Series sorting in Appian

Did you try to use todatasubset()?

We have tried todatasubset but it was not sorting as expected. Thanks

Using the sort in a datasubset IS sorting correctly,. The array is a list of string (not number - 1.1.1 is not a valid numeric value) and sorting as as list of string works correctly, but obviously doesn't meet your requirements. You'll have to implement your own custom sort. I suspect (although haven't done the work) that you'll have to examine each item, determine how many 'dots' the item contains, split out the numbers between the dots and implement a nested sort based upon how many levels of dot an item contains.

Using the sort in a datasubset IS sorting correctly,. The array is a list of string (not number - 1.1.1 is not a valid numeric value) and sorting as as list of string works correctly, but obviously doesn't meet your requirements. You'll have to implement your own custom sort. I suspect (although haven't done the work) that you'll have to examine each item, determine how many 'dots' the item contains, split out the numbers between the dots and implement a nested sort based upon how many levels of dot an item contains.

Thanks for the response Stewart ! Can you provide any sample example for the custom sort? it would be helpful to implement the logics.

I can't provide sample code as it's likely to be quite time consuming. I can recommend an approach, which would be to extract strings that are logically at the same level e.g. '1.1.1.1' and '3.4.56.8' both have 4 numbers (with 3 "dots"). Once extracted to its own list this will sort correctly. 

Assuming that for every level below level 1 each item always has a parent present (e.g. '1.1' will have a parent of '1', '3.4.6' will have a parent of '3.4', and '3.4' will have a parent of '3') then you can partition your sorted lists at each level (e.g. all '1.1.x' items in one sorted list, all '1.2.x' items in a sorted list) you should be able to inject your sorted partition lists in the right position in the result list. 

a!localVariables(
  local!values: {"1",
    "1.1",
    "1.1.1",
    "4",
    "1.2",
    "3",
    "2",
    "2.1",
    "2.2",
    "10",
    "13",
    "12"
  },
  local!prepare: a!forEach(
    items: local!values,
    expression: a!map(
      first: text(split(fv!item, ".")[1], "0000"),
      second: text(index(split(fv!item, "."), 2, null), "0000"),
      third: text(index(split(fv!item, "."), 3, null), "0000"),
      index: fv!index
    )
  ),
  local!sorted: todatasubset(
    local!prepare,
    a!pagingInfo(
      startIndex: 1,
      batchSize: -1,
      sort: {
        a!sortInfo(field: "first", ascending: true),
        a!sortInfo(field: "second", ascending: true),
        a!sortInfo(field: "third", ascending: true),
      }
    )
  ),
  index(local!values, local!sorted.data.index)
)

Gah! I'd completely forgotten that the sort now supports nested sortInfo()...that makes the problem WAYYYYY easier...

Thanks Stefan for the sample. we will take it forward from here.