/**
 * Pure deterministic title-cleaning rules for time-sensitive queries in Browse.
 * Lightweight, synchronous, 100% deterministic with zero external dependencies.
 *
 * CONSTRAINTS:
 * - Public API: applyHeuristicTitleRules(originalUserMessage: string): string
 * - Max length: 50 characters (including trailing "...")
 * - No LLM/NLP libraries, no external dependencies, no external side effects.
 */

export interface TitleRule {
  name: string;
  apply(title: string): string;
}

// ─────────────────────────────────────────────────────────────────────────────
// Rule 1: normalizeWhitespace
// ─────────────────────────────────────────────────────────────────────────────
export const normalizeWhitespaceRule: TitleRule = {
  name: "normalizeWhitespace",
  apply(title: string): string {
    return title.replace(/[\s\r\n\t]+/g, " ").trim();
  },
};

// ─────────────────────────────────────────────────────────────────────────────
// Rule 2: simplifyComparisons ("difference between X and Y" -> "X vs Y")
// ─────────────────────────────────────────────────────────────────────────────
export const simplifyComparisonsRule: TitleRule = {
  name: "simplifyComparisons",
  apply(title: string): string {
    const text = title.replace(/[?!.]+$|\s+$/g, "").trim();

    const diffMatch = text.match(/^(?:what\s+is\s+)?(?:the\s+)?difference\s+between\s+(.+?)\s+and\s+(.+)$/i);
    if (diffMatch && diffMatch[1] && diffMatch[2]) {
      return `${diffMatch[1].trim()} vs ${diffMatch[2].trim()}`;
    }

    const betterMatch = text.match(/^(?:whether\s+|if\s+)?(.+?)\s+(?:is|are)\s+better\s+than\s+(.+)$/i);
    if (betterMatch && betterMatch[1] && betterMatch[2]) {
      return `${betterMatch[1].trim()} vs ${betterMatch[2].trim()}`;
    }

    return title;
  },
};

// ─────────────────────────────────────────────────────────────────────────────
// Rule 3: timeSensitiveScaffolding (Strip question/request openers safely)
// ─────────────────────────────────────────────────────────────────────────────
export const timeSensitiveScaffoldingRule: TitleRule = {
  name: "timeSensitiveScaffolding",
  apply(title: string): string {
    let res = title.trim();

    // Weather ("What's the weather?" -> "Weather", "What's the weather today?" -> "Weather Today")
    const weather = res.match(/^what(?:'s|\s+is)\s+(?:the\s+)?weather(?:\s+like)?(?:\s+(in\s+[^?!.]+?))?(?:\s+(today|tomorrow|yesterday|this\s+week))?[?!.]?$/i);
    if (weather) {
      const loc = weather[1] ? ` ${weather[1].trim()}` : "";
      const time = weather[2] ? ` ${weather[2].trim()}` : "";
      return `Weather${loc}${time}`;
    }

    // Office Holders ("who is currently the CEO of OpenAI?" -> "Current CEO of OpenAI")
    const currentlyOffice = res.match(/^who\s+is\s+currently\s+(?:the\s+)?(.+?)[?!.]?$/i);
    if (currentlyOffice && currentlyOffice[1]) {
      return `Current ${currentlyOffice[1].trim()}`;
    }

    const currentOffice = res.match(/^who\s+is\s+(?:the\s+)?current\s+(.+?)[?!.]?$/i);
    if (currentOffice && currentOffice[1]) {
      return `Current ${currentOffice[1].trim()}`;
    }

    // Markets & Events ("what happened in the markets today?" -> "Markets Today")
    const eventIn = res.match(/^what\s+happened\s+in\s+(?:the\s+)?(.+?)\s+(today|yesterday|this week)[?!.]?$/i);
    if (eventIn && eventIn[1] && eventIn[2]) {
      return `${eventIn[1].trim()} ${eventIn[2].trim()}`;
    }

    const eventTimeIn = res.match(/^what\s+happened\s+(today|yesterday|this week)\s+in\s+(?:the\s+)?(.+?)[?!.]?$/i);
    if (eventTimeIn && eventTimeIn[1] && eventTimeIn[2]) {
      return `${eventTimeIn[2].trim()} ${eventTimeIn[1].trim()}`;
    }

    // Conversational & Request Opener Stripping
    res = res.replace(/^(?:bring|get|show|tell|find|look\s+up|search\s+for)\s+(?:me\s+)?(?:about\s+)?(?:the\s+)?/i, "");

    // Question Opener Stripping (safety check: preserve short queries like "What is IT?")
    const questionMatch = res.match(/^(?:what(?:'s|\s+is)|who\s+is|where\s+is|what\s+are)\s+(?:the\s+)?(.+)$/i);
    if (questionMatch && questionMatch[1]) {
      const remainder = questionMatch[1].replace(/[?!.]+$|\s+$/g, "").trim();
      const words = remainder.split(/\s+/).filter(Boolean);
      if (words.length >= 2 || (remainder.length >= 4 && !/^[A-Z]{1,3}$/.test(remainder))) {
        res = remainder;
      }
    }

    res = res.replace(/^(?:how\s+do\s+I|how\s+to)\s+/i, "");

    return res.trim();
  },
};

// ─────────────────────────────────────────────────────────────────────────────
// Rule 4: essentialCasing (High-value entities, domains & capitalization)
// ─────────────────────────────────────────────────────────────────────────────
const ESSENTIAL_CASE_MAP: Record<string, string> = {
  openai: "OpenAI",
  nvidia: "NVIDIA",
  techcrunch: "TechCrunch",
  qwen: "Qwen",
  gemma: "Gemma",
  deepseek: "DeepSeek",
  gpu: "GPU",
  cpu: "CPU",
  api: "API",
  http: "HTTP",
  tcp: "TCP",
  udp: "UDP",
  cds: "CDS",
  ceo: "CEO",
};

export const essentialCasingRule: TitleRule = {
  name: "essentialCasing",
  apply(title: string): string {
    if (!title) return title;
    let text = title.replace(/[?!.]+$|\s+$/g, "").trim();

    // Clean domain names: e.g. techcrunch.com -> TechCrunch
    text = text.replace(/\b([a-z0-9-]+)\.(?:com|org|net|io|ai)\b/gi, (m, base) => {
      const lower = base.toLowerCase();
      return ESSENTIAL_CASE_MAP[lower] || (base.charAt(0).toUpperCase() + base.slice(1));
    });

    const tokens = text.split(/(\s+|[-/])/);
    for (let i = 0; i < tokens.length; i++) {
      const token = tokens[i];
      const lower = token.toLowerCase();

      // Possessive match e.g. openai's -> OpenAI's
      const possessive = token.match(/^(.+?)(['’]s)$/i);
      if (possessive && ESSENTIAL_CASE_MAP[possessive[1].toLowerCase()]) {
        tokens[i] = ESSENTIAL_CASE_MAP[possessive[1].toLowerCase()] + possessive[2].toLowerCase();
        continue;
      }

      if (ESSENTIAL_CASE_MAP[lower]) {
        tokens[i] = ESSENTIAL_CASE_MAP[lower];
      }
    }

    let result = tokens.join("").trim();
    if (result.length > 0 && /^[a-z]/.test(result)) {
      result = result.charAt(0).toUpperCase() + result.slice(1);
    }
    return result;
  },
};

// ─────────────────────────────────────────────────────────────────────────────
// Rule 5: shortenTitle (Strict max 50 chars including "...")
// ─────────────────────────────────────────────────────────────────────────────
export const shortenTitleRule: TitleRule = {
  name: "shortenTitle(50)",
  apply(title: string): string {
    const clean = title.trim();
    if (clean.length <= 50) return clean;

    const targetLen = 47; // 50 - 3 for "..."
    let truncated = clean.substring(0, targetLen);

    const lastSpace = truncated.lastIndexOf(" ");
    if (lastSpace > 15) {
      truncated = truncated.substring(0, lastSpace);
    }
    truncated = truncated.replace(/\s+(?:with|and|or|for|in|of|the|a|an|to|from|by|at|on|as|is|are|vs|about)$/i, "");
    truncated = truncated.replace(/[:\-\s,.!?]+$/, "");
    return `${truncated}...`;
  },
};

/** Ordered list of heuristic rules applied during title cleaning. */
export const defaultTitleRules: readonly TitleRule[] = [
  normalizeWhitespaceRule,
  simplifyComparisonsRule,
  timeSensitiveScaffoldingRule,
  essentialCasingRule,
  shortenTitleRule,
];

/**
 * Pure function that transforms a raw user message into a clean, normalized conversation title.
 * 100% deterministic, zero side effects, no external dependencies.
 */
export function applyHeuristicTitleRules(originalUserMessage: string): string {
  if (!originalUserMessage || !originalUserMessage.trim()) {
    return "Conversation";
  }

  const rawMessage = originalUserMessage.trim();
  let currentTitle = rawMessage;

  for (const rule of defaultTitleRules) {
    try {
      currentTitle = rule.apply(currentTitle);
    } catch {
      // Retain title state if an unexpected edge case occurs
    }
  }

  currentTitle = currentTitle.replace(/[\s\r\n\t]+/g, " ").trim();

  // Fallback safety check: return clean rawMessage if result is empty or too short
  if (!currentTitle || currentTitle === "..." || currentTitle.length < 2) {
    currentTitle = rawMessage.replace(/[\s\r\n\t]+/g, " ").trim();
    if (currentTitle.length > 50) {
      let truncated = currentTitle.substring(0, 47);
      const lastSpace = truncated.lastIndexOf(" ");
      if (lastSpace > 15) truncated = truncated.substring(0, lastSpace);
      currentTitle = `${truncated.replace(/[:\-\s,.!?]+$/, "")}...`;
    }
  }

  return currentTitle;
}

/**
 * Pure deterministic title-cleaning rules for time-sensitive queries in Browse.
 * Lightweight, synchronous, 100% deterministic with zero external dependencies.
 *
 * CONSTRAINTS:
 * - Public API: applyHeuristicTitleRules(originalUserMessage: string): string
 * - Max length: 50 characters (including trailing "...")
 * - No LLM/NLP libraries, no external dependencies, no external side effects.
 */

export interface TitleRule {
  name: string;
  apply(title: string): string;
}

// ─────────────────────────────────────────────────────────────────────────────
// Rule 1: normalizeWhitespace
// ─────────────────────────────────────────────────────────────────────────────
export const normalizeWhitespaceRule: TitleRule = {
  name: "normalizeWhitespace",
  apply(title: string): string {
    return title.replace(/[\s\r\n\t]+/g, " ").trim();
  },
};

// ─────────────────────────────────────────────────────────────────────────────
// Rule 2: simplifyComparisons ("difference between X and Y" -> "X vs Y")
// ─────────────────────────────────────────────────────────────────────────────
export const simplifyComparisonsRule: TitleRule = {
  name: "simplifyComparisons",
  apply(title: string): string {
    const text = title.replace(/[?!.]+$|\s+$/g, "").trim();

    const diffMatch = text.match(/^(?:what\s+is\s+)?(?:the\s+)?difference\s+between\s+(.+?)\s+and\s+(.+)$/i);
    if (diffMatch && diffMatch[1] && diffMatch[2]) {
      return `${diffMatch[1].trim()} vs ${diffMatch[2].trim()}`;
    }

    const betterMatch = text.match(/^(?:whether\s+|if\s+)?(.+?)\s+(?:is|are)\s+better\s+than\s+(.+)$/i);
    if (betterMatch && betterMatch[1] && betterMatch[2]) {
      return `${betterMatch[1].trim()} vs ${betterMatch[2].trim()}`;
    }

    return title;
  },
};

// ─────────────────────────────────────────────────────────────────────────────
// Rule 3: timeSensitiveScaffolding (Strip question/request openers safely)
// ─────────────────────────────────────────────────────────────────────────────
export const timeSensitiveScaffoldingRule: TitleRule = {
  name: "timeSensitiveScaffolding",
  apply(title: string): string {
    let res = title.trim();

    // Weather ("What's the weather?" -> "Weather", "What's the weather today?" -> "Weather Today")
    const weather = res.match(/^what(?:'s|\s+is)\s+(?:the\s+)?weather(?:\s+like)?(?:\s+(in\s+[^?!.]+?))?(?:\s+(today|tomorrow|yesterday|this\s+week))?[?!.]?$/i);
    if (weather) {
      const loc = weather[1] ? ` ${weather[1].trim()}` : "";
      const time = weather[2] ? ` ${weather[2].trim()}` : "";
      return `Weather${loc}${time}`;
    }

    // Office Holders ("who is currently the CEO of OpenAI?" -> "Current CEO of OpenAI")
    const currentlyOffice = res.match(/^who\s+is\s+currently\s+(?:the\s+)?(.+?)[?!.]?$/i);
    if (currentlyOffice && currentlyOffice[1]) {
      return `Current ${currentlyOffice[1].trim()}`;
    }

    const currentOffice = res.match(/^who\s+is\s+(?:the\s+)?current\s+(.+?)[?!.]?$/i);
    if (currentOffice && currentOffice[1]) {
      return `Current ${currentOffice[1].trim()}`;
    }

    // Markets & Events ("what happened in the markets today?" -> "Markets Today")
    const eventIn = res.match(/^what\s+happened\s+in\s+(?:the\s+)?(.+?)\s+(today|yesterday|this week)[?!.]?$/i);
    if (eventIn && eventIn[1] && eventIn[2]) {
      return `${eventIn[1].trim()} ${eventIn[2].trim()}`;
    }

    const eventTimeIn = res.match(/^what\s+happened\s+(today|yesterday|this week)\s+in\s+(?:the\s+)?(.+?)[?!.]?$/i);
    if (eventTimeIn && eventTimeIn[1] && eventTimeIn[2]) {
      return `${eventTimeIn[2].trim()} ${eventTimeIn[1].trim()}`;
    }

    // Conversational & Request Opener Stripping
    res = res.replace(/^(?:bring|get|show|tell|find|look\s+up|search\s+for)\s+(?:me\s+)?(?:about\s+)?(?:the\s+)?/i, "");

    // Question Opener Stripping (safety check: preserve short queries like "What is IT?")
    const questionMatch = res.match(/^(?:what(?:'s|\s+is)|who\s+is|where\s+is|what\s+are)\s+(?:the\s+)?(.+)$/i);
    if (questionMatch && questionMatch[1]) {
      const remainder = questionMatch[1].replace(/[?!.]+$|\s+$/g, "").trim();
      const words = remainder.split(/\s+/).filter(Boolean);
      if (words.length >= 2 || (remainder.length >= 4 && !/^[A-Z]{1,3}$/.test(remainder))) {
        res = remainder;
      }
    }

    res = res.replace(/^(?:how\s+do\s+I|how\s+to)\s+/i, "");

    return res.trim();
  },
};

// ─────────────────────────────────────────────────────────────────────────────
// Rule 4: essentialCasing (High-value entities, domains & capitalization)
// ─────────────────────────────────────────────────────────────────────────────
const ESSENTIAL_CASE_MAP: Record<string, string> = {
  openai: "OpenAI",
  nvidia: "NVIDIA",
  techcrunch: "TechCrunch",
  qwen: "Qwen",
  gemma: "Gemma",
  deepseek: "DeepSeek",
  gpu: "GPU",
  cpu: "CPU",
  api: "API",
  http: "HTTP",
  tcp: "TCP",
  udp: "UDP",
  cds: "CDS",
  ceo: "CEO",
};

export const essentialCasingRule: TitleRule = {
  name: "essentialCasing",
  apply(title: string): string {
    if (!title) return title;
    let text = title.replace(/[?!.]+$|\s+$/g, "").trim();

    // Clean domain names: e.g. techcrunch.com -> TechCrunch
    text = text.replace(/\b([a-z0-9-]+)\.(?:com|org|net|io|ai)\b/gi, (m, base) => {
      const lower = base.toLowerCase();
      return ESSENTIAL_CASE_MAP[lower] || (base.charAt(0).toUpperCase() + base.slice(1));
    });

    const tokens = text.split(/(\s+|[-/])/);
    for (let i = 0; i < tokens.length; i++) {
      const token = tokens[i];
      const lower = token.toLowerCase();

      // Possessive match e.g. openai's -> OpenAI's
      const possessive = token.match(/^(.+?)(['’]s)$/i);
      if (possessive && ESSENTIAL_CASE_MAP[possessive[1].toLowerCase()]) {
        tokens[i] = ESSENTIAL_CASE_MAP[possessive[1].toLowerCase()] + possessive[2].toLowerCase();
        continue;
      }

      if (ESSENTIAL_CASE_MAP[lower]) {
        tokens[i] = ESSENTIAL_CASE_MAP[lower];
      }
    }

    let result = tokens.join("").trim();
    if (result.length > 0 && /^[a-z]/.test(result)) {
      result = result.charAt(0).toUpperCase() + result.slice(1);
    }
    return result;
  },
};

// ─────────────────────────────────────────────────────────────────────────────
// Rule 5: shortenTitle (Strict max 50 chars including "...")
// ─────────────────────────────────────────────────────────────────────────────
export const shortenTitleRule: TitleRule = {
  name: "shortenTitle(50)",
  apply(title: string): string {
    const clean = title.trim();
    if (clean.length <= 50) return clean;

    const targetLen = 47; // 50 - 3 for "..."
    let truncated = clean.substring(0, targetLen);

    const lastSpace = truncated.lastIndexOf(" ");
    if (lastSpace > 15) {
      truncated = truncated.substring(0, lastSpace);
    }
    truncated = truncated.replace(/\s+(?:with|and|or|for|in|of|the|a|an|to|from|by|at|on|as|is|are|vs|about)$/i, "");
    truncated = truncated.replace(/[:\-\s,.!?]+$/, "");
    return `${truncated}...`;
  },
};

/** Ordered list of heuristic rules applied during title cleaning. */
export const defaultTitleRules: readonly TitleRule[] = [
  normalizeWhitespaceRule,
  simplifyComparisonsRule,
  timeSensitiveScaffoldingRule,
  essentialCasingRule,
  shortenTitleRule,
];

/**
 * Pure function that transforms a raw user message into a clean, normalized conversation title.
 * 100% deterministic, zero side effects, no external dependencies.
 */
export function applyHeuristicTitleRules(originalUserMessage: string): string {
  if (!originalUserMessage || !originalUserMessage.trim()) {
    return "Conversation";
  }

  const rawMessage = originalUserMessage.trim();
  let currentTitle = rawMessage;

  for (const rule of defaultTitleRules) {
    try {
      currentTitle = rule.apply(currentTitle);
    } catch {
      // Retain title state if an unexpected edge case occurs
    }
  }

  currentTitle = currentTitle.replace(/[\s\r\n\t]+/g, " ").trim();

  // Fallback safety check: return clean rawMessage if result is empty or too short
  if (!currentTitle || currentTitle === "..." || currentTitle.length < 2) {
    currentTitle = rawMessage.replace(/[\s\r\n\t]+/g, " ").trim();
    if (currentTitle.length > 50) {
      let truncated = currentTitle.substring(0, 47);
      const lastSpace = truncated.lastIndexOf(" ");
      if (lastSpace > 15) truncated = truncated.substring(0, lastSpace);
      currentTitle = `${truncated.replace(/[:\-\s,.!?]+$/, "")}...`;
    }
  }

  return currentTitle;
}