Yeah, I think it makes sense to add, but I've just been doing this when needed

xs.flatMap((x, i) => i ? [sql`, `, x] : x)

You can fairly quickly make a function to use instead if you feel like it - eg 😋

sql.join = (xs, joiner) => xs.flatMap((x, i) => i ? [joiner, x] : x)

I'll keep the issue open for now.

sql.join = (xs, joiner) => xs.flatMap((x, i) => i ? [joiner, x] : x)

oh interesting! i would never have come up with that on my own

do you have any reservations about adding this as a core feature?

and if no reservations, would you accept a PR adding this as a core feature?

It would be great to at least add the work-around to the docs. It wasn't obvious to me how to join sql fragments with a comma. Glad I found this.

i'm working on a PR

anyone want to bike-shed over join vs concat to avoid confusion with sql joins?
or concat_ws to use actual postgres api 😁

Yeah, I think it makes sense to add, but I've just been doing this when needed

xs.flatMap((x, i) => i ? [sql`, `, x] : x)

You can fairly quickly make a function to use instead if you feel like it - eg 😋

sql.join = (xs, joiner) => xs.flatMap((x, i) => i ? [joiner, x] : x)

I'll keep the issue open for now.

This was insanely helpful, I have a really nested structure so my version is slightly different, and I am monkey patching the sql/database variable.

// Dumb placeholder to shorten code
export type PendingQuery<T extends postgres.Row = postgres.Row> =
  postgres.PendingQuery<T[]>;

// Hack to add the "join" method.
export type sql = ReturnType<typeof postgres> & {
  join: (a: Array<PendingQuery | null>, b: PendingQuery) => PendingQuery;
};

const db = postgres(config) as sql;

db["join"] = (fragments: (PendingQuery | null)[], joiner: PendingQuery) =>
  fragments.flatMap((x, i) => {
    if (!i) return x;

    if (Array.isArray(x)) {
      // already an array, means we're nested, so lets spread it
      return [joiner, ...x];
    }

    return [joiner, x];
  }) as unknown as PendingQuery;

It's typescript, but you can strip the types and it should work, usage is:

db.join([
  db`name = 'test1'`,
  db`name = 'test2'`,
  db`name = 'test3'`,
  db.join([
    db`updated_at >= 2024-01-01`,
    db`updated_at <= '2024-12-31`
  ],db`AND`)
], db`OR`)

Hopefully this can make it into somewhere in the codebase as it's super useful for report building.

@DanielFGray hopefully this code can help if you haven't figured out this part.