picoclaw/pkg/providers/http_provider.go

// PicoClaw - Ultra-lightweight personal AI agent
// Inspired by and based on nanobot: https://github.com/HKUDS/nanobot
// License: MIT
//
// Copyright (c) 2026 PicoClaw contributors

package providers

import (
	"bytes"
	"context"
	"encoding/json"
	"fmt"
	"io"
	"net/http"
	"net/url"
	"strings"
	"time"
)

type HTTPProvider struct {
	apiKey         string
	apiBase        string
	maxTokensField string // Field name for max tokens (e.g., "max_completion_tokens" for o1/glm models)
	httpClient     *http.Client
}

func NewHTTPProvider(apiKey, apiBase, proxy string) *HTTPProvider {
	return NewHTTPProviderWithMaxTokensField(apiKey, apiBase, proxy, "")
}

func NewHTTPProviderWithMaxTokensField(apiKey, apiBase, proxy, maxTokensField string) *HTTPProvider {
	client := &http.Client{
		Timeout: 120 * time.Second,
	}

	if proxy != "" {
		proxyURL, err := url.Parse(proxy)
		if err == nil {
			client.Transport = &http.Transport{
				Proxy: http.ProxyURL(proxyURL),
			}
		}
	}

	return &HTTPProvider{
		apiKey:         apiKey,
		apiBase:        strings.TrimRight(apiBase, "/"),
		maxTokensField: maxTokensField,
		httpClient:     client,
	}
}

func (p *HTTPProvider) Chat(ctx context.Context, messages []Message, tools []ToolDefinition, model string, options map[string]interface{}) (*LLMResponse, error) {
	if p.apiBase == "" {
		return nil, fmt.Errorf("API base not configured")
	}

	// Strip provider prefix from model name (e.g., moonshot/kimi-k2.5 -> kimi-k2.5, groq/openai/gpt-oss-120b -> openai/gpt-oss-120b, ollama/qwen2.5:14b -> qwen2.5:14b)
	if idx := strings.Index(model, "/"); idx != -1 {
		prefix := model[:idx]
		if prefix == "moonshot" || prefix == "nvidia" || prefix == "groq" || prefix == "ollama" || prefix == "qwen" || prefix == "cerebras" {
			model = model[idx+1:]
		}
	}

	requestBody := map[string]interface{}{
		"model":    model,
		"messages": messages,
	}

	if len(tools) > 0 {
		requestBody["tools"] = tools
		requestBody["tool_choice"] = "auto"
	}

	if maxTokens, ok := options["max_tokens"].(int); ok {
		// Use configured max_tokens_field if specified, otherwise fallback to model-based detection
		fieldName := p.maxTokensField
		if fieldName == "" {
			// Fallback: detect from model name for backward compatibility
			lowerModel := strings.ToLower(model)
			if strings.Contains(lowerModel, "glm") || strings.Contains(lowerModel, "o1") {
				fieldName = "max_completion_tokens"
			} else {
				fieldName = "max_tokens"
			}
		}
		requestBody[fieldName] = maxTokens
	}

	if temperature, ok := options["temperature"].(float64); ok {
		lowerModel := strings.ToLower(model)
		// Kimi k2 models only support temperature=1
		if strings.Contains(lowerModel, "kimi") && strings.Contains(lowerModel, "k2") {
			requestBody["temperature"] = 1.0
		} else {
			requestBody["temperature"] = temperature
		}
	}

	jsonData, err := json.Marshal(requestBody)
	if err != nil {
		return nil, fmt.Errorf("failed to marshal request: %w", err)
	}

	req, err := http.NewRequestWithContext(ctx, "POST", p.apiBase+"/chat/completions", bytes.NewReader(jsonData))
	if err != nil {
		return nil, fmt.Errorf("failed to create request: %w", err)
	}

	req.Header.Set("Content-Type", "application/json")
	if p.apiKey != "" {
		req.Header.Set("Authorization", "Bearer "+p.apiKey)
	}

	resp, err := p.httpClient.Do(req)
	if err != nil {
		return nil, fmt.Errorf("failed to send request: %w", err)
	}
	defer resp.Body.Close()

	body, err := io.ReadAll(resp.Body)
	if err != nil {
		return nil, fmt.Errorf("failed to read response: %w", err)
	}

	if resp.StatusCode != http.StatusOK {
		return nil, fmt.Errorf("API request failed:\n  Status: %d\n  Body:   %s", resp.StatusCode, string(body))
	}

	return p.parseResponse(body)
}

func (p *HTTPProvider) parseResponse(body []byte) (*LLMResponse, error) {
	var apiResponse struct {
		Choices []struct {
			Message struct {
				Content   string `json:"content"`
				ToolCalls []struct {
					ID       string `json:"id"`
					Type     string `json:"type"`
					Function *struct {
						Name             string `json:"name"`
						Arguments        string `json:"arguments"`
						ThoughtSignature string `json:"thought_signature"`
					} `json:"function"`
				} `json:"tool_calls"`
			} `json:"message"`
			FinishReason string `json:"finish_reason"`
		} `json:"choices"`
		Usage *UsageInfo `json:"usage"`
	}

	if err := json.Unmarshal(body, &apiResponse); err != nil {
		return nil, fmt.Errorf("failed to unmarshal response: %w", err)
	}

	if len(apiResponse.Choices) == 0 {
		return &LLMResponse{
			Content:      "",
			FinishReason: "stop",
		}, nil
	}

	choice := apiResponse.Choices[0]

	toolCalls := make([]ToolCall, 0, len(choice.Message.ToolCalls))
	for _, tc := range choice.Message.ToolCalls {
		arguments := make(map[string]interface{})
		name := ""
		thoughtSignature := ""
		argsStr := ""

		if tc.Function != nil {
			name = tc.Function.Name
			thoughtSignature = tc.Function.ThoughtSignature
			argsStr = tc.Function.Arguments
			if argsStr != "" {
				if err := json.Unmarshal([]byte(argsStr), &arguments); err != nil {
					arguments["raw"] = argsStr
				}
			}
		}

		toolCalls = append(toolCalls, ToolCall{
			ID:   tc.ID,
			Type: tc.Type,
			Function: &FunctionCall{
				Name:             name,
				Arguments:        argsStr,
				ThoughtSignature: thoughtSignature,
			},
			Name:      name,
			Arguments: arguments,
		})
	}

	return &LLMResponse{
		Content:      choice.Message.Content,
		ToolCalls:    toolCalls,
		FinishReason: choice.FinishReason,
		Usage:        apiResponse.Usage,
	}, nil
}

func (p *HTTPProvider) GetDefaultModel() string {
	return ""
}