JSLT: removing duplicates by key

When you end up with an array that has duplicate entries by some key (e.g. id, tid), you can dedupe in JSLT by turning the array into an object keyed by that field, then turning the object’s values back into an array. Duplicate keys overwrite, so you keep one entry per key.

Pattern

1. Object from array: { for ($array) .keyField : . }
   Builds an object: key = .keyField, value = the whole element. Same key appears more than once → later value wins → unique keys.

2. Array from object: [ for ($object) .value ]
   Iterating over an object gives key/value; take .value and you get the original elements back, now unique by that key.

Together:

let withDuplicates = [
  { "id": "a", "name": "Alpha" },
  { "id": "b", "name": "Beta" },
  { "id": "a", "name": "Alpha again" }
]
let byId = { for ($withDuplicates) .id : . }
[ for ($byId) .value ]

Result: two objects (one per distinct id), not three. The second "a" overwrote the first in byId.

In a function

Same idea in a reusable helper:

def dedupe-by-key(array, keyField)
  let obj = { for ($array) .[$keyField] : . }
  [ for ($obj) .value ]

Use it like dedupe-by-key($myArray, "tid") (or whatever your key field is). If JSLT doesn’t support dynamic field access, inline the key:

def dedupe-by-tid(array)
  let obj = { for ($array) .tid : . }
  if ($obj)
    [ for ($obj) .value ]
  else
    []

Then dedupe-by-tid($sharedDevicesWithDuplicates) gives you devices unique by tid.

Where this fits

Typical flow:

1. Build an array that might have duplicates (e.g. from nested loops, flattened links, or merging two lists by a key).
2. Dedupe with the object trick: let unique = [ for ({ for ($withDuplicates) .tid : . }) .value ].
3. Use $unique (or guard with if ($unique) $unique else [] if you need an empty array when there’s nothing).

Try it: JSLT demo playground.