aragrp/lib/blog-utils.ts

import {
  discoverBlocks,
  mergeHeadingIdsIntoContentSegments,
  type ContentSegment,
} from "@/lib/blog-blocks";
import { getFullSizeWpImageUrl } from "@/lib/wp-api";

export type { ContentSegment } from "@/lib/blog-blocks";

const AUTHOR_NAME_OVERRIDES: Record<string, string> = {
  adm: "احمدرضا بدیعی",
};

/** Maps WordPress author display names to site-facing labels. */
export function formatAuthorName(name: string | null | undefined): string | null {
  if (!name?.trim()) return null;
  const normalized = name.trim().toLowerCase();
  return AUTHOR_NAME_OVERRIDES[normalized] ?? name.trim();
}

/**
 * Creates a URL-safe slug from heading text (strip HTML, normalize, Persian-friendly).
 */
function slugify(text: string): string {
  const stripped = text.replace(/<[^>]*>/g, "").trim();
  return stripped
    .replace(/\s+/g, "-")
    .replace(/[^\p{L}\p{N}-]/gu, "")
    .replace(/-+/g, "-")
    .replace(/^-|-$/g, "")
    .toLowerCase() || "section";
}

export interface TocItem {
  id: string;
  text: string;
  level: number; // 2 = h2, 3 = h3, 4 = h4
}

const HEADING_REGEX = /<(h[234])[^>]*>([\s\S]*?)<\/\1>/gi;
const usedIds = new Map<string, number>();

function ensureUniqueId(baseId: string): string {
  const count = usedIds.get(baseId) ?? 0;
  usedIds.set(baseId, count + 1);
  return count === 0 ? baseId : `${baseId}-${count}`;
}

/**
 * پارس محتوا: ابتدا بلاک‌ها کشف و به JSON تبدیل می‌شوند (blog-blocks)،
 * سپس id سرنویس‌ها تزریق می‌شود و سگمنت‌های نهایی برای رندر برمی‌گردند.
 */
export function parseContentHeadings(html: string): {
  contentWithIds: string;
  toc: TocItem[];
  segments: ContentSegment[];
} {
  const { htmlWithPlaceholders, segments: rawSegments } = discoverBlocks(
    enhanceContentImages(html)
  );

  usedIds.clear();
  const toc: TocItem[] = [];
  let lastIndex = 0;
  let match: RegExpExecArray | null;
  const parts: string[] = [];
  const regex = new RegExp(HEADING_REGEX.source, "gi");

  while ((match = regex.exec(htmlWithPlaceholders)) !== null) {
    const [fullMatch, tag, inner] = match;
    const level = parseInt(tag.charAt(1), 10) as 2 | 3 | 4;
    const text = inner.replace(/<[^>]*>/g, "").trim();
    const baseId = slugify(text);
    const id = ensureUniqueId(baseId);

    toc.push({ id, text, level });

    const openTag = fullMatch.slice(0, fullMatch.indexOf(">") + 1);
    const hasId = /^\s*<h[234][^>]*\bid\s*=/i.test(openTag);
    const newOpenTag = hasId
      ? openTag
      : openTag.replace(/^(<\s*h[234])(\s|>)/i, `$1 id="${id}"$2`);

    parts.push(htmlWithPlaceholders.slice(lastIndex, match.index));
    parts.push(newOpenTag + inner + `</${tag}>`);
    lastIndex = regex.lastIndex;
  }

  parts.push(htmlWithPlaceholders.slice(lastIndex));
  const contentWithIds = parts.join("");
  const segments = mergeHeadingIdsIntoContentSegments(
    contentWithIds,
    rawSegments
  );

  return { contentWithIds, toc, segments };
}

/** Average reading speed in words per minute (Persian/English mix). */
const WPM = 200;

/**
 * Estimates reading time in minutes from HTML content.
 */
export function getReadingTimeMinutes(html: string): number {
  const text = html.replace(/<[^>]*>/g, " ").replace(/\s+/g, " ").trim();
  const wordCount = text.split(/\s+/).filter(Boolean).length;
  return Math.max(1, Math.ceil(wordCount / WPM));
}

/** Use full-size WordPress uploads and drop responsive srcset that picks smaller variants. */
export function enhanceContentImages(html: string): string {
  return html.replace(/<img\b([^>]*)>/gi, (tag, attrs: string) => {
    let next = attrs
      .replace(/\s+srcset=(["'])[^"']*\1/gi, "")
      .replace(/\s+sizes=(["'])[^"']*\1/gi, "")
      .replace(/\s+src=(["'])([^"']+)\1/gi, (_m, quote: string, src: string) => {
        return ` src=${quote}${getFullSizeWpImageUrl(src)}${quote}`;
      });
    if (!/\bloading=/i.test(next)) next += ' loading="lazy"';
    if (!/\bdecoding=/i.test(next)) next += ' decoding="async"';
    return `<img${next}>`;
  });
}