Skip to main content
Puts a value into a map assign under a key worked out at runtime — the way to build a map up one entry at a time inside a loop. Every other assign node needs its key typed in at design time. This one computes it from the data, so a Flow Each over a flat list can sort its results into buckets as it goes: expense rows into their report, time reports into their employee, invoice rows into their cost centre.

​
Configuration

  • Target assign — the map to grow. Starts as an empty map if unset.
  • Key source — where the entry key comes from. Resolved against normal assigns, so inside a loop point it at the current item.
  • Value source — what to store.
  • Mode — Put replaces whatever is at that key. Append treats the key as holding a list and adds to it (creating the list on first write) — this is the one you want for “collect all rows for this identifier”. Concat spreads a list onto the one already there.
  • On nil key — Put under nil (default), Drop the entry, or Halt flow. Prefer one of the latter two when a missing key means the value can’t be placed meaningfully.

​
Tips

  • Follow the loop with a Flow Map Entries node to iterate what you built — Flow Each takes lists, not maps.
  • Reset the map with a Flow Assign (literal, empty map) before the loop if the flow can run the loop more than once in a single run.
  • If the whole list is available up front, Flow Group By does this in one node and needs no accumulator.

​
Ports