SQL Developer 19.2.1 - Custom format: line break on hints, columns/argument alignment

1. Comments (and hints) are filtered away by parser, and are consequently ignored by formatter. To be able to process them with Custom Format, the  input text must be made accessible as java object in Nashorn. I made this amendment to 20.1 as per your earlier request to be able to update the formatting timestamp. Then, a hint  can be recognized in the input and, consequently, new line can be appended. However you would likely want to indent the first column with the others -- this doesn't look easy at the moment.

2. You have disabled double breaks after statements in the Advanced Options, but you need to enable them back to make the rule brkX2 working. If you don't need double line breaks under the certain conditions, then you just remove those disjuncts. Assuming you don't want any of them, here is the amended rule:

brkX2:

  --brkX2_1 | brkX2_2

  [node) basic_decl_item_list & [node+1) full_cursor_body

->;

The formatted output:

DECLARE

   c_defaultbsstrno  CONSTANT store.str_no%TYPE := 0;

   c_defaultbsstrno  CONSTANT store.str_no%TYPE := 0;

   CURSOR l_rtltndrdtlcur IS

   SELECT *

     FROM dual;

   CURSOR l_rtltndrdtlcur IS

   SELECT *

     FROM dual;

BEGIN

   NULL;

END;

3.

simpleIndentConditions: 

...

| [node) privilege_audit_clause

--| [node) parm_list_opt &  [node^) fml_part 

--| [node) prm_spec &       [node^) fml_part 

--| [node) query_partition_clause --& [node^) analytic_clause

...

_paddedIdsInScope1: (

-- types

  :alignTypeDecl & [id) identifier & [id+1) datatype & [scope) relational_properties

--| :alignTypeDecl & ( :breaksAfterComma | :breaksBeforeComma ) & [id) decl_id & ([id+1) prm_spec_unconstrained_type | [id+1) mode /*30067569:*/| [id+1) 'OUT' | [id+1) 'IN' ) & [scope) fml_part

| :alignTypeDecl & ( :breaksAfterComma | :breaksBeforeComma ) & [id) decl_id & ([id+1) constrained_type | [id+1) object_d_rhs) & [scope) adt_definition

...

_extraBrkBefore:

...

| :breaksBeforeComma &  [node+1) field                 & [node) ','

--| :breaksBeforeComma &  [node+1) prm_spec              & [node) ','

| :breaksBeforeComma &  [node+1) XML_attributes_clause        & [node) ','

...

Output:

CREATE OR REPLACE PACKAGE BODY emp_mgmt AS

   tot_emps   NUMBER;

   tot_depts  NUMBER;

   FUNCTION hire (last_name VARCHAR2,job_id VARCHAR2,manager_id NUMBER,salary NUMBER,department_id NUMBER) RETURN NUMBER IS

      new_empno NUMBER;

   BEGIN

...

4.

5.

simpleIndentConditions: 

...

| [node) alter_method_spec

--| [node) and_expr & ![node) rel

| .breaksProcArgs & [node) arg & [node^) paren_expr_list & [node+1) arg_list

...

_extraBrkBefore:

[node) 'AND' & [node+1) arith_expr |

  [node) physical_properties

...

Output:

   BEGIN

      IF (l_prevstrno = l_tndrcrctnchgrec.str_no)

      AND (l_prevtrmlno = l_tndrcrctnchgrec.trml_no)

      AND (l_prevtxno = l_tndrcrctnchgrec.tx_no)

      AND (l_prevtxdte = l_tndrcrctnchgrec.tx_dte)

      AND (l_prevtrmltypcd = l_tndrcrctnchgrec.trml_typ_cd)

      AND (l_prevrundte = l_tndrcrctnchgrec.run_dte)

      THEN ilevel := 'Low Income';

P.S. Tested everything in 19.4, not 19.2.

1. Advanced Format -> Line Breaks -> IF/CASE/WHILE  = Terse (line breaks only after actions)

2. Custom Format:

simpleIndentConditions:

...

   | [node) case_expression & ![node) select_term & ![node^) "aliased_expr" & ![node^) comparison_condition

...

  | [node) on_using_condition

...

pairwiseAlignments1: predecessor = node-1 & (

    [predecessor) group_by_list[5,12)            & [node) ','

  | [predecessor) order_by_clause[31,38)          & [node) ','

  | [predecessor) select_list                    & [node) ','

  | [predecessor) cartesian_product              & [node) ','

  | [predecessor) merge_insert_clause[26,33)      & [node) ','

  | [predecessor) column                          & [node) merge_insert_clause[26,33)

  | [predecessor) "expr_list_def"[16,34)          & [node) ','

  | [predecessor) "expr_list"                    & [node) ','          -- leading commas

  | [predecessor) column                          & [node) "expr_list_def"[16,34)

  | [predecessor) merge_update_clause[36,56)      & [node) ','

  | !:alignRight & [predecessor) condition        & [node) AND_OR

  | [predecessor) expr                  & [node) compound_expression[25,44)

  | [predecessor) '>'                    & [node) expr  & [node^) arg

  | [predecessor) arith_expr            & [node)  binary_add_op

| [predecessor) 'CASE'      & [node)  searched_case_expression

);

pairwiseAlignments2: predecessor=node-1-1  & (

    [predecessor) select_list    & [node) select_term & [node-1) ','

  | [predecessor) cartesian_product      & [node) table_reference  & [node-1) ','

  | [predecessor) condition      & [node) condition  & [node-1) AND_OR

  | [predecessor) adt_field_list & [node) adt_field  & [node-1) ','

  | [predecessor) adt_field      & [node) adt_field  & [node-1) ','

  | [predecessor) "expr_list"    & [node) expr        & [node-1) ','

  | [predecessor) expr          & [node) expr      

  | !:breakOnSubqueries  & [predecessor) '('    & [node) ')'  & [node-1) query_block & ![node^) par_subquery

  | [predecessor) 'CASE'      & [node)  'END'

  | [predecessor-1) 'CASE'      & [node)  'END'

  | [predecessor) 'CASE'      & [node)  else_clause

);

additional2Spaces:

   [node) searched_case_expression

| [node) else_clause

-> { padNode(2) }

Output:

SELECT rtd.col1

      ,rtd.col2

      ,CASE

          WHEN rtd.col1 = 'Y' THEN (rtd.col1 / 100) * - 1

          ELSE (rtd.col1 / 100) * 1

        END AS prod_prc

      ,t99.col1

      ,trunc (sysdate) AS run_dte

  FROM table1 rtd

  JOIN table2 t99

   ON rtd.str_no = t99.str_no

      AND CASE

            WHEN rtd.col1 = 'N' THEN rtd.col1 * - 1

            ELSE rtd.col1 / 100 * 1

          END = rtd.extd_prc

      AND CASE

            WHEN rtd.can_key_fg = 'N' THEN 'A'

            ELSE 'V'

          END = t99.void_fg

  JOIN table2 t98

   ON rtd.str_no = t98.str_no

WHERE t99.run_dte BETWEEN p_rundte AND p_rundte + 1;