SAP的校验和替代(validations and substitutions)
SAP的校验和替代(validations and substitutions)
(2011-04-20 23:06:16)
I. Creating, activating and transporting validations and substitutions
1. Which transactions are required to use validation or substitution functions?
GGB0 Validation Maintenance GGB1 Substitution Maintenance GCVZ Rule Maintenance GGB4 Action selection for validation/substitution, for example, activation of application areas, use of all messages or substituted fields GCX2 Definition of the Includes for user exits
OB28 Activation of FI validations OBBH Activation of FI substitutions OBBZ Activation of cost of sales accounting substitution (in the FI area) OKC7 Activation of CO validations OKC9 Activation of CO substitutions
GCT9 Transport: Validations GCT0 Transport: Substitutions
2. What must I take into account when using the formula editor?
a) Validation GGB0
If you create the first validation for a callup point, for example 'financial accounting - document header', you have to define a message class. This is then valid for ALL validations for this callup point.
Not all fields can be used for validations on certain callup points, for example 'Complete document in FI'. This is because there are several rows there and it is not clear on which row the check is to be carried out. You can only use numeric fields here. (Note 413956) You should use a user exit if you need more checks on this callup point.
b) Substitution GGB1
If you create a substitution, for example 'financial accounting document line item', only certain fields are authorized for the substitution. This is because problems can occur with the documents or inconsistencies can occur between different applications. You can release more fields for substitution but this can be critical as serious problems can occur as a result. This is why you should regard the following change as a modification (Note 170183): The GB01 table contains information about which table fields can or cannot be substituted for which callup point. You can make a field substitutable by changing the BEXCLUDE field to ' '.
For example: 'Complete document' callup point in the FI In accordance with the default, no field is released for substitution on this callup point. You can see this in the following entries:
BOOLCLASS CLASSTYPE BCLTAB BCLFIELD BEXCLUDE 015 S BKPF * X 015 S BSEG * X
BOOLCLASS - Specifies the Boolean class. This is determined in the application area and callup point in each case, for example: 008 - Document header (FI) 009 - Line item (FI) 015 - Complete document (FI) 016 - Cost of sales accounting 100 - Document header (CO) CLASSTYPE Specifies if the field can be used in the case of validations (B), substitutions (S) or both BCLTAB - Table that is used BLCFIELD - Field of this table BEXCLUDE - Can be used ' ' or cannot be used 'X'.
As mentioned above, you should equate this change with a modification. Problems that occur as a result of this change cannot be handled in the 'standard' support.
3. What must I take into account when activating validations or substitutions?
You can activate validations and substitutions in two ways. Activation level: 1 - Active 2 - Active, except during 'batch input' processing
Because the activation in the financial accounting (FI) is company-code-specific, the validations or substitutions should be activated in both company codes in the case of cross-company code transactions.
4. What must I take into account when transporting validations or substitutions?
Usually, only table entries are transported when validations or substitutions are transported. The generated source code should not be transported, since this can cause problems. After the import into the target system, the source code is generated automatically when the RGUGBR00 report is started.
If you would like to delete a validation/substitution, and do not want to do this in each system individually, proceed as described in Note 840981.
II. Using user exits in validations or substitutions
1. Where can I maintain user exits for validations and substitutions?
The exits for validations and substitutions are stored in an Include of your choice. The following Includes are delivered as examples by default: RGGBR000 Val/Sub: Exits for rules RGGBS000 Val/Sub: Exits for substitutions You should copy these into the customer namespace, for example ZGGBR000. The length of the name you choose should not exceed 8 characters. This is because the name is also generated in the source code, and so a certain line length may be exceeded as a result. You can also create a separate Include per client as this can improve performance.
In Transaction GCX2, you should then define where your Include is used. To do this, enter the name of your Include as a form pool in the relevant application area.
2. How should I create a user exit?
Define your user exit in the FORM Routine GET_EXIT_TITLES with the correct exit category (EXITS-PARAM). You can use the existing sample entries as a model for this.
The exit types required in the FORM Routine GET_EXIT_TITLES are organized as follows:
· C_EXIT_PARAM_NONE No parameter
Apart from the return value (B_TRUE or B_FALSE), you do not need any parameters for the validation. If you do not want to substitute a field, use this parameter in the substitution (select "Only exit").
However, you must then use a TABLE statement in the formpool (not in the actual form) to identify the structure concerned.
· C_EXIT_PARAM_FIELD A Field is transferred as a parameter
You only use this type in the substitution. Exactly one field is substituted. Enter user exits of this type in the substitution maintenance instead of assigning a constant value.
· C_EXIT_PARAM_CLASS A complete type is transferred as a parameter
Matrix validations/substitutions ("Only exit"). This exit type can only be used in the case of callup points with complete documents, for example in the FI and the LC. You cannot use this exit category on any other callup points. The transferred type contains all information of a class. An example is the use of the document header and all corresponding line items in the FI area.
In the FORM routine GET_EXIT_TITLES, change the NAME and TEXT field of the EXITS internal table to the values you want.
Use one of the sample routines as a model and copy these to the names you selected. You can now implement the functions you want in this form routine. After you have finished creating the user exit, you can use it in your validations or substitutions.
3. What must I take into account when implementing the user exit?
No dialog boxes, warning messages, information messages and error messages can be issued in an exit nor can a "COMMIT WORK" be rejected. This would, for example, interrupt the batch processing or cause problems with the documents to be posted. If you find a serious error in the field contents in the exit, send a termination message (A or X), which will ensure that the transaction is terminated.
Do not use any ABAP commands that cause the exit to be left directly, for example 'LEAVE (TO) SCREEN ...' or 'SUBMIT ...' or 'CALL TRANSACTION ...'
In an exit, use only structure fields that belong to the Boolean class of the calling substitution, validation or rule. The other structure fields are not filled or are not filled properly on the callup point of the calls of the substitution, validation or rule.
Changes to field contents may only be made in a substitution exit if this field can be substituted in the Boolean class. You can display such fields by selecting the "Add entry" function in the substitution maintenance.
In the Includes of the substitution exits, you must not use the commands MODIFY, INSERT or DELETE in the internally used structures such as BSEG or BKPF. These structures are interpreted internally as database tables because they are defined by a TABLES statement. As a result, database records are written, deleted or changed when you use the commands mentioned above. This can cause serious inconsistencies and problems in the document processing.
If you want to change field contents in Exit Types C_EXIT_PARAM_CLASS, you should make the changes in the internal table BOOL_DATA (for example BOLL_DATA-BSEG).
III. Problem analysis
1. What must I take into account before I analyze the problem?
Before you analyze a problem with validations/substitutions, make sure that the generated source code is up-to-date. You can use Report 'RGUGBR00' to do this. Call this report in Transaction SE38. You receive the following selection screen:
Application area ' ' Application callup point ' '
' ' Regenerate main program ' ' Generate table file ' ' Generate rules ' ' Generate validations ' ' Generate substitutions ' ' Generate sets ' ' Generate rule use ' ' Generate validation use ' ' Generate substitution usage ' ' Generate substitution export routines ' ' Generate substitution routines in all clients ' ' Generate trace program calls
Now update the source code for the application area and application callup point in which the validation or substitution was created. To do this, enter the application area, for example FI, and the application callup point (0003/complete document), and set all indicators except for 'Generate substitution routine in all clients' and 'Generate trace program calls'. Then execute the report.
The source code is now at its current status.
2. How can the validation and substitution trace be used?
a) Validation trace
You created and activated a validation. You can now activate the trace, for example in Transaction GGBO. To do this, place the cursor on the validation and select 'Extras -> Activate trace'. Example: You created an FI validation on the 'Line item' callup point. To test the validation, execute, for example, Transaction FB01. The trace is displayed as soon as the validation is run. You can now see different things in the trace, for example the values contained in the fields of the various structures. From this, you can deduce whether additional checks are required, or tell what the cause of a problem is.
b) Substitution trace
Activate the trace in the same way as you activate validations. Example: You created a substitution in the same way as you activated the trace in the case of the validation. The trace for substitutions has an additional function, in that you can view the field contents before the substitution as well as after the substitution. There, problems can be reproduced in the same way.
3. How can I use the DEBUGGER to analyze a problem?
a) Basic information
All Includes for validations and substitutions are created in this form: GBTmmaa0. mm stands for two digit client key. You can use the G_CLIENT_ENCODE function module to convert a three-digit client (>099) to a two-digit client, for example: client 100 --> AA client 800 --> TQ AA stands for the application area, for example FI --> Financial Accounting CO --> Cost Accounting You will find this abbreviation in the maintenance transaction. Therefore, GBTAAFI0 is the Include in client 100 for FI validations and substitutions.
b) Structure of the Includes and form routines for validations and substitutions and tips on how to 'debug'
The above-mentioned Includes are structured in the following way (sample Include in FI, Client 800):
INCLUDE GBTTQFIA. " Rules INCLUDE GBTTQFIB. " Substitutions INCLUDE GBTTQFIC. " Substitution conditions INCLUDE GBTTQFID. " Validations INCLUDE GBTTQFIE. " Validation conditions INCLUDE GBTTQFIF. " Validation checks INCLUDE GBTTQFII. " Application rules INCLUDE GBTTQFIJ. " Substitution export routines INCLUDE GBTTQFIM. " Trace/simulation INCLUDE GBTTQFIS. " Substitution import routines
c) The form routines for validations have the following sample structure (sample validation with the name VAL01):
FORM VAL_VAL01. G_BFOUND = B_TRUE. DATA: COND LIKE D_BOOL, CHECK LIKE D_BOOL, NVALSEVEREST LIKE SY-SUBRC VALUE -1.
G_STEP = '001'. PERFORM VALCOND_1VAL01##001 CHANGING COND.
IF COND = B_TRUE . PERFORM VALCHCK_2VAL01##001 CHANGING CHECK.
IF CHECK = B_FALSE . PERFORM MESSAGE_OUTPUT USING " ...Nachrichtenparameter CHANGING NVALSEVEREST. ENDIF. ENDIF. ENDFORM. " VAL_VAL01
Statement G_STEP = '001' indicates the step that is executed. The prerequisite of the step is defined in Form Routine VALCOND_..., the form routine VALCHCK_... checks the validation step.
To analyze this step, you should create breakpoints on these two PERFORM statements and also on the code line 'IF CHECK = FALSE'. In doing so, you can determine whether your prerequisite or the check is causing the problem you have, for example a field that is to be checked does not contain the value that you expect.
d) The form routines for substitutions have the following sample structure (sample substitution with the name SUB01):
FORM SUB_SUB01. G_BFOUND = B_TRUE. DATA: COND LIKE D_BOOL. G_CLASSID = '009'. G_SUBST_PERFORMED = B_FALSE.
IF G_TAB_FLAG-BSEG <> B_TRUE . PERFORM READ_MASTER_DATA USING 'BSEG' G_BFOUND. IF G_BFOUND = B_FALSE . MESSAGE E083 WITH 'BSEG'. ENDIF. ENDIF.
G_STEP = '001'. PERFORM SUBCON_3SUB01##001 CHANGING COND.
IF COND = B_TRUE . BSEG-SGTXT = 'This is a Test'. " Text G_SUBST_PERFORMED = B_TRUE. ENDIF. ENDFORM. " SUB_SUB01
As mentioned above for validations, 'G_STEP = '001' indicates the step, and the prerequisite of the substitution is defined in Form Routine SUBCON_.... The actual substitution is carried out by the statements in the 'IF COND = B_TRUE --> ENDIF' structure. These statements vary depending on the 'type' of substitution created (field allocation, constant value, user exit).
In the substitutions, you should also set a breakpoint on the PERFORM on the prerequisite and the statement 'IF COND = B_TRUE'.
4. Function modules that call validations and substitutions
Validations - Function module G_VSR_VALIDATION_CALL The different structures/tables, the name of the validation, callup point and application area are assigned to this function module. Substitutions - Function module G_VSR_SUBSTITUTION_CALL The parameters transferred to validations are also transferred to this function module. However, structures/tables that contain the changed values are also returned.
You could set a breakpoint to analyze both function modules. In doing so, you could see which data was transferred, which data the validation or substitution attains, and which values are returned.
5. What must I take into account during the set usage?
When sets are used in validations or substitutions, source code is generated from the values in the sets. Performance problems can occur if a set is of a certain size, since the generated source code causes longer runtimes. A runtime error can also occur.
To reduce the size of the source code of the sets used, you can proceed as follows:
Use a smaller set in the validation, substitution or rule. To do so, group single values in the set into intervals (for example, instead of the single values 0001, 0002 and 0003, use the interval 0001 - 0003)
Instead of " IN SET", use the rule "( IN SET1 OR IN SET2)" in which SET1 and SET2 contain the values from SET (for example, subhierarchies on the highest hierarchy level).
The simplest way to determine the number of single values or intervals in a set hierarchy is to execute Program RGSEX000. Enter the name of the set and execute the program. All contained sets and values/intervals are listed and the sequential number of the entries is contained in the first column.
6. Why does a validation or substitution work in the development system, but not work in another system after a transport?
You should make sure that validations and substitutions are activated in the target system. You can also use Report RGUGBR00 to execute the regeneration of the source code again.
There could be another cause in the case of substitutions: You try to substitute a field that is not released as a default by SAP. To substitute a field, you have adjusted Table GB01 in your development system. However, this did not occur in the target system. The generated source code is correct and the value was changed in the substitution, but the change was not returned to the calling application via the export routines.
