Extract substring of specific values with JQ

This is probably something very specific to my use case, but I was working with some JSON that looked something like this:

json
{
  "something": {
    "Identifying Key": [
      {
        "foo": "a.b.c",
        "bar": "First Three"
      },
      {
        "foo": "a.b.d",
        "bar": "Second Three"
      }
    ],
    "Another Key": [
      {
        "foo": "z.b.c",
        "bar": "First Three, Take Two"
      },
      {
        "foo": "z.b.d",
        "bar": "Second Three, Take Two"
      }
    ]
  }
}

From this, I wanted to extract everything before the first period in every instance of the foo key. Like I said, very specific use case.

To do it, I built the following jq expression.

bash
cat data.json | jq -r '.something | map(.[].foo | split(".")[0]) | unique | join("\n")'

Breaking that down:

bash
# Select the top level namespace
cat data.json | jq -r '.something'
# Map over everything as key => value
cat data.json | jq -r '.something | map(.)'
# From each of those elements, step in to each array
cat data.json | jq -r '.something | map(.[])'
# And extract the foo key
cat data.json | jq -r '.something | map(.[].foo)'
# Inside the map, use "split" to split the string on a "."
cat data.json | jq -r '.something | map(.[].foo | split("."))'
# And only select the first element in the returned arrays
cat data.json | jq -r '.something | map(.[].foo | split(".")[0])'
# We only want to know about unique keys
cat data.json | jq -r '.something | map(.[].foo | split(".")[0]) | unique'
# And we want it as a string, not an array
cat data.json | jq -r '.something | map(.[].foo | split(".")[0]) | unique | join("\n")'

And with that, I have a unique list of values extracted from a specific key in an object. As I mentioned, it's very specific to my use case but it might help someone in the future