Probably every ABAP developer was already in the situation to search for an existing data element to be used in a new program/table/structure. The search help in SE11 only covers the basic needs for searching an existing data element. While you are able to find a data element using one of the existing texts (and there are five texts for each data element) and/or the data type/length, at least the lower case flag is not visible in the search result. The UI of the search result is also really old fashioned.

To solve this issue we built a report which will make a select to the DDIC tables holding the data element values. This select basically just consumes one DDL view:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23

@AbapCatalog.sqlViewName: 'ZBC_DTEL_SEARCHV'
@AbapCatalog.compiler.compareFilter: true
@AbapCatalog.preserveKey: true
@AccessControl.authorizationCheck: #CHECK
@EndUserText.label: 'Dataelement Search'

define view ZBC_DTEL_SEARCH as select from dd04l 
  left outer join dd04t
    on dd04t.rollname = dd04l.rollname 
    and dd04t.ddlanguage = $session.system_language  
{
    key dd04l.rollname,
    dd04l.domname,   
    dd04l.datatype,
    dd04l.leng,
    dd04l.decimals,
    dd04l.lowercase,
    dd04t.ddtext,
    dd04t.reptext,
    dd04t.scrtext_s,
    dd04t.scrtext_m,
    dd04t.scrtext_l
}

Using the following select will enable searching for all different test fields (but case sensitive, since casting the fields to lower case is not possible with our basis release):

1
2
3
4
5
6
7
8
9
10
11
12
13

SELECT * 
  FROM zbc_dtel_searchv 
  INTO TABLE @lt_result
  WHERE rollname IN @s_rolnam
    AND datatype IN @s_type
    AND leng IN @s_leng
    AND decimals IN @s_deci
    AND ( lowercase IN @s_text
      OR ddtext IN @s_text
      OR reptext IN @s_text
      OR scrtext_s IN @s_text
      OR scrtext_m IN @s_text
      OR scrtext_l IN @s_text ).

The problems here are that the search is being done case sensitive (since only the field lowercase is guaranteed to be actually lower case - whereas the others are not) and we do have a lack of fuzzyness in the search. Fortunately HANA gives us the option to implement a fuzzy search with AMDB code pushdown.

Implementing this fuzzy search with AMDB has been done with the following method:

1
2
3
4
5
6
7
8
9
10
11
12
13
14

METHOD search_fuzzy BY DATABASE PROCEDURE
                    FOR HDB LANGUAGE SQLSCRIPT
                    OPTIONS READ-ONLY
                    USING zbc_dtel_searchv.

  rt_result = SELECT * FROM zbc_dtel_searchv
                WHERE CONTAINS (ddtext, :iv_text, FUZZY( 0.5 ))
                  OR CONTAINS (reptext, :iv_text, FUZZY( 0.9 ))
                  OR CONTAINS (scrtext_s, :iv_text, FUZZY( 0.9 ))
                  OR CONTAINS (scrtext_m, :iv_text, FUZZY( 0.9 ))
                  OR CONTAINS (scrtext_l, :iv_text, FUZZY( 0.9 ))
                ORDER BY score() DESC;

ENDMETHOD.

Which can be called by our report:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

zcl_dtel_search_helper=>search_fuzzy(
  EXPORTING
    iv_text   = p_text
  IMPORTING
    rt_result = lt_result
).

LOOP AT lt_result ASSIGNING FIELD-SYMBOL(<row>) 
  WHERE rollname IN s_rolnam
    AND datatype IN s_type
    AND leng IN s_leng
    AND decimals IN s_deci
    AND lowercase IN s_lower.

  APPEND <row> TO gt_outtab.

ENDLOOP.

By putting the results into an ALV grid we got a super fast and flexible data element search:

ALV result