REUSE_ALV_FIELDCATALOG_MERGE takes not all the fields - abap

I have a problem with REUSE_ALV_FIELDCATALOG_MERGE FM. I need to add 5 fields into ALV.
DATA: BEGIN OF gt_collective7 OCCURS 0,
selected(1),
dat LIKE resb-bdter,
overdue TYPE c,
werks LIKE resb-werks,
extwg LIKE mara-extwg,
vd LIKE fmfincode-fincode,
...
TM like afvc-tplnr,
ARBPL_Z like pat106-entries,
ARBPL_R like pat106-entries,
PLTXT like iflo-pltxt,
lgort like resb-lgort,
END OF gt_collective7,
gt_collective7 defined. Next i use
CALL FUNCTION 'REUSE_ALV_FIELDCATALOG_MERGE'
EXPORTING
i_program_name = sy-repid
i_internal_tabname = 'GT_COLLECTIVE7'
i_inclname = sy-repid
i_bypassing_buffer = 'X'
CHANGING
ct_fieldcat = fieldcat
EXCEPTIONS
inconsistent_interface = 1
program_error = 2
OTHERS = 3.
and inside fieldcat there's no new fields appear. What could be a problem in this case? I thought that it could be 'type' inside gt_collective7 definition, but no - it worked just fine previously.

Don't know how exactly that helped, but declaring all additional fields of gt_collective7 in lowercase and reactivating the program somehow resolved situation.

Related

Date format from 2020.11.20 to 11/2020

I'm trying to change the date format from 2020.11.20 to 11/2020
My objective is to remove the day and leave just month/year.
If I change the type of the field EDATU from vbep-EDATU to string it doesn't work.
Any tips on how to achieve my goal?
DATA: GR_COLUMNS TYPE REF TO CL_SALV_COLUMNS_TABLE,
GR_TABLE TYPE REF TO CL_SALV_TABLE.
TYPES: BEGIN OF IT_STR,
EDATU TYPE VBEP-EDATU,
vbeln type vbep-vbeln,
END OF IT_STR.
DATA: IT_FINAL TYPE STANDARD TABLE OF IT_STR.
FIELD-SYMBOLS: <F_DAT> TYPE IT_STR.
SELECT EDATU vbeln FROM VBEP INTO TABLE IT_FINAL up to 10 rows.
LOOP AT IT_FINAL ASSIGNING <F_DAT>.
<F_DAT>-EDATU = <F_DAT>-EDATU+4(2) && '/' && <F_DAT>-EDATU(4).
ENDLOOP.
TRY.
CALL METHOD CL_SALV_TABLE=>FACTORY
EXPORTING
LIST_DISPLAY = IF_SALV_C_BOOL_SAP=>FALSE
IMPORTING
R_SALV_TABLE = GR_TABLE
CHANGING
T_TABLE = IT_FINAL.
CATCH CX_SALV_MSG .
ENDTRY.
GR_COLUMNS = GR_TABLE->GET_COLUMNS( ).
CALL METHOD GR_TABLE->DISPLAY.
Even though a type D is really just a char based data type of length 8 it has some special behavior when outputting it. If you check in the debugger you will see the data in IT_FINAL is what you want, it's just that the ALV processes this data as a date regardless of the value in it.
So, for example date 20221019 got changed in IT_FINAL-EDATU to '10/2022'. Now, when you display it in the ALV it gets interpreted as a date and gets displayed depending on your user settings, assuming yours are yyyy.mm.dd the output would be: '10/2.02.2'.
You can get around this in two ways:
Add a new field to your table with a char-like type to hold the month value as suggested by Skin, and suppress the display of the EDATU field in the ALV
gr_columns->get_column( columnname = 'EDATU' )->set_technical( abap_true ).
Change the behavior of the EDATU column in the ALV by changing the edit mask with:
gr_columns->get_column( columnname = 'EDATU' )->set_edit_mask('_______').
Option 1 is clearly the better way to avoid unintended mishaps in the future.
I created a new colum type C and changed the loop to
LOOP AT IT_FINAL ASSIGNING <F_DAT>.
data(month) = <F_DAT>-EDATU+4(2).
data(year) = <F_DAT>-EDATU(4).
<f_dat>-char = month && '/' && year.
ENDLOOP.
Thanks for everyone who helped me.

Parameter type error on BAPI_OBJCL_GETDETAIL call?

lv_objectkey2 = ls_mseg-matnr.
"Transport Category
CALL FUNCTION 'BAPI_OBJCL_GETDETAIL'
EXPORTING
objectkey = lv_objectkey2
objecttable = 'MARA'
classnum = 'Z_MATERIAL_CLASS'
classtype = '001'
TABLES
allocvaluesnum = lt_allocvaluesnum2
allocvalueschar = lt_allocvalueschar2
allocvaluescurr = lt_allocvaluescurr2
return = lt_return2.
READ TABLE lt_allocvaluesnum2 INTO ls_valnum2 WITH KEY charact= 'Z_ADR_QUANTITY'.
IF sy-subrc = 0.
lv_adr_quan = ls_valnum2-value_from + lv_adr_quan.
WRITE: lv_adr_quan TO ls_item-ADR_QUAN EXPONENT 0 DECIMALS 2.
* CONDENSE ls_item-ADR_QUAN.
ENDIF.
Here is my problem : The program gives me that problem : "ADR_QUAN" must be a character-type field (data type C, N, D or T).
I need your opinions to fix the issue or solutions.
CONDENSE Statement works only for the character-like variables. Here, ls_item-ADR_QUAN field is of type Float, That's why you're getting that error.
You can go through the attached documentation link for CONDENSE statement.
CONDENSE Documentation
Even if you want to perform CONDENSE then first, you've to assign ls_item-ADR_QUAN to a character-like variable.

ITXEX field cleared when HR_INFOTYPE_OPERATION is called

We got into difficulties in maintaining the ITXEX field (Long text indication) of an Infotype record.
Say we got an existing record in an Infotype database table with a long text filled (ITXEX field value in that record is set to 'X').
Some process updates the record through HR_CONTROL_INFTY_OPERATION like this:
CALL FUNCTION 'HR_CONTROL_INFTY_OPERATION'
EXPORTING
infty = '0081'
number = '12345678'
subtype = '01'
validityend = '31.12.9999'
validitybegin = '19.05.2019'
record = ls_0081 " ( ITXEX = 'X' )
operation = 'MOD'
tclas = 'A'
nocommit = abap_true
IMPORTING
return = ls_return.
This call does update the record but clearing it's ITXEX field.
It's important to say that making the same action through PA30 does update the record and maintain ITXEX field as it was.
The described problem seems similar to that question. Trying the solutions given there didn't solve the problem.
Why the two approaches (PA30 and function module) don't work the same? How to fix this?
First of all, FM parameters you use are incorrect. How do you want the infotype to be updated if you set nocommit = TRUE?
Also, you are missing the correct sequence which must be used for the update procedure:
Lock the Employee
Read the infotype
Update the infotype
Unlock the Employee
The correct snippet for your task would be
DATA: ls_return TYPE bapireturn1.
DATA: l_infty_tab TYPE TABLE OF p0002.
CALL FUNCTION 'HR_READ_INFOTYPE'
EXPORTING
pernr = '00000302'
infty = '0002'
TABLES
infty_tab = l_infty_tab.
READ TABLE l_infty_tab ASSIGNING FIELD-SYMBOL(<infotype>) INDEX 1.
<infotype>-midnm = 'Shicklgruber'. " updating the field of infotype
CALL FUNCTION 'ENQUEUE_EPPRELE'
EXPORTING
pernr = '00000302'
infty = '0002'.
CALL FUNCTION 'HR_CONTROL_INFTY_OPERATION'
EXPORTING
infty = <infotype>-infty
number = <infotype>-pernr
subtype = <infotype>-subty
validityend = <infotype>-endda
validitybegin = <infotype>-begda
record = <infotype>
operation = 'MOD'
tclas = 'A'
IMPORTING
return = ls_return.
CALL FUNCTION 'DEQUEUE_EPPRELE'
EXPORTING
pernr = '00000302'
infty = '0002'.
This way itxex field is treated correctly and if existed on that record, will remain intact. However, this method will not work for updating the long text itself, for that you must use object-oriented way, methods of class CL_HRPA_INFOTYPE_CONTAINER.

How to use REUSE_ALV_FIELDCATALOG_MERGE function module?

I'm trying to use the function module REUSE_ALV_FIELDCATALOG_MERGE to pass the field label in ddic to display in the column header of the alv report.
But, that didn't work.
If I comment the I_STRUCTURE_NAME = 'TY_YNAH_CUS_OBJ_REQ' line, it give me runtime error state
The ABAP program lines are wider than the internal table.
But if I uncomment it ,the program still did not work
REPORT YALV_TEST.
tables sscrfields.
type-pools : slis.
"CREATE STRUCTURE -1
TYPES: BEGIN OF TY_YNAH_CUS_OBJ_REQ,
REQID TYPE YNAH_REQ_ID,
REQUESTOR TYPE YNAH_REQUESTOR,
BUSINESS_OWNER TYPE YNAH_BUS_OWN,
FUNCTIONAL_OWNER TYPE YNAH_FUNC_OWN,
REQNUM TYPE YNAH_SERVICE_REQ_NUM,
PROJECT_ID TYPE YNAH_PRO_ID,
SYSTEM_ID TYPE YNAH_SYS_ID,
FUNCTIONAL_AREA TYPE YNAH_FUNC_AREA,
REQUEST_DATE TYPE YNAH_REQ_DATE,
REQUEST_TIME TYPE YNAH_REQ_TIME,
END OF TY_YNAH_CUS_OBJ_REQ.
"defining internal table -2
DATA: IT_YNAH_CUS_OBJ_REQ type TABLE OF TY_YNAH_CUS_OBJ_REQ
* WA_YNAH_CUS_OBJ_REQ type TY_YNAH_CUS_OBJ_REQ.
DATA: it_fcat TYPE slis_t_fieldcat_alv ,
wa_fcat TYPE slis_fieldcat_alv,
gd_layout TYPE slis_layout_alv.
SELECTION-SCREEN BEGIN OF BLOCK menu WITH FRAME TITLE text-001.
SELECT-OPTIONS: s_proid FOR IT_YNAH_CUS_OBJ_REQ-PROJECT_ID.
PARAMETER p_sysid type TY_YNAH_CUS_OBJ_REQ-SYSTEM_ID.
SELECTION-SCREEN: BEGIN OF LINE,
pushbutton 33(8) BUT user-command search.
SELECTION-SCREEN END OF LINE.
SELECTION-SCREEN END OF BLOCK menu.
initialization.
BUT = 'SEARCH'. END-OF-SELECTION.
"execute search function when user click search button
at selection-screen. "after processing user input
case SSCRFIELDS.
when 'SEARCH'.
SSCRFIELDS-UCOMM = 'ONLI'.
endcase.
"fetch data using select-4 START-OF-SELECTION.
SELECT *
FROM YNAH_CUS_OBJ_REQ "Database
INTO CORRESPONDING FIELDS OF TABLE IT_YNAH_CUS_OBJ_REQ "Into internal table
WHERE
PROJECT_ID in s_proid and
SYSTEM_ID eq p_sysid.
CALL FUNCTION 'REUSE_ALV_FIELDCATALOG_MERGE'
EXPORTING
I_PROGRAM_NAME = sy-repid
I_INTERNAL_TABNAME ='TY_YNAH_CUS_OBJ_REQ'
* I_STRUCTURE_NAME = 'TY_YNAH_CUS_OBJ_REQ'
* I_CLIENT_NEVER_DISPLAY = 'X'
I_INCLNAME = sy-repid
* I_BYPASSING_BUFFER = 'X'
* I_BUFFER_ACTIVE = CHANGING CT_FIELDCAT = it_fcat.
* EXCEPTIONS
* INCONSISTENT_INTERFACE = 1
* PROGRAM_ERROR = 2
* OTHERS = 3
* .
IF SY-SUBRC <> 0.
** Implement suitable error handling here
ENDIF.
The REUSE_*ALV* function modules are unsupported. I'd suggest switching to the CL_SALV_* classes. The documentation is better, there are more sample programs (DEMO_SALV_*) and you get support.
You need a dictionary structure if you want to get dictionary-based field descriptions (duh). If you assemble a structure type on the ABAP level using TYPE ... BEGIN OF ... END OF ..., as far as I know, the dictionary types for the individual fields are converted to ABAP types first and only then assembled into a structure type. Anyway, the dictionary reference of the original fields is lost. Instead of defining the structure of the output table in your code, use a dictionary structure.
You have some mistakes you might have not known (SAP is very confusing sometimes
and not transparent with error-messages). I got for you a working example of mine, have a look on it, especially on the comments.
First, data definition:
TYPE-POOLS slis. "import you need for REUSE_ALV_FIELDCATALOG_MERGE
DATA:
lt_fieldcat TYPE slis_t_fieldcat_alv,
BEGIN OF G_IT_MATERIAL occurs 0,
MATNR LIKE MARA-MATNR,
MTART LIKE MARA-MTART,
MAKTX_DE LIKE MAKT-MAKTX,
MAKTX_FR LIKE MAKT-MAKTX,
MAKTX_IT LIKE MAKT-MAKTX,
ERNAM LIKE MARA-ERNAM,
ERSDA LIKE MARA-ERSDA,
LAEDA LIKE MARA-LAEDA,
END OF G_IT_MATERIAL.
It is absolutely necessary that you define your local structure directly with LIKE, otherwise the parser from REUSE_ALV_FIELDCATALOG_MERGE will not find it.
Select your stuff:
SELECT ma~matnr ma~mtart ma~ernam ma~ersda ma~laeda
de~maktx as maktx_de fr~maktx as maktx_fr it~maktx as maktx_it
FROM mara as ma
LEFT JOIN MAKT as de ON de~matnr = ma~matnr AND de~spras = 'DE'
LEFT JOIN MAKT as fr ON fr~matnr = ma~matnr AND fr~spras = 'FR'
LEFT JOIN MAKT as it ON it~matnr = ma~matnr AND it~spras = 'IT'
INTO CORRESPONDING FIELDS OF TABLE g_it_material
WHERE ...
Create a field catalog dynamically:
CALL FUNCTION 'REUSE_ALV_FIELDCATALOG_MERGE'
EXPORTING
I_PROGRAM_NAME = sy-repid
I_INTERNAL_TABNAME = 'G_IT_MATERIAL'
I_INCLNAME = sy-repid
CHANGING
ct_fieldcat = lt_fieldcat
EXCEPTIONS
inconsistent_interface = 1
program_error = 2
OTHERS = 3.
IF sy-subrc <> 0.
* MESSAGE ID SY-MSGID TYPE SY-MSGTY NUMBER SY-MSGNO
* WITH SY-MSGV1 SY-MSGV2 SY-MSGV3 SY-MSGV4.
ENDIF.
Now display the ALV grid:
CALL FUNCTION 'REUSE_ALV_GRID_DISPLAY'
EXPORTING
it_fieldcat = lt_fieldcat "you could also give a structure
"i_structure_name = 'ZMM_SMATERIAL' "here instead of the fieldcat
TABLES
t_outtab = g_it_material
EXCEPTIONS
program_error = 1
OTHERS = 2.
Note that the parser also needs a max linesize of 72 chars.
There are several different text components provided by structure slis_fieldcat_alv that are used as column labels. The chosen text depends on the current column width (which itself usually depends on the length of the data displayed). Make sure that you change them all accordingly!
The usual technique is: By passing the I_STRUCTURE_NAME, you get a field catalog corresponding to this DDIC structure (the changing parameter ct_fieldcat). You then modify this internal table according to your needs and pass that modified table to the REUSE_ALV_GRID_DISPLAY.
In cases where I don't distinguish the different-sized text versions, I use the following macros to set all the text fields to the same value. The macros require a local work area ls_fieldcat (with the linetype of ct_fieldcat) and a local string variablelv_text` in order to work.
define set_field.
* Feld &1 für Anzeigefeld &2 den Wert &3 zuweisen
ls_fieldcat-&1 = &3.
modify ct_fieldcat from ls_fieldcat
transporting &1
where fieldname cp '&2'.
end-of-definition.
define set_text_direct.
lv_text = &2.
set_field seltext_s &1 lv_text.
lv_text = &2.
set_field seltext_m &1 lv_text.
lv_text = &2.
set_field seltext_l &1 lv_text.
end-of-definition.

WITH statement for structures

In some languages there exist syntax to simplify method and property access of a single object. As an example from VB.NET:
With foo
.bar()
.reset(true)
myVar = .getName()
End With
Or as Jagger points out, in Pascal:
With myBookRec do
Begin
Title := 'Some Book';
Author := 'Victor John Saliba';
ISBN := '0-12-345678-9';
Price := 25.5;
End;
Now I often find myself setting a ton of properties on a given structure. Take for instance the classic case of setting a field catalog:
LOOP AT lt_fieldcat INTO ls_fieldcat.
CASE ls_fieldcat-fieldname.
WHEN 'COLUMN_1'.
ls_fieldcat-edit = ''.
WHEN 'COLUMN_2'.
ls_fieldcat-edit = 'X'.
ls_fieldcat-outputlen = 20.
WHEN 'COLUMN_3'.
ls_fieldcat-edit = abap_true.
ls_fieldcat-drdn_hndl = '2'.
ls_fieldcat-drdn_alias = 'X'.
ls_fieldcat-outputlen = 5.
ls_fieldcat-col_pos = 1.
ls_fieldcat-scrtext_s = text-f01.
ls_fieldcat-scrtext_m = text-f01.
ls_fieldcat-scrtext_l = text-f01.
ENDCASE.
MODIFY lt_fieldcat FROM ls_fieldcat.
ENDLOOP.
Is there any way to replicate WITH-like functionality in ABAP?
If not, is there a better way to replicate this behaviour?
I've actually considered replicating the "return this" method daisy chain for a field catalog class. This seems like a WITH-alternative but I'm not entirely sold on its usefulness, given that you'd have to maintain a class for every structure.
ls_fieldcat = zcl_fieldcat=>new( ls_fieldcat )->edit('')->outputlen(20)->col_post(1)->get().
I don't think the daisy-chain approach is that useful - it's not much easier to read, and if anything goes wrong, it's awfully hard to debug. Personally, I'd probably use a "multi-setter" with multiple optional attributes:
lr_my_field->set_attributes( edit = abap_true
title = 'foo'
* ...
bar = 'baz' ).
That being said, I don't have a problem with the code in your question at all. Clearly readable, simple data handling. Just put it away into a method named PREPARE_FIELDCATALOG and everone knows what to expect.
With the new ABAP syntax you could simplify your code a bit with VALUE #( ... ) statement (or VALUE #( BASE ls_fieldcat ...) for modifying a structure with already assigned fields) which can be used for emulating WITH in ABAP. Not a perfect solution, especially if one has a CASE structure like yours but surely neat for initalizing a structured variable.
LOOP AT lt_fieldcat INTO ls_fieldcat.
CASE ls_fieldcat-fieldname.
WHEN 'COLUMN_1'.
ls_fieldcat = VALUE #(
BASE ls_fieldcat
edit = ''
).
WHEN 'COLUMN_2'.
ls_fieldcat = VALUE #(
BASE ls_fieldcat
edit = 'X'
outputlen = 20
).
WHEN 'COLUMN_3'.
ls_fieldcat = VALUE #(
BASE ls_fieldcat
edit = abap_true
drdn_hndl = '2'
drdn_alias = 'X'
outputlen = 5
col_pos = 1
scrtext_s = text-f01
scrtext_m = text-f01
scrtext_l = text-f01
).
ENDCASE.
MODIFY lt_fieldcat FROM ls_fieldcat.
ENDLOOP.