numolus/app/frontend/engine/refs.ts

/**
 * Reference helpers: A1 <-> (col,row) conversion and range expansion.
 *
 * All coordinates the engine uses internally are 0-based. The A1 notation
 * exposed to the user is 1-based for rows and letter-based for columns
 * (A=0, B=1, ..., Z=25, AA=26, ...).
 */

import { tokenize, type Token } from './lexer';

const CELL_RE = /^(\$?)([A-Za-z]{1,3})(\$?)([0-9]{1,7})$/;

/** "A" -> 0, "B" -> 1, "Z" -> 25, "AA" -> 26 ... */
export function colToNumber(letters: string): number {
  let n = 0;
  const up = letters.toUpperCase();
  for (let i = 0; i < up.length; i++) {
    n = n * 26 + (up.charCodeAt(i) - 64); // 'A' -> 1
  }
  return n - 1;
}

/** 0 -> "A", 25 -> "Z", 26 -> "AA" ... */
export function numberToCol(n: number): string {
  let s = '';
  let x = n + 1;
  while (x > 0) {
    const rem = (x - 1) % 26;
    s = String.fromCharCode(65 + rem) + s;
    x = Math.floor((x - 1) / 26);
  }
  return s;
}

/** True if `text` looks like a bare cell reference (e.g. "A1", "$B$2", "aa10"). */
export function looksLikeCell(text: string): boolean {
  return CELL_RE.test(text);
}

/**
 * Parse "A1" / "$A1" / "A$1" / "$A$1" -> 0-based coords plus which axes are
 * `$`-locked (absolute). A locked axis is exempt from the reference shift a
 * copy/fill applies (see `shiftFormula` in the store); an unlocked one is
 * "relative" and always shifts with it. Throws for malformed input.
 */
export function parseA1(
  text: string,
): { col: number; row: number; colAbs: boolean; rowAbs: boolean } {
  const m = CELL_RE.exec(text);
  if (!m) throw new Error(`Bad cell reference: ${text}`);
  return { col: colToNumber(m[2]), row: Number(m[4]) - 1, colAbs: m[1] === '$', rowAbs: m[3] === '$' };
}

/** Build "A1" from 0-based coordinates; `colAbs`/`rowAbs` add the `$` lock markers. */
export function toA1(col: number, row: number, colAbs = false, rowAbs = false): string {
  return `${colAbs ? '$' : ''}${numberToCol(col)}${rowAbs ? '$' : ''}${row + 1}`;
}

/**
 * Excel/Numbers-style F4: cycle every cell reference through
 * relative -> absolute -> row-locked -> column-locked -> relative, each
 * from its own current state.
 *
 * With a collapsed selection (`selStart === selEnd`, the common case -- just
 * a caret), this toggles the single reference at/adjacent to it. With a real
 * text selection, it toggles *every* reference the selection overlaps at
 * once -- select a stretch of formula text spanning several references and
 * F4 adjusts them all in one press, same as Excel.
 *
 * `selStart`/`selEnd` are 0-based offsets into `formula` (which must start
 * with "="), i.e. straight from an `<input>`/`<textarea>`'s
 * `selectionStart`/`selectionEnd`. Returns the rewritten formula and a
 * selection re-spanning the (possibly resized) references, so a repeated F4
 * keeps cycling the same ones; `null` when none were found.
 */
export function cycleReferenceLocks(
  formula: string,
  selStart: number,
  selEnd: number,
): { text: string; start: number; end: number } | null {
  if (!formula.startsWith('=')) return null;
  const body = formula.slice(1);
  const lo = Math.min(selStart, selEnd) - 1;
  const hi = Math.max(selStart, selEnd) - 1;
  const collapsed = lo === hi;
  let toks: Token[];
  try {
    toks = tokenize(body);
  } catch {
    return null;
  }
  const targets: Token[] = [];
  for (let i = 0; i < toks.length; i++) {
    const t = toks[i];
    if (t.type !== 'IDENT' || !looksLikeCell(t.value)) continue;
    const next = toks[i + 1];
    if (next && (next.type === 'BANG' || next.type === 'DBLCOLON')) continue; // a table-name qualifier, not a cell
    const tEnd = t.pos + t.value.length;
    const hit = collapsed ? lo >= t.pos && lo <= tEnd : t.pos < hi && tEnd > lo;
    if (hit) targets.push(t);
  }
  if (!targets.length) return null;

  let out = '';
  let last = 0;
  let start = -1;
  let end = -1;
  for (const tok of targets) {
    const { col, row, colAbs, rowAbs } = parseA1(tok.value);
    const [nColAbs, nRowAbs] = nextLock(colAbs, rowAbs);
    const rewritten = toA1(col, row, nColAbs, nRowAbs);
    out += body.slice(last, tok.pos) + rewritten;
    if (start === -1) start = out.length - rewritten.length;
    end = out.length;
    last = tok.pos + tok.value.length;
  }
  out += body.slice(last);
  // a collapsed caret in -> a collapsed caret out (after the one toggled
  // reference), same as before; a real selection in keeps spanning every
  // reference it touched, so a repeated F4 keeps cycling all of them.
  return collapsed
    ? { text: '=' + out, start: 1 + end, end: 1 + end }
    : { text: '=' + out, start: 1 + start, end: 1 + end };
}

/** A1 -> $A$1 -> A$1 -> $A1 -> A1 -> ... (the classic F4 cycle). */
function nextLock(colAbs: boolean, rowAbs: boolean): [boolean, boolean] {
  if (!colAbs && !rowAbs) return [true, true];
  if (colAbs && rowAbs) return [false, true];
  if (!colAbs && rowAbs) return [true, false];
  return [false, false]; // colAbs && !rowAbs
}

/**
 * If the caret sits inside (anywhere within, or right at the end of) a bare
 * identifier being typed in a formula -- not yet a resolved cell reference,
 * not the start of an already-open function call -- return the identifier's
 * text *up to the caret* and its start offset in `formula` (which still
 * includes the leading "="). Drives function-name autocomplete; `null`
 * means "nothing to suggest here" (not a formula at all, a cell reference
 * like "A1", the caret sitting in whitespace or right after a completed
 * token rather than still touching it, ...).
 *
 * Same shape as {@link cycleReferenceLocks} above: strip the leading "=",
 * tokenize, and inspect the token(s) around the caret -- except here only
 * the prefix up to the caret is tokenized (nothing after it is ever looked
 * at, same as how autocomplete works in most editors), so the "identifier"
 * found is always exactly what's to the *left* of the caret, regardless of
 * whether more of the same word continues to its right.
 */
export function currentIdentifierAtCaret(
  formula: string,
  caret: number,
): { text: string; start: number } | null {
  if (!formula.startsWith('=')) return null;
  const pos = caret - 1; // caret position relative to body = formula.slice(1)
  if (pos < 0) return null;
  let toks: Token[];
  try {
    toks = tokenize(formula.slice(1, caret));
  } catch {
    return null;
  }
  // tokenize() appends a trailing EOF marker -- skip it to get the last
  // *real* token.
  const real = toks.filter((t) => t.type !== 'EOF');
  const last = real[real.length - 1];
  if (!last || last.type !== 'IDENT') return null;
  // the token must reach all the way to the caret -- if it doesn't, the
  // caret has moved past it (trailing whitespace, e.g. "=SUM |") and
  // there's no word actively being typed right at the caret any more.
  if (last.pos + last.value.length !== pos) return null;
  if (looksLikeCell(last.value)) return null; // looks like a cell reference, not a function name
  return { text: last.value, start: last.pos + 1 };
}

/** One cell/range reference found in a formula's text -- see {@link formulaRefSpans}. */
export interface RefSpan {
  /** offsets into the formula (including its leading "="): `start` is where
   *  the qualifier (if any) begins, `refStart` where the A1 part begins */
  start: number;
  refStart: number;
  end: number;
  qualifier: string | null;
  sheetQualifier?: string;
  /** both corners as written (`end` equals `start` for a single cell) */
  a: { col: number; row: number; colAbs: boolean; rowAbs: boolean };
  b: { col: number; row: number; colAbs: boolean; rowAbs: boolean };
  /** a whole-column/whole-row range (`B:B`, `2:5`); its open axis runs to
   *  {@link WHOLE_LINE_END} */
  whole?: LineAxis;
}

/**
 * Every cell/range reference in a formula, in text order, with where it sits
 * in the text -- what the grid colours while a formula is edited (and lets
 * the user drag to change it, Excel-style). Same token-level reading of the
 * `reference` grammar rule the parser applies (optional `Blatt!Tabelle::` /
 * `Tabelle!` qualifier, `A1`, `A1:B2`, `B:B`), but tolerant: an incomplete
 * formula still yields the references written so far. Named ranges are left
 * out (they have no position of their own to drag); a formula that doesn't
 * even tokenize (open "..." literal) yields nothing.
 */
export function formulaRefSpans(formula: string): RefSpan[] {
  if (!formula.startsWith('=')) return [];
  let toks: Token[];
  try {
    toks = tokenize(formula.slice(1));
  } catch {
    return [];
  }
  const isName = (t: Token | undefined) => !!t && (t.type === 'IDENT' || t.type === 'QUOTED');
  const out: RefSpan[] = [];
  let i = 0;
  while (i < toks.length) {
    const first = toks[i];
    let j = i;
    let qualifier: string | null = null;
    let sheetQualifier: string | undefined;
    if (isName(toks[j]) && toks[j + 1]?.type === 'BANG' && isName(toks[j + 2]) && toks[j + 3]?.type === 'DBLCOLON') {
      sheetQualifier = toks[j].value;
      qualifier = toks[j + 2].value;
      j += 4;
    } else if (isName(toks[j]) && (toks[j + 1]?.type === 'BANG' || toks[j + 1]?.type === 'DBLCOLON')) {
      qualifier = toks[j].value;
      j += 2;
    }
    const line = wholeLineAt(toks, j);
    const cell = toks[j];
    if (line) {
      const max = WHOLE_LINE_END;
      const isCol = line.axis === 'col';
      out.push({
        start: first.pos + 1,
        refStart: line.a.pos + 1,
        end: line.b.pos + line.b.value.length + 1,
        qualifier,
        sheetQualifier,
        a: { col: isCol ? line.start : 0, row: isCol ? 0 : line.start, colAbs: isCol && line.startAbs, rowAbs: !isCol && line.startAbs },
        b: { col: isCol ? line.end : max, row: isCol ? max : line.end, colAbs: isCol && line.endAbs, rowAbs: !isCol && line.endAbs },
        whole: line.axis,
      });
      i = j + 3;
      continue;
    }
    if (cell?.type === 'IDENT' && looksLikeCell(cell.value) && toks[j + 1]?.type !== 'LPAREN') {
      const a = parseA1(cell.value);
      let last = cell;
      let b = a;
      const other = toks[j + 2];
      if (toks[j + 1]?.type === 'COLON' && other?.type === 'IDENT' && looksLikeCell(other.value)) {
        b = parseA1(other.value);
        last = other;
        j += 2;
      }
      out.push({
        start: first.pos + 1,
        refStart: cell.pos + 1,
        end: last.pos + last.value.length + 1,
        qualifier,
        sheetQualifier,
        a,
        b,
      });
      i = j + 1;
      continue;
    }
    i++;
  }
  return out;
}

// ---- whole-column / whole-row references (`B:B`, `A:C`, `1:1`, `$2:$5`) ----

/** Stand-in for "the last row/column" in a whole-column/whole-row range's
 *  open end. Every consumer clamps it to the actual table size (see
 *  {@link rangeCells}); a consumer that merely clamps to its table (the
 *  Name Box selection) gets the right answer for free. */
export const WHOLE_LINE_END = 9_999_998;

export type LineAxis = 'col' | 'row';

const COL_PART_RE = /^(\$?)([A-Za-z]{1,3})$/;
const ROW_PART_RE = /^(\$?)([0-9]{1,7})$/;

/** One side of a whole-line reference: `"B"`/`"$B"` (a column) or
 *  `"3"`/`"$3"` (a 1-based row) -> its 0-based index, or `null`. */
export function parseLinePart(text: string): { axis: LineAxis; index: number; abs: boolean } | null {
  let m = COL_PART_RE.exec(text);
  if (m) return { axis: 'col', index: colToNumber(m[2]), abs: m[1] === '$' };
  m = ROW_PART_RE.exec(text);
  if (m && Number(m[2]) >= 1) return { axis: 'row', index: Number(m[2]) - 1, abs: m[1] === '$' };
  return null;
}

export function formatLinePart(axis: LineAxis, index: number, abs: boolean): string {
  return (abs ? '$' : '') + (axis === 'col' ? numberToCol(index) : String(index + 1));
}

/** If `toks[i]`, `toks[i+1]`, `toks[i+2]` spell a whole-line range (`B:B`,
 *  `A:C`, `1:1`, `$2:$5` -- both sides the same axis), return its parts.
 *  Shared by the parser and every token-level formula rewriter (copy,
 *  row/column insert/delete, table rename), so they all agree on the shape. */
export function wholeLineAt(
  toks: Token[],
  i: number,
): { axis: LineAxis; a: Token; b: Token; start: number; end: number; startAbs: boolean; endAbs: boolean } | null {
  const a = toks[i];
  const colon = toks[i + 1];
  const b = toks[i + 2];
  if (!a || !colon || !b || colon.type !== 'COLON') return null;
  if ((a.type !== 'IDENT' && a.type !== 'NUMBER') || (b.type !== 'IDENT' && b.type !== 'NUMBER')) return null;
  const pa = parseLinePart(a.value);
  const pb = parseLinePart(b.value);
  if (!pa || !pb || pa.axis !== pb.axis) return null;
  return { axis: pa.axis, a, b, start: pa.index, end: pb.index, startAbs: pa.abs, endAbs: pb.abs };
}

/** The cells a range covers, with a whole-column/whole-row range's open
 *  end clamped to its table's current size (`size` null = unknown table ->
 *  no cells). An ordinary range is just {@link expandRange}. */
export function rangeCells(
  r: { start: { col: number; row: number }; end: { col: number; row: number }; whole?: LineAxis },
  size: { cols: number; rows: number } | null,
): Array<{ col: number; row: number }> {
  if (!r.whole) return expandRange(r.start, r.end);
  if (!size) return [];
  const clamp = (v: number, n: number) => Math.min(v, n - 1);
  const start = { col: Math.min(r.start.col, r.end.col), row: Math.min(r.start.row, r.end.row) };
  const end = { col: Math.max(r.start.col, r.end.col), row: Math.max(r.start.row, r.end.row) };
  // a whole line entirely beyond the table covers nothing
  if (r.whole === 'col' && start.col >= size.cols) return [];
  if (r.whole === 'row' && start.row >= size.rows) return [];
  return expandRange(start, { col: clamp(end.col, size.cols), row: clamp(end.row, size.rows) });
}

/**
 * Expand a rectangular range (given by two corners, any order) into the list of
 * cell coordinates it covers, row-major.
 */
export function expandRange(
  a: { col: number; row: number },
  b: { col: number; row: number },
): Array<{ col: number; row: number }> {
  const c0 = Math.min(a.col, b.col);
  const c1 = Math.max(a.col, b.col);
  const r0 = Math.min(a.row, b.row);
  const r1 = Math.max(a.row, b.row);
  const out: Array<{ col: number; row: number }> = [];
  for (let r = r0; r <= r1; r++) {
    for (let c = c0; c <= c1; c++) {
      out.push({ col: c, row: r });
    }
  }
  return out;
}