import os
from collections import defaultdict
from typing import Dict, List

ROOT = os.path.abspath(os.path.join(os.path.dirname(__file__), '..'))

KEYWORDS: Dict[str, List[str]] = {
    'GraphDB (Neo4j)': ['neo4j', 'py2neo'],
    'Vector DB (FAISS/Milvus)': ['faiss', 'milvus', 'weaviate', 'vector_store'],
    'Neuro-symbolic (PyTorch models)': ['import torch', 'torch.nn', 'torch_geometric'],
    'Quantum (Qiskit/Cirq/cuQuantum)': ['qiskit', 'cirq', 'cuquantum', 'tensornetwork'],
    'Formal provers (Lean/Coq)': ['.lean', '.v', 'lean-toolchain', 'coq'],
    'Orchestration (Kubernetes/Ray/Dask)': ['kubernetes', 'ray', 'dask'],
    'gRPC/GraphQL': ['grpc', 'graphql', 'graphene'],
}

IMPORT_CHECKS = {
    'neo4j': 'neo4j',
    'py2neo': 'py2neo',
    'faiss': 'faiss',
    'torch': 'torch',
    'qiskit': 'qiskit',
    'cirq': 'cirq',
    'tensornetwork': 'tensornetwork',
}

EXTS = {'.py', '.md', '.yaml', '.yml', '.toml', '.json', '.lean', '.v', '.ipynb'}


def scan_repo(root: str) -> Dict[str, int]:
    counts = defaultdict(int)
    for dirpath, dirnames, filenames in os.walk(root):
        if any(skip in dirpath for skip in ['.git', '__pycache__', 'venv', 'env', 'node_modules', '.venv']):
            continue
        for fn in filenames:
            _, ext = os.path.splitext(fn)
            if ext.lower() not in EXTS and fn not in ('Dockerfile',):
                continue
            p = os.path.join(dirpath, fn)
            try:
                with open(p, 'r', encoding='utf-8', errors='ignore') as f:
                    txt = f.read().lower()
            except Exception:
                continue
            for area, kws in KEYWORDS.items():
                for kw in kws:
                    if kw.lower() in txt:
                        counts[area] += txt.count(kw.lower())
    return counts


def try_imports(mapping: Dict[str, str]) -> Dict[str, bool]:
    # Use importlib.util.find_spec to discover packages without executing
    # their top-level import code (safer and faster for heavy packages).
    import importlib.util

    res = {}
    for k, mod in mapping.items():
        try:
            spec = importlib.util.find_spec(mod)
            res[k] = spec is not None
        except Exception:
            res[k] = False
    return res


def find_formal_files(root: str) -> Dict[str, int]:
    c = {'lean': 0, 'coq': 0}
    for dirpath, dirnames, filenames in os.walk(root):
        for fn in filenames:
            if fn.endswith('.lean'):
                c['lean'] += 1
            if fn.endswith('.v'):
                c['coq'] += 1
    return c


def print_summary(counts, imports, formal):
    print('Repo root:', ROOT)
    print('\nSummary:')
    for area in KEYWORDS:
        cnt = counts.get(area, 0)
        status = 'Missing'
        evidence = []
        if cnt > 0:
            status = 'Partial (code references)'
            evidence.append(f'{cnt} mentions')
        if area.startswith('GraphDB') and (imports.get('neo4j') or imports.get('py2neo')):
            status = 'Present (neo4j driver importable)'
        if area.startswith('Vector DB') and imports.get('faiss'):
            status = 'Present (faiss importable)'
        if area.startswith('Neuro-symbolic') and imports.get('torch'):
            status = 'Present (torch importable)'
        print(f'- {area}: {status}')
        for e in evidence:
            print('   -', e)

    print('\nFormal prover artifacts:')
    print('- Lean files:', formal.get('lean', 0))
    print('- Coq files (.v):', formal.get('coq', 0))

    print('\nImport checks:')
    for k, v in sorted(imports.items()):
        print('-', k + ':', 'importable' if v else 'not importable')


def main():
    counts = scan_repo(ROOT)
    imports = try_imports(IMPORT_CHECKS)
    formal = find_formal_files(ROOT)
    print_summary(counts, imports, formal)


if __name__ == '__main__':
    main()