Skip to main content
Question

Need an Idescriptor function that can search for multiple substrings and return a boolean list

  • September 10, 2024
  • 4 replies
  • 3 views
Shawn Waldie
Forum|alt.badge.img

Here is a record sample:

The TESTING column is a multi-valued list, and the pipes represent actual SVMs for readability.

I'm looking for a function that will return an associated boolean list indicating whether or not either '1400' or '9101' exists in TESTING, so...

In this sample, if one is present so is the other, but that's not always the case. So a 1 should be returned if either exists in what is actually a subvalued list.

EDIT 1: A subroutine would make this very easy, I know, but I don't have time for the administrative overhead that comes with developing/delivering one to PROD, so I need this to be "idescriptor" language.

Also, there will be subsequent processing, based on a SUM() of the Y/N list for each record, which will be compared to DCOUNT(TESTING,@VM).

EDIT 2: I think I found a solution. I didn't realize I could use the INDEX function on lists. Though it will require another step or 2 to get the final result, I think it will work.

    4 replies

    Manu Fernandes
    Forum|alt.badge.img+2
    • Manu Fernandes
    • Inspiring
    • 257 replies
    • September 10, 2024

    Here is a record sample:

    The TESTING column is a multi-valued list, and the pipes represent actual SVMs for readability.

    I'm looking for a function that will return an associated boolean list indicating whether or not either '1400' or '9101' exists in TESTING, so...

    In this sample, if one is present so is the other, but that's not always the case. So a 1 should be returned if either exists in what is actually a subvalued list.

    EDIT 1: A subroutine would make this very easy, I know, but I don't have time for the administrative overhead that comes with developing/delivering one to PROD, so I need this to be "idescriptor" language.

    Also, there will be subsequent processing, based on a SUM() of the Y/N list for each record, which will be compared to DCOUNT(TESTING,@VM).

    EDIT 2: I think I found a solution. I didn't realize I could use the INDEX function on lists. Though it will require another step or 2 to get the final result, I think it will work.

    hi

    SUBROUTINE TEST(RESULT, VALS)
* RESULT MV'ED 1/0 if sv contains VALS comma separated
CONVERT ',' TO @AM IN VALS 
TESTINGFIELD = 1 ;* fixme 
TESTING = @RECORD<TESTINGFIELD>
NBMV = DCOUNT( TESTING,@VM) ;* define the qty of mv value 
NBV = DCOUNT(VALS,#AM) ;* define qty of VALS 
RESULT = '' 
FOR I = 1 TO NBMV ;* loop on all mv 
    RES = ''
    FOR II = 1 TO NBV ;* loop all vals on this mv 
        LOCATE(VALS<II>, TESTING, 1, NBMV ; RG ) THEN 
            RES += 1 ;* ok this is found 
        END ELSE 
            EXIT ;* get out the current loop, at leat one value is not found 
        END 
    NEXT II 
    RESULT<NBMV> = (RES = NBV)  ;* if all values are found ... set result to 1 else 0 
NEXT NBMV 
RETURN 
*---------------------------------

    for the itype, defined it MV'ed and on the same ASSOCNAME than TESTING FIELD 

    call method 

    SUBR("TEST", "1400,1901") 
    if you want set 1400,1901 as part of the query like LIST FILE WHEN TEST = "1400,1901" 
     
    you get the statement from @SENTENCE 
    retrieve the condition on TEST field into VALS 
    you set the RESULT as a string (concat all VALS comma separate) "1400,1901" instead 1/0  
     
    I hope this help
    manu
    Manu Fernandes
    Forum|alt.badge.img+2
    • Manu Fernandes
    • Inspiring
    • 257 replies
    • September 10, 2024

    Here is a record sample:

    The TESTING column is a multi-valued list, and the pipes represent actual SVMs for readability.

    I'm looking for a function that will return an associated boolean list indicating whether or not either '1400' or '9101' exists in TESTING, so...

    In this sample, if one is present so is the other, but that's not always the case. So a 1 should be returned if either exists in what is actually a subvalued list.

    EDIT 1: A subroutine would make this very easy, I know, but I don't have time for the administrative overhead that comes with developing/delivering one to PROD, so I need this to be "idescriptor" language.

    Also, there will be subsequent processing, based on a SUM() of the Y/N list for each record, which will be compared to DCOUNT(TESTING,@VM).

    EDIT 2: I think I found a solution. I didn't realize I could use the INDEX function on lists. Though it will require another step or 2 to get the final result, I think it will work.

    the version where testing VALS are on the statement,

    SUBROUTINE TEST(RESULT)

COMMON VALS 
IF UNASSIGNED(VALS) THEN 
    RES = TRIM(@SENTENCE)
    I = INDEX('WHEN TEST')
    VALS = FIELD(RES[I,999],' ',4) 
    CONVERT '"' TO '' IN VALS 
    CONVERT "'" TO '' IN VALS 
    CONVERT ',' TO @AM IN VALS 
END 

TESTINGFIELD = 1 
TESTING = @RECORD<TESTINGFIELD>

NBMV = DCOUNT( TESTING,@VM)
NBV = DCOUNT(VALS,#AM) 
RESULT = '' 
FOR I = 1 TO NBMV ;* LOOP ALL MV
    RES = ''
    FOR II = 1 TO NBV ;* LOOP ALL VALS 
        LOCATE(VALS<II>, TESTING, 1, NBMV ; RG ) THEN 
            RES<-1> = VALS<II>
        END ELSE 
            EXIT ;* OUT OF CURRENT FORLOOP THIS VALUE IS NOT FOUND ... 
        END 
    NEX TII 
    RESULT<NBMV> = CONVERT(@AM,',',RES)
NEXT NBMV 
RETURN 
*---------------------------------

ITYPE defined as MV on the same ASSOCNAME than TESTING FIELD 
ITYPE TESTING
SUBR("TEST") 

*-----------
LIST FILE WHEN TEST = "1400,1901"
    Tyrel Marak
    Forum|alt.badge.img
    • Tyrel Marak
    • Participating Frequently
    • 56 replies
    • September 11, 2024

    Here is a record sample:

    The TESTING column is a multi-valued list, and the pipes represent actual SVMs for readability.

    I'm looking for a function that will return an associated boolean list indicating whether or not either '1400' or '9101' exists in TESTING, so...

    In this sample, if one is present so is the other, but that's not always the case. So a 1 should be returned if either exists in what is actually a subvalued list.

    EDIT 1: A subroutine would make this very easy, I know, but I don't have time for the administrative overhead that comes with developing/delivering one to PROD, so I need this to be "idescriptor" language.

    Also, there will be subsequent processing, based on a SUM() of the Y/N list for each record, which will be compared to DCOUNT(TESTING,@VM).

    EDIT 2: I think I found a solution. I didn't realize I could use the INDEX function on lists. Though it will require another step or 2 to get the final result, I think it will work.

    Shawn, try this: two I-Types (I'll just include the 2nd line of each)

    1. TESTING_FLAT
       CATS(CONVERT(@SM,'|',TESTING),REUSE('|'))

    2. TESTING_MATCH
       ORS(NES(INDEXS(TESTING_FLAT,'1400|',1),REUSE('0')),NES(INDEXS(TESTING_FLAT,'9101|',1),REUSE('0')))

    I think that gives you what you want.
    Tyrel

    Shawn Waldie
    Forum|alt.badge.img
    • Shawn Waldie
    • Author
    • 54 replies
    • September 11, 2024

    Here is a record sample:

    The TESTING column is a multi-valued list, and the pipes represent actual SVMs for readability.

    I'm looking for a function that will return an associated boolean list indicating whether or not either '1400' or '9101' exists in TESTING, so...

    In this sample, if one is present so is the other, but that's not always the case. So a 1 should be returned if either exists in what is actually a subvalued list.

    EDIT 1: A subroutine would make this very easy, I know, but I don't have time for the administrative overhead that comes with developing/delivering one to PROD, so I need this to be "idescriptor" language.

    Also, there will be subsequent processing, based on a SUM() of the Y/N list for each record, which will be compared to DCOUNT(TESTING,@VM).

    EDIT 2: I think I found a solution. I didn't realize I could use the INDEX function on lists. Though it will require another step or 2 to get the final result, I think it will work.

    Thank you, @Tyrel Marak.  That is precisely what I ended up doing yesterday:

    EVAL "

    <<F(SAVEDLISTS,BEG.DT000,1)>>;

    <<F(SAVEDLISTS,END.DT000,1)>>;

    SUBR('S.GET.SCS.ID', '', @ID, @1, @2, 'I':@VM:'E', 'N', 'Y');

    SUBR('S.TRANS','STUDENT.ACAD.CRED',@3,'STC.SCS.COURSE.SECTION','X');

    SUBR('S.TRANS','COURSE.SECTIONS',@4,'SEC.COURSE.LEVELS','X');

    CONVERT(@SM,'|', @5);

    SUBR('-INDEXS',@6,'1400',1);

    SUBR('-INDEXS',@6,'9101',1);

    SUBR('-GTS',@7,REUSE(0));

    SUBR('-GTS',@8,REUSE(0));

    SUM(SUBR('-ORS',@9,@10));

    DCOUNT(@3,@VM);

    IF @11 EQ 0 THEN 'NO DEU' ELSE (IF @11 EQ @12 THEN 'ALL DEU' ELSE 'SOME DEU')

    "

    And I'll mention something else that had me stumped for about half-hour. That is, I initially had the EVAL expression wrapped in single quotes. This was fine until I added that last statement and the system returned a syntax error message. Stepping back and taking a broader view, I noticed that the system wrapped @1 and @2 in single quotes, so I'm surprised the error didn't appear earlier. In any case, another take away for me was wrap your EVAL expression in double quotes.


    Terms & ConditionsAccessibility statement