// Text helpers shared by the digest, the tools and the commands.

// A rough token estimate. Good enough for budgets; the exact count depends on the model.
export function estimateTokens(text: string): number {
  return Math.ceil(text.length / 4);
}

export function oneLine(text: string): string {
  return text.replace(/\s+/g, " ").trim();
}

export function clip(text: string, maxChars: number): string {
  const t = oneLine(text);
  if (t.length <= maxChars) return t;
  const cut = t.slice(0, maxChars - 1);
  const space = cut.lastIndexOf(" ");
  return (space > maxChars * 0.6 ? cut.slice(0, space) : cut) + "…";
}

const STOPWORDS = new Set("a an and are as at be by for from in into is it of on or so than that the then this to with".split(" "));

// Lowercase words joined by hyphens, for project names that are easy to type. Small words
// are left out so names stay short and distinctive.
export function slugify(text: string, maxChars = 40): string {
  const words = text
    .toLowerCase()
    .normalize("NFKD")
    .replace(/[\u0300-\u036f]/g, "")
    .replace(/[^a-z0-9]+/g, " ")
    .trim()
    .split(" ")
    .filter((w) => w && !STOPWORDS.has(w));
  let slug = "";
  for (const w of words) {
    const next = slug ? `${slug}-${w}` : w;
    if (next.length > maxChars) break;
    slug = next;
    if (slug.split("-").length >= 6) break;
  }
  return slug || "project";
}

// Removes blocks throughLine added to a message, so text taken from history is the user's own.
export function stripThroughLineBlocks(text: string): string {
  return text.replace(/\n*\[throughLine[^\]]*\][\s\S]*?\[end throughLine\]\n*/g, "\n").replace(/\n*\[throughLine:[^\]]*\]\s*$/g, "").trim();
}

// Text helpers shared by the digest, the tools and the commands.

// A rough token estimate. Good enough for budgets; the exact count depends on the model.
export function estimateTokens(text: string): number {
  return Math.ceil(text.length / 4);
}

export function oneLine(text: string): string {
  return text.replace(/\s+/g, " ").trim();
}

export function clip(text: string, maxChars: number): string {
  const t = oneLine(text);
  if (t.length <= maxChars) return t;
  const cut = t.slice(0, maxChars - 1);
  const space = cut.lastIndexOf(" ");
  return (space > maxChars * 0.6 ? cut.slice(0, space) : cut) + "…";
}

const STOPWORDS = new Set("a an and are as at be by for from in into is it of on or so than that the then this to with".split(" "));

// Lowercase words joined by hyphens, for project names that are easy to type. Small words
// are left out so names stay short and distinctive.
export function slugify(text: string, maxChars = 40): string {
  const words = text
    .toLowerCase()
    .normalize("NFKD")
    .replace(/[\u0300-\u036f]/g, "")
    .replace(/[^a-z0-9]+/g, " ")
    .trim()
    .split(" ")
    .filter((w) => w && !STOPWORDS.has(w));
  let slug = "";
  for (const w of words) {
    const next = slug ? `${slug}-${w}` : w;
    if (next.length > maxChars) break;
    slug = next;
    if (slug.split("-").length >= 6) break;
  }
  return slug || "project";
}

// Removes blocks throughLine added to a message, so text taken from history is the user's own.
export function stripThroughLineBlocks(text: string): string {
  return text.replace(/\n*\[throughLine[^\]]*\][\s\S]*?\[end throughLine\]\n*/g, "\n").replace(/\n*\[throughLine:[^\]]*\]\s*$/g, "").trim();
}