AI+Go 打造你的智能办公助手|果fx

+-------------------+     +-------------------+     +-------------------+|   用户界面层      | <-> |   应用服务层      | <-> |   数据存储层      |+-------------------+     +-------------------+     +-------------------+        ^                           |                           |        |                           v                           v+-------------------+     +-------------------+     +-------------------+|   用户交互模块    |     |   文档处理模块    |     |   文档存储模块    ||   （Web/API）     |     |   （LangChain）   |     |   （数据库/文件） |+-------------------+     +-------------------+     +-------------------+

gopackage main import (	"context"	"fmt"	"log" 	"github.com/sashabaranov/go-openai"	"github.com/tmc/langchaingo/documentloaders"	"github.com/tmc/langchaingo/schema") func loadPDFDocument(filePath string) ([]schema.Document, error) {	loader := documentloaders.NewUnstructuredPDFLoader(filePath)	docs, err := loader.Load(context.Background())	if err != nil {		return nil, fmt.Errorf("failed to load PDF document: %v", err)	}	return docs, nil} func main() {	docs, err := loadPDFDocument("example.pdf")	if err != nil {		log.Fatal(err)	}	for _, doc := range docs {		fmt.Printf("Document content: %s\n", doc.PageContent)	}}

gopackage main import (	"context"	"fmt"	"log" 	"github.com/tmc/langchaingo/documentloaders"	"github.com/tmc/langchaingo/schema"	"github.com/tmc/langchaingo/textsplitters") func splitText(text string, chunkSize int) ([]schema.Document, error) {	splitter := textsplitters.NewCharacterTextSplitter(chunkSize)	docs, err := splitter.SplitDocuments(context.Background(), []schema.Document{{PageContent: text}})	if err != nil {		return nil, fmt.Errorf("failed to split text: %v", err)	}	return docs, nil} func main() {	longText := "This is a long text that needs to be split into smaller chunks for better processing. " +		"Each chunk should have a reasonable size to ensure efficient vectorization and similarity search. " +		"..." // 假设这是一个很长的文本 	docs, err := splitText(longText, 500) // 按 500 个字符分割	if err != nil {		log.Fatal(err)	}	for i, doc := range docs {		fmt.Printf("Chunk %d content: %s\n", i+1, doc.PageContent)	}}

gopackage main import (	"context"	"fmt"	"log" 	"github.com/sashabaranov/go-openai"	"github.com/tmc/langchaingo/documentloaders"	"github.com/tmc/langchaingo/embeddings"	"github.com/tmc/langchaingo/schema"	"github.com/tmc/langchaingo/stores/vectorstore"	"github.com/tmc/langchaingo/stores/vectorstore/memstore"	"github.com/tmc/langchaingo/textsplitters") func main() {	// 1. 加载文档	loader := documentloaders.NewUnstructuredPDFLoader("example.pdf")	docs, err := loader.Load(context.Background())	if err != nil {		log.Fatal(err)	} 	// 2. 文本分割	splitter := textsplitters.NewCharacterTextSplitter(500)	splitDocs, err := splitter.SplitDocuments(context.Background(), docs)	if err != nil {		log.Fatal(err)	} 	// 3. 向量化（这里使用 OpenAI 的嵌入模型作为示例，实际项目中可以使用其他向量模型）	client := openai.NewClient("your-openai-api-key")	embeddingsClient := embeddings.NewOpenAIEmbeddings(client)	embeddings, err := embeddingsClient.EmbedDocuments(context.Background(), splitDocs)	if err != nil {		log.Fatal(err)	} 	// 4. 构建向量存储（这里使用内存存储作为示例，实际项目中可以使用向量数据库）	store := memstore.NewVectorStore()	for i, doc := range splitDocs {		err = store.Add(context.Background(), string(embeddings[i]), doc.Metadata, doc.PageContent)		if err != nil {			log.Fatal(err)		}	} 	// 5. 用户提问	question := "What is the main idea of the document?"	questionEmbedding, err := embeddingsClient.EmbedQuery(context.Background(), question)	if err != nil {		log.Fatal(err)	} 	// 6. 相似度搜索	results, err := store.SimilaritySearch(context.Background(), string(questionEmbedding), 3) // 搜索最相似的 3 个文档	if err != nil {		log.Fatal(err)	} 	// 7. 生成回答（这里简单地将相关文档内容拼接作为回答，实际项目中可以使用更复杂的链式处理）	answer := ""	for _, result := range results {		answer += result.PageContent + "\n\n"	}	fmt.Printf("Answer: %s\n", answer)}