Files
geo/server/internal/tenant/app/article_imitation_service.go
T

687 lines
22 KiB
Go
Raw Normal View History

package app
import (
"context"
"encoding/json"
"fmt"
"net/url"
"strings"
"time"
"github.com/jackc/pgx/v5/pgxpool"
"github.com/geo-platform/tenant-api/internal/shared/auth"
sharedcache "github.com/geo-platform/tenant-api/internal/shared/cache"
"github.com/geo-platform/tenant-api/internal/shared/config"
"github.com/geo-platform/tenant-api/internal/shared/llm"
"github.com/geo-platform/tenant-api/internal/shared/messaging/rabbitmq"
"github.com/geo-platform/tenant-api/internal/shared/response"
"github.com/geo-platform/tenant-api/internal/shared/stream"
"github.com/geo-platform/tenant-api/internal/tenant/prompts"
"github.com/geo-platform/tenant-api/internal/tenant/repository"
)
const (
articleImitationTaskType = "imitation"
articleImitationSourceType = "imitation"
articleImitationFallbackTitle = "仿写文章生成中"
maxImitationSourceContentRunes = 24000
defaultImitationWebSearchLimit = int32(5)
)
type ArticleImitationService struct {
pool *pgxpool.Pool
llm llm.Client
rabbitMQ *rabbitmq.Client
knowledge *KnowledgeService
streams *stream.GenerationHub
streamEnabled bool
articleTimeout time.Duration
maxOutputTokens int64
cache sharedcache.Cache
}
type GenerateImitationRequest struct {
SourceURL string `json:"source_url" binding:"required"`
SourceTitle string `json:"source_title"`
Locale string `json:"locale"`
Industry string `json:"industry"`
BrandName string `json:"brand_name"`
Region string `json:"region"`
TargetAudience string `json:"target_audience"`
ContentGoal string `json:"content_goal"`
Tone string `json:"tone"`
LengthGoal string `json:"length_goal"`
Keywords []string `json:"keywords"`
PreservePoints string `json:"preserve_points"`
AvoidPoints string `json:"avoid_points"`
ExtraRequirements string `json:"extra_requirements"`
EnableWebSearch bool `json:"enable_web_search"`
WebSearchLimit *int32 `json:"web_search_limit"`
KnowledgeGroupIDs []int64 `json:"knowledge_group_ids"`
}
type GenerateImitationResponse struct {
ArticleID int64 `json:"article_id"`
TaskID int64 `json:"task_id"`
}
type articleImitationJob struct {
OperatorID int64
TenantID int64
ArticleID int64
TaskID int64
ReservationID int64
InputParams map[string]interface{}
InitialTitle string
WebSearch llm.WebSearchOptions
}
func NewArticleImitationService(
pool *pgxpool.Pool,
llmClient llm.Client,
rabbitMQClient *rabbitmq.Client,
knowledge *KnowledgeService,
streams *stream.GenerationHub,
cfg config.GenerationConfig,
llmMaxOutputTokens int64,
) *ArticleImitationService {
articleTimeout := cfg.ArticleTimeout
if articleTimeout <= 0 {
articleTimeout = 8 * time.Minute
}
return &ArticleImitationService{
pool: pool,
llm: llmClient,
rabbitMQ: rabbitMQClient,
knowledge: knowledge,
streams: streams,
streamEnabled: cfg.StreamEnabled,
articleTimeout: articleTimeout,
maxOutputTokens: llmMaxOutputTokens,
}
}
func (s *ArticleImitationService) WithCache(c sharedcache.Cache) *ArticleImitationService {
s.cache = c
return s
}
func (s *ArticleImitationService) Generate(ctx context.Context, req GenerateImitationRequest) (*GenerateImitationResponse, error) {
actor := auth.MustActor(ctx)
if err := s.llm.Validate(); err != nil {
return nil, response.ErrServiceUnavailable(50311, "llm_unavailable", err.Error())
}
sourceURL, err := normalizeImitationSourceURL(req.SourceURL)
if err != nil {
return nil, err
}
quotaRepo := repository.NewQuotaRepository(s.pool)
balance, err := quotaRepo.GetCurrentBalance(ctx, actor.TenantID, "article_generation")
if err != nil || balance < 1 {
return nil, response.ErrForbidden(40301, "quota_insufficient", "generation quota insufficient, please upgrade your plan")
}
inputParams := buildImitationInputParams(sourceURL, req)
inputJSON, _ := json.Marshal(inputParams)
initialTitle := buildImitationInitialTitle(req.SourceTitle)
wizardStateJSON, _ := json.Marshal(map[string]interface{}{
"title": initialTitle,
"source_url": sourceURL,
"source_title": strings.TrimSpace(req.SourceTitle),
"current_step": 0,
})
tx, err := s.pool.Begin(ctx)
if err != nil {
return nil, response.ErrInternal(50012, "create_failed", "failed to begin generation transaction")
}
defer func() {
_ = tx.Rollback(ctx)
}()
quotaTx := repository.NewQuotaRepository(tx)
auditTx := repository.NewAuditRepository(tx)
newBalance := balance - 1
reason := "generation_reserve"
referenceType := "generation_task"
_, err = quotaTx.InsertQuotaLedger(ctx, repository.QuotaLedgerInput{
TenantID: actor.TenantID,
OperatorID: actor.UserID,
QuotaType: "article_generation",
Delta: -1,
BalanceAfter: newBalance,
Reason: &reason,
ReferenceType: &referenceType,
})
if err != nil {
return nil, response.ErrInternal(50012, "create_failed", "failed to reserve generation quota")
}
reservationID, err := quotaTx.CreateQuotaReservation(ctx, repository.QuotaReservationInput{
TenantID: actor.TenantID,
OperatorID: actor.UserID,
QuotaType: "article_generation",
ResourceType: "article",
ResourceID: nil,
ReservedAmount: 1,
ExpireAt: time.Now().Add(time.Hour),
})
if err != nil {
return nil, response.ErrInternal(50012, "create_failed", "failed to create quota reservation")
}
var articleID int64
err = tx.QueryRow(ctx, `
INSERT INTO articles (tenant_id, source_type, generate_status, publish_status, wizard_state_json)
VALUES ($1, $2, 'generating', 'unpublished', $3)
RETURNING id
`, actor.TenantID, articleImitationSourceType, wizardStateJSON).Scan(&articleID)
if err != nil {
return nil, response.ErrInternal(50012, "create_failed", "failed to create imitation article")
}
if err := quotaTx.UpdateQuotaReservationResource(ctx, reservationID, actor.TenantID, articleID); err != nil {
return nil, response.ErrInternal(50012, "create_failed", "failed to update quota reservation resource")
}
taskID, err := auditTx.CreateGenerationTask(ctx, repository.GenerationTaskInput{
OperatorID: actor.UserID,
TenantID: actor.TenantID,
ArticleID: &articleID,
QuotaReservationID: &reservationID,
TaskType: articleImitationTaskType,
InputParamsJSON: inputJSON,
})
if err != nil {
return nil, response.ErrInternal(50012, "create_failed", "failed to create generation task")
}
if err := tx.Commit(ctx); err != nil {
return nil, response.ErrInternal(50012, "create_failed", "failed to commit generation task")
}
invalidateArticleCaches(ctx, s.cache, actor.TenantID, &articleID)
job := articleImitationJob{
OperatorID: actor.UserID,
TenantID: actor.TenantID,
ArticleID: articleID,
TaskID: taskID,
ReservationID: reservationID,
InputParams: inputParams,
InitialTitle: initialTitle,
WebSearch: llm.WebSearchOptions{
Enabled: req.EnableWebSearch,
},
}
if req.WebSearchLimit != nil && *req.WebSearchLimit > 0 {
job.WebSearch.Limit = *req.WebSearchLimit
} else if req.EnableWebSearch {
job.WebSearch.Limit = defaultImitationWebSearchLimit
}
if err := publishArticleGenerationTask(ctx, s.rabbitMQ, articleGenerationTaskEnvelope{
TaskID: taskID,
TaskType: articleImitationTaskType,
TenantID: actor.TenantID,
ArticleID: articleID,
}); err != nil {
s.failGeneration(context.Background(), job, "queue_enqueue", err)
return nil, response.ErrServiceUnavailable(50310, "generation_queue_unavailable", "generation queue is busy, please retry")
}
if s.streamEnabled && s.streams != nil {
s.streams.Start(articleID, taskID, initialTitle)
}
return &GenerateImitationResponse{ArticleID: articleID, TaskID: taskID}, nil
}
func (s *ArticleImitationService) ProcessQueuedTask(ctx context.Context, task ClaimedGenerationTask) {
inputParams := cloneInputParams(task.InputParams)
job := articleImitationJob{
OperatorID: task.OperatorID,
TenantID: task.TenantID,
ArticleID: task.ArticleID,
TaskID: task.ID,
ReservationID: task.QuotaReservationID,
InputParams: inputParams,
InitialTitle: buildImitationInitialTitle(extractString(inputParams, "source_title")),
WebSearch: llm.WebSearchOptions{
Enabled: extractBool(inputParams, "enable_web_search"),
},
}
if limit := extractInt(inputParams, "web_search_limit"); limit > 0 {
job.WebSearch.Limit = int32(limit)
} else if job.WebSearch.Enabled {
job.WebSearch.Limit = defaultImitationWebSearchLimit
}
s.executeGeneration(ctx, job)
}
func (s *ArticleImitationService) HandleTaskFailure(ctx context.Context, task ClaimedGenerationTask, stage string, taskErr error) {
job := articleImitationJob{
OperatorID: task.OperatorID,
TenantID: task.TenantID,
ArticleID: task.ArticleID,
TaskID: task.ID,
ReservationID: task.QuotaReservationID,
InputParams: cloneInputParams(task.InputParams),
InitialTitle: buildImitationInitialTitle(extractString(task.InputParams, "source_title")),
}
s.failGeneration(ctx, job, stage, taskErr)
}
func (s *ArticleImitationService) executeGeneration(ctx context.Context, job articleImitationJob) {
now := time.Now()
title := strings.TrimSpace(job.InitialTitle)
if title == "" {
title = articleImitationFallbackTitle
}
auditRepo := repository.NewAuditRepository(s.pool)
_ = auditRepo.UpdateGenerationTaskStatus(ctx, repository.GenerationTaskStatusInput{
ID: job.TaskID,
TenantID: job.TenantID,
Status: "running",
StartedAt: &now,
})
sourceURL := strings.TrimSpace(extractString(job.InputParams, "source_url"))
if sourceURL == "" {
s.failGeneration(ctx, job, "task_load", fmt.Errorf("source_url is required"))
return
}
if s.knowledge == nil {
s.failGeneration(ctx, job, "source_fetch", fmt.Errorf("knowledge service is not configured"))
return
}
parsed, err := s.knowledge.parseWebsiteKnowledge(ctx, sourceURL)
if err != nil {
s.failGeneration(ctx, job, "source_fetch", err)
return
}
sourceContent := truncateRunes(strings.TrimSpace(parsed.Markdown), maxImitationSourceContentRunes)
if sourceContent == "" {
sourceContent = truncateRunes(strings.TrimSpace(parsed.Text), maxImitationSourceContentRunes)
}
if sourceContent == "" {
s.failGeneration(ctx, job, "source_fetch", fmt.Errorf("source article content is empty"))
return
}
knowledgePrompt := ""
if knowledgeGroupIDs := extractKnowledgeGroupIDs(job.InputParams["knowledge_group_ids"]); len(knowledgeGroupIDs) > 0 {
if s.knowledge == nil {
s.failGeneration(ctx, job, "knowledge_resolve", fmt.Errorf("knowledge service is not configured"))
return
}
knowledgeContext, resolveErr := s.knowledge.ResolveContext(
ctx,
job.TenantID,
knowledgeGroupIDs,
buildImitationKnowledgeQuery(job.InputParams),
)
if resolveErr != nil {
s.failGeneration(ctx, job, "knowledge_resolve", resolveErr)
return
}
knowledgePrompt = s.knowledge.RenderPromptSection(knowledgeContext)
}
prompt := buildImitationGenerationPrompt(job.InputParams, sourceContent, knowledgePrompt)
req := llm.GenerateRequest{
Prompt: prompt,
Timeout: s.articleTimeout,
MaxOutputTokens: s.maxOutputTokens,
}
if job.WebSearch.Enabled {
req.WebSearch = &job.WebSearch
}
result, err := s.llm.Generate(ctx, req, func(delta string) {
if s.streamEnabled && s.streams != nil && delta != "" {
s.streams.AppendDelta(job.ArticleID, job.TaskID, title, delta)
}
})
if err != nil {
s.failGeneration(ctx, job, "llm_generate", err)
return
}
content := strings.TrimSpace(result.Content)
if content == "" {
s.failGeneration(ctx, job, "llm_generate", fmt.Errorf("模型返回空内容"))
return
}
title = resolveGeneratedImitationTitle(content, job.InputParams)
wordCount := estimateWordCount(content)
tx, err := s.pool.Begin(ctx)
if err != nil {
s.failGeneration(ctx, job, "persist_begin_tx", err)
return
}
defer rollbackGenerationTx(tx)
failWithRollback := func(stage string, err error) {
rollbackGenerationTx(tx)
s.failGeneration(ctx, job, stage, err)
}
articleRepo := repository.NewArticleRepository(tx)
quotaRepo := repository.NewQuotaRepository(tx)
auditTx := repository.NewAuditRepository(tx)
versionID, err := articleRepo.CreateArticleVersion(ctx, repository.CreateArticleVersionInput{
ArticleID: job.ArticleID,
VersionNo: 1,
Title: title,
HTMLContent: "",
MarkdownContent: content,
WordCount: wordCount,
SourceLabel: result.Model,
})
if err != nil {
failWithRollback("persist_create_version", err)
return
}
if err := articleRepo.UpdateArticleCurrentVersion(ctx, job.ArticleID, job.TenantID, versionID); err != nil {
failWithRollback("persist_current_version", err)
return
}
if err := articleRepo.UpdateArticleGenerateStatus(ctx, job.ArticleID, job.TenantID, "completed"); err != nil {
failWithRollback("persist_article_status", err)
return
}
completed := time.Now()
if err := auditTx.UpdateGenerationTaskStatus(ctx, repository.GenerationTaskStatusInput{
ID: job.TaskID,
TenantID: job.TenantID,
Status: "completed",
StartedAt: &now,
CompletedAt: &completed,
}); err != nil {
failWithRollback("persist_task_status", err)
return
}
if err := quotaRepo.ConfirmReservation(ctx, job.ReservationID, job.TenantID); err != nil {
failWithRollback("confirm_reservation", err)
return
}
if err := tx.Commit(ctx); err != nil {
failWithRollback("persist_commit", err)
return
}
cleanupCtx, cancel := newGenerationCleanupContext()
defer cancel()
invalidateArticleCaches(cleanupCtx, s.cache, job.TenantID, &job.ArticleID)
if s.streamEnabled && s.streams != nil {
s.streams.Complete(job.ArticleID, job.TaskID, title, content)
}
}
func (s *ArticleImitationService) failGeneration(ctx context.Context, job articleImitationJob, stage string, genErr error) {
cleanupCtx, cancel := newGenerationCleanupContext()
defer cancel()
errMsg := formatGenerationFailure(stage, genErr)
now := time.Now()
title := strings.TrimSpace(job.InitialTitle)
if title == "" {
title = articleImitationFallbackTitle
}
auditRepo := repository.NewAuditRepository(s.pool)
articleRepo := repository.NewArticleRepository(s.pool)
quotaRepo := repository.NewQuotaRepository(s.pool)
_ = auditRepo.UpdateGenerationTaskStatus(cleanupCtx, repository.GenerationTaskStatusInput{
ID: job.TaskID,
TenantID: job.TenantID,
Status: "failed",
ErrorMessage: &errMsg,
CompletedAt: &now,
})
if job.ArticleID > 0 {
_ = articleRepo.UpdateArticleGenerateStatus(cleanupCtx, job.ArticleID, job.TenantID, "failed")
_, _ = s.pool.Exec(cleanupCtx, `
UPDATE articles
SET wizard_state_json = jsonb_set(COALESCE(wizard_state_json, '{}'::jsonb), '{title}', to_jsonb($1::text), true),
updated_at = NOW()
WHERE id = $2 AND tenant_id = $3
`, title, job.ArticleID, job.TenantID)
}
if job.ReservationID > 0 {
balance, balanceErr := quotaRepo.GetCurrentBalance(cleanupCtx, job.TenantID, "article_generation")
if balanceErr == nil {
reason := "generation_refund"
referenceType := "generation_task"
taskReferenceID := job.TaskID
_, _ = quotaRepo.InsertQuotaLedger(cleanupCtx, repository.QuotaLedgerInput{
TenantID: job.TenantID,
OperatorID: job.OperatorID,
QuotaType: "article_generation",
Delta: 1,
BalanceAfter: balance + 1,
Reason: &reason,
ReferenceType: &referenceType,
ReferenceID: &taskReferenceID,
})
}
_ = quotaRepo.RefundReservation(cleanupCtx, job.ReservationID, job.TenantID)
}
if job.ArticleID > 0 {
invalidateArticleCaches(cleanupCtx, s.cache, job.TenantID, &job.ArticleID)
}
if s.streamEnabled && s.streams != nil && job.ArticleID > 0 {
s.streams.Fail(job.ArticleID, job.TaskID, title, errMsg)
}
}
func normalizeImitationSourceURL(raw string) (string, error) {
trimmed := strings.TrimSpace(raw)
if trimmed == "" {
return "", response.ErrBadRequest(40001, "source_url_required", "source_url is required")
}
parsed, err := url.Parse(trimmed)
if err != nil || parsed.Scheme == "" || parsed.Host == "" {
return "", response.ErrBadRequest(40001, "invalid_source_url", "source_url must be a valid URL")
}
if parsed.Scheme != "http" && parsed.Scheme != "https" {
return "", response.ErrBadRequest(40001, "invalid_source_url", "source_url must use http or https")
}
return parsed.String(), nil
}
func buildImitationInputParams(sourceURL string, req GenerateImitationRequest) map[string]interface{} {
params := map[string]interface{}{
"source_url": sourceURL,
"source_title": strings.TrimSpace(req.SourceTitle),
"locale": normalizeImitationLocale(req.Locale),
"industry": strings.TrimSpace(req.Industry),
"brand_name": strings.TrimSpace(req.BrandName),
"region": strings.TrimSpace(req.Region),
"target_audience": strings.TrimSpace(req.TargetAudience),
"content_goal": strings.TrimSpace(req.ContentGoal),
"tone": strings.TrimSpace(req.Tone),
"length_goal": strings.TrimSpace(req.LengthGoal),
"keywords": normalizeImitationKeywords(req.Keywords),
"preserve_points": strings.TrimSpace(req.PreservePoints),
"avoid_points": strings.TrimSpace(req.AvoidPoints),
"extra_requirements": strings.TrimSpace(req.ExtraRequirements),
"enable_web_search": req.EnableWebSearch,
"knowledge_group_ids": normalizeInt64IDs(req.KnowledgeGroupIDs),
}
if req.WebSearchLimit != nil && *req.WebSearchLimit > 0 {
params["web_search_limit"] = *req.WebSearchLimit
}
return params
}
func buildImitationInitialTitle(sourceTitle string) string {
title := strings.TrimSpace(sourceTitle)
if title == "" {
return articleImitationFallbackTitle
}
return "仿写:" + title
}
func buildImitationGenerationPrompt(params map[string]interface{}, sourceContent string, knowledgePrompt string) string {
locale := normalizeImitationLocale(extractString(params, "locale"))
sourceTitle := strings.TrimSpace(extractString(params, "source_title"))
sourceURL := strings.TrimSpace(extractString(params, "source_url"))
keywords := extractStringList(params["keywords"], 16)
knowledgePrompt = strings.TrimSpace(knowledgePrompt)
var sourceMeta strings.Builder
if sourceTitle != "" {
appendRawPromptLine(&sourceMeta, "标题", sourceTitle)
}
if sourceURL != "" {
appendRawPromptLine(&sourceMeta, "URL", sourceURL)
}
var settings strings.Builder
appendPromptLine(&settings, "industry", extractString(params, "industry"))
appendPromptLine(&settings, "brand_name", extractString(params, "brand_name"))
appendPromptLine(&settings, "region", extractString(params, "region"))
appendPromptLine(&settings, "target_audience", extractString(params, "target_audience"))
appendPromptLine(&settings, "content_goal", extractString(params, "content_goal"))
appendPromptLine(&settings, "tone", extractString(params, "tone"))
appendPromptLine(&settings, "length_goal", extractString(params, "length_goal"))
if len(keywords) > 0 {
appendPromptLine(&settings, "keywords", strings.Join(keywords, "、"))
}
appendPromptLine(&settings, "preserve_points", extractString(params, "preserve_points"))
appendPromptLine(&settings, "avoid_points", extractString(params, "avoid_points"))
appendPromptLine(&settings, "extra_requirements", extractString(params, "extra_requirements"))
knowledgeSection := ""
if knowledgePrompt != "" {
knowledgeSection = "## 知识库参考\n" + knowledgePrompt
}
return prompts.ArticleImitationPrompt(
locale,
sourceMeta.String(),
settings.String(),
knowledgeSection,
sourceContent,
)
}
func buildImitationKnowledgeQuery(params map[string]interface{}) string {
if len(params) == 0 {
return ""
}
parts := make([]string, 0, 8)
appendValue := func(value string) {
value = strings.TrimSpace(value)
if value == "" {
return
}
parts = append(parts, value)
}
appendValue(extractString(params, "source_title"))
appendValue(extractString(params, "industry"))
appendValue(extractString(params, "brand_name"))
appendValue(extractString(params, "region"))
appendValue(extractString(params, "target_audience"))
appendValue(extractString(params, "content_goal"))
appendValue(extractString(params, "tone"))
appendValue(strings.Join(extractStringList(params["keywords"], 16), "\n"))
appendValue(extractString(params, "extra_requirements"))
return strings.Join(parts, "\n")
}
func appendPromptLine(b *strings.Builder, label, value string) {
value = strings.TrimSpace(value)
if value == "" {
return
}
b.WriteString("- " + prompts.ArticleImitationSettingLabel(label) + "" + value + "\n")
}
func appendRawPromptLine(b *strings.Builder, label, value string) {
value = strings.TrimSpace(value)
if value == "" {
return
}
b.WriteString("- " + label + "" + value + "\n")
}
func resolveGeneratedImitationTitle(markdown string, params map[string]interface{}) string {
for _, line := range strings.Split(markdown, "\n") {
line = strings.TrimSpace(line)
if strings.HasPrefix(line, "#") {
title := strings.TrimSpace(strings.TrimLeft(line, "#"))
if title != "" {
return title
}
}
}
return buildImitationInitialTitle(extractString(params, "source_title"))
}
func normalizeImitationLocale(value string) string {
if strings.EqualFold(strings.TrimSpace(value), "en-US") {
return "en-US"
}
return "zh-CN"
}
func normalizeImitationKeywords(values []string) []string {
if len(values) == 0 {
return []string{}
}
seen := make(map[string]struct{}, len(values))
keywords := make([]string, 0, len(values))
for _, item := range values {
item = strings.TrimSpace(item)
if item == "" {
continue
}
key := strings.ToLower(item)
if _, exists := seen[key]; exists {
continue
}
seen[key] = struct{}{}
keywords = append(keywords, item)
if len(keywords) >= 16 {
break
}
}
return keywords
}
func truncateRunes(value string, maxRunes int) string {
if maxRunes <= 0 {
return value
}
runes := []rune(value)
if len(runes) <= maxRunes {
return value
}
return string(runes[:maxRunes])
}