"""HTTP client for the Nous inference chat completions API."""

from __future__ import annotations

import os

import httpx

_DEFAULT_UA = "osint-dashboard-news-summarizer"
_DEFAULT_BASE = "https://inference-api.nousresearch.com/v1"
_JSON_SYSTEM = (
    "You are an OSINT executive briefer. Reply with a single complete JSON object. "
    "Never truncate mid-sentence. If you run out of room, drop the lowest-priority item."
)


def chat(prompt, *, api_key, model, base_url, json_mode=False) -> str:
    resolved = (base_url or os.environ.get("NOUS_BASE_URL", _DEFAULT_BASE)).rstrip("/")
    url = f"{resolved}/chat/completions"
    headers = {
        "Authorization": f"Bearer {api_key}",
        "User-Agent": os.environ.get("OSINT_USER_AGENT") or _DEFAULT_UA,
    }
    max_tokens = 8192 if json_mode else 4096
    timeout = 120.0 if json_mode else 60.0
    messages = [{"role": "user", "content": prompt}]
    if json_mode:
        messages = [
            {"role": "system", "content": _JSON_SYSTEM},
            {"role": "user", "content": prompt},
        ]
    payload = {
        "model": model,
        "messages": messages,
        "temperature": 0.2,
        "max_tokens": max_tokens,
    }
    if json_mode:
        payload["response_format"] = {"type": "json_object"}
    last_content = ""
    try:
        for attempt in range(2):
            with httpx.Client(timeout=timeout) as client:
                resp = client.post(url, headers=headers, json=payload)
            if resp.status_code == 401 or resp.status_code >= 500:
                return ""
            data = resp.json()
            choice = (data.get("choices") or [{}])[0]
            last_content = (choice.get("message") or {}).get("content") or ""
            finish = choice.get("finish_reason")
            if finish == "length" and attempt == 0:
                payload["max_tokens"] = min(int(payload["max_tokens"]) * 2, 16384)
                continue
            return last_content
        return last_content
    except Exception:
        return ""