Mycontent/soul-api/internal/handler/miniprogram_ai_article.go

package handler

import (
	"bytes"
	"context"
	"encoding/json"
	"fmt"
	"io"
	"net/http"
	"net/url"
	"path"
	"strings"
	"time"
	"unicode"
	"unicode/utf8"

	"soul-api/internal/config"
	"soul-api/internal/database"

	"github.com/gin-gonic/gin"
)

const aiArticleSystemPrompt = `你是「卡若创业派对」的内容创作助手，帮助超级个体写出真实、有深度的商业与创业内容文章。

写作要求：
- 标题：精炼有力，≤30字，有吸引力，点明核心价值
- 正文：600-1000字，有真实故事或案例，有观点洞察，语气自然真诚
- 风格：像朋友分享经验，而非说教；有画面感，接地气
- 不要使用"首先、其次、最后"等模板化结构词

严格输出 JSON，不要有任何额外文字：
{"title":"文章标题","content":"文章正文"}`

const (
	maxArticleMaterialFetchBody = 512 * 1024 // 单素材最大读取
	maxCombinedReferenceRunes   = 60000      // 合并后截断上限（含提示与素材摘要）
)

// MiniprogramSuperArticleGenerate POST /api/miniprogram/super/articles/generate
// 根据提示词 + 可选参考文本 / 素材文件 URL（本站上传链接） / 图片 URL，调用 AI 生成草稿（仅超级个体）。
func MiniprogramSuperArticleGenerate(c *gin.Context) {
	var req struct {
		UserID         string   `json:"userId"`
		Description    string   `json:"description"`    // 提示词（与原字段兼容）
		ReferenceText  string   `json:"referenceText"` // 小程序端可读出的纯文本（多文件拼接）
		MaterialURLs   []string `json:"materialUrls"`  // 上传后的文件 URL（服务端在安全域名内抓取文本）
		ImageURLs      []string `json:"imageUrls"`
	}
	if err := c.ShouldBindJSON(&req); err != nil {
		c.JSON(http.StatusOK, gin.H{"success": false, "error": "参数错误"})
		return
	}
	req.UserID = strings.TrimSpace(req.UserID)
	req.Description = strings.TrimSpace(req.Description)
	req.ReferenceText = strings.TrimSpace(req.ReferenceText)
	if req.UserID == "" {
		c.JSON(http.StatusOK, gin.H{"success": false, "error": "缺少 userId"})
		return
	}

	hasPrompt := req.Description != ""
	hasRefText := req.ReferenceText != ""

	var mats []string
	for _, u := range req.MaterialURLs {
		if u = strings.TrimSpace(u); u != "" {
			mats = append(mats, u)
		}
	}
	var imgs []string
	for _, u := range req.ImageURLs {
		if u = strings.TrimSpace(u); u != "" {
			imgs = append(imgs, u)
		}
	}

	if !hasPrompt && !hasRefText && len(mats) == 0 && len(imgs) == 0 {
		c.JSON(http.StatusOK, gin.H{"success": false, "error": "请填写提示词，或上传参考文件"})
		return
	}

	db := database.DB()
	if _, ok := miniprogramSuperIndividualPerson(db, req.UserID); !ok {
		c.JSON(http.StatusOK, gin.H{"success": false, "error": "仅超级个体可使用 AI 生成", "forbidden": true})
		return
	}

	cfg := config.Get()
	if cfg == nil || strings.TrimSpace(cfg.OpenAIAPIKey) == "" {
		c.JSON(http.StatusOK, gin.H{"success": false, "error": "AI 功能暂未开放，请联系管理员配置"})
		return
	}

	// 过滤空 URL
	var validImageURLs []string
	for _, u := range imgs {
		validImageURLs = append(validImageURLs, u)
	}

	combined, errMsg := assembleSuperArticlePrompt(cfg, req.Description, req.ReferenceText, mats, len(validImageURLs))
	if errMsg != "" {
		c.JSON(http.StatusOK, gin.H{"success": false, "error": errMsg})
		return
	}

	title, content, err := aiGenerateArticle(cfg, combined, validImageURLs)
	if err != nil {
		c.JSON(http.StatusOK, gin.H{"success": false, "error": "AI 生成失败：" + err.Error()})
		return
	}

	c.JSON(http.StatusOK, gin.H{"success": true, "data": gin.H{"title": title, "content": content}})
}

// assembleSuperArticlePrompt 合并提示词、端上传入的参考正文、以及安全域名内可抓取的素材 URL 文本。
func assembleSuperArticlePrompt(cfg *config.Config, prompt, refText string, materialURLs []string, imageCount int) (combined string, errMsg string) {
	var sb strings.Builder
	if prompt != "" {
		sb.WriteString("用户写作要求（提示词）：\n")
		sb.WriteString(prompt)
		sb.WriteString("\n\n")
	}
	if refText != "" {
		sb.WriteString("以下为用户提供的参考正文（可能来自多个文本文件）：\n")
		sb.WriteString(refText)
		sb.WriteString("\n\n")
	}
	for _, mu := range materialURLs {
		if !materialFetchURLAllowed(cfg, mu) {
			return "", "素材链接不在允许范围，请使用小程序内上传生成的链接"
		}
		snippet, note := fetchMaterialText(cfg, mu)
		name := snippetNameFromURL(mu)
		if snippet != "" {
			sb.WriteString("--- 参考文件（")
			sb.WriteString(name)
			sb.WriteString("）---\n")
			sb.WriteString(snippet)
			sb.WriteString("\n\n")
		} else if note != "" {
			sb.WriteString("--- 参考附件（")
			sb.WriteString(name)
			sb.WriteString("）---\n")
			sb.WriteString(note)
			sb.WriteString("\n\n")
		}
	}
	combined = strings.TrimSpace(sb.String())
	if combined == "" && imageCount > 0 {
		combined = "请根据用户上传的图片，结合创业与个体成长语境，写一篇自然真诚、有观点的短文。"
	}
	if len([]rune(combined)) > maxCombinedReferenceRunes {
		rs := []rune(combined)
		combined = string(rs[:maxCombinedReferenceRunes]) + "\n\n（参考资料后文已截断）"
	}
	return combined, ""
}

func snippetNameFromURL(raw string) string {
	u, err := url.Parse(strings.TrimSpace(raw))
	if err != nil {
		return "素材"
	}
	base := path.Base(u.Path)
	if base == "" || base == "." || base == "/" {
		return "素材"
	}
	return base
}

func materialFetchURLAllowed(cfg *config.Config, raw string) bool {
	u, err := url.Parse(strings.TrimSpace(raw))
	if err != nil || u.Hostname() == "" {
		return false
	}
	switch strings.ToLower(u.Scheme) {
	case "http", "https":
	default:
		return false
	}
	host := strings.ToLower(u.Hostname())
	escPath := u.EscapedPath()
	if !strings.Contains(escPath, "/uploads/") {
		return false
	}
	if cfg != nil && strings.TrimSpace(cfg.BaseURL) != "" {
		if bu, err := url.Parse(strings.TrimSpace(cfg.BaseURL)); err == nil && bu.Hostname() != "" {
			if host == strings.ToLower(bu.Hostname()) {
				return true
			}
		}
	}
	if strings.Contains(host, ".aliyuncs.com") {
		return true
	}
	if strings.HasPrefix(host, "127.0.0.1") || host == "localhost" {
		return true
	}
	return false
}

func fetchMaterialText(cfg *config.Config, rawURL string) (text string, note string) {
	if !materialFetchURLAllowed(cfg, rawURL) {
		return "", ""
	}
	ctx, cancel := context.WithTimeout(context.Background(), 25*time.Second)
	defer cancel()
	req, err := http.NewRequestWithContext(ctx, http.MethodGet, strings.TrimSpace(rawURL), nil)
	if err != nil {
		return "", "无法下载该素材"
	}
	resp, err := http.DefaultClient.Do(req)
	if err != nil {
		return "", "下载素材失败，请稍后重试"
	}
	defer resp.Body.Close()
	if resp.StatusCode != http.StatusOK {
		return "", fmt.Sprintf("下载素材失败（HTTP %d）", resp.StatusCode)
	}
	limited := io.LimitReader(resp.Body, maxArticleMaterialFetchBody)
	body, err := io.ReadAll(limited)
	if err != nil {
		return "", "读取素材内容失败"
	}
	u, _ := url.Parse(strings.TrimSpace(rawURL))
	ext := ""
	if u != nil {
		ext = strings.ToLower(path.Ext(path.Base(u.Path)))
	}
	ct := strings.ToLower(resp.Header.Get("Content-Type"))
	if s, ok := bodyAsUTF8Reference(ct, ext, body); ok && strings.TrimSpace(s) != "" {
		return strings.TrimSpace(s), ""
	}
	return "", "该文件未能解析为可读文本（如 PDF/Office 请改用 .txt/.md 说明要点，或与图片一起上传由模型看图理解）"
}

func bodyAsUTF8Reference(contentType, extHint string, body []byte) (string, bool) {
	body = bytes.TrimPrefix(body, []byte{0xEF, 0xBB, 0xBF})
	ct := strings.ToLower(strings.TrimSpace(contentType))
	if strings.HasPrefix(ct, "text/") || strings.Contains(ct, "json") || ct == "" {
		if looksLikeUTF8ArticleRef(body) {
			return string(body), true
		}
	}
	switch extHint {
	case ".txt", ".md", ".json", ".csv", ".log", ".yaml", ".yml":
		if looksLikeUTF8ArticleRef(body) {
			return string(body), true
		}
	}
	if looksLikeUTF8ArticleRef(body) {
		return string(body), true
	}
	return "", false
}

func looksLikeUTF8ArticleRef(b []byte) bool {
	if len(b) == 0 {
		return false
	}
	if !utf8.Valid(b) {
		return false
	}
	s := string(b)
	printable := 0
	total := 0
	for _, r := range s {
		if r == '\n' || r == '\r' || r == '\t' {
			printable++
			total++
			continue
		}
		total++
		if unicode.IsPrint(r) || unicode.IsSpace(r) {
			printable++
		}
	}
	if total == 0 {
		return false
	}
	return float64(printable)/float64(total) > 0.90
}

// aiGenerateArticle 调用 OpenAI 兼容 API 生成文章标题和正文。
// 若有图片 URL 且模型支持视觉，以多模态格式传入；否则退化为纯文本。
func aiGenerateArticle(cfg *config.Config, description string, imageURLs []string) (title, content string, err error) {
	apiKey := strings.TrimSpace(cfg.OpenAIAPIKey)
	baseURL := strings.TrimSuffix(strings.TrimSpace(cfg.OpenAIBaseURL), "/")
	if baseURL == "" {
		baseURL = "https://api.openai.com/v1"
	}
	model := strings.TrimSpace(cfg.OpenAIModel)
	if model == "" {
		model = "gpt-4o-mini"
	}

	// 判断模型是否支持视觉（OpenAI gpt-4 系列 / claude-3 系列）
	supportsVision := len(imageURLs) > 0 && (
		strings.Contains(model, "gpt-4") ||
		strings.Contains(model, "claude-3") ||
		strings.Contains(model, "vision"))

	var userContent interface{}
	if supportsVision {
		parts := []map[string]interface{}{
			{"type": "text", "text": buildUserPrompt(description, imageURLs, true)},
		}
		for _, u := range imageURLs {
			parts = append(parts, map[string]interface{}{
				"type":      "image_url",
				"image_url": map[string]string{"url": u, "detail": "low"},
			})
		}
		userContent = parts
	} else {
		userContent = buildUserPrompt(description, imageURLs, false)
	}

	reqBody, _ := json.Marshal(map[string]interface{}{
		"model": model,
		"messages": []map[string]interface{}{
			{"role": "system", "content": aiArticleSystemPrompt},
			{"role": "user", "content": userContent},
		},
		"temperature": 0.8,
		"max_tokens":  2000,
	})

	ctx, cancel := context.WithTimeout(context.Background(), 90*time.Second)
	defer cancel()

	httpReq, err := http.NewRequestWithContext(ctx, "POST", baseURL+"/chat/completions", bytes.NewReader(reqBody))
	if err != nil {
		return "", "", fmt.Errorf("构建请求失败: %v", err)
	}
	httpReq.Header.Set("Content-Type", "application/json")
	httpReq.Header.Set("Authorization", "Bearer "+apiKey)

	resp, err := http.DefaultClient.Do(httpReq)
	if err != nil {
		return "", "", fmt.Errorf("请求 AI 服务失败: %v", err)
	}
	defer resp.Body.Close()
	body, _ := io.ReadAll(resp.Body)

	var apiResp struct {
		Choices []struct {
			Message struct {
				Content string `json:"content"`
			} `json:"message"`
		} `json:"choices"`
		Error *struct {
			Message string `json:"message"`
		} `json:"error"`
	}
	if err := json.Unmarshal(body, &apiResp); err != nil {
		return "", "", fmt.Errorf("解析 AI 响应失败")
	}
	if apiResp.Error != nil {
		return "", "", fmt.Errorf("%s", apiResp.Error.Message)
	}
	if len(apiResp.Choices) == 0 || apiResp.Choices[0].Message.Content == "" {
		return "", "", fmt.Errorf("AI 无返回内容")
	}

	return parseArticleJSON(apiResp.Choices[0].Message.Content)
}

func buildUserPrompt(description string, imageURLs []string, visionMode bool) string {
	var sb strings.Builder
	sb.WriteString("根据以下素材，创作一篇精彩的文章：\n\n")
	sb.WriteString("用户描述：")
	sb.WriteString(description)
	if len(imageURLs) > 0 {
		if visionMode {
			sb.WriteString(fmt.Sprintf("\n\n（附%d张图片，请结合图片内容写作）", len(imageURLs)))
		} else {
			sb.WriteString(fmt.Sprintf("\n\n（用户上传了%d张图片作为参考素材，请在文章中适当体现图片场景）", len(imageURLs)))
		}
	}
	sb.WriteString("\n\n请严格输出 JSON 格式：{\"title\":\"标题\",\"content\":\"正文\"}")
	return sb.String()
}

func parseArticleJSON(raw string) (title, content string, err error) {
	raw = strings.TrimSpace(raw)
	// 去掉 markdown 代码块包裹
	if idx := strings.Index(raw, "{"); idx >= 0 {
		if end := strings.LastIndex(raw, "}"); end > idx {
			raw = raw[idx : end+1]
		}
	}

	var article struct {
		Title   string `json:"title"`
		Content string `json:"content"`
	}
	if jsonErr := json.Unmarshal([]byte(raw), &article); jsonErr != nil {
		// 降级：第一行当标题，其余为正文
		lines := strings.SplitN(raw, "\n", 2)
		if len(lines) >= 2 {
			return strings.TrimSpace(lines[0]), strings.TrimSpace(lines[1]), nil
		}
		return "", "", fmt.Errorf("解析 AI 输出失败，请重试")
	}

	title = strings.TrimSpace(article.Title)
	content = strings.TrimSpace(article.Content)
	if title == "" || content == "" {
		return "", "", fmt.Errorf("AI 生成内容不完整，请重试")
	}

	// 截断超限内容
	if tr := []rune(title); len(tr) > 40 {
		title = string(tr[:40])
	}
	if cr := []rune(content); len(cr) > 5000 {
		content = string(cr[:5000])
	}

	return title, content, nil
}