Is it possible to sort a structure by the col_pos from a field catalog table?

[walkerist]

Is it possible that the X_FINAL-EINDT will be concatenated first instead of X_FINAL-WERKS? Is there a way to it without hard coding the field names?

<code>DATA: begin of t_final,
   werks like ekpo-werks, 
   eindt like ekes-eindt, 
   uzeit like ekes-uzeit, 
   slfdt TYPE eket-slfdt, 
   ekgrp like ekko-ekgrp, 
END of t_final. 

DATA: t_fieldcat TYPE slis_fieldcat_alv,
      x_final  LIKE LINE OF t_final.

IN T_FIELDCAT, the first column is EIND

WHile in the T_FINAL and X_FINAL, the first column is WERKS.

Is it possible or useful to align the sorting orber of t_final from t_fieldcat?

The reason I'm asking because the X_FINAL is being used in a CONCATENATE code:

CONCATENATE
x_final-werks x_final-eindt x_final-uzeit x_final-ekgrp

Instead of WERKS being always the first to be concatenated, I want to concatenate the first field according to COL_POS from T_FIELDCAT.

Example: T_FIELDCAT-FIELDNAME: EIDNT

T_FIELDCAT-COL_POS: 1

T_FIELDCAT-FIELDNAME: WERKS

T_FIELDCAT-COL_POS: 2

The problem with the code below is sometimes the EIDNT has a col_pos of 1 but sometimes it has a value of like 15. can I concatenate without hard coding?

CONCATENATE
X_FINAL-EIDNT X_FINAL-WERKS

<br>

[jens_michaelsen]

Try it like this

FIELD-SYMBOLS: <fs_value> TYPE ANY. 
DATA: x_line type STRING. 

DATA: x_value TYPE STRING.
LOOP AT t_fieldcat ASSIGNING FIELD-SYMBOL(<fs_fieldcat>). 
    ASSIGN COMPONENT <fs_fieldcat>-fieldname OF STRUCTURE x_final TO <fs_value>. 

    x_value = <fs_value>. 
    CONCATENATE x_line x_value INTO x_line. 
ENDLOOP. <br>

[jens_michaelsen]

But be sure, that you can convert every part of x_final to a string.