sql_mode[] != :athena
                 new_metadata = get_table_metadata(sq.db, string($(esc(join_table))))
             else
-                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), $(esc(athena_params)))
+                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), sq.athena_params)
             end
             sq.metadata = vcat(sq.metadata, new_metadata)
         else
@@ -155,7 +155,7 @@ end
 """
 $docstring_full_join
 """
-macro full_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=nothing)
+macro full_join(sqlquery, join_table, lhs_column, rhs_column)
     # Convert column references to string
     lhs_col_str = string(lhs_column)
     rhs_col_str = string(rhs_column)
@@ -186,10 +186,10 @@ macro full_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=noth
                 sq.from *= join_clause
             end
 
-            if $(esc(athena_params)) === nothing
+            if current_sql_mode[] != :athena
                 new_metadata = get_table_metadata(sq.db, string($(esc(join_table))))
             else
-                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), $(esc(athena_params)))
+                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), sq.athena_params)
             end
             sq.metadata = vcat(sq.metadata, new_metadata)
         else
@@ -203,7 +203,7 @@ end
 """
 $docstring_semi_join
 """
-macro semi_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=nothing)
+macro semi_join(sqlquery, join_table, lhs_column, rhs_column)
     # Convert column references to string
     lhs_col_str = string(lhs_column)
     rhs_col_str = string(rhs_column)
@@ -234,10 +234,10 @@ macro semi_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=noth
                 sq.from *= join_clause
             end
 
-            if $(esc(athena_params)) === nothing
+            if current_sql_mode[] != :athena
                 new_metadata = get_table_metadata(sq.db, string($(esc(join_table))))
             else
-                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), $(esc(athena_params)))
+                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), sq.athena_params)
             end
             sq.metadata = vcat(sq.metadata, new_metadata)
         else
@@ -251,7 +251,7 @@ end
 """
 $docstring_anti_join
 """
-macro anti_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=nothing)
+macro anti_join(sqlquery, join_table, lhs_column, rhs_column)
     # Convert column references to string
     lhs_col_str = string(lhs_column)
     rhs_col_str = string(rhs_column)
@@ -282,10 +282,10 @@ macro anti_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=noth
                 sq.from *= join_clause
             end
 
-            if $(esc(athena_params)) === nothing
+            if current_sql_mode[] != :athena
                 new_metadata = get_table_metadata(sq.db, string($(esc(join_table))))
             else
-                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), $(esc(athena_params)))
+                new_metadata = get_table_metadata_athena(sq.db, string($(esc(join_table))), sq.athena_params)
             end
             sq.metadata = vcat(sq.metadata, new_metadata)
         else
@@ -294,4 +294,3 @@ macro anti_join(sqlquery, join_table, lhs_column, rhs_column, athena_params=noth
         sq
     end
 end
-