# -*- coding: utf-8 -*-
"""건설품셈 기계설비 제7~13장 md 읽기 (build_소요량_건설품셈_기계설비.py 가 씀).

항(###) · 목(####) · 표 · 표 머리 깊이(주석) · 표 앞 글 · [주] [비고] [계산예] 글을 읽음.
"""

import re

from _md표_읽기 import unesc
from _토목_읽기 import ROOT

BASE = ROOT / "resources/knowledge/original/원가계산/건설공사_표준품셈/본문/04_기계설비부문"
DIV = "건설품셈 기계설비 "
NOTE0 = re.compile(r"^[①-⑳]")
BLOCK = re.compile(r"^\[(주|비고|계산예|참고|참고자료|별표)\]\s*(.*)$")


class Tab:
    def __init__(self, ident, title, mok, base, rows, k, hdr, pre):
        self.ident, self.title, self.mok, self.base = ident, title, mok, base
        self.rows, self.k, self.hdr, self.pre = rows, k, hdr, pre
        self.notes: list[str] = []


def cells(line):
    return [c.strip() for c in line.strip().strip("|").split("|")]


def title_of(s):
    return re.sub(r"\s*\((?:['‘’]\d\d[^)]*)\)\s*$", "", s).strip()


def clean(ln):
    ln = re.sub(r"<!--.*?-->", "", ln).rstrip()
    return re.sub(r"^>\s?", "", ln)


def read_chapter(chdir):
    """장 폴더 → 표 목록 (표마다 항 · 목 · 기준 · 머리 깊이 · 앞 글 · 주)."""
    tabs = []
    for p in sorted(next(BASE.glob(chdir + "*")).glob("*.md")):
        L = p.read_text(encoding="utf-8").split("\n")
        L = L[[i for i, x in enumerate(L) if x.strip() == "---"][1] + 1 :]
        ident = title = mok = base = ""
        pre: list[str] = []
        k, mode, i = 0, "", 0
        while i < len(L):
            raw = L[i]
            ln = clean(raw)
            m = re.match(r"^(#{2,4})\s+(.*)$", ln)
            if m:
                mode, base, pre = "", "", []
                d = len(m.group(1))
                if d == 2:
                    mm = re.match(r"^(\d+-\d+)\s+(.*)$", m.group(2))
                    ident, title, mok, k = mm.group(1), title_of(mm.group(2)), "", 0
                elif d == 3:
                    mm = re.match(r"^(\d+-\d+-\d+)\s+(.*)$", m.group(2))
                    ident, title, mok, k = mm.group(1), title_of(mm.group(2)), "", 0
                elif d == 4:
                    mk = re.match(r"^(?:\d+|[가-힣])\.\s*(.*)$", m.group(2))
                    mok = title_of(mk.group(1) if mk else m.group(2))
                i += 1
                continue
            if not ident:
                i += 1
                continue
            if ln.startswith("|"):
                j, rows = i, []
                while j < len(L) and L[j].startswith("|"):
                    c = cells(L[j])
                    if not all(re.fullmatch(r":?-+:?", x) for x in c):
                        rows.append(c)
                    j += 1
                hint = ""
                while j < len(L) and (L[j].startswith("<!--") or not L[j].strip()):
                    hint += L[j] if L[j].startswith("<!--") else ""
                    if not L[j].strip() and not (j + 1 < len(L) and L[j + 1].startswith("<!--")):
                        break
                    j += 1
                mh = re.search(r"몸 (\d+)행까지 머리", hint)
                b = base[1:-1] if base else ""
                tabs.append(
                    Tab(ident, title, mok, b, rows, k, 1 + int(mh.group(1)) if mh else 1, pre)
                )
                k += 1
                base, pre, mode = "", [], ""
                i = j
                continue
            s = ln.strip()
            i += 1
            if not s or s.startswith("!["):
                continue
            mb = BLOCK.match(s)
            if mb:
                kind, rest = mb.groups()
                mode_before = mode
                mode = "note" if kind in ("주", "계산예", "비고") else ""
                if kind in ("참고", "참고자료", "별표"):
                    if "참조" in rest and tabs and tabs[-1].ident == ident:
                        mode = mode_before  # 딴 곳을 가리키는 글 = 앞 주 이어짐
                        tabs[-1].notes.append(unesc(f"{kind} {rest}"))
                        continue
                    pre = [s] if not rest else [f"[{kind}] {rest}"]
                    mode = ""
                    continue
                first = rest if kind == "주" else f"{kind} {rest}".strip()
                if first and tabs and tabs[-1].ident == ident:
                    tabs[-1].notes.append(unesc(first))
                mode = "note" if kind in ("주", "계산예") else ""
                if kind == "계산예":
                    mode = "calc"
                continue
            if re.fullmatch(r"\(.*\)", s) and not s.startswith("(주"):
                base, mode = s, ""
                continue
            if mode in ("note", "calc") and tabs and tabs[-1].ident == ident:
                t = tabs[-1]
                if mode == "calc":
                    t.notes.append("계산예 " + unesc(s)) if not t.notes or not t.notes[
                        -1
                    ].startswith("계산예") else t.notes.__setitem__(
                        -1, t.notes[-1] + " " + unesc(s)
                    )
                elif NOTE0.match(s) or not t.notes:
                    t.notes.append(unesc(s))
                else:
                    t.notes[-1] += " " + unesc(s)
            else:
                pre.append(s)
    return tabs