from __future__ import annotations

import re
from dataclasses import dataclass, asdict
from pathlib import Path
from typing import List

PROJECT_DIR = Path('/opt/obsidian-vault/Agent-Solutions/Kundenprojekte')

@dataclass
class CustomerProject:
    project_id: str
    title: str
    status: str
    description: str
    target_customer: str
    target_industries: List[str]
    customer_problem: str
    acquisition: str
    references: str
    potential_score: float | None
    confidence: str
    source_file: str

    def as_dict(self) -> dict:
        return asdict(self)
def _frontmatter(text: str) -> dict:
    if not text.startswith('---'):
        return {}
    _, raw, _ = text.split('---', 2)
    out = {}
    for line in raw.splitlines():
        if ':' not in line:
            continue
        key, value = line.split(':', 1)
        out[key.strip()] = value.strip().strip('"')
    return out


def _section(text: str, heading: str) -> str:
    pattern = rf'^## {re.escape(heading)}\s*$\n(.*?)(?=^## |\Z)'
    match = re.search(pattern, text, flags=re.M | re.S)
    return match.group(1).strip() if match else ''


def _industries(raw: str) -> List[str]:
    first = raw.splitlines()[0] if raw else ''
    return [x.strip() for x in re.split(r'[·|,;]', first) if x.strip()]
def load_project(path: Path) -> CustomerProject:
    text = path.read_text(encoding='utf-8')
    fm = _frontmatter(text)
    eval_section = _section(text, 'Bewertungsmodell / Evidenzstärke')
    score_match = re.search(r'Gesamtpotenzial:\s*\*\*(\d+(?:[,.]\d+)?)/10', eval_section)
    conf_match = re.search(r'Konfidenz:\s*\*\*([^*]+)\*\*', eval_section)
    score = float(score_match.group(1).replace(',', '.')) if score_match else None
    return CustomerProject(
        project_id=fm.get('project_id', path.stem),
        title=fm.get('title', path.stem),
        status=fm.get('status', ''),
        description=_section(text, 'Projektbeschreibung'),
        target_customer=_section(text, 'Zielkunde'),
        target_industries=_industries(_section(text, 'Interessant für / Zielbranchen')),
        customer_problem=_section(text, 'Kundenproblem'),
        acquisition=_section(text, 'Kundengewinnung / Akquisewege'),
        references=_section(text, 'Vorhandene Demos / Referenzen / Agenten / Tools'),
        potential_score=score,
        confidence=conf_match.group(1).strip() if conf_match else '',
        source_file=str(path),
    )


def load_all_projects(project_dir: Path = PROJECT_DIR) -> List[CustomerProject]:
    return [load_project(p) for p in sorted(project_dir.glob('*.md'))]