FpML Issues Tracker

619: cd-5 has an incomplete scenario and unnecessary condition

March 1, 2008

closed

Minor

Always

Validation Rules

Admin

lyteck

Summary

Problem 1: Validation rule cd-5 does not cater for the scenario where either or both of //element(*, GeneralTerms)/(effectiveDate|scheduledTerminationDate) do not exist.

Problem 2: The existence of //element(*, GeneralTerms)/scheduledTerminationDate is sufficient for the existence of //element(*, GeneralTerms)/scheduledTerminationDate/adjustableDate. Or informally: if you have a scheduledTermination then you always have a adjustableDate because it makes adjustableDate's existence mandatory. You don't need to test for the prescence of adjustableDate, just scheduledTerminationDate is enough.

The current text of the rule is: " Context: GeneralTerms (complex type) cd-5 (Mandatory) If element scheduledTerminationDate/adjustableDate exists, then effectiveDate/unadjustedDate < scheduledTerminationDate/adjustableDate/unadjustedDate. " I propose it is changed to: " Context: GeneralTerms (complex type) cd-5 (Mandatory) The existence of scheduledTerminationDate is sufficient for the existence of effectiveDate, and effectiveDate/unadjustedDate less than scheduledTerminationDate/adjustableDate/unadjustedDate " btw - we agreed to use XPath paths and operators. "lt" (less than) has slightly different semantics to "<", but for this case they are equivalent. The corrected text more closely fits the business meaning, which is why it will be more tolerant of the MTF's refactoring of this to AdjustableDate2 in FpML 5.0.

Notes:

  • andrew

    04/22/08 1:56 pm

    Or

    Context: GeneralTerms (complex type)
    cd-5 (Mandatory)
    If effectiveDate and scheduledTerminationDate both exist then effectiveDate/unadjustedDate < scheduledTerminationDate/adjustableDate/unadjustedDate.

  • mgratacos

    04/22/08 1:58 pm

    Validation WG April 22, 2008: agreed to implement Andrew’s suggested implementation.

  • matthewdr

    04/29/08 2:38 pm

    Please implement.

  • lyteck

    04/29/08 4:04 pm

    fixed as suggested in notes.

  • matthewdr

    04/30/08 4:01 pm

    I just checked the rule in https://dedicated.fpml.org/svn/fpml/trunk/src/validation/rules-english-cd.xml

    The file has not been changed.

    Lyteck, please change the file as VWG agreed, and commit it so SVN.

  • lyteck

    05/08/08 1:28 pm

    I fixed the rule using Andrew’s Suggestion at revision 3731 of the file (4/29/2008). Please advise.

  • matthewdr

    05/08/08 2:13 pm

    The issue is fixed.

    Perhaps it is the incorrect last update value again that the proxies act on?

  • Leave an update

    You must be logged in to post an update.