"""Get file outline - symbols in a specific file."""

import json
import os
import time
from typing import Optional

from ..storage import IndexStore, record_savings, estimate_savings, cost_avoided
from ..parser import build_symbol_tree
from ._utils import resolve_repo


def _get_file_outline_single(
    file_path: str,
    index,
    owner: str,
    name: str,
    store: IndexStore,
    start: float,
) -> dict:
    """Core logic for a single file_path query. Returns the original flat shape."""
    if not index.has_source_file(file_path):
        return {
            "repo": f"{owner}/{name}",
            "file": file_path,
            "language": "",
            "file_summary": "",
            "symbols": [],
        }

    # Filter symbols to this file
    file_symbols = [s for s in index.symbols if s.get("file") == file_path]
    language = index.file_languages.get(file_path, "")
    file_summary = index.file_summaries.get(file_path, "")

    # Token savings: raw file size vs outline response size
    raw_bytes = 0
    try:
        raw_file = store._content_dir(owner, name) / file_path
        raw_bytes = os.path.getsize(raw_file)
    except OSError:
        pass

    if not file_symbols:
        elapsed = (time.perf_counter() - start) * 1000
        tokens_saved = estimate_savings(raw_bytes, 0)
        total_saved = record_savings(tokens_saved, tool_name="get_file_outline", repo_key=f"{owner}/{name}")
        return {
            "repo": f"{owner}/{name}",
            "file": file_path,
            "language": language,
            "file_summary": file_summary,
            "symbols": [],
            "_meta": {
                "timing_ms": round(elapsed, 1),
                "symbol_count": 0,
                "tokens_saved": tokens_saved,
                "total_tokens_saved": total_saved,
                **cost_avoided(tokens_saved, total_saved),
                "tip": "Tip: use file_paths=[...] to query multiple files in one call.",
            },
        }

    # Build symbol tree
    from ..parser import Symbol
    symbol_objects = [_dict_to_symbol(s) for s in file_symbols]
    tree = build_symbol_tree(symbol_objects)

    # Convert to output format
    symbols_output = [_node_to_dict(n) for n in tree]

    elapsed = (time.perf_counter() - start) * 1000
    response_bytes = len(json.dumps(symbols_output).encode("utf-8"))
    tokens_saved = estimate_savings(raw_bytes, response_bytes)
    total_saved = record_savings(tokens_saved, tool_name="get_file_outline", repo_key=f"{owner}/{name}")

    return {
        "repo": f"{owner}/{name}",
        "file": file_path,
        "language": language,
        "file_summary": file_summary,
        "symbols": symbols_output,
        "_meta": {
            "timing_ms": round(elapsed, 1),
            "symbol_count": len(symbols_output),
            "tokens_saved": tokens_saved,
            "total_tokens_saved": total_saved,
            **cost_avoided(tokens_saved, total_saved),
            "tip": "Tip: use file_paths=[...] to query multiple files in one call.",
        },
    }


def _get_file_outline_batch(
    file_paths: list[str],
    index,
    owner: str,
    name: str,
    store: IndexStore,
    start: float,
) -> dict:
    """Batch logic: loop over file_paths, return grouped results array."""
    results = []

    for file_path in file_paths:
        result = _get_file_outline_single(file_path, index, owner, name, store, start)
        # Strip tip from batch results to keep them clean
        if "_meta" in result and "tip" in result["_meta"]:
            del result["_meta"]["tip"]
        results.append(result)

    elapsed = (time.perf_counter() - start) * 1000
    return {
        "repo": f"{owner}/{name}",
        "results": results,
        "_meta": {"timing_ms": round(elapsed, 1)},
    }


def get_file_outline(
    repo: str,
    file_path: Optional[str] = None,
    storage_path: Optional[str] = None,
    file_paths: Optional[list[str]] = None,
) -> dict:
    """Get symbols in a file with hierarchical structure.

    Supports two modes:
    - Singular: pass ``file_path`` to get the original flat response shape.
    - Batch: pass ``file_paths`` (list) to query multiple files at once,
      returning a grouped ``results`` array.

    Args:
        repo: Repository identifier (owner/repo or just repo name)
        file_path: Path to file within repository (singular mode)
        storage_path: Custom storage path
        file_paths: List of file paths (batch mode)

    Returns:
        Singular mode: dict with file, language, file_summary, symbols, _meta.
        Batch mode: dict with ``results`` array (one entry per input file_path).

    Raises:
        ValueError: if neither or both of file_path and file_paths are provided.
    """
    # Normalize: some MCP clients send file_paths=[] alongside file_path when they mean singular mode
    if file_path is not None and file_paths is not None and len(file_paths) == 0:
        file_paths = None
    if (file_path is None and file_paths is None) or (file_path is not None and file_paths is not None):
        raise ValueError("Provide exactly one of 'file_path' or 'file_paths', not both and not neither.")

    start = time.perf_counter()

    try:
        owner, name = resolve_repo(repo, storage_path)
    except ValueError as e:
        return {"error": str(e)}

    # Load index ONCE for both modes
    store = IndexStore(base_path=storage_path)
    index = store.load_index(owner, name)

    if not index:
        return {"error": f"Repository not indexed: {owner}/{name}"}

    if file_paths is not None:
        return _get_file_outline_batch(file_paths, index, owner, name, store, start)
    else:
        return _get_file_outline_single(file_path, index, owner, name, store, start)


def _dict_to_symbol(d: dict) -> "Symbol":
    """Convert dict back to Symbol dataclass."""
    from ..parser import Symbol
    return Symbol(
        id=d["id"],
        file=d["file"],
        name=d["name"],
        qualified_name=d["qualified_name"],
        kind=d["kind"],
        language=d["language"],
        signature=d["signature"],
        docstring=d.get("docstring", ""),
        summary=d.get("summary", ""),
        decorators=d.get("decorators", []),
        keywords=d.get("keywords", []),
        parent=d.get("parent"),
        line=d["line"],
        end_line=d["end_line"],
        byte_offset=d["byte_offset"],
        byte_length=d["byte_length"],
        content_hash=d.get("content_hash", ""),
    )


def _node_to_dict(node) -> dict:
    """Convert SymbolNode to output dict."""
    result = {
        "id": node.symbol.id,
        "kind": node.symbol.kind,
        "name": node.symbol.name,
        "signature": node.symbol.signature,
        "summary": node.symbol.summary,
        "line": node.symbol.line,
    }

    if node.symbol.decorators:
        result["decorators"] = node.symbol.decorators

    if node.children:
        result["children"] = [_node_to_dict(c) for c in node.children]

    return result