RemoveValue ( Value ; ValueList )
A recursive function that removes all matching values from the supplied list
Average rating: 3.8 (45 votes) Log in to vote
|
Caleb Ruth - Show more from this author
Data Performance http://www.datap.co |
Sample input:
RemoveValue ( "rabbit" ; "mouse¶jackrabbit¶rabbit¶lynx" )
Sample output:
mouse
jackrabbit
lynx
jackrabbit
lynx
Function definition: (Copy & paste into FileMaker's Edit Custom Function window)
This function removes all matching values from a list. Unlike other functions it does not require any padding nor adds nor removes any extra paragraph chars to the resulting list (save one at the end). Recursion limits apply to the size of the list.
Comments
 |
F. Osman Cabi, Turkey Jun 9, 2010 |
|
| There's no need for recursion for this function. This does it in a single line of code: Substitute ( "mouse¶rabbit¶lynx¶snake" ; ["rabbit";""] ; ["¶¶";"¶"] ) |
||
|
Cynthia, USA Jun 16, 2010 |
|
| I like the non-recursive one line solution suggested in the comment above, but it may remove parts of a value, not the whole value. Example: Substitute ( "mouse¶rabbit¶lynx¶snake¶jackrabbit" ; ["rabbit";""] ; ["¶¶";"¶"] ) results in "mouse¶lynx¶snake¶jack" Adding the usual "¶" at the beginning and end of the ValueList and at the beginning and end of Value helps, but there may still be leftover returns depending on what the original list looked like. Rather than reinventing the wheel for dealing with extraneous "¶", I used the excellent CullNulls CF to do the cleanup and ended up with the following: CullNulls( Substitute ( ¶ & ValueList & ¶ ; ¶ & Value & ¶ ; ¶ ) ) Note that unlike the original function presented, using CullNulls will remove leading/trailing returns originally in ValueList, so if retaining leading/trailing returns is important, additional logic is needed. |
||
|
Carter Brooks, San Francisco Jan 16, 2012 |
|
| Here's an edit to the code that removes the trailing return by using the List function instead of an '& "¶"' Also converted If statements to Case statements for brevity. /* RemoveValue custom function by Caleb Ruth, Data Performance LLC Edited by Carter Brooks, developer.carterbrooks.com. Converted to use List function to remove trailing return parameters: Value ValueList */ Let ( [ the_value_count = ValueCount ( ValueList ) ; the_existing_value = GetValue ( ValueList ; 1 ) ] ; Case ( the_value_count > 0 ; List( Case ( not Exact ( Value ; the_existing_value ) ; the_existing_value ) ; RemoveValue ( Value ; RightValues ( ValueList ; the_value_count - 1 ) ) ) ) ) //let |
||
|
Daniel A. Shockley, New York, NY Jan 27, 2012 |
|
| Here's a succinct non-rescursive function that neither adds nor removes leading or trailing line returns, only removes entire values (no partial line issues). It adds its own line returns to make the substitute work, and so knows it can remove them before returning the new list: // ValueRemove ( valueList; itemValue ) // version 1.0, Daniel A. Shockley Let( cleanedListPadded = Substitute( "¶" & valueList & "¶"; "¶" & itemValue & "¶"; "¶") ; Middle( cleanedListPadded; 2; Length( cleanedListPadded ) - 2 ) ) |
||
|
Justin Pakes, Virginia, USA Jul 17, 2012 |
|
| I liked Daniel's version, but found it still fails if valueList starts with itemValue twice E.g. ValueRemove ( "a¶a¶b¶c¶a" ; "a" ) returns "a¶b¶c" It then fails to catch one of those two. |
||
|
eeGuler, Istanbul Jan 6, 2013 |
|
| I guess this one is better. // RemoveValue ( valueList ; theValue ) // version 1.0, eeGuler Let ( result = Substitute ( Let ( firstStep = Substitute ( valueList ; ¶ & theValue ; "" ) ; Substitute ( firstStep ; theValue & ¶ ; "" ) ) ; theValue ; "" ) ; result ) |
||
|
rob 'jesus Land Tidd' lewis, wawaka Indiana USA Feb 7, 2013 |
|
| yes !! but how do you remove values such as 2-5 or 1-7 or 5-10 ? | ||
|
Jack James, London Apr 22, 2013 |
|
| How about just: Let ( a = Substitute ( valueList & ¶ ; theValue & ¶ ; "" ) ; Left ( a ; Length ( a - 1 ) ) ) |
||
|
Jack James, London Apr 22, 2013 |
|
| Sorry put a bracket in the wrong place, should be: Let ( a = Substitute ( valueList & ¶ ; theValue & ¶ ; "" ) ; Left ( a ; Length ( a ) - 1 ) ) |
||
|
Mattias Thorslund, Sweden Sep 8, 2016 |
|
| custom function: RemoveValue parameters: values, value Let( text = Substitute ( ¶ & values & ¶ ; ¶ & value & ¶ ; ¶ ); Middle ( text ; 2 ; Length( text )-2 ) ) |
||
 |
RealGrouchy Nov 7, 2016 |
|
| I didn't see the hidden comments on this page with additional suggestions until after I wrote up this variant. This one is ExclueValue*s* because you can specify multiple items to be excluded. I needed this in order to compare two checklist sets against each other to find which one had been added/removed. It also has checks the parameters for trailing line breaks. - RG> // Custom function: ExcludeValues ( ValueList ; ValuesToExclude ) // 2016-11-07 - RG> // Based on discussions at https://www.briandunning.com/filemaker-custom-functions/detail.php?fn_id=1175&comments=all // e.g. ExcludeValues ( "a¶b¶c¶c¶d" ; "a¶c" ) will return "b¶d" Let ( [ ValuesToExclude = ValuesToExclude & If ( Right ( ValuesToExclude ; 1 ) = ¶ ; "" ; ¶ ) ; ValueListPlusBreak = ValueList & If ( Right ( ValueList ; 1 ) = ¶ ; "" ; ¶ ) ; FirstValue = Left ( ValuesToExclude ; Position ( ValuesToExclude ; ¶ ; 1 ; 1 ) ) ; RemainingValuesPlusBreak = Right ( ValuesToExclude ; Length ( ValuesToExclude ) - Length ( FirstValue ) ) ; RemainingValues = Left ( RemainingValuesPlusBreak ; Length ( RemainingValuesPlusBreak ) - If ( Right ( RemainingValuesPlusBreak ; 1 ) = ¶ ; 1 ; 0 ) ) ; FilterOutFirstValue = Substitute ( ValueListPlusBreak ; FirstValue ; "" ) ; Result = Left ( FilterOutFirstValue ; Length ( FilterOutFirstValue ) - If ( Right ( FilterOutFirstValue ; 1 ) = ¶ ; 1 ; 0 ) ) ] ; If ( Length (RemainingValues ) = 0 ; Result ; ExcludeValues ( Result ; RemainingValues ) ) ) //let |
||
 |
Dan Shockley Feb 8, 2019 |
|
| Here's an updated version of what we use in my office: Note that, while it is recursive, even in the worst-case scenario where the entire value list is the same item you are removing, each recursion knocks out half of them, so if the value list was the word Testing repeated 1 million times, the recursion would only be 30 levels deep. Note also that, unlike some of FileMaker's built-in "Value List" functions (and some custom functions out there), this ONLY removes the specified item without other side effects. It does not tack on a trailing line return (like FM's MiddleValues, RightValues, etc. functions). It does not strip off any leading or trailing whitespace. It does exactly what it is defined to do and nothing else: removes every instance of exactly the specified value from the list. // ValueRemove ( valueList; itemValue ) // version 1.2, Erik Shagdar /* Removes every occurrence of itemValue from valueList. HISTORY: 1.2 - 2016-04-21 ( eshagdar ): fixed memory leak bug if both the list and value are empty 1.1 - 2015-10-16 ( eshagdar ): fixed bug where consecutive items were not removed 1.0 - xxxx-xx-xx ( dshockley ): first created */ If ( IsEmpty ( valueList ) and IsEmpty ( itemValue ); ""; Let ( [ cleanedListPadded = Substitute ( "¶" & valueList & "¶"; [ "¶" & itemValue & "¶"; "¶" ] ); cleanedList = Middle ( cleanedListPadded; 2; Length ( cleanedListPadded ) - 2 ) ]; If ( Position ( ¶ & cleanedList & ¶; ¶ & itemValue & ¶; 1; 1 ); ValueRemove ( cleanedList; itemValue ); cleanedList ) ) ) |
||
Note: these functions are not guaranteed or supported by BrianDunning.com. Please contact the individual developer with any questions or problems.