Returning an XmlDocument from an external function?

Hi,
In an attempt to simplify my XQueries, I have started looking at writing some of the more complex functions in Java and calling them via an XmlExternalFunction.
This seems fine, except that I have an issue with returning a constructed document (or any node set). The only way I have found to return a document without it being destroyed is to put it into a temporary database first. Here is the end of my function. I have written a convenience function that converts a DOM Xml Document into a dbxml XmlDocument (DomToXmlDbDoc).
     XmlResults results = mgr.createResults();
     XmlDocument result = DomToXmlDbDoc(doc);
     XmlContainer tmp = mgr.openContainer(txn, "_tmp.dbxml");
     result.setName(UUID.randomUUID().toString());
     putDocument(tmp, result);
     tmp.close();
     XmlValue value = new XmlValue(result);
     results.add(value);
     return results;
I would prefer not to have to use the temporary storage in the middle as it is inefficient and rather untidy, but the following just gives a blank result.
     XmlResults results = mgr.createResults();
     XmlDocument result = DomToXmlDbDoc(doc);
     XmlValue value = new XmlValue(result);
     results.add(value);
     return results;
Best Regards
Pete.

Hi,
this is a bug in DB XML, namely in Java bindings. I could try to search a patch for this (which is not very well tested though) and send you by email if you tell me your address.
Vyacheslav

Similar Messages

  • Return multiple columns from an analytic function with a window

    Hello,
    Is it possible to obtain multiple columns from an analytic function with a window?
    I have a table with 4 columns, an id, a test id, a date, and the result of a test. I'm using an analytic function to obtain, for each row, the current test value, and the maximum test value in the next 2 days like so:
    select
    id,
    test_id,
    date,
    result,
    MAX ( result ) over ( partition BY id, test_id order by date RANGE BETWEEN CURRENT ROW AND INTERVAL '2' DAY FOLLOWING ) AS max_result_next_two_day
    from table
    This is working fine, but I would like to also obtain the date when the max result occurs. I can see that this would be possible using a self join, but I'd like to know if there is a better way? I cannot use the FIRST_VALUE aggregate function and order by result, because the window function needs to be ordered by the date.
    It would be a great help if you could provide any pointers/suggestions.
    Thanks,
    Dan
    http://danieljamesscott.org

    Assuming RESULT is a positive integer that has a maximum width of, say 10,
    and assuming date has no time-component:
    select
       id
      ,test_id
      ,date
      ,result
      ,to_number(substr(max_result_with_date,1,10)) as max_result_next_two_day
      ,to_date(substr(max_result_with_date,11),'YYYYMMDD') as date_where_max_result_occurs
    from (select
            id
           ,test_id
           ,date
           ,result
           ,MAX(lpad(to_char(result),10,'0')||to_char(date,'YYYYMMDD'))
               over (partition BY id, test_id
                     order by date
                     RANGE BETWEEN CURRENT ROW AND INTERVAL '2' DAY FOLLOWING )
            AS max_result_with_date
          from table)

  • How to return custJava Object from Java Stored Function/Procedure to Java Application

    I WILL PAY $100 FOR SOLUTION.
    Oracle JDBC documentation gives step by step on how to return Oracle Type Object to Java Application. No question here.
    But is it possible to create Java Stored Function [getMyClass() ] which returns CUSTOM Java based object [MyClass] to Java application like:
    //java stored function
    class Foo
    public static MyClass getMyClass()
    return new MyClass();
    //java class I want to return to Java
    class MyClass
    public String getGreetings(String name)
    return "Hello " + name;
    I tried to do it using intermittent Oracle
    Type Object and OracleConnection map to make binding between Java->Oracle->Java. I could
    not get throgh meaningless run time ORA Errors. Any kind of help would be greatly
    appreciated. If you know the answer please
    post it or e-mail [email protected]
    null

    You don't say which version of the database you are on. There's an interesting looking article on OTN on handling CLOBs in JDBC in 10g. It may well be useful for earlier versions too (but no guarantees).
    Cheers, APC

  • Problem in the External Function returned data RAW

    Hi,
    I have a problem, I would want to create an external function in Oracle 10g, than given in input a data RAW it gives back in output the value in format raw.
    I have realized such function through language C.
    these are the prototypes of the function
    ORACLE side:
    CREATE OR REPLACE FUNCTION F_RETURN_RAW (inputText IN RAW)
    RETURN RAW                                                                  -- I do not know to return RAW !!!
    AS LANGUAGE C
    NAME "f_Return_Raw"
    LIBRARY C_Library_DAN
    PARAMETERS
    (inputText RAW,
    inputText INDICATOR,
    inputText LENGTH);
    C side:
    ???? f_Return_Raw(CK_BYTE_PTR in_Text,
    short *in_TextInd,
    int *in_TextLen)
    I do not know what to put in place of ????
    someone can help me with some example?
    thanks to all!
    Regards
    Daniele

    Hello I had already seen the package UTL_RAW,
    in particular is CAST_TO_VARCHAR2 that CAST_TO_RAW,
    but these me are useful Oracle side,
    my problem are to give back to Oracle, from side C, one RAW.
    Perhaps it is not possible to give back a RAW directly?
    or I have more probably not understood your suggestion!
    I would want to create a function that he concurred me to make that
    insert into table_example values(f_Return_Raw(RAW_INPUT))
    that the function gives back RAW directly
    Regards
    Daniele

  • The value returned from the load function is not of type numeric  errors after migration to Coldfusion 11

    I am currently testing our website with CF11. It is currently working with CF8 however after migrating it to a new server running CF11 I have encountered the following error.
    The value returned from the load function is not of type numeric.
    The error occurred in
    D:/Applications/CFusion/CustomTags/nec/com/objects/address.cfc: line 263
    Called from D:/Applications/CFusion/CustomTags/nec/com/objects/contact.cfc: line 331
    Called from D:/Applications/CFusion/CustomTags/nec/com/objects/user.cfc: line 510
    Called from D:/Applications/CFusion/CustomTags/nec/com/objects/user.cfc: line 1675
    Called from D:/website/NECPhase2/action.validate.cfm: line 54
    261 : <cfif isNumeric(get.idCountry)>
    262 : <cfset rc = this.objCountry.setID(get.idCountry)>
    263 : <cfset rc = this.objCountry.load()>
    264 : </cfif>
    265 : <cfset this.sPostcode = get.sPostcode>
    Have there been any changes between CF8 and CF11 that could  cause this error?
    Does anyone have ideas?

    This is the code in file object file country.cfc (nec.com.objects.country):
    <cfcomponent displayname="Country object" hint="This is a Country object, it allows you to access and set values in the Country.">
    <!---
    // Construct this object
    --->
    <cfset this.objFunctions = CreateObject( 'component', 'nec.com.system.functions' )>
    <cfscript>
      this.idCountryID = 0;
      this.sCountryName = "";
      this.sISOCode = "";
      this.sDHLCode = "";
      this.iErrorID = "";
    </cfscript>
    <!---
    // The following functions are the setters and getters. offering us a better way to get
    // at the contents of the object
    --->
    <!---
    // Getters
    --->
    <cffunction name="getID" displayname="Get ID" returntype="numeric" output="false" hint="This returns the ID of the current item.">
      <cfreturn this.idCountryID>
    </cffunction>
    <cffunction name="getsCountryName" displayname="Get sCountryName" returntype="string" output="false" hint="This gets the sCountryName value of this item.">
      <cfreturn this.sCountryName>
    </cffunction>
    <cffunction name="getsISOCode" displayname="Get sISOCode" returntype="string" output="false" hint="This gets the sISOCode value of this item.">
      <cfreturn this.sISOCode>
    </cffunction>
    <cffunction name="getsDHLCode" displayname="Get sDHLCode" returntype="string" output="false" hint="This gets the sDHLCode value of this item.">
      <cfreturn this.sDHLCode>
    </cffunction>
    <cffunction name="iError" displayname="Get iError" returntype="numeric" output="false" hint="This returns the iError of the current item.">
      <cfreturn this.iError>
    </cffunction>
    <!---
    // Setters
    --->
    <cffunction name="setID" displayname="Set ID" returntype="boolean" output="false" hint="This sets the ID value of this item.">
      <cfargument name="idCountryID" required="true" type="numeric" displayname="ID" hint="The ID to use.">
      <cfset this.idCountryID = arguments.idCountryID>
      <cfreturn true>
    </cffunction>
    <cffunction name="setsCountryName" displayname="Set sCountryName" returntype="boolean" output="false" hint="This sets the sCountryName value of this item.">
      <cfargument name="sCountryName" required="true" type="string" displayname="sCountryName" hint="The sCountryName to use.">
      <cfset this.sCountryName = arguments.sCountryName>
      <cfreturn true>
    </cffunction>
    <cffunction name="setsISOCode" displayname="Set sISOCode" returntype="boolean" output="false" hint="This sets the sISOCode value of this item.">
      <cfargument name="sISOCode" required="true" type="string" displayname="sISOCode" hint="The sISOCode to use.">
      <cfset this.sISOCode = arguments.sISOCode>
      <cfreturn true>
    </cffunction>
    <cffunction name="setsDHLCode" displayname="Set sDHLCode" returntype="boolean" output="false" hint="This sets the sDHLCode value of this item.">
      <cfargument name="sDHLCode" required="true" type="string" displayname="sDHLCode" hint="The sDHLCode to use.">
      <cfset this.sDHLCode = arguments.sDHLCode>
      <cfreturn true>
    </cffunction>
    <!---
    // Clear, to empty out the contents of this object
    --->
    <cffunction name="clear" displayname="Clear items Details" returntype="boolean" output="false" hint="Clears out all of the items details.">
      <cfscript>
       this.sCountryName = "";
       this.sISOCode = "";
       this.sDHLCode = "";
       this.iErrorID = "";
      </cfscript>
      <cfreturn true>
    </cffunction>
    <!---
    // The following functions deal with the load, save and deleting of objects
    --->
    <!---
    // Load
    --->
    <cffunction name="load" displayname="Load items details" returntype="numeric" output="false" hint="This loads in all the information about an item.">
      <cfset rc = this.clear()>
      <!---
      // First of all we need to get the name of the data source we are going to be using
      --->
      <cfscript>
      objDS = CreateObject("component","nec.com.system.settings");
      sDatasource = objDS.getDatasource();
    </cfscript>
      <!---
      // Check to see if it exists
      --->
      <cftry>
       <cfquery name="checkID" datasource="#sDatasource#">
        SELECT idCountryID
        FROM tblCountry
        WHERE idCountryID = #this.idCountryID#
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("load: checkID: '#this.idCountryID#' #cfcatch.detail#");
        </cfscript>
        <cfset this.iErrorID = iErrorID>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <cfif not checkID.recordCount>
       <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         if(isDefined("session.afr")){
          whichOne = "#session.afr.getsAFRNumber()#";
         } else {
          whichOne = "";
         iErrorID = objError.addError("A Country with that id doesn't exists.[#this.idCountryID#][#whichOne#]");
        </cfscript>
       <cfset this.iErrorID = iErrorID>
       <cfreturn iErrorID>
      </cfif>
      <!---
      // If we got past all then then load in the details
      --->
      <cftry>
       <cfquery name="get" datasource="#sDatasource#">
        SELECT idCountryID, RTRIM(sCountryName) as sCountryName, RTRIM(sISOCode) as sISOCode, RTRIM(sDHLCode) as sDHLCode
        FROM tblCountry
        WHERE idCountryID = #this.idCountryID#
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("load: get: #cfcatch.detail#");
        </cfscript>
        <cfset this.iErrorID = iErrorID>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <cfset this.idCountryID = get.idCountryID>
      <cfset this.sCountryName = get.sCountryName>
      <cfset this.sISOCode = get.sISOCode>
      <cfset this.sDHLCode = get.sDHLCode>
      <cfset this.iErrorID = "">
      <cfreturn true>
    </cffunction>
    <!---
    // Save
    --->
    <cffunction name="save" displayname="Save items Details" returntype="numeric" output="false" hint="Saves (to some source) the current details for the ID of the item.">
      <!---
      // First of all we need to get the name of the data source we are going to be using
      --->
      <cfscript>
      objDS = CreateObject("component","nec.com.system.settings");
      sDatasource = objDS.getDatasource();
    </cfscript>
      <!---
      // Now check to see if ithat ID exists
      --->
      <cftry>
       <cfquery name="checkID" datasource="#sDatasource#">
        SELECT idCountryID
        FROM tblCountry
        WHERE idCountryID = #this.idCountryID#
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("save: checkID: #cfcatch.detail#");
        </cfscript>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <!---
      // If it doesn't exist, then add the record, otherwise update the record
      --->
      <cfif not checkID.recordCount>
       <cfreturn this.add()>
      <cfelse>
       <cfreturn this.update()>
      </cfif>
    </cffunction>
    <!---
    // Add
    --->
    <cffunction name="add" displayname="Add Country" returntype="numeric" output="false" hint="This adds a Country.">
      <!---
      // Check to see if that a different item isn't already using the same unique details
      --->
      <cftry>
       <cfquery name="checkUnique" datasource="#sDatasource#">
        SELECT idCountryID
        FROM tblCountry
        WHERE sCountryName = '#this.objFunctions.scrubText(this.sCountryName)#'
        OR sISOCOde = '#this.objFunctions.scrubText(this.sISOcode)#'
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("add: checkUnique: #cfcatch.detail#");
        </cfscript>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <cfif checkUnique.recordCount>
       <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("A Country with that name or ISO code already exists. idCountryID=#checkUnique.idCountryID#");
        </cfscript>
       <cfreturn iErrorID>
      </cfif>
      <cftry>
       <cfquery name="add" datasource="#sDatasource#">
        SET nocount on
        INSERT INTO tblCountry(sCountryName, sISOCode, sDHLCode)
        VALUES('#this.objFunctions.scrubText(this.sCountryName)#','#this.objFunctions.scrubText(t his.sISOCode)#','#this.objFunctions.scrubText(this.sDHLCode)#')
        SELECT @@identity as autoID
        SET nocount off  
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("add: add: #cfcatch.detail#");
        </cfscript>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <cfreturn add.autoID>
    </cffunction>
    <!---
    // Update
    --->
    <cffunction name="update" displayname="Update Country" returntype="numeric" output="false" hint="This updates a Country record.">
      <!---
      // Check to see if that a different item isn't already using the same unique details
      --->
      <cftry>
       <cfquery name="checkUnique" datasource="#sDatasource#">
        SELECT idCountryID
        FROM tblCountry
        WHERE (sCountryName = '#this.objFunctions.scrubText(this.sCountryName)#'
        OR sISOCOde = '#this.objFunctions.scrubText(this.sISOcode)#')
        AND idCountryID <> #this.idCountryID#
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("update: checkUnique: #cfcatch.detail#");
        </cfscript>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <cfif checkUnique.recordCount>
       <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("Another Country with that name already exists. idCountryID=#checkUnique.idCountryID#");
        </cfscript>
       <cfreturn iErrorID>
      </cfif>
      <!---
      // Attempt to update the record to the datasource
      // if this fails for any reason then we submit an error message
      // to the error component and return the ID of the error
      --->
      <cftry>
       <cfquery name="update" datasource="#sDatasource#">
        UPDATE tblCountry
        SET sCountryName = '#this.objFunctions.scrubText(this.sCountryName)#',
        sISOCode = '#this.objFunctions.scrubText(this.sISOCode)#',
        sDHLCode = '#this.objFunctions.scrubText(this.sDHLCode)#'
        WHERE idCountryID = #this.idCountryID#
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("update: update: #cfcatch.detail#");
        </cfscript>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <cfreturn this.idCountryID> 
    </cffunction>
    <!---
    // Delete
    --->
    <cffunction name="delete" displayname="Delete Country" returntype="numeric" output="false" hint="This deletes a Country record.">
      <!---
      // First of all we need to get the name of the data source we are going to be using
      --->
      <cfscript>
      objDS = CreateObject("component","nec.com.system.settings");
      sDatasource = objDS.getDatasource();
    </cfscript>
      <!---
      // Now check to see if ithat ID exists
      --->
      <cftry>
       <cfquery name="checkID" datasource="#sDatasource#">
        SELECT idCountryID
        FROM tblCountry
        WHERE idCountryID = #this.idCountryID#
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("delete: checkID: #cfcatch.detail#");
        </cfscript>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <cfif not checkID.recordCount>
       <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("A Country with that id doesn't exists, delete failed.");
        </cfscript>
       <cfreturn iErrorID>
      </cfif>
      <!---
      // Now check to see if there are any dependancies, if so we can't delete the item
      --->
      <cftry>
       <cfquery name="checkDependancies" datasource="#sDatasource#">
        SELECT idCountry
        FROM tblAddress
        WHERE idCountry = #this.idCountryID#
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("delete: checkDependancies: idCountry: #cfcatch.detail#");
        </cfscript>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <cfif checkDependancies.recordCount>
       <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("That Country is being used by an address, delete failed.");
        </cfscript>
       <cfreturn iErrorID>
      </cfif>
      <!---
      // Now attempt to remove the record.
      // if this fails for any reason then we submit an error message
      // to the error component and return the ID of the error
      --->
      <cftry>
       <cfquery name="delete" datasource="#sDatasource#">
        DELETE FROM tblCountry
        WHERE idCountryID = #this.idCountryID#
       </cfquery>
       <cfcatch>
        <cfscript>
         objError = CreateObject("component","nec.com.system.errors");
         iErrorID = objError.addError("delete: delete: #cfcatch.detail#");
        </cfscript>
        <cfreturn iErrorID>
       </cfcatch>
      </cftry>
      <cfreturn this.idCountryID>
    </cffunction>
    </cfcomponent>

  • Can I use the value returned from a Text Function in another Formula?

    I'm writing a report in Hyperion System 9 BI + Financial Reporting Studio version 9.2. I have 2 grids in my report.
    Grid1 Column A is set up as a text function using the function type - <<GetCell("Grid2", 1, a, 1)>>. I would like to use the values returned from this text function in Column A (Grid 1) in a formula in Column B (Grid 1).
    Is it possible to use the values returned in Column A of the text function in another formula? My report does not seem to recognize Column A as numerical values, even though the values to be returned are numerical.
    If so, how do I recognize the values in Column A Grid 1 as numerical values and not text?
    Thanks for any help you can offer!

    Hi Edson,
    Yes you need to use the CALC_ERROR macro function to be able to test whether the last macro function returned an error. CALC_ERROR will return an 'X' if there an error occured during the execution of the last macro function.
    You can use a macro similar to the following:
    IF
      CALC_ERROR( )
      = 'X'
          DO SOMETHING HERE
    ENDIF
    Let me explain how this works internally. The SAP system maintains a global variable g_flg_calc_error during the execution of macros in the planning book. The g_flg_calc_error variable will contain the value of f_calc_error that was set by the last macro function which executed. The ABAP coding of a planning book is something like this:
    data: g_flg_calc_error type /SAPAPO/FLAG.
    * SAP will pass g_flg_calc_error variable to all macro
    * functions. When SAP calls a macro function, it does
    * something like this.
    call function '/SAPAPO/MACRO_FUNCTION_HERE'
            exporting
              plob_values      = i_s_adv_plob_values
              sdp_book         = g_c_advf_sdp_book
              sdp_view         = g_c_advf_sdp_view
            tables
              cols_index       = i_t_cols
              value_tab        = l_t_value_tab
            changing
              f_calc_error     = g_flg_calc_error
    As you can see, the g_flg_calc_error variable
    is passed in the "changing" part of the call. The macro  function being called can then use the f_calc_error
    variable to change the value of the global
    g_flg_calc_error variable. In fact, the macro function being called can also check (by looking at the f_calc_error variable) if the last macro function reported an error.  The CALC_ERROR macro function just checks the value of f_calc_error parameter (w/c in fact is the value of the g_flg_calc_error variable) and returns "true/X" if the f_calc_error was set to true by the last macro function.
    Hope this helps in clearing things out

  • How to return a table from function

    Hi,
    I want to return a table from the function. But i'm unable to do it.
    can any one give me a peice of code.
    Thanks in advance.
    Regards,
    Alok

    Are you referring to a pipelined table function?
    Here is an example of such a function returning a "hailstone" sequence.. a number range that always end in 4,2,1. (see the [url http://en.wikipedia.org/wiki/Collatz_conjecture]Collatz Conjecture for more details on the maths used).
    SQL> create or replace type TNumbers is table of number;
    2 /
    Type created.
    SQL>
    SQL> create or replace function hailstone( nValue number ) return TNumbers pipelined is
    2 n number;
    3 begin
    4 n := nValue;
    5 loop
    6 PIPE ROW( n );
    7 exit when n = 1;
    8
    9 if MOD(n,2) != 0 then
    10 n := (n * 3) + 1;
    11 else
    12 n := n / 2;
    13
    14 end if;
    15 end loop;
    16 end;
    17 /
    Function created.
    SQL> show errors
    No errors.
    SQL>
    SQL>
    SQL> select * from TABLE( hailstone(10) );
    COLUMN_VALUE
    10
    5
    16
    8
    4
    2
    1
    7 rows selected.
    SQL> select * from TABLE( hailstone(1923413) );
    ..etc.

  • Transaction Launcher : How to get return parameters from an external URL

    Hello,
    We need to launch an external URL in the Web Client UI. After the URL is launched, the user does some processing which results in some output values. These value have to be passed back to the Web Client UI and become a part of the quotation he is creating.
    We are able to launch the URL using the Transaction Launcher. But we are not able to find a way to get the output values from the external URL to the Web client UI.
    Please suggest any approach for this. It is OK even if we need to use an intermediate BSP to launch the external URL.
    Regards
    Ajay

    hi ajay
    I had a similar requirement , when the user clicks on button I want to launch the  External URL , can you please let me know how to launch the external url when the user cliks on button
    Thanks & Regards
    madhu

  • How to call external functions without a .DLL (just using a .H and .LIB file)?

    Hello everyone,
    actually I'm facing little difficulties on how to get an external function getting called from within CVI (Version 2009).
    I was supplied with a .H file and a .LIB file to call an external function from within my CVI project. The .H file looks like this:
    void exportedFunction(double *parameter);
    As far as I know, the external function was written with MS Visual C++ 6.
    So I tried to statically link to the extern al function like this:
    - Add the .H file and the .LIB file to the CVI project
    - #include the .H file where I needed to call the external function
    - do the external function call
    When building I get an unresolved external function call error from CVI so this seems not to be working.
    I made some searches around and got up with two possible issues. Maybe one of you can help me get a bit further and get things working.
    1) The "real" function code is located in the DLL file which was not delivered to me. Or is there a way to get things done (call external functions) just with a .H and a .LIB file (without any .DLL file included)?
    2) The external function does not export according to "C-Style" rules. The function signature in the .H file shows no things like
    extern "C" __declspec(dllexport) void __stdcall ...
     Or maybe it's a combination of both issues (missing .DLL + wrong export style of function)?
    I guess I could get around the wrong export style of the function when I manage to write a wrapper around the original function that actually uses C-Style exporting. But I guess I need the .DLL file for this try as well.
    Thank you for your answers.
    Best regards,
    Bernd
    Solved!
    Go to Solution.

    There is no need  for the dllexport stuff. There is also the option for a static library without any DLL.  But the 'extern "C"' is essential, because it forces the C++  compiler, which was probably used to compile the library , to use C calling convention.
    If you can't ask the provider of the library to provide a version which was compiled using C calling convention the way to go is to write a wrapper with VC++6 around that library which reexports the functions using C calling convertion. Something like
    extern "C" type0 myfunc1(type1 arg1, ...) {
           return func1( arg1,...);
    for every function , you need to use.
    BTW. "unresolved symbol" is the linker error message, you can expect if you try to link C code against a library build with C++ calling convention.

  • User Defined External Function - How to work?

    I have defined some external functions in java by following the description in
    C:\OraBPELPM_3\integration\orabpel\samples\demos\XSLMapper\ExtensionFunctions
    I can use my functions in JDeveloper, but they do not work on the Oracle BPEL engine.
    Is it enough just to place the jar-file in <OC4J_HOME>\j2ee\home\applib and add the jar-file in class-path (as described)?
    What about the xml-file specifying the functions (called SampleExtentionFunctions.xml in the documentation)?
    Should this file be copied to <OC4J_HOME>\... ?
    Here is the error message I get by running my service:
    XPath expression failed to execute.
    Error while processing xpath expression, the expression is "ora:processXSLT("TransformationInput.xsl", bpws:getVariableData("inputVariable", "request"))", the reason is java.lang.NoSuchMethodException: For extension function, could not find method org.apache.xpath.objects.XNodeSet.leadingZeros([ExpressionContext,] #NUMBER)..
    Please verify the xpath query.
    ______________ MY JAVA CODE EXAMPLE: _____________________________
    package extentionfunctions;
    This is a sample XSL Mapper User Defined Extension Functions implementation class.
    public class JavaExtensionFunctionsBpel
    * Inserts leading zeros to a text, if this starts with a digit.
    * Else the return value will be the same as the given text.
    * The return value will have the specified length.
    public static String leadingZeros(String text, int len)
    {  String retur = text;
    char c = (text == ""?'0':text.charAt(0));
    if ('0'<=c && c<='0'+9) { // Is first char a digit?
    retur = "";
    int n = len - (text == ""?0:text.length());
    for (int i=0; i<n; i++) { // Insert zeros:
    retur = '0' + retur;
    retur = retur + text;
    return retur;
    * Removes leading zeros from a text.
    public static String removeLeadingZeros(String text)
    {  String retur = text;
    int pos = 0;
    int len = (text == ""?0:text.length());
    for (int i=0; i<len; i++) {
    if (text.charAt(i)=='0') { // Is char a digit?
    pos++;
    return retur;
    public static void main(String[] args)
    { // Basic test of functions:
    int len = 5;
    String s = "1234";
    String r;
    System.out.println("leadingZeros("+s+","+len+")="+(r=leadingZeros(s,len)));
    System.out.println("removeLeadingZeros("+r+")="+removeLeadingZeros(s));
    Regards
    Flemming Als

    Flemming, it looks like somthing is wrong in the xsl that it still goes to org.apache.xpath.objects.XNodeSet package instead of yours ..
    I created a sample that illustrates the usage of this functions (even with 2 params, different types).
    Java class (com.otn.samples.xslt.CustomXSLTFunctions)
    package com.otn.samples.xslt;
    public class CustomXSLTFunctions
    <function name="extension:multiplyStringAndInt" as="number">
    <param name="base" as="String"/>
    <param name="multiplier" as="number"/>
    </function>
    public int multiplyStringAndInt (String pString, int pMultiplier)
    int base = Integer.parseInt(pString);
    return base * pMultiplier;
    XML descriptor:
    Note the types (in the java class and the xml descriptor)
    for java:base -> string and for xslt:base -> string
    for java:multiplier -> int and for xslt:multiplier -> number
    <?xml version="1.0" encoding="UTF-8"?>
    <extension-functions>
    <functions extension:multiplyStringAndInt="http://www.oracle.com/XSL/Transform/java/com.otn.samples.xslt.CustomXSLTFunctions">
    <function name="extension:multiplyStringAndInt" as="number">
    <param name="base" as="string"/>
    <param name="multiplier" as="number"/>
    </function>
    </functions>
    </extension-functions>
    Definition of variables in the process:
    (as you can see, the base is a string - I do the conversion in my method, and the return value
    is converted to a string by the engine)
    <!-- Input -->
    <element name="CustomXSLTFunctionProcessRequest">
    <complexType>
    <sequence>
    <element name="base" type="string"/>
    <element name="multiplier" type="int"/>
    </sequence>
    </complexType>
    </element>
    <!-- Output -->
    <element name="CustomXSLTFunctionProcessResponse">
    <complexType>
    <sequence>
    <element name="result" type="string"/>
    </sequence>
    </complexType>
    </element>
    Using the new function in the XSL transformation:
    - watch out for the namespace declaration (xmlns:sample="...")
    - and the usage (sample:multiplyStringAndInt)
    <xsl:stylesheet version="1.0"
    xmlns:ldap="http://schemas.oracle.com/xpath/extension/ldap"
    xmlns:xp20="http://www.oracle.com/XSL/Transform/java/oracle.tip.pc.services.functions.Xpath20"
    xmlns:bpws="http://schemas.xmlsoap.org/ws/2003/03/business-process/"
    xmlns:plnk="http://schemas.xmlsoap.org/ws/2003/05/partner-link/"
    xmlns:extension="http://www.oracle.com/XSL/Transform/java/com.otn.samples.xslt.CustomXSLTFunctions"
    xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
    xmlns:ora="http://schemas.oracle.com/xpath/extension"
    xmlns:ns0="http://www.w3.org/2001/XMLSchema"
    xmlns:orcl="http://www.oracle.com/XSL/Transform/java/oracle.tip.pc.services.functions.ExtFunc"
    xmlns:client="http://xmlns.oracle.com/CustomXSLTFunction"
    exclude-result-prefixes="xsl plnk ns0 client ldap xp20 bpws extension ora orcl">
    <xsl:template match="/">
    <client:CustomXSLTFunctionProcessResponse>
    <client:result>
    <xsl:value-of select="extension:multiplyStringAndInt(/client:CustomXSLTFunctionProcessRequest/client:base,/client:CustomXSLTFunctionProcessRequest/client:multiplier)"/>
    </client:result>
    </client:CustomXSLTFunctionProcessResponse>
    </xsl:template>
    </xsl:stylesheet>
    Then I created a jar file with the class (and for testing the xml descriptor in it)...
    and put it into j2ee/home/applib directory
    hth clemens

  • External function specification

    Hi All!
    (Envs: Centos 5.2, Oracle 10.2.0.4)
    According to documentation a developer should provide RETURN clause in the external function parameters declaration.
    But I faced a conflict with additional extproc specific parameters , INDICATOR for example...
    Option 1 (no INDICATOR variable in specification)
    CREATE OR REPLACE FUNCTION MyTest(Lang IN VARCHAR2, Form IN VARCHAR2) RETURN VARCHAR2
    AS EXTERNAL
    LIBRARY Lemma_Ext
    NAME "MyExtTest"
    LANGUAGE "C"
    WITH CONTEXT
    PARAMETERS(CONTEXT,Lang string, Form string, RETURN string)
    and it works OK with it's "C" prototype
    +char * MyExtTest (char Lang, char Form);+
    Option 2 (with INDICATOR, doesn't work):
    CREATE OR REPLACE FUNCTION MyTest(Lang IN VARCHAR2, Form IN VARCHAR2) RETURN VARCHAR2
    AS EXTERNAL
    LIBRARY Lemma_Ext
    NAME "MyExtTest"
    LANGUAGE "C"
    WITH CONTEXT
    PARAMETERS(CONTEXT, Lang string, Form string, RETURN INDICATOR short, RETURN string)
    and it is not works with following "C" prototype:
    +char * MyExtTest (char Lang, char Form, short *return_ind);+
    Option 3 (with INDICATOR, works!):
    CREATE OR REPLACE FUNCTION MyTest(Lang IN VARCHAR2, Form IN VARCHAR2) RETURN VARCHAR2
    AS EXTERNAL
    LIBRARY Lemma_Ext
    NAME "MyExtTest"
    LANGUAGE "C"
    WITH CONTEXT
    PARAMETERS(CONTEXT, Lang string, Form string, RETURN INDICATOR short)
    and it works like a charm with the same "C" prototype as in the Option 2:
    +char * MyExtTest (char Lang, char Form, short *return_ind);+
    Just removing "RETURN string" was a correct solution!
    And I found a couple of examples of successful external functions (with extra parameters) from Metalink. All these examples don't contain RETURN <type> in function specification.
    So, where is the source of misunderstanding?
    Regards,
    Andrew.

    Hi AndresG,
    Are you using O2A cartridges or not?
    If you are not, by default, the fulfillment states are in the following location --
    a) external fulfillment state -- /ControlData/Functions/<Function>/orderItem/ExternalFulfillmentState
    b) composite fulfillment state (i.e. output of FS composition) -- /ControlData/OrderItem/OrderItemFulfillmentState for order items, and /ControlData/OrderFulfillmentState for overall order.
    Note that mapped fulfillment states are not persisted.
    These defaults are configurable by creating FulfillmentStateModule.xquery file, and have XML catalog include the file. In fact, this is how O2A cartridges override these defaults; the fulfillment states are stored in /OrderLifeCycleManagement area.
    Yes, you can update the external fulfillment state during function execution -- in your automation of the function, simply return the external fulfillment state in your xquery automator.
    To send composite fulfillment state to CRM, yes, you need to create a Data Changed Event Notification -- set it to trigger by Composite Fulfillment State change.
    Cheers,
    Daniel Ho
    OSM Product Management

  • [C++] External functions

    Hi, I'm new of this forum...hallo world!
    I'm trying to include function from external files to manage to develop clearer project. At the moment I compile with g++ from the terminal, could it be a problem? I tried to use Eclipse but I'm not able to understand how to create a project and how to compile it.
    I use three files:
    - a header, called simply "header";
    - a .cpp containing the main, called Programma.cpp;
    - a .cpp containind the function called Funzione.cpp.
    "header":
    #include <iostream>
    using namespace std;
    int somma(int,int);
    "Programma.cpp":
    #include "header"
    int main()
        int a,b,c;
        a=3;
        b=1;
        c=somma(a,b);
        cout << a << b << c << endl;
        return 0;
    "Funzione.cpp":
    #include "header"
    int somma(int a, int b)
        int risultato;
        risultato=a+b;
        return risultato;
    When I compile using:
    "g++ -o Programma Programma.cpp"
    it give me the following error:
    "Undefined symbols:
      "somma(int, int)", referenced from:
          _main in ccBWPWka.o
    ld: symbol(s) not found
    collect2: ld returned 1 exit status"
    What is the problem?
    Thank you in advance

    Perfect, thank you very much.
    I have two other question:
    1) Ones I compiled and linked the main program (in this case Programma.cpp) with all my function, and I have to modify something only in the main, can I compile only the main (and if yes, how can I do it)?
    For instance...now this is my project(?) *I'm not sure it's the correct name*:
    "header.h":
    int somma(int,int);
    "Programma.cpp":
    #include <iostream>
    #include "header.h"
    using namespace std;
    int main()
        int a,b,c;
        a=3;
        b=1;
        c=somma(a,b);
        cout << a << endl;
        cout << b << endl;
        cout << c << endl;
        return 0;
    "Funzione.cpp":
    #include <iostream>
    #include "header.h"
    using namespace std;
    int somma(int a, int b)
        int risultato;
        risultato=a+b;
        return risultato;
    And it works perfectly.
    If now I want to change the value of b from b=1 to b=24 in the main, can I only compile the main? I ask that because usually a project has a lot of "standard" external function that are fixed and are never modified, so it will be complicated if I have to follow the process you suggest me every time I change the main:
    g++ -o Programma Programma.cpp Funzione1.cpp Funzione2.cpp ... FunzioneN.cpp
    2) Second question, it seems to be equal if I call the header "header.h" or just "header", what is the difference?
    I didn't use the extension .h because also the iostream is without .h in my programs, and if I try to include <iostream.h>, I have the following error:
    /usr/include/c++/4.2.1/backward/backward_warning.h:32:2: warning: #warning This file includes at least one deprecated or antiquated header. Please consider using one of the 32 headers found in section 17.4.1.2 of the C++ standard. Examples include substituting the <X> header for the <X.h> header for C++ includes, or <iostream> instead of the deprecated header <iostream.h>. To disable this warning use -Wno-deprecated.
    Thanks again

  • Creating external function

    We are trying to create an external function that will be linked at run- time. How is such a function compiled?
    Does it need to be compiled into a Dynamic library? Or would the function be called?
    The operating system we are trying to do this is Sun UNIX Solaris. However any insight would be of great
    assistance.

    I have a nasty feeling there is in XQuery expressions, but I'm not going there. ;-)
    michaels>  WITH t AS
         (SELECT 1 ID, 1 x, 2 y FROM DUAL UNION ALL
          SELECT 2,    3,   4   FROM DUAL UNION ALL
          SELECT 3,    5,   6   FROM DUAL)
    SELECT   *
        FROM XMLTable('declare function local:a($a,$b)
                         ($a + $b)
                       }; (: eof :)
                       for $i in /ROWSET/ROW
                       return <ROW>
                              <ID>{$i/ID}</ID>
                              <X>{$i/X}</X>
                              <Y>{$i/Y}</Y>
                              <A>{local:a($i/X,$i/Y)}</A>
                              <B>{local:a($i/X,$i/Y) + 2}</B>
                              </ROW>' PASSING XMLTYPE(CURSOR(SELECT * FROM t))
                       COLUMNS ID NUMBER PATH 'ID',
                                X NUMBER PATH 'X',
                                Y NUMBER PATH 'Y',
                                A NUMBER PATH 'A',
                                B NUMBER PATH 'B'
            ID          X          Y          A          B
             1          1          2          3          5
             2          3          4          7          9
             3          5          6         11         13

  • [UIX] How To: Return multiple values from a LOV

    Hi gang
    I've been receiving a number of queries via email on how to return multiple items from a LOV using UIX thanks to earlier posts of mine on OTN. I'm unfortunately aware my previous posts on this are not that clear thanks to the nature of the forums Q&A type approach. So I thought I'd write one clear post, and then direct any queries to it from now on to save me time.
    Following is my solution to this problem. Please note it's just one method of many in skinning a cat. It's my understanding via chatting to Oracle employees that LOVs are to be changed in a future release of JDeveloper to be more like Oracle Forms LOVs, so my skinning skills may be rather bloody & crude very soon (already?).
    I'll base my example on the hr schema supplied with the standard RDBMS install.
    Say we have an UIX input-form screen to modify an employees record. The employees record has a department_id field and a fk to the departments table. Our requirement is to build a LOV for the department_id field such that we can link the employees record to any department_id in the database. In turn we want the department_name shown on the employees input form, so this must be returned via the LOV too.
    To meet this requirement follow these steps:
    1) In your ADF BC model project, create 2 EOs for employees and departments.
    2) Also in your model, create 2 VOs for the same EOs.
    3) Open your employees VO and create a new attribute DepartmentName. Check “selected in query”. In expressions type “(SELECT dept.department_name FROM departments dept WHERE dept.department_id = employees.department_id)”. Check Updateable “always”.
    4) Create a new empty UIX page in your ViewController project called editEmployees.uix.
    5) From the data control palette, drag and drop EmployeesView1 as an input-form. Notice that the new field DepartmentName is also included in the input-form.
    6) As the DepartmentName will be populated either from querying existing employees records, or via the LOV, disable the field as the user should not have the ability to edit it.
    7) Select the DepartmentId field and delete it. In the UI Model window delete the DepartmentId binding.
    8) From the data controls palette, drag and drop the DepartmentId field as a messageLovInput onto your page. Note in your application navigator a new UIX page lovWindow0.uix (or similar) has been created for you.
    9) While the lovWindow0.uix is still in italics (before you save it), rename the file to departmentsLov.uix.
    10) Back in your editEmployees.uix page, your messageLovInput source will look like the following:
    <messageLovInput
        model="${bindings.DepartmentId}"
        id="${bindings.DepartmentId.path}"
        destination="lovWindow0.uix"/>Change it to be:
    <messageLovInput
        model="${bindings.DepartmentId}"
        id="DepartmentId"
        destination="departmentsLov.uix"
        partialRenderMode="multiple"
        partialTargets="_uixState DepartmentName"/>11) Also change your DepartmentName source to look like the following:
    <messageTextInput
        id=”DepartmentName”
        model="${bindings.DepartmentName}"
        columns="10"
        disabled="true"/>12) Open your departmentsLov.uix page.
    13) In the data control palette, drag and drop the DepartmentId field of the DepartmentView1 as a LovTable into the Results area on your page.
    14) Notice in the UI Model window that the 3 binding controls have been created for you, an iterator, a range and a binding for DepartmentId.
    15) Right click on the DepartmentsLovUIModel node in the UI Model window, then create binding, display, and finally attribute. The attribute binding editor will pop up. In the select-an-iterator drop down select the DepartmentsView1Iterator. Now select DepartmentName in the attribute list and then the ok button.
    16) Note in the UI Model you now have a new binding called DCDefaultControl. Select this, and in the property palette change the Id to DepartmentName.
    17) View the LOV page’s source, and change the lovUpdate event as follows:
    <event name="lovSelect">
        <compound>
            <set value="${bindings.DepartmentId.inputValue}" target="${sessionScope}" property="MyAppDepartmentId" />
            <set value="${bindings.DepartmentName.inputValue}" target="${sessionScope}" property="MyAppDepartmentName" />
        </compound>
    </event>18) Return to editEmployees.uix source, and modify the lovUpdate event to look as follows:
    <event name="lovUpdate">
        <compound>
            <set value="${sessionScope.MyAppDepartmentId}" target="${bindings.DepartmentId}" property="inputValue"/>
            <set value="${sessionScope.MyAppDepartmentName}" target="${bindings.DepartmentName}" property="inputValue"/>     
        </compound>
    </event>That’s it. Now when you select a value in your LOV, it will return 2 (multiple!) values.
    A couple things to note:
    1) In the messageLovInput id field we don’t use the “.path” notation. This is mechanism for returning 1 value from the LOV and is useless for us.
    2) Again in the messageLovInput we supply “_uixState” as an entry in the partialTargets.
    3) We are relying on partial-page-refresh functionality to update multiple items on the screen.
    I’m not going to take the time out to explain these 3 points, but it’s worthwhile you learning more about them, especially the last 2, as a separate exercise.
    One other useful thing to do is, in your messageLovInput, include as a last entry in the partialTargets list “MessageBox”. In turn locate the messageBox control on your page (if any), and supply an id=”MessageBox”. This will allow the LOV to place any errors raised in the MessageBox and show them to the user.
    I hope this works for you :)
    Cheers,
    CM.

    Thanks Chris,
    It took me some time to find the information I needed, how to use return multiple values from a LOV popup window, then I found your post and all problems were solved. Its working perfectly, well, almost perfectly.
    Im always fighting with ADF-UIX, it never does the thing that I expect it to do, I guess its because I have a hard time letting go of the total control you have as a developer and let the framework take care of a few things.
    Anyway, I'm using your example to fill 5 fields at once, one of the fields being a messageChoice (a list with countries) with a LOV to a lookup table (id , country).
    I return the countryId from the popup LOV window, that works great, but it doesn't set the correct value in my messageChoice . I think its because its using the CountryId for the listbox index.
    So how can I select the correct value inside my messageChoice? Come to think of it, I dont realy think its LOV related...
    Can someone help me out out here?
    Kind regards
    Ido

  • Returning a table from a stored procedure

    hi, i need to return a table from a stored procedure and show it, and come to this, but a don't know hoy to run it, so i don't know if it is right, can anyone help me?
    uTable out objects_uptime%rowtype
    as
    begin
    select * into uTable from objects_uptime;
    end;

    well, i finally discovered how to do the trick
    this is the code for the function:
    CREATE OR REPLACE FUNCTION FN_GET_RECORDS RETURN UPTIME PIPELINED IS
    CURSOR cUptime is select * from objects_uptime;
    p refcur.refcur_t;
    temp p%ROWTYPE;
    temp2 OUPTIME := OUPTIME(null, null, null, null, null, null, null, null, null);
    BEGIN
    OPEN cUptime;
    LOOP
    FETCH cUptime into temp;
    temp2.OBJ_NAME := temp.OBJ_NAME;
    temp2.IP := temp.IP;
    temp2.STATUS := temp.STATUS;
    temp2.DOE := temp.DOE;
    temp2.ENABLED := temp.ENABLED;
    temp2.COMMENT00000 := temp.COMMENT00000;
    temp2.USERID := temp.USERID;
    temp2.OBJECTID := temp.OBJECTID;
    temp2.MNTID := temp.MNTID;
    pipe row(temp2);
    EXIT WHEN cUptime%NOTFOUND;
    END LOOP;
    RETURN;
    END FN_GET_RECORDS;
    and this for the auxiliar package, object and table:
    CREATE OR REPLACE PACKAGE REFCUR
    as
    TYPE refcur_t IS REF CURSOR RETURN objects_uptime%ROWTYPE;
    end REFCUR;
    CREATE OR REPLACE TYPE OUPTIME AS OBJECT ( "OBJ_NAME"
    VARCHAR2(255), "IP" VARCHAR2(20), "STATUS" VARCHAR2(10),
    "DOE" DATE, "ENABLED" NUMBER(10, 1), "COMMENT00000"
    VARCHAR2(1000), "USERID" VARCHAR2(50), "OBJECTID" NUMBER(10,
    1), "MNTID" NUMBER(10, 1) )
    CREATE TYPE TUPTIME AS
    TABLE OF OUPTIME
    i call it this way:
    select * from table(FN_GET_RECORDS ())
    if anyone knows how to do the same in a shorter manner, tell me please.
    thanks to everybody for the help.

Maybe you are looking for

  • Can't Drag Songs to Change Order??

    All the playlists I've ever created on my iPod I can just drag songs up and down to change the order they play in. I just created a couple of new playlists, and they won't let me change the order of the songs. I can't drag them up and down. I tried r

  • EPMA dimension alias

    Hello, I am working on EPM 11.1.2.3, and notice that I encountered an issue with EPMA Planning application. When I create the dimension (for example, Product dimension), in the system property, there is a dimension alias field. However, when I put an

  • Screen shot (Adobe Pro 9.0.0)

    1st issue: after using camera symbol to create screen shot, I cannot paste it into the same or other PDF document (e.g. for comparison due overlapping). In the previous Acrobat version Pro 8.1.2 it was possible to do this. 2nd issue: if I export this

  • Redistribution in MP-BGP

    Hello, I have a question regarding the redistribution process in MPLS network on the SAME PE router. I have PE-CE RIPv2, and in the MPLS is MP-BGPv4. Between MPLS Core and LAN Core is OSPF. I redistribute RIP to BGP on PE, so I have CE routes in the

  • ITunes Keeps Crashing with Mavericks

    Ever since updating my iMac from Snow Leopard to Mavericks, I'm finding iTunes crashes after several hours of running. I often leave iTunes running over night and then sync my podcasts up before heading to work. This has never been a problem, going b