Replacing hex characters problem

[Former Member]

I have a text box where users can paste text into. When processing I notice a bunch of special characters.

To remove them, I am converting to Hex using

  
CALL METHOD cl_abap_codepage=>convert_to
    EXPORTING
      source      = iv_text
      codepage    = `UTF-8`
      replacement = ''
      ignore_cerr = abap_false
    RECEIVING
      result      = buffer.

Then I replace the special character posing as an apostrophe with the hex equivalant of an apostrophe.

  DATA:lv_old_ap TYPE x VALUE '99',
            lv_new_ap TYPE x VALUE '27',
  
  REPLACE ALL OCCURRENCES OF lv_old_ap IN buffer WITH lv_new_ap IN BYTE MODE .

Then convert it back to a string.

  CALL METHOD cl_abap_codepage=>convert_from
    EXPORTING
      source      = buffer
      codepage    = `UTF-8`
      replacement = ''
      ignore_cerr = abap_true
    RECEIVING
      result      = iv_text.

However, for some reason it's also inserting two spaces when converting the apostrophe.

Input text:

u201CI donu2019t have any skulls and crossbones on me.u201D

Output text:

u201CI don 't have any skulls and crossbones on me.u201D

Any idea why it's inserting two spaces?

[Former Member]

Just use the statement

If you know the Hex value then use the below class to get the special char..

CL_ABAP_CONV_IN_CE =>UCCP

exporting

uccp = '00A0' " Hex value

receiving

char = buffer.

your replace statement need to change. like below

REPLACE ALL OCCURRENCES OF buffer IN lv_old_ap WITH lv_new_ap IN BYTE MODE .

Thanks

Subhankar

[Sandra_Rossi]

Hi,

Could you provide us with the original hex value of iv_text please? (for the text you mentioned, do it with the debugger )

I would like to make sure about the original encoding (because 99 doesn't correspond to an apostrophe, so it is probably accompanied with something else)

Moreover, do you have a Unicode system?

Thx

Sandra

[Former Member]

yes, we have a unicode system.

I just solved the issue by running the text through FM 'SCP_REPLACE_STRANGE_CHARS' instead of messing around with the hex replacement.

thanks for the help

[Former Member]

used FM 'SCP_REPLACE_STRANGE_CHARS'

[fabriciocaete]

Thank you very much. It solved the problem I had in Longtext.

[Subhankar]

Just use the statement

If you know the Hex value then use the below class to get the special char..

CL_ABAP_CONV_IN_CE =>UCCP

exporting

uccp = '00A0' " Hex value

receiving

char = buffer.

your replace statement need to change. like below

REPLACE ALL OCCURRENCES OF buffer IN lv_old_ap WITH lv_new_ap IN BYTE MODE .

Thanks

Subhankar