import re
import json
from pathlib import Path
from modules.term_workspace import TermWorkspace
from config import CITATION_DENSITY_TARGET, MIN_TOTAL_CITATIONS, MIN_STUDY_WORDS

# Section markers for splitting the study
SECTION_MARKERS = [
    ("tahdid", "تمهيد"),
    ("lugha", "المبحث الأول"),
    ("tarikh", "المبحث الثاني"),
    ("madloul", "المبحث الثالث"),
    ("muqarana", "المبحث الرابع"),
    ("khatima", "المبحث الخامس"),
    ("masadir", "قائمة المصادر"),
]

SECTION_NAMES = {
    "tahdid": "التمهيد",
    "lugha": "المبحث الأول: الدراسة اللغوية",
    "tarikh": "المبحث الثاني: الدراسة الاصطلاحية",
    "madloul": "المبحث الثالث: استعمال القطان",
    "muqarana": "المبحث الرابع: الدراسة التطبيقية المقارنة",
    "khatima": "المبحث الخامس: الخاتمة",
    "masadir": "قائمة المصادر",
}

# Broad citation patterns — catches all common inline formats
CITATION_INLINE_PATTERN = (
    r"\([^)]*(?:"
    r"ت\s*\d+"  # death year: (ت 852 هـ)
    r"|ج\s*[:_]?\s*\d+"  # volume: (ج1، ص10) or (ج:1)
    r"|ص\s*[:_]?\s*\d+"  # page: (ص15) or (ص:15)
    r"|🔗"  # link symbol
    r"|https?://"  # URL
    r"|يحتاج توثيقاً"  # needs documentation
    r"|المصدر السابق"  # previous source reference
    r")[^)]*\)"
)


class StudyAuditor:
    """Audit a completed study for quality and compliance — section-aware."""

    def __init__(self, term: str, category: str):
        self.term = term
        self.category = category
        self.ws = TermWorkspace(term, category)
        self.issues = []
        self.stats = {}

    def audit(self, text: str, draft: dict = None) -> dict:
        self.issues = []
        self.stats = {}

        # Strip any existing audit report appendix first to ensure clean stats
        text = re.split(r"\n## ملحق: تقرير الجودة", text)[0].strip()

        self._count_citations(text)
        self._check_citation_density(text)
        self._check_hallucinated_links(text)
        self._check_structure(text)
        self._check_length(text)
        self._check_language(text)

        # New: section-aware analysis
        sections = self._split_into_sections(text)
        self._audit_sections(text, sections)

        # New: raw material utilization
        if draft:
            self._check_raw_utilization(text, draft)

        # New: cross-reference validation (book titles + narrators mentioned in study vs sources)
        self._check_cross_references(text, draft)

        passed = len(self.issues) == 0
        severity = "PASS" if passed else "FAIL"

        if self.issues:
            severity = (
                "WARN"
                if all(i["severity"] == "warning" for i in self.issues)
                else "FAIL"
            )

        result = {
            "term": self.term,
            "category": self.category,
            "severity": severity,
            "stats": self.stats,
            "issues": self.issues,
            "passed": passed,
        }

        self.ws.save_meta({"audit": result})
        return result

    def _split_into_sections(self, text: str) -> dict[str, str]:
        """Split study text into sections based on ## headings."""
        sections = {}
        lines = text.split("\n")
        current_key = None
        current_lines = []

        for line in lines:
            stripped = line.strip()
            if stripped.startswith("## "):
                # Save previous section
                if current_key:
                    sections[current_key] = "\n".join(current_lines)

                # Identify section
                header_text = stripped[3:].strip()
                current_key = None
                for key, marker in SECTION_MARKERS:
                    if marker in header_text:
                        current_key = key
                        break

                if current_key:
                    current_lines = [line]
                else:
                    current_lines = [line]
            elif current_key:
                current_lines.append(line)

        # Save last section
        if current_key:
            sections[current_key] = "\n".join(current_lines)

        return sections

    def _audit_sections(self, full_text: str, sections: dict[str, str]):
        """Per-section citation density and quality analysis."""
        section_stats = {}

        for key, marker in SECTION_MARKERS:
            sec_text = sections.get(key, "")
            if not sec_text:
                section_stats[key] = {
                    "name": SECTION_NAMES.get(key, key),
                    "exists": False,
                    "words": 0,
                    "citations": 0,
                    "paragraphs": 0,
                    "paragraphs_without_citation": 0,
                    "density": 0.0,
                }
                continue

            # Count citations in this section
            citations = len(re.findall(CITATION_INLINE_PATTERN, sec_text))

            # Count paragraphs
            body = re.split(r"\n##\s*(?:قائمة\s+)?المصادر", sec_text)[0].strip()
            paragraphs = [
                p.strip()
                for p in body.split("\n\n")
                if p.strip()
                and not p.strip().startswith("#")
                and p.strip() not in ["---", "***", "___"]
                and not p.strip().startswith("|")
            ]

            paras_without = 0
            for p in paragraphs:
                if not re.search(CITATION_INLINE_PATTERN, p):
                    paras_without += 1

            words = len(sec_text.split())
            density = 1 - (paras_without / len(paragraphs)) if paragraphs else 0.0

            section_stats[key] = {
                "name": SECTION_NAMES.get(key, key),
                "exists": True,
                "words": words,
                "citations": citations,
                "paragraphs": len(paragraphs),
                "paragraphs_without_citation": paras_without,
                "density": round(density, 2),
            }

            # Flag sections with low citation density
            if paragraphs and density < CITATION_DENSITY_TARGET and key != "masadir":
                self.issues.append(
                    {
                        "type": "section_low_density",
                        "severity": "error",
                        "message": f"كثافة التوثيق منخفضة في {SECTION_NAMES.get(key, key)}: {density:.0%} ({paras_without}/{len(paragraphs)} فقرة بلا توثيق)",
                        "section": key,
                    }
                )

            # Flag sections that are too short
            if key != "masadir" and 0 < words < 200:
                self.issues.append(
                    {
                        "type": "section_too_short",
                        "severity": "warning",
                        "message": f"القسم {SECTION_NAMES.get(key, key)} قصير جداً: {words} كلمة",
                        "section": key,
                    }
                )

        self.stats["sections"] = section_stats

    def _check_raw_utilization(self, text: str, draft: dict):
        """Check what percentage of raw material links appear in the study — per section."""
        study_sections = self._split_into_sections(text)
        section_util = {}
        all_raw_links = set()

        for sec_key, sec_data in draft.get("by_section", {}).items():
            sec_links = set()
            for item in sec_data.get("results", []):
                link = item.get("link", "")
                if link:
                    sec_links.add(link)
                    all_raw_links.add(link)

            if not sec_links:
                section_util[sec_key] = {"total": 0, "used": 0, "ratio": 0.0}
                continue

            sec_text = study_sections.get(sec_key, "")
            sec_study_links = set(
                re.findall(r"https://shamela\.ws/book/\d+/\d+", sec_text)
            )
            used = sec_links & sec_study_links
            ratio = len(used) / len(sec_links) if sec_links else 0.0

            section_util[sec_key] = {
                "total": len(sec_links),
                "used": len(used),
                "ratio": round(ratio, 2),
            }

        self.stats["section_utilization"] = section_util

        # Global stats for backward compatibility
        all_study_links = set(re.findall(r"https://shamela\.ws/book/\d+/\d+", text))
        global_used = all_raw_links & all_study_links
        global_util = len(global_used) / len(all_raw_links) if all_raw_links else 0.0

        self.stats["raw_links_total"] = len(all_raw_links)
        self.stats["raw_links_used"] = len(global_used)
        self.stats["raw_utilization"] = round(global_util, 2)

        # Average per-section utilization for warning threshold
        non_empty = [s for s in section_util.values() if s["total"] > 0]
        avg_util = (
            sum(s["ratio"] for s in non_empty) / len(non_empty) if non_empty else 0.0
        )

        if avg_util < 0.3 and non_empty:
            self.issues.append(
                {
                    "type": "low_raw_utilization",
                    "severity": "warning",
                    "message": f"نسبة استخدام المادة الخام منخفضة: {avg_util:.0%} (متوسط عبر الأقسام، {len(global_used)}/{len(all_raw_links)} رابط مستخدم)",
                    "details": f"تم استخدام {len(global_used)} رابط من أصل {len(all_raw_links)} في المادة الخام",
                }
            )

    def _check_cross_references(self, text: str, draft: dict = None):
        """Validate that narrator names and book titles in the study match the raw material."""
        if not draft:
            return

        # Collect known narrators from draft
        known_narrators = set()
        for sec_data in draft.get("by_section", {}).values():
            for item in sec_data.get("results", []):
                narrator = item.get("narrator", "").strip()
                if narrator:
                    known_narrators.add(narrator)

        # Collect known book titles from draft
        known_books = set()
        for sec_data in draft.get("by_section", {}).values():
            for item in sec_data.get("results", []):
                book = item.get("book", "").strip()
                if book:
                    known_books.add(book)

        # Check for narrator names mentioned in study but not in raw material
        # Arabic name pattern: 2-4 words of Arabic chars
        study_narrators = set()
        for match in re.finditer(
            r"(?:الراوي|الرواة|يقول|قال)\s+([^\s,،]+(?:\s+[^\s,،]+){0,2})", text
        ):
            name = match.group(1).strip()
            if len(name) > 3 and not any(
                stop in name for stop in ["هذا", "هذه", "ذلك", "تلك", "الذي", "التي"]
            ):
                study_narrators.add(name)

        unvalidated_narrators = []
        for name in study_narrators:
            # Check if this name appears in any known narrator
            found = any(name in kn or kn in name for kn in known_narrators)
            if not found and len(name) > 5:
                unvalidated_narrators.append(name)

        if unvalidated_narrators:
            self.stats["unvalidated_narrators"] = len(unvalidated_narrators)
            self.issues.append(
                {
                    "type": "unvalidated_narrators",
                    "severity": "warning",
                    "message": f"{len(unvalidated_narrators)} أسماء رواة في الدراسة لا تتطابق مع المادة الخام",
                    "details": unvalidated_narrators[:5],
                }
            )

        self.stats["known_narrators"] = len(known_narrators)
        self.stats["known_books"] = len(known_books)

    def _count_citations(self, text: str):
        numbered_pattern = r"\(\d+\)"

        inline_citations = re.findall(CITATION_INLINE_PATTERN, text)
        numbered_citations = re.findall(numbered_pattern, text)

        total_inline = len(inline_citations)
        total_numbered = len(numbered_citations)

        self.stats["inline_citations"] = total_inline
        self.stats["numbered_citations"] = total_numbered
        self.stats["total_citations"] = total_inline + total_numbered

    def _check_citation_density(self, text: str):
        # Exclude the bibliography section (and anything after it) from paragraph density checks
        body_text = re.split(r"\n##\s*(?:قائمة\s+)?المصادر", text)[0].strip()
        paragraphs = [
            p.strip()
            for p in body_text.split("\n\n")
            if p.strip()
            and not p.strip().startswith("#")
            and p.strip() not in ["---", "***", "___"]
            and not p.strip().startswith("|")
        ]
        total_paragraphs = len(paragraphs)

        paragraphs_without_citation = 0
        for p in paragraphs:
            if not re.search(CITATION_INLINE_PATTERN, p):
                paragraphs_without_citation += 1

        self.stats["total_paragraphs"] = total_paragraphs
        self.stats["paragraphs_without_citation"] = paragraphs_without_citation

        if total_paragraphs > 0:
            density = 1 - (paragraphs_without_citation / total_paragraphs)
            self.stats["citation_density"] = round(density, 2)

            if density < CITATION_DENSITY_TARGET:
                self.issues.append(
                    {
                        "type": "low_citation_density",
                        "severity": "error",
                        "message": f"كثافة التوثيق منخفضة: {density:.0%} (المطلوب ≥ {CITATION_DENSITY_TARGET:.0%})",
                        "details": f"{paragraphs_without_citation}/{total_paragraphs} فقرات بلا توثيق",
                    }
                )

        if self.stats["total_citations"] < MIN_TOTAL_CITATIONS:
            self.issues.append(
                {
                    "type": "insufficient_citations",
                    "severity": "error",
                    "message": f"عدد التواقيع غير كافٍ: {self.stats['total_citations']} (الحد الأدنى: {MIN_TOTAL_CITATIONS})",
                }
            )

    def _check_hallucinated_links(self, text: str):
        links = re.findall(r"https://shamela\.ws/book/\d+/\d+", text)
        valid_links = self.ws.raw_dir / "search_results.json"

        if valid_links.exists():
            draft = json.loads(valid_links.read_text(encoding="utf-8"))
            known_links = set()
            for sec_data in draft.get("by_section", {}).values():
                for item in sec_data.get("results", []):
                    known_links.add(item.get("link", ""))

            hallucinated = [l for l in links if l not in known_links]
            self.stats["total_links"] = len(links)
            self.stats["hallucinated_links"] = len(hallucinated)

            if hallucinated:
                self.issues.append(
                    {
                        "type": "hallucinated_links",
                        "severity": "error",
                        "message": f"تم اكتشاف {len(hallucinated)} روابط مختلقة",
                        "details": hallucinated[:5],
                    }
                )

    def _check_structure(self, text: str):
        required = [
            "تمهيد",
            "المبحث الأول",
            "المبحث الثاني",
            "المبحث الثالث",
            "المبحث الرابع",
            "المبحث الخامس",
            "قائمة المصادر",
        ]

        missing = []
        for section in required:
            found = any(
                section in line
                for line in text.split("\n")
                if line.strip().startswith("#")
            )
            if not found:
                missing.append(section)

        self.stats["required_sections"] = len(required)
        self.stats["found_sections"] = len(required) - len(missing)

        if missing:
            self.issues.append(
                {
                    "type": "missing_sections",
                    "severity": "error",
                    "message": f"أقسام مفقودة: {missing}",
                }
            )

    def _check_length(self, text: str):
        words = len(text.split())
        chars = len(text)
        lines = len(text.split("\n"))

        self.stats["words"] = words
        self.stats["chars"] = chars
        self.stats["lines"] = lines

        if words < MIN_STUDY_WORDS:
            self.issues.append(
                {
                    "type": "too_short",
                    "severity": "warning",
                    "message": f"الدراسة قصيرة جداً: {words} كلمة (المطلوب ≥ {MIN_STUDY_WORDS})",
                }
            )

    def _check_language(self, text: str):
        english_words = re.findall(r"\b[a-zA-Z]{3,}\b", text)
        exclude_words = {
            "sha",
            "html",
            "http",
            "com",
            "https",
            "shamela",
            "book",
            "url",
            "www",
            "org",
            "net",
        }
        english_words = [w for w in english_words if w.lower() not in exclude_words]

        self.stats["english_words"] = len(english_words)

        if english_words:
            self.issues.append(
                {
                    "type": "english_content",
                    "severity": "warning",
                    "message": f"تم اكتشاف {len(english_words)} كلمة إنجليزية",
                    "details": english_words[:10],
                }
            )

    def print_report(self):
        result = self.audit(
            (self.ws.processed_dir / "study.md").read_text(encoding="utf-8")
            if (self.ws.processed_dir / "study.md").exists()
            else ""
        )

        print(f"\n{'=' * 60}")
        print(f"تقرير Audit لللفظ: {self.term}")
        print(f"{'=' * 60}")
        print(f"الحالة: {'✅ PASS' if result['passed'] else '❌ FAIL'}")
        print(f"{'=' * 60}")

        print(f"\n📊 الإحصائيات العامة:")
        for k, v in result["stats"].items():
            if k != "sections":
                print(f"  {k}: {v}")

        # Print section-level stats
        sections = result["stats"].get("sections", {})
        if sections:
            print(f"\n📊 تفاصيل الأقسام:")
            for key, sec in sections.items():
                status = "✅" if sec["exists"] else "❌"
                density_str = f"{sec['density']:.0%}" if sec["exists"] else "N/A"
                print(
                    f"  {status} {sec['name']}: {sec['words']} كلمة | {sec['citations']} توثيق | كثافة: {density_str}"
                )

        if result["issues"]:
            print(f"\n⚠️ المشاكل ({len(result['issues'])}):")
            for issue in result["issues"]:
                icon = "❌" if issue["severity"] == "error" else "⚠️"
                print(f"  {icon} [{issue['type']}] {issue['message']}")
                if "details" in issue:
                    for d in (
                        issue["details"][:3]
                        if isinstance(issue["details"], list)
                        else [issue["details"]]
                    ):
                        print(f"      → {d}")
        else:
            print(f"\n✅ لا توجد مشاكل!")

        print(f"\n{'=' * 60}")
        return result


def format_audit_report_markdown(audit_res: dict) -> str:
    stats = audit_res.get("stats", {})
    severity = audit_res.get("severity", "PASS")
    severity_map = {
        "PASS": "ناجح (مطابق للمنهجية)",
        "WARN": "تنبيه (ملاحظات منهجية)",
        "FAIL": "راسب (مخالف للمنهجية)",
    }
    severity_arabic = severity_map.get(severity, severity)

    total_citations = stats.get("total_citations", 0)
    density = stats.get("citation_density", 0.0)
    density_percent = int(density * 100)
    words = stats.get("words", 0)
    total_paragraphs = stats.get("total_paragraphs", 0)
    paragraphs_without_citation = stats.get("paragraphs_without_citation", 0)
    found_sections = stats.get("found_sections", 0)
    required_sections = stats.get("required_sections", 7)

    issues = audit_res.get("issues", [])
    if not issues:
        issues_list = "* ✓ لا توجد أي مخالفات منهجية أو لغوية."
    else:
        issues_list = ""
        for issue in issues:
            icon = "❌" if issue.get("severity") == "error" else "⚠️"
            issues_list += (
                f"\n- {icon} **[{issue.get('type')}]** {issue.get('message')}"
            )
            if "details" in issue:
                details = issue["details"]
                if isinstance(details, list):
                    for d in details:
                        issues_list += f"\n  - {d}"
                else:
                    issues_list += f"\n  - {details}"

    # Section-level breakdown — no tables, use bullet lists
    sections = stats.get("sections", {})
    section_rows = ""
    if sections:
        section_rows = "\n### تفاصيل الأقسام:\n"
        for key, sec in sections.items():
            status = "✓" if sec.get("exists") else "✗"
            density_val = f"{sec.get('density', 0):.0%}" if sec.get("exists") else "غير متاح"
            section_rows += (
                f"- **{sec.get('name', key)}**: الحالة {status}، "
                f"الكلمات {sec.get('words', 0)}، "
                f"التوثيقات {sec.get('citations', 0)}، "
                f"الكثافة {density_val}\n"
            )

    # Raw utilization
    raw_util = stats.get("raw_utilization")
    raw_section = ""
    if raw_util is not None:
        raw_used = stats.get("raw_links_used", 0)
        raw_total = stats.get("raw_links_total", 0)
        raw_section = f"\n### استخدام المادة الخام:\n- نسبة الاستخدام: {raw_util:.0%} ({raw_used}/{raw_total} رابط)\n"

        # Per-section breakdown — bullet list
        sec_util = stats.get("section_utilization", {})
        if sec_util:
            for key, su in sec_util.items():
                name = SECTION_NAMES.get(key, key)
                ratio_str = f"{su['ratio']:.0%}"
                raw_section += f"- **{name}**: المجموع {su['total']}، المستخدم {su['used']}، النسبة {ratio_str}\n"

    report_md = f"""

## ملحق: تقرير الجودة المنهجية والأكاديمية

* **حالة التدقيق**: {severity_arabic}
* **إجمالي التوثيقات**: {total_citations}
* **كثافة التوثيق في الفقرات**: {density_percent}%
* **عدد الكلمات**: {words} كلمة
* **عدد الفقرات**: {total_paragraphs} فقرة
* **الفقرات الخالية من التوثيق**: {paragraphs_without_citation} فقرة
* **الأقسام المكتملة**: {found_sections} من {required_sections}
{section_rows}{raw_section}
### تفاصيل ملاحظات التدقيق المنهجي:
{issues_list}
"""
    return report_md


if __name__ == "__main__":
    import sys

    if len(sys.argv) < 3:
        print("Usage: python3 study_audit.py <term> <category>")
        print("  category: tadeel or jarh")
        sys.exit(1)

    term = sys.argv[1]
    category = sys.argv[2]
    auditor = StudyAuditor(term, category)
    auditor.print_report()