Enhancing mbdoc's classical datapart

[Former Member]

Hi all,

We have a requirement that needs enhancing of the an mbdoc's classical datastructure with an additional field (to influence data distribution).

Adding the field to the bdoc is nog big deal, but can anybody tell me where to fill it? At the moment, I'm using an additional function, added to the standard flow context of the bdoc, but maybe there's a better place to do this.

Does anybody have any experience with this?

Thx in advance,

Michael.

[Former Member]

Hi Michael,

Adding a field to an MBDoc requires the following changes.

1. The place at which the SBDoc is mapped to an MBDoc should be enhanced to get the z field of MBDoc to be mapped with data.

2. If step 1 is done, the validation module will have data for the Z field. Now modification should be done to the save function present inside the validation module so that it updates the correponding online tables with value for Zfield.

Also the SBDoc created for Process notification should be filled with the zfield manually so that data is passed to mobile bridge.

(Maybe for thes modifications check if there is some user exit available inside SBDoc to MBDoc mapping and validatio n mapping. Then that is the right place to add values to the fields.)

I don't think putting an additional call such as 'MOA' will be of help. As additionals calls are used for functions like sending data to BW or sending data to some other systems.

I hope this is of some help to you.

MBDoc classic part.

To fill MBDoc classic part the function used is

CALL METHOD cl_smw_mflow=>set_header_fields

EXPORTING

in_header = in_header

in_bdoc_type = bdoc_type

in_msg_type = msg_type

in_root_id = root_id

IMPORTING

out_header = ls_header.

you can modify but please be careful as this function is used by all MBDocs.

best regards

Middleware developer

[Former Member]

Hi,

Thx for the info, but it doesn't really help me any further.

The field is only necessary in the mbdoc header: no mapping to sbdoc or mobile bridge enhancements are necessary.

As said, I've enhanced the flow context of the bdoc to fill up the Z-field, but the function needs to be inserted before the other ones: if not (eg MOA), the outbound adapters will already be called before the information is actually there.

I'm just wondering whether there is a better place in the standard to influence the header-fields of an mbdoc.

Regards,

Michael.