application file with more than one #, error while splitting into fields.

[Former Member]

Data file at application server is like this.

RAJU#SOM#1701 AVENUE#DELI#IN#EN#12345 --->record1

ARTHUR##1701 NW 23 AVENUE####12345 -

---

>record2

-

---

Read file into work_area.

SPLIT work_area AT '#' INTO f1

f2

f3

f4

f5

f6

f7.

if the work_area is seperated by one # then it is splitting and sy-subrc is 0. That is first record is spliting fine.

for the second record. sy-subrc is 4, i.e., this split is not splitting work_area into the 7 fields.

Please help out with this urgent.

[Former Member]

Why not try the following:

REPLACE ALL OCCURRENCES OF '#' IN work_area WITH '|'.

SPLIT work_area AT '|' INTO f1

f2

f3

f4

f5

f6

f7.

Often on your dataloads we use the pipe field to delimit data

Hope this helps

J

[Former Member]

Hi J,

Replace ALL OCCURANCE IS NOT WORKING.

THAT IS, # is not replacing with |.

and if there is only one # in between 2 fields its replacing.

if there are two # in between its not replacing.

Can you please explain why

[Former Member]

Hi J,

My requirement is to add a custom button in the ALV output and handle that button after user presses.

So, can any body please tell in which method of BADI: CATS_REPORTING, a custom button can be added in the ALV output of tcode CATS_APPR_LITE and in which method the custom button can be handled.

This tcode is related to Cats Time Approve.

Thanks in advance

[Former Member]

My6 guess is that ## is being treated as a Unicode character. Looking at Tabs, Line Feeds and Carriage returned these are displayed as #.

If the | works why worry.

J

[Former Member]

Hey J,

For two # i.e., ## replace all occurance and split is not working so can you please tell me any solution to seperate the fields.

Thanks.

[Former Member]

This works in ECC6.0

Otherwise the old and trusted method.

Do 6 times

replace '#' with '|' into work_area.

enddo.

J

[Former Member]

oh no J

DO 6 Times is also not working

[Former Member]

Your code works on our system so you are not going mad, I cut and pasted it into a test program.

test

RAJU SOM 1701 AVENU DELI IN EN 12345

ARTHUR 1701 NW 23 12345

sorry can't add the screen print to this

I have checked it out - what version of SAP are you on? May be that is the difference.

Does the file have to be delimited by # what if it is delimited by |. My concerns are unicode.

J

[Former Member]

Hi J,

I am using ECC 6.0

and yes the file is delimited by # only

RAJU#SOM#1701 AVENUE#DELI#IN#EN#12345 --->record1

ARTHUR##EUM###12345 -

---

>record2

record 1 is working fine.

record 2 giving problem because fields are delimited by more than one #

[Former Member]

This has got to be something to do with unicode. What does the file look like before it is loaded to the application server. I have seen ',' transfered to the application server as # due to Unicode

J

[Former Member]

the file is tab delimited .txt file

which is uploaded into application server.

That is the reason the fields in the application server is seperated by #

Edited by: Jalaaluddin Syed on Jul 16, 2008 3:10 PM

[Former Member]

Definitely a unicode thing

try.

Constants: Tab_mark(2) type c value cl_abap_char_utilities=>HORIZONTAL_TAB,

Split at tab_mark.......................etc.

Or

Also when the file is being transfered to the application servce I assume by FTP try different settings maybe IBM.( this is a guess I as Don't have access to FTP stuff'.

J

[Former Member]

Hey J

there are 2 transfer format 1. is BIN and other is ASC

so i have used both but its delimiting with # only

and can you please explain me wht you have sent in ur last forum

***************************************************

Constants: Tab_mark(2) type c value cl_abap_char_utilities=>HORIZONTAL_TAB,

Split at tab_mark.......................etc.

Or

Also when the file is being transfered to the application servce I assume by FTP try different settings maybe IBM.( this is a guess I as Don't have access to FTP stuff'.

[Former Member]

This works so you must have missed a step.

J

REPORT ZJKSW_TREST.

types w_area(252) type c.

data work_area type standard table of w_area with header line.

data: f1(20) type c,

f2(20) type c,

f3(20) type c,

f4(20) type c,

f5(20) type c,

f6(20) type c,

f7(20) type c.

work_area = 'RAJU#SOM#1701 AVENUE#DELI#IN#EN#12345'.

append work_area.

work_area = 'ARTHUR##1701 NW 23 AVENUE####12345'.

append work_area.

Loop at work_area.

replace all occurrences of '#' in work_area with '|'.

SPLIT work_area AT '#' INTO f1

f2

f3

f4

f5

f6

f7.

endloop.

[Former Member]

Ops spotted my typo

SPLIT work_area AT '|' INTO f1

J

[Former Member]

Hey J its moving only the first field.

and as i said replace all occurance is not working at all

[Former Member]

hi syed,

it will not give sy-subrc = 4.

what it will do is that it will split on every # and your string in f2, f4,f5,f6 will be empty as there is no space between #### in second record

check this code

data: a type string,

b type string.

data: f1(20),f2(20),f3(20),f4(20),f5(20),f6(20),f7(20).

a = 'RAJU#SOM#1701 AVENUE#DELI#IN#EN#12345'.

b = 'ARTHUR##1701 NW 23 AVENUE####12345'.

SPLIT a AT '#' INTO f1

f2

f3

f4

f5

f6

f7.

write:/ f1,f2,f3,f4,f5,f6,f7.

SPLIT b AT '#' INTO f1

f2

f3

f4

f5

f6

f7.

write:/ f1,f2,f3,f4,f5,f6,f7.

Sachin

[Former Member]

ya sachin its filling on the first field i.e., f1.

rest of the fields are empty.

Can you please tel me the solution of this

[Former Member]

Wipee,

This appears to work.

J

REPORT ZJKSW_TREST1.

Constants: Tab_mark type string value cl_abap_char_utilities=>HORIZONTAL_TAB.

Data: c_hash type string,

c_pipe type string.

c_pipe = '|'.

c_hash = '#'.

data : a type string,

b type string,

c type string.

data: f1(10),f2(10),f3(10),f4(10),f5(10),f6(10),f7(10).

concatenate 'RAJU' tab_mark tab_mark tab_mark tab_mark 'SPRING' tab_mark 'Smith' tab_mark into c.

replace tab_mark with c_pipe into c.

replace tab_mark with c_pipe into c.

replace tab_mark with c_pipe into c.

replace tab_mark with c_pipe into c.

replace tab_mark with c_pipe into c.

replace tab_mark with c_pipe into c.

SPLIT c AT '|' INTO f1

f2

f3

f4

f5

f6

f7.

write:/ f1,f2,f3,f4,f5,f6,f7.

[Former Member]

Hi Julie Waller,

it solved my problem.

Thanks alot

Syed

[Former Member]

Good puzzle - one for the memory banks.

Enjoy.

J