md_to_gost/md2gost/profiles.py

"""Document type profiles and preprocess helpers for MIREA TZ."""

from __future__ import annotations

import re
from dataclasses import dataclass
from datetime import datetime


DOC_TYPES = ("coursework", "practice", "vkr", "PIS_custom", "APID_coursework")

# auto  — Word multilevel list numbers headings; strip digits from md text
# manual — digits stay in md (# 1 … / ## 1.1 …); Word auto-numbering disabled
HEADING_NUMBERING_MODES = ("auto", "manual")
DEFAULT_HEADING_NUMBERING = "manual"

# manual — content built by md2gost with page numbers from layout tracker
# native — Word TOC field (обновить поле при открытии в Word)
TOC_MODES = ("manual", "native")
DEFAULT_TOC_MODE = "native"

# off     — одна таблица Word, пагинацию делает Word (без автоподписи)
# legacy  — режем по оценке высоты + «Продолжение…» с page_break_before
# caption — режем по оценке + явный PageBreak + «Продолжение…»
# soft    — синоним off (автоподпись mid-page без точной вёрстки Word невозможна в DOCX)
# word    — как off при рендере; после save Word COM режет по реальной пагинации + «Продолжение…»
TABLE_CONTINUATION_MODES = ("off", "legacy", "caption", "soft", "word")
DEFAULT_TABLE_CONTINUATION = "word"

# Same modes as tables (word = post-process via Word COM after save).
LISTING_CONTINUATION_MODES = TABLE_CONTINUATION_MODES
DEFAULT_LISTING_CONTINUATION = DEFAULT_TABLE_CONTINUATION

# section — Рисунок 1.1 / 2.1 (ТЗ МИРЭА); continuous — Рисунок 1, 2, 3 (ПИС)
NUMBERING_SCOPES = ("section", "continuous")
STYLE_PRESETS = ("mirea", "pis_custom")


@dataclass
class DocProfile:
    name: str
    min_sources: int
    max_sources: int | None  # None = no upper bound
    sectional_biblio: bool
    require_graphic_appendix: bool
    max_source_age_years: int = 5
    # mirea — ГОСТ/методичка МИРЭА; pis_custom — итоговый отчёт по практическим работам
    style_preset: str = "mirea"
    numbering_scope: str = "section"
    # Structure expectations for checker
    require_intro_conclusion: bool = True
    require_bibliography: bool = True
    require_practical_works: bool = False
    # КР по «Архитектуре приложений и данных» (методичка Аншиной/Лагуновой)
    require_apid_kr_structure: bool = False


PROFILES = {
    "coursework": DocProfile("coursework", 5, 7, False, False),
    "practice": DocProfile("practice", 5, 7, False, False),
    "vkr": DocProfile("vkr", 10, None, True, True),
    # Курсовая АПИД: оформление ГОСТ/Кириллина, источники 7–20 по методичке дисциплины
    "APID_coursework": DocProfile(
        "APID_coursework",
        min_sources=7,
        max_sources=20,
        sectional_biblio=False,
        require_graphic_appendix=False,
        max_source_age_years=5,
        style_preset="mirea",
        numbering_scope="section",
        require_intro_conclusion=True,
        require_bibliography=True,
        require_practical_works=False,
        require_apid_kr_structure=True,
    ),
    # Итоговый отчёт по практическим работам (чек-лист ПИС)
    "PIS_custom": DocProfile(
        "PIS_custom",
        min_sources=0,
        max_sources=None,
        sectional_biblio=False,
        require_graphic_appendix=False,
        max_source_age_years=10,
        style_preset="pis_custom",
        numbering_scope="continuous",
        require_intro_conclusion=False,
        require_bibliography=False,
        require_practical_works=True,
    ),
}


def get_profile(doc_type: str) -> DocProfile:
    if doc_type not in PROFILES:
        raise ValueError(f"Unknown document type: {doc_type}. Expected one of {DOC_TYPES}")
    return PROFILES[doc_type]


# --- text autofixes ---

def fix_russian_quotes(text: str) -> str:
    """Replace "..." with «...» for Russian text segments (heuristic)."""
    def repl(m):
        inner = m.group(1)
        if inner.lstrip().startswith("%"):
            return m.group(0)
        return f"«{inner}»"
    return re.sub(r'"([^"\n]+)"', repl, text)


def fix_dashes(text: str) -> str:
    """Replace spaced hyphen used as dash with em dash; keep range hyphens."""
    # "слово - слово" → "слово — слово"
    text = re.sub(r"(\S) - (\S)", r"\1 — \2", text)
    return text


def replace_emdash_with_hyphen(text: str) -> str:
    """Replace typographic em dash U+2014 with ASCII hyphen-minus."""
    return text.replace("—", "-")


# Runtime switch for — vs - (set from CLI / Converter)
_EMDASH_TO_HYPHEN = False


def set_emdash_to_hyphen(enabled: bool) -> None:
    global _EMDASH_TO_HYPHEN
    _EMDASH_TO_HYPHEN = bool(enabled)


def emdash_to_hyphen_enabled() -> bool:
    return _EMDASH_TO_HYPHEN


def dash_separator() -> str:
    """Caption / prose dash: ' - ' or ' — ' depending on CLI flag."""
    return " - " if _EMDASH_TO_HYPHEN else " — "


def preprocess_markdown(text: str, emdash_to_hyphen: bool | None = None) -> str:
    if emdash_to_hyphen is not None:
        set_emdash_to_hyphen(emdash_to_hyphen)
    text = fix_russian_quotes(text)
    if _EMDASH_TO_HYPHEN:
        # Keep ASCII hyphens; flatten any em dashes from the source
        text = replace_emdash_with_hyphen(text)
    else:
        text = fix_dashes(text)
    text = separate_biblio_lines(text)
    return text


def separate_biblio_lines(text: str) -> str:
    """Ensure each [n]: bibliography line is its own paragraph (Marko soft-breaks otherwise)."""
    lines = text.splitlines(keepends=True)
    out: list[str] = []
    in_biblio = False
    for line in lines:
        stripped = line.strip()
        if re.match(r"^#\s*\*?\s*СПИСОК\s+ИСПОЛЬЗОВАНН?ЫХ\s+ИСТОЧНИКОВ\s*$", stripped, re.I):
            in_biblio = True
            out.append(line)
            continue
        if in_biblio and re.match(r"^#\s+", stripped):
            in_biblio = False
        if in_biblio and re.match(r"^\[\d+(?:\.\d+)?\]:", stripped):
            # blank line before entry if previous wasn't blank
            if out and out[-1].strip():
                out.append("\n")
            out.append(line if line.endswith("\n") else line + "\n")
            out.append("\n")
            continue
        out.append(line)
    return "".join(out)


def find_formula_refs(text: str) -> set[str]:
    """Labels of formulas that are cited, e.g. @Формула:euler or (1.1) after %eq."""
    refs = set()
    for m in re.finditer(r"@Формула:(\w+)", text, re.IGNORECASE):
        refs.add(m.group(1))
    for m in re.finditer(r"@Formula:(\w+)", text, re.IGNORECASE):
        refs.add(m.group(1))
    return refs


def current_year() -> int:
    return datetime.now().year