regarding string operations

[Former Member]

Hi All,

i've a string holding the value as given below.

AA,17,2/19/2003,"9,999.00",USD,00,10,318,"193,275.31"

by performing some string operations i want the result string in the format as given below:

AA,17,2/19/2003,"9999.00",USD,00,10,318,"193,275.31"

i.e., i want to remove all the commas(,)that are included in between a pair of " " only.

can anyone provide me a sample code for the same

[christian_wohlfahrt]

Hi Vijay!

Maybe you have to do it the hard way:

- Move the input-line to a char field (l_text type char255)

- start a do-loop.

- assign one char after each other to a field symbol

- check, if content is equal ": if yes, switch a flag (on or off)

- check, if content is equal ',' and flag is on: replace content with space (because of assign, you change l_text)

- enddo.

- condense l_text -> all space are gone

Here the help for assign:

DATA text(10) TYPE c VALUE '0123456789'. 

FIELD-SYMBOLS <char> TYPE c. 

DATA off TYPE i. 

DO.
  off = sy-index - 1. 
  ASSIGN text+off(1) TO <char>. 
  IF sy-subrc ne 0.
    EXIT.
  ENDIF.
  WRITE / <char>. 
ENDDO. 

Regards,

Christian

[RichHeilman]

Vijay, this may not look pretty but it does work. Check it out.

report zrich_0003 .

data: pos1 type i,
      pos2 type i,
      diff type i.
data: string type string.
data: c(1000) type c.


string = 'AA,17,2/19/2003,"9,999.00",USD,00,10,318,"193,275.31"'.


c = string.

while sy-subrc = 0.
  search c for '"'.
  if sy-subrc = 0.
    pos1 = sy-fdpos + 2.
    search c for '"' starting at pos1.
    if sy-subrc = 0.
      pos2 = sy-fdpos + 2.
      pos1 = pos1 - 2.
      while sy-subrc = 0.
        replace ',' with space into c+pos1(pos2).
      endwhile.
      sy-subrc = 0.
      while sy-subrc = 0.
        replace '"' with space into c+pos1(pos2).
      endwhile.
      condense c+pos1(pos2) no-gaps.
      sy-subrc = 0.
    endif.
  endif.
endwhile.
condense c no-gaps.

write:/ c.

Regards,

Rich Heilman

[Former Member]

Hi Vijay,

Try this code.

data: text(255).

data: v_x.

text = 'AA,17,2/19/2003,"9,999.00",USD,00,10,318,"193,275.31"'.

do.

v_x = ' '.

search text for '"' and mark.

if sy-subrc = 0.

replace first OCCURRENCE OF '"' in section offset sy-fdpos of text with

''.

replace first OCCURRENCE OF ',' in section offset sy-fdpos of text with

''.

replace first OCCURRENCE OF '",' in section offset sy-fdpos of text with

','.

v_x = 'X'.

endif.

if v_x = ' '.

exit.

endif.

enddo.

write:/ text.

[Former Member]

Hi Phani,

ur logic worked great, but if i've have more than one comma in between a pair of " " it replaces only the first one. how can i handle this scenario ?

[Former Member]

you can use replace all occurences of ','.

regards

vijay

[Former Member]

Dear Vijay ,

Please try this code. Hope this would give the required result.

Report rpt .

Data: strtest type c length 100.

data: i type i.

data: j type i.

data: k type c. " Indicator flag for "

strtest = 'AA,17,2/19/2003,"9,999.00",USD,00,10,318,"193,275.31"'.

message strtest type 'I'.

i = 0.

j = strlen( strtest ).

while i < j.

if strtest+i(1) = '"' and k = ''.

k = 'X'.

elseif strtest+i(1) = '"' and k = 'X'.

k = space.

else.

endif.

if strtest+i(1) = ',' and k = 'X'.

strtest+i(1) = ''.

endif.

i = i + 1.

endwhile.

message strtest type 'I'.

Regards,

Shankar

[Former Member]

Hi vijay,

A bit complex but works for sure, check the following logic,

REPORT zsritest.

DATA: gs_string TYPE string.


gs_string = 'AA,17,2/19/2003,"9,999.00",USD,00,10,318,"193,275.31"'.

WRITE: / gs_string.

PERFORM string_trim CHANGING gs_string.

*---------------------------------------------------------------------*
*       FORM string_trim                                              *
*---------------------------------------------------------------------*
*       ........                                                      *
*---------------------------------------------------------------------*
*  -->  LS_STRING                                                     *
*---------------------------------------------------------------------*
FORM string_trim CHANGING ls_string.

  DATA: lt_string TYPE string OCCURS 0 WITH HEADER LINE,
        lv_tabix TYPE i,
        lv_start.

  SPLIT gs_string AT '"' INTO TABLE lt_string.
  CHECK sy-subrc EQ 0.
  CLEAR gs_string.
  LOOP AT lt_string.
    lv_tabix = sy-tabix MOD 2.
    IF lv_tabix EQ 0.
      TRANSLATE lt_string USING ', '.
      CONDENSE lt_string NO-GAPS.
    ENDIF.

    IF lv_tabix EQ 0 OR lv_start EQ 'X'.
      CONCATENATE gs_string lt_string INTO gs_string SEPARATED BY '"'.
      IF lv_start EQ 'X'.
        CLEAR lv_start.
      ELSE.
        lv_start = 'X'.
      ENDIF.
    ELSE.
      CONCATENATE gs_string lt_string INTO gs_string.
    ENDIF.
  ENDLOOP.

  IF lv_start EQ 'X'.
    CONCATENATE gs_string '"' INTO gs_string.
  ENDIF.

  WRITE: / gs_string.

ENDFORM.

Hope this helps..

Sri

[Former Member]

Hi vijay,

As you know that if you split this string with delimiter as ,then you will get 11 strings. right.

In that these strings you donot need , between 4 and 5

& 10 and 11. Right.

Use the bellow code.

data str1(20),
     str2(20),
     str3(20),
     str4(20),
     str5(20),
     str6(20),
     str7(20),
     str8(20),
     str9(20),
     str10(20),
     str11(20).

SPLIT source_string AT ',' INTO
str1,str2,str3,str4,str5,str6,str7,str8,str9,str10,str11.

concatenate str1, ',',
            str2, ',',
            str3, ',',
            str4,
            str5, ',',
            str6, ',',
            str7, ',',
            str8, ',',
            str9, ',',
            str10,
            str11.

hope it helps.

Regards,

Maheswaran.B

Message was edited by: Maheswaran B

[Former Member]

Hi Vijay,

u can do this using offset functionality.read each letter from the string when ever u encounter ".pass the letters after that inverted comma (") until another one is encountered in between this if u encounter a comma delete it.and finally concatenate that string

Regards,

Suresh

Message was edited by: suresh B