Method _make_request

web/pgadmin/llm/providers/docker.py:222–272

Make an HTTP request to the Docker Model Runner API.

(self, payload: dict)

        ]

    def _make_request(self, payload: dict) -> dict:
        """Make an HTTP request to the Docker Model Runner API."""
        headers = {
            'Content-Type': 'application/json'
        }

        # Docker Model Runner uses /engines/v1 path for OpenAI-compatible API
        url = f'{self._api_url}/engines/v1/chat/completions'

        request = urllib.request.Request(
            url,
            data=json.dumps(payload).encode('utf-8'),
            headers=headers,
            method='POST'
        )

        try:
            # Use longer timeout for local models which can be slower
            with urllib.request.urlopen(
                request, timeout=300, context=SSL_CONTEXT
            ) as response:
                return json.loads(response.read().decode('utf-8'))
        except urllib.error.HTTPError as e:
            error_body = e.read().decode('utf-8')
            try:
                error_data = json.loads(error_body)
                error_msg = error_data.get('error', {}).get('message', str(e))
            except json.JSONDecodeError:
                error_msg = error_body or str(e)

            raise LLMClientError(LLMError(
                message=error_msg,
                code=str(e.code),
                provider=self.provider_name,
                retryable=e.code in (429, 500, 502, 503, 504)
            ))
        except urllib.error.URLError as e:
            raise LLMClientError(LLMError(
                message=f"Connection error: {e.reason}. "
                        f"Is Docker Model Runner running at {self._api_url}?",
                provider=self.provider_name,
                retryable=True
            ))
        except socket.timeout:
            raise LLMClientError(LLMError(
                message="Request timed out. Local models can be slow - "
                        "try a smaller model or wait for the response.",
                code='timeout',
                provider=self.provider_name,
                retryable=True
            ))

    def _parse_response(self, data: dict) -> LLMResponse:
        """Parse the API response into an LLMResponse."""