Some Comma Delimited CSV files will contain commas in the data fields and this causes problems in loading the file into our system. I don't think there is a standard function module to do this but I will share a small code snippet which will save someone to re-invent the wheel.
Here is a sample of the comma delimited CSV text file, you can copy below text then paste in notepad and save as c:\Temp\Book1.csv file.
10,Shafiq,"Sacramento, CA"
11,Willie,"New York, NY"
12,Conner,"Seattle, WA"
Here is the sample program which will read this file and store in a table that will preserve the commas in the city field:
Upload from Local PC:
TYPES: BEGIN OF kcde_intern_struc.
INCLUDE STRUCTURE kcde_cells.
TYPES: END OF kcde_intern_struc.
DATA: l_intern TYPE TABLE OF kcde_intern_struc WITH HEADER LINE.
DATA: BEGIN OF it_output OCCURS 0,
id(10) TYPE c,
name(20) TYPE c,
location(20) TYPE c,
END OF it_output.
DATA: wa_output LIKE LINE OF it_output.
DATA: v_index TYPE i.
FIELD-SYMBOLS : <fs>.
CALL FUNCTION 'KCD_CSV_FILE_TO_INTERN_CONVERT'
EXPORTING
i_filename = 'C:\Temp\Book1.csv'
i_separator = ','
TABLES
e_intern = l_intern
EXCEPTIONS
upload_csv = 1
upload_filetype = 2.
LOOP AT l_intern.
MOVE : l_intern-col TO v_index.
ASSIGN COMPONENT v_index OF STRUCTURE wa_output TO <fs>.
MOVE : l_intern-value TO <fs>.
AT END OF row.
APPEND wa_output TO it_output.
CLEAR wa_output.
ENDAT.
ENDLOOP.
Upload from Unix Server
DATA: BEGIN OF itab_upload OCCURS 0,
str(4096),
END OF itab_upload.
DATA: l_intern TYPE TABLE OF kcde_intern_struc WITH HEADER LINE.
OPEN DATASET p_path FOR INPUT ENCODING UTF-8 IN TEXT MODE.
DO.
READ DATASET p_path INTO itab_upload.
IF sy-subrc NE 0.
EXIT.
ENDIF.
APPEND itab_upload.
ENDDO.
CLOSE DATASET p_path.
PERFORM separated_to_intern_convert IN PROGRAM saplkcde
TABLES itab_upload l_intern USING ','.
LOOP AT l_intern.
MOVE : l_intern-col TO v_index.
ASSIGN COMPONENT v_index OF STRUCTURE wa_output TO <fs>.
MOVE : l_intern-value TO <fs>.
AT END OF row.
APPEND wa_output TO it_output.
CLEAR wa_output.
ENDAT.
ENDLOOP.