(** Pre-scan phase for file-local declarations.

    Scans the entire input for syntax-affecting keywords like #+TODO,
    #+SEQ_TODO, and #+TYP_TODO before the main parse. These declarations may
    appear anywhere in the file and affect how headings are interpreted. *)

let lstrip = String_util.lstrip
let strip = String_util.strip

(** Parse a TODO keyword line value into a todo_sequence. Format: "KEYWORD1
    KEYWORD2 | DONE1 DONE2" If no | is present, all keywords are active and
    there are no done keywords. *)
let parse_todo_value value =
  let value = strip value in
  if String.length value = 0 then None
  else
    let words =
      String.split_on_char ' ' value |> List.filter (fun s -> s <> "")
    in
    let rec split_at_bar acc = function
      | [] -> (List.rev acc, [])
      | "|" :: rest -> (List.rev acc, rest)
      | w :: rest -> split_at_bar (w :: acc) rest
    in
    let active, done_ = split_at_bar [] words in
    if active = [] && done_ = [] then None else Some { Config.active; done_ }

(** Extract the value part after #+KEY: from a line. *)
let extract_keyword_value line =
  (* Line format: #+KEY: VALUE *)
  let line = lstrip line in
  if String.length line < 2 || line.[0] <> '#' || line.[1] <> '+' then None
  else
    let rest = String.sub line 2 (String.length line - 2) in
    match String.index_opt rest ':' with
    | None -> None
    | Some colon_pos ->
        let key = String.uppercase_ascii (String.sub rest 0 colon_pos) in
        let value =
          if colon_pos + 1 < String.length rest then
            String.sub rest (colon_pos + 1) (String.length rest - colon_pos - 1)
          else ""
        in
        Some (key, strip value)

(** Scan the input for TODO keyword declarations. Returns a [Config.t] with all
    discovered TODO sequences. *)
let scan input =
  let lines = String.split_on_char '\n' input in
  let sequences =
    List.filter_map
      (fun line ->
        match extract_keyword_value line with
        | Some (key, value)
          when key = "TODO" || key = "SEQ_TODO" || key = "TYP_TODO" ->
            parse_todo_value value
        | _ -> None)
      lines
  in
  match sequences with
  | [] -> Config.default
  | seqs -> { Config.todo_sequences = seqs }