# File: src/resume_rag.py
# Purpose: Resume chunking, embedding (sentence-transformers), ChromaDB indexing,
#          and skill extraction via Groq (no OpenAI required)

import json
import re
from typing import List

import chromadb
from chromadb.utils import embedding_functions
from groq import Groq

from config import (
    GROQ_API_KEY, GROQ_LLM_MODEL,
    EMBEDDING_MODEL, CHROMA_PERSIST_DIR,
    CHUNK_SIZE, CHUNK_OVERLAP, TOP_K_RESUME_CHUNKS,
)

# Local sentence-transformers embeddings — no API key needed
_st_ef = embedding_functions.SentenceTransformerEmbeddingFunction(
    model_name=EMBEDDING_MODEL
)

# Groq client
_groq = Groq(api_key=GROQ_API_KEY)


def _chunk_text(text: str, size: int = CHUNK_SIZE, overlap: int = CHUNK_OVERLAP) -> List[str]:
    words = text.split()
    chunks, i = [], 0
    while i < len(words):
        chunks.append(" ".join(words[i: i + size]))
        i += size - overlap
    return chunks


def build_resume_index(session_id: int, resume_text: str) -> chromadb.Collection:
    client = chromadb.PersistentClient(path=CHROMA_PERSIST_DIR)
    collection_name = f"resume_{session_id}"
    try:
        client.delete_collection(collection_name)
    except Exception:
        pass
    collection = client.create_collection(
        name=collection_name,
        embedding_function=_st_ef,
    )
    chunks = _chunk_text(resume_text)
    collection.add(
        documents=chunks,
        ids=[f"chunk_{i}" for i in range(len(chunks))],
    )
    return collection


def retrieve_resume_context(session_id: int, query: str, top_k: int = TOP_K_RESUME_CHUNKS) -> str:
    client = chromadb.PersistentClient(path=CHROMA_PERSIST_DIR)
    collection_name = f"resume_{session_id}"
    try:
        collection = client.get_collection(
            name=collection_name,
            embedding_function=_st_ef,
        )
    except Exception:
        return ""
    results = collection.query(
        query_texts=[query],
        n_results=min(top_k, collection.count()),
    )
    docs = results.get("documents", [[]])[0]
    return "\n".join(docs)


def extract_skills_from_resume(resume_text: str) -> List[str]:
    prompt = (
        "Extract a concise list of technical skills, tools, and frameworks mentioned in the "
        "following resume. Return ONLY a valid JSON array of strings. "
        "No explanation, no markdown, no extra text.\n\n"
        f"Resume:\n{resume_text[:3000]}"
    )
    response = _groq.chat.completions.create(
        model=GROQ_LLM_MODEL,
        messages=[{"role": "user", "content": prompt}],
        temperature=0,
        max_tokens=400,
    )
    raw = response.choices[0].message.content.strip()
    raw = re.sub(r"```json|```", "", raw).strip()
    try:
        return json.loads(raw)
    except Exception:
        return []