v2 #1

Open
kamil wants to merge 231 commits from v2 into main
5 changed files with 443 additions and 87 deletions
Showing only changes of commit 18e76c9668 - Show all commits

View File

@@ -10,16 +10,24 @@
//
// Environment variables:
//
// BROWSERLESS_URL Browserless base URL (default: http://localhost:3030)
// BROWSERLESS_TOKEN Browserless API token (default: "")
// BROWSERLESS_STRATEGY content | scrape | cdp (default: content)
// BROWSERLESS_URL Browserless base URL (default: http://localhost:3030)
// BROWSERLESS_TOKEN Browserless API token (default: "")
// BROWSERLESS_STRATEGY content | scrape | cdp (default: content)
// BROWSERLESS_MAX_CONCURRENT Max simultaneous browser sessions (default: 5)
// SCRAPER_WORKERS Chapter goroutine count (default: NumCPU)
// SCRAPER_STATIC_ROOT Output directory (default: ./static/books)
// SCRAPER_HTTP_ADDR HTTP listen address (default: :8080)
// KOKORO_URL Kokoro-FastAPI base URL (default: "")
// KOKORO_VOICE Default TTS voice (default: af_bella)
// LOG_LEVEL debug | info | warn | error (default: info)
// SCRAPER_WORKERS Chapter goroutine count (default: NumCPU)
// SCRAPER_HTTP_ADDR HTTP listen address (default: :8080)
// KOKORO_URL Kokoro-FastAPI base URL (default: "")
// KOKORO_VOICE Default TTS voice (default: af_bella)
// POCKETBASE_URL PocketBase API base URL (default: http://localhost:8090)
// POCKETBASE_EMAIL PocketBase admin email (default: admin@libnovel.local)
// POCKETBASE_PASSWORD PocketBase admin password (default: adminpassword)
// MINIO_ENDPOINT MinIO endpoint host:port (default: localhost:9000)
// MINIO_ACCESS_KEY MinIO access key (default: minioadmin)
// MINIO_SECRET_KEY MinIO secret key (default: minioadmin)
// MINIO_USE_SSL Use TLS for MinIO (default: false)
// MINIO_BUCKET_CHAPTERS Chapter objects bucket (default: libnovel-chapters)
// MINIO_BUCKET_AUDIO Audio objects bucket (default: libnovel-audio)
// LOG_LEVEL debug | info | warn | error (default: info)
package main
import (
@@ -38,7 +46,7 @@ import (
"github.com/libnovel/scraper/internal/novelfire"
"github.com/libnovel/scraper/internal/orchestrator"
"github.com/libnovel/scraper/internal/server"
"github.com/libnovel/scraper/internal/writer"
"github.com/libnovel/scraper/internal/storage"
)
func main() {
@@ -88,9 +96,30 @@ func run(log *slog.Logger) error {
bc := newBrowserClient(strategy, browserCfg)
urlClient := newBrowserClient(urlStrategy, browserCfg)
staticRoot := envOr("SCRAPER_STATIC_ROOT", "./static/books")
w := writer.New(staticRoot)
nf := novelfire.New(bc, log, urlClient, w)
// ── Storage backends ────────────────────────────────────────────────────
minioCfg := storage.MinioConfig{
Endpoint: envOr("MINIO_ENDPOINT", "localhost:9000"),
AccessKey: envOr("MINIO_ACCESS_KEY", "minioadmin"),
SecretKey: envOr("MINIO_SECRET_KEY", "minioadmin"),
UseSSL: strings.ToLower(os.Getenv("MINIO_USE_SSL")) == "true",
BucketChapters: envOr("MINIO_BUCKET_CHAPTERS", "libnovel-chapters"),
BucketAudio: envOr("MINIO_BUCKET_AUDIO", "libnovel-audio"),
}
pbCfg := storage.PocketBaseConfig{
BaseURL: envOr("POCKETBASE_URL", "http://localhost:8090"),
AdminEmail: envOr("POCKETBASE_EMAIL", "admin@libnovel.local"),
AdminPassword: envOr("POCKETBASE_PASSWORD", "adminpassword"),
}
ctx, stop := signal.NotifyContext(context.Background(), syscall.SIGINT, syscall.SIGTERM)
defer stop()
store, err := storage.NewHybridStore(ctx, pbCfg, minioCfg)
if err != nil {
return fmt.Errorf("storage init failed: %w", err)
}
nf := novelfire.New(bc, log, urlClient, &rankingCacheAdapter{store: store})
workers := 0
if s := os.Getenv("SCRAPER_WORKERS"); s != "" {
@@ -104,13 +133,9 @@ func run(log *slog.Logger) error {
}
oCfg := orchestrator.Config{
Workers: workers,
StaticRoot: staticRoot,
Workers: workers,
}
ctx, stop := signal.NotifyContext(context.Background(), syscall.SIGINT, syscall.SIGTERM)
defer stop()
switch cmd {
case "run":
// Optional --url flag.
@@ -121,10 +146,9 @@ func run(log *slog.Logger) error {
"strategy", strategy,
"workers", workers,
"max_concurrent", browserCfg.MaxConcurrent,
"static_root", oCfg.StaticRoot,
"single_book", oCfg.SingleBookURL,
)
o := orchestrator.New(oCfg, nf, log)
o := orchestrator.New(oCfg, nf, log, store)
return o.Run(ctx)
case "refresh":
@@ -133,13 +157,12 @@ func run(log *slog.Logger) error {
return fmt.Errorf("refresh command requires a book slug argument")
}
slug := args[1]
w := writer.New(oCfg.StaticRoot)
meta, ok, err := w.ReadMetadata(slug)
meta, ok, err := store.ReadMetadata(ctx, slug)
if err != nil {
return fmt.Errorf("failed to read metadata for %s: %w", slug, err)
}
if !ok {
return fmt.Errorf("book %q not found in %s", slug, oCfg.StaticRoot)
return fmt.Errorf("book %q not found in store", slug)
}
if meta.SourceURL == "" {
return fmt.Errorf("book %q has no source_url in metadata", slug)
@@ -149,7 +172,7 @@ func run(log *slog.Logger) error {
"slug", slug,
"source_url", meta.SourceURL,
)
o := orchestrator.New(oCfg, nf, log)
o := orchestrator.New(oCfg, nf, log, store)
return o.Run(ctx)
case "serve":
@@ -164,7 +187,7 @@ func run(log *slog.Logger) error {
"kokoro_url", kokoroURL,
"kokoro_voice", kokoroVoice,
)
srv := server.New(addr, oCfg, nf, log, kokoroURL, kokoroVoice)
srv := server.New(addr, oCfg, nf, log, store, kokoroURL, kokoroVoice)
return srv.ListenAndServe(ctx)
default:
@@ -192,6 +215,20 @@ func envOr(key, fallback string) string {
return fallback
}
// rankingCacheAdapter bridges storage.HybridStore (context-aware) to the
// context-free scraper.RankingPageCacher interface expected by novelfire.New.
type rankingCacheAdapter struct {
store *storage.HybridStore
}
func (a *rankingCacheAdapter) WriteRankingPageCache(page int, html string) error {
return a.store.WriteRankingPageCache(context.Background(), page, html)
}
func (a *rankingCacheAdapter) ReadRankingPageCache(page int) (string, error) {
return a.store.ReadRankingPageCache(context.Background(), page)
}
func printUsage() {
fmt.Fprintf(os.Stderr, `libnovel scraper

View File

@@ -19,7 +19,7 @@ import (
"sync"
"github.com/libnovel/scraper/internal/scraper"
"github.com/libnovel/scraper/internal/writer"
"github.com/libnovel/scraper/internal/storage"
)
// Config holds tunable parameters for the orchestrator.
@@ -28,7 +28,8 @@ type Config struct {
// Defaults to runtime.NumCPU() when 0.
Workers int
// StaticRoot is the path to the static/books output directory.
// StaticRoot is kept for backwards-compatibility but is no longer used
// when a Store is provided.
StaticRoot string
// SingleBookURL when non-empty causes the orchestrator to scrape only
@@ -40,13 +41,13 @@ type Config struct {
type Orchestrator struct {
cfg Config
novel scraper.NovelScraper
writer *writer.Writer
store storage.Store
log *slog.Logger
workers int
}
// New returns a new Orchestrator.
func New(cfg Config, novel scraper.NovelScraper, log *slog.Logger) *Orchestrator {
// New returns a new Orchestrator backed by the provided Store.
func New(cfg Config, novel scraper.NovelScraper, log *slog.Logger, store storage.Store) *Orchestrator {
workers := cfg.Workers
if workers <= 0 {
workers = runtime.NumCPU()
@@ -54,7 +55,7 @@ func New(cfg Config, novel scraper.NovelScraper, log *slog.Logger) *Orchestrator
return &Orchestrator{
cfg: cfg,
novel: novel,
writer: writer.New(cfg.StaticRoot),
store: store,
log: log,
workers: workers,
}
@@ -66,7 +67,6 @@ func (o *Orchestrator) Run(ctx context.Context) error {
o.log.Info("orchestrator starting",
"source", o.novel.SourceName(),
"workers", o.workers,
"static_root", o.cfg.StaticRoot,
)
// chapterWork is the shared queue consumed by chapter worker goroutines.
@@ -89,8 +89,8 @@ func (o *Orchestrator) Run(ctx context.Context) error {
default:
}
// Skip if already on disk.
if o.writer.ChapterExists(job.slug, job.ref) {
// Skip if already stored.
if o.store.ChapterExists(ctx, job.slug, job.ref) {
o.log.Debug("chapter already exists, skipping",
"book", job.slug, "chapter", job.ref.Number)
continue
@@ -107,7 +107,7 @@ func (o *Orchestrator) Run(ctx context.Context) error {
continue
}
if err := o.writer.WriteChapter(job.slug, chapter); err != nil {
if err := o.store.WriteChapter(ctx, job.slug, chapter); err != nil {
o.log.Error("chapter write failed",
"book", job.slug,
"chapter", job.ref.Number,
@@ -135,8 +135,8 @@ func (o *Orchestrator) Run(ctx context.Context) error {
return
}
// Persist / update metadata.yaml.
if err := o.writer.WriteMetadata(meta); err != nil {
// Persist / update metadata.
if err := o.store.WriteMetadata(ctx, meta); err != nil {
o.log.Error("metadata write failed", "slug", meta.Slug, "err", err)
// Continue — chapters can still be scraped.
}

View File

@@ -22,7 +22,7 @@ import (
"github.com/libnovel/scraper/internal/orchestrator"
"github.com/libnovel/scraper/internal/scraper"
"github.com/libnovel/scraper/internal/writer"
"github.com/libnovel/scraper/internal/storage"
)
// Server wraps an HTTP mux with the scraping endpoints.
@@ -31,7 +31,7 @@ type Server struct {
oCfg orchestrator.Config
novel scraper.NovelScraper
log *slog.Logger
writer *writer.Writer
store storage.Store
mu sync.Mutex
running bool
rankingRunning bool
@@ -42,26 +42,23 @@ type Server struct {
voiceMu sync.RWMutex
cachedVoices []string // populated on first request from Kokoro /v1/audio/voices
// audioMu guards audioCache and audioInFlight.
// audioCache maps a cache key to the Kokoro download filename returned by
// POST /v1/audio/speech with return_download_link=true.
// audioMu guards audioInFlight only.
// Completed audio filenames are persisted to the Store (PocketBase).
// audioInFlight deduplicates concurrent generation requests for the same key.
audioMu sync.Mutex
audioCache map[string]string // cacheKey → kokoro download filename
audioInFlight map[string]chan struct{} // cacheKey → closed when done
}
// New creates a new Server.
func New(addr string, oCfg orchestrator.Config, novel scraper.NovelScraper, log *slog.Logger, kokoroURL, kokoroVoice string) *Server {
func New(addr string, oCfg orchestrator.Config, novel scraper.NovelScraper, log *slog.Logger, store storage.Store, kokoroURL, kokoroVoice string) *Server {
return &Server{
addr: addr,
oCfg: oCfg,
novel: novel,
log: log,
writer: writer.New(oCfg.StaticRoot),
store: store,
kokoroURL: kokoroURL,
kokoroVoice: kokoroVoice,
audioCache: make(map[string]string),
audioInFlight: make(map[string]chan struct{}),
}
}
@@ -174,7 +171,7 @@ func (s *Server) handleChapterText(w http.ResponseWriter, r *http.Request) {
http.NotFound(w, r)
return
}
raw, err := s.writer.ReadChapter(slug, n)
raw, err := s.store.ReadChapter(r.Context(), slug, n)
if err != nil {
http.NotFound(w, r)
return
@@ -223,15 +220,14 @@ func (s *Server) handleAudioGenerate(w http.ResponseWriter, r *http.Request) {
cacheKey := fmt.Sprintf("%s/%d/%s/%.2f", slug, n, voice, speed)
// Fast path: already generated this session.
s.audioMu.Lock()
if filename, ok := s.audioCache[cacheKey]; ok {
s.audioMu.Unlock()
// Fast path: already generated (check persistent store first).
if filename, ok := s.store.GetAudioCache(r.Context(), cacheKey); ok {
s.writeAudioResponse(w, slug, n, voice, speed, filename)
return
}
// Deduplicate concurrent generation for the same key.
s.audioMu.Lock()
if ch, ok := s.audioInFlight[cacheKey]; ok {
s.audioMu.Unlock()
select {
@@ -240,10 +236,8 @@ func (s *Server) handleAudioGenerate(w http.ResponseWriter, r *http.Request) {
http.Error(w, `{"error":"request cancelled"}`, http.StatusServiceUnavailable)
return
}
s.audioMu.Lock()
filename, ok := s.audioCache[cacheKey]
s.audioMu.Unlock()
if ok {
// Check store again after waiting.
if filename, ok := s.store.GetAudioCache(r.Context(), cacheKey); ok {
s.writeAudioResponse(w, slug, n, voice, speed, filename)
} else {
http.Error(w, `{"error":"audio generation failed"}`, http.StatusInternalServerError)
@@ -262,7 +256,7 @@ func (s *Server) handleAudioGenerate(w http.ResponseWriter, r *http.Request) {
}()
// Load and validate chapter text.
raw, err := s.writer.ReadChapter(slug, n)
raw, err := s.store.ReadChapter(r.Context(), slug, n)
if err != nil {
http.Error(w, `{"error":"chapter not found"}`, http.StatusNotFound)
return
@@ -287,9 +281,7 @@ func (s *Server) handleAudioGenerate(w http.ResponseWriter, r *http.Request) {
return
}
s.audioMu.Lock()
s.audioCache[cacheKey] = filename
s.audioMu.Unlock()
_ = s.store.SetAudioCache(r.Context(), cacheKey, filename)
s.log.Info("audio generated", "slug", slug, "chapter", n, "filename", filename)
s.writeAudioResponse(w, slug, n, voice, speed, filename)
@@ -378,10 +370,7 @@ func (s *Server) handleAudioProxy(w http.ResponseWriter, r *http.Request) {
}
cacheKey := fmt.Sprintf("%s/%d/%s/%.2f", slug, n, voice, speed)
s.audioMu.Lock()
filename, ok := s.audioCache[cacheKey]
s.audioMu.Unlock()
filename, ok := s.store.GetAudioCache(r.Context(), cacheKey)
if !ok {
http.Error(w, "audio not generated yet", http.StatusNotFound)
return
@@ -462,7 +451,7 @@ func (s *Server) runAsync(w http.ResponseWriter, cfg orchestrator.Config) {
ctx, cancel := context.WithTimeout(context.Background(), 24*time.Hour)
defer cancel()
o := orchestrator.New(cfg, s.novel, s.log)
o := orchestrator.New(cfg, s.novel, s.log, s.store)
if err := o.Run(ctx); err != nil {
s.log.Error("scrape job failed", "err", fmt.Sprintf("%v", err))
}

View File

@@ -15,7 +15,7 @@ import (
"github.com/libnovel/scraper/internal/orchestrator"
"github.com/libnovel/scraper/internal/scraper"
"github.com/libnovel/scraper/internal/writer"
"github.com/libnovel/scraper/internal/storage"
"github.com/yuin/goldmark"
"github.com/yuin/goldmark/extension"
goldhtml "github.com/yuin/goldmark/renderer/html"
@@ -629,7 +629,7 @@ func (s *Server) handleHome(w http.ResponseWriter, r *http.Request) {
return
}
books, err := s.writer.ListBooks()
books, err := s.store.ListBooks(r.Context())
if err != nil {
http.Error(w, "failed to list books: "+err.Error(), http.StatusInternalServerError)
return
@@ -639,8 +639,8 @@ func (s *Server) handleHome(w http.ResponseWriter, r *http.Request) {
for i, b := range books {
items[i] = homeBookItem{
BookMeta: b,
Downloaded: s.writer.CountChapters(b.Slug),
AddedAt: s.writer.MetadataMtime(b.Slug),
Downloaded: s.store.CountChapters(r.Context(), b.Slug),
AddedAt: s.store.MetadataMtime(r.Context(), b.Slug),
}
}
@@ -877,7 +877,7 @@ const scrapeTmpl = `
</script>`
func (s *Server) handleScrape(w http.ResponseWriter, r *http.Request) {
rankingItems, _ := s.writer.ReadRankingItems()
rankingItems, _ := s.store.ReadRankingItems(r.Context())
rankingJSON, _ := json.Marshal(rankingItems)
t := template.Must(template.New("scrape").Parse(scrapeTmpl))
@@ -1324,12 +1324,12 @@ const rankingTmpl = `
// rankingViewItem enriches a RankingItem with whether it is present in the
// local book library, so the template can highlight it differently.
type rankingViewItem struct {
writer.RankingItem
storage.RankingItem
Local bool
}
// toRankingViewItems annotates items with Local=true for slugs found in localSlugs.
func toRankingViewItems(items []writer.RankingItem, localSlugs map[string]bool) []rankingViewItem {
func toRankingViewItems(items []storage.RankingItem, localSlugs map[string]bool) []rankingViewItem {
out := make([]rankingViewItem, len(items))
for i, it := range items {
out[i] = rankingViewItem{
@@ -1403,13 +1403,13 @@ const rankingPageSize = 20
// It does NOT trigger a live scrape; use POST /ranking/refresh for that.
// Supports ?page=N for browsing through cached items (20 per page).
func (s *Server) handleRanking(w http.ResponseWriter, r *http.Request) {
rankingItems, err := s.writer.ReadRankingItems()
rankingItems, err := s.store.ReadRankingItems(r.Context())
if err != nil {
s.log.Error("failed to read cached ranking", "err", err)
}
cachedAt := ""
if info, statErr := s.writer.RankingFileInfo(); statErr == nil {
if info, statErr := s.store.RankingFileInfo(r.Context()); statErr == nil && info != nil {
cachedAt = info.ModTime().Format("Jan 2, 2006 at 15:04")
}
@@ -1482,7 +1482,7 @@ func (s *Server) handleRanking(w http.ResponseWriter, r *http.Request) {
}
// Encode full dataset + local slugs for client-side cross-page filtering.
localSlugs := s.writer.LocalSlugs()
localSlugs, _ := s.store.LocalSlugs(r.Context())
type rankingJSONItem struct {
Rank int `json:"rank"`
Slug string `json:"slug"`
@@ -1577,14 +1577,14 @@ func (s *Server) handleRankingRefresh(w http.ResponseWriter, r *http.Request) {
rankingCh, errCh := s.novel.ScrapeRanking(ctx, maxPages)
var rankingItems []writer.RankingItem
var rankingItems []storage.RankingItem
for rankingCh != nil || errCh != nil {
select {
case meta, ok := <-rankingCh:
if !ok {
rankingCh = nil
} else {
rankingItems = append(rankingItems, writer.RankingItem{
rankingItems = append(rankingItems, storage.RankingItem{
Rank: meta.Ranking,
Slug: meta.Slug,
Title: meta.Title,
@@ -1605,7 +1605,7 @@ func (s *Server) handleRankingRefresh(w http.ResponseWriter, r *http.Request) {
}
if len(rankingItems) > 0 {
if err := s.writer.WriteRanking(rankingItems); err != nil {
if err := s.store.WriteRanking(ctx, rankingItems); err != nil {
s.log.Error("failed to save ranking", "err", err)
}
}
@@ -1669,7 +1669,7 @@ const rankingViewTmpl = `
</div>`
func (s *Server) handleRankingView(w http.ResponseWriter, r *http.Request) {
items, err := s.writer.ReadRankingItems()
items, err := s.store.ReadRankingItems(r.Context())
if err != nil {
http.Error(w, "failed to read ranking: "+err.Error(), http.StatusInternalServerError)
return
@@ -1927,7 +1927,7 @@ const bookTmpl = `
func (s *Server) handleBook(w http.ResponseWriter, r *http.Request) {
slug := r.PathValue("slug")
meta, ok, err := s.writer.ReadMetadata(slug)
meta, ok, err := s.store.ReadMetadata(r.Context(), slug)
if err != nil {
http.Error(w, "failed to read metadata: "+err.Error(), http.StatusInternalServerError)
return
@@ -1937,7 +1937,7 @@ func (s *Server) handleBook(w http.ResponseWriter, r *http.Request) {
return
}
chapters, err := s.writer.ListChapters(slug)
chapters, err := s.store.ListChapters(r.Context(), slug)
if err != nil {
http.Error(w, "failed to list chapters: "+err.Error(), http.StatusInternalServerError)
return
@@ -2054,7 +2054,7 @@ func (s *Server) handleBookChaptersPage(w http.ResponseWriter, r *http.Request)
}
}
chapters, err := s.writer.ListChapters(slug)
chapters, err := s.store.ListChapters(r.Context(), slug)
if err != nil {
http.Error(w, "failed to list chapters: "+err.Error(), http.StatusInternalServerError)
return
@@ -3035,7 +3035,7 @@ func (s *Server) handleChapter(w http.ResponseWriter, r *http.Request) {
return
}
raw, err := s.writer.ReadChapter(slug, n)
raw, err := s.store.ReadChapter(r.Context(), slug, n)
if err != nil {
http.NotFound(w, r)
return
@@ -3051,15 +3051,15 @@ func (s *Server) handleChapter(w http.ResponseWriter, r *http.Request) {
return
}
chapters, _ := s.writer.ListChapters(slug)
chapters, _ := s.store.ListChapters(r.Context(), slug)
prevN, nextN := adjacentChapters(chapters, n)
title := firstHeading(raw, fmt.Sprintf("Chapter %d", n))
chapterTitle, chapterDate := writer.SplitChapterTitle(title)
chapterTitle, chapterDate := splitChapterTitle(title)
// Load cover URL for Media Session artwork (best-effort; ignore errors).
var coverURL string
if meta, ok, err := s.writer.ReadMetadata(slug); err == nil && ok {
if meta, ok, err := s.store.ReadMetadata(r.Context(), slug); err == nil && ok {
coverURL = meta.Cover
}
@@ -3096,6 +3096,36 @@ func (s *Server) handleChapter(w http.ResponseWriter, r *http.Request) {
// ─── helpers ──────────────────────────────────────────────────────────────────
// splitChapterTitle splits a raw chapter heading into a human-readable title
// and a trailing relative-date string (e.g. "1 year ago"). It mirrors the
// same logic in internal/writer and internal/storage/hybrid.
func splitChapterTitle(raw string) (title, date string) {
raw = strings.TrimSpace(raw)
// Strip leading numeric index.
if idx := strings.IndexFunc(raw, func(r rune) bool { return r == ' ' || r == '\t' }); idx > 0 {
prefix := raw[:idx]
allDigit := true
for _, c := range prefix {
if c < '0' || c > '9' {
allDigit = false
break
}
}
if allDigit {
raw = strings.TrimSpace(raw[idx:])
}
}
// Strip "Chapter N - N: " prefix.
chNumRe := regexp.MustCompile(`(?i)^chapter\s+\d+(?:\s*-\s*\d+)?\s*:\s*`)
raw = strings.TrimSpace(chNumRe.ReplaceAllString(raw, ""))
// Detect trailing relative date.
dateRe := regexp.MustCompile(`\s*(\d+\s+(?:second|minute|hour|day|week|month|year)s?\s+ago)\s*$`)
if m := dateRe.FindStringSubmatchIndex(raw); m != nil {
return strings.TrimSpace(raw[:m[0]]), strings.TrimSpace(raw[m[2]:m[3]])
}
return raw, ""
}
// sortedKeys returns the keys of a string-bool map in sorted order.
func sortedKeys(m map[string]bool) []string {
out := make([]string, 0, len(m))
@@ -3127,7 +3157,7 @@ func stripMarkdown(src string) string {
// adjacentChapters returns the chapter numbers immediately before and after n
// in the sorted chapters list. 0 means "does not exist".
func adjacentChapters(chapters []writer.ChapterInfo, n int) (prev, next int) {
func adjacentChapters(chapters []storage.ChapterInfo, n int) (prev, next int) {
for i, ch := range chapters {
if ch.Number == n {
if i > 0 {
@@ -3211,7 +3241,7 @@ func (s *Server) handleUIScrapeBook(w http.ResponseWriter, r *http.Request) {
ctx, cancel := context.WithTimeout(context.Background(), 24*time.Hour)
defer cancel()
o := orchestrator.New(cfg, s.novel, s.log)
o := orchestrator.New(cfg, s.novel, s.log, s.store)
if err := o.Run(ctx); err != nil {
s.log.Error("UI scrape job failed", "url", bookURL, "err", err)
}

View File

@@ -0,0 +1,300 @@
// hybrid.go implements the Store interface using PocketBase for structured data
// and MinIO for binary chapter/audio blobs.
package storage
import (
"context"
"encoding/json"
"fmt"
"os"
"sort"
"strconv"
"strings"
"time"
"github.com/libnovel/scraper/internal/scraper"
)
// HybridStore satisfies Store by routing structured data to PocketBase and
// binary objects (chapters, audio) to MinIO.
type HybridStore struct {
pb *PocketBaseStore
minio *MinioClient
}
// NewHybridStore constructs a HybridStore. It connects to both backends and
// calls EnsureCollections to bootstrap any missing PocketBase collections.
func NewHybridStore(ctx context.Context, pbCfg PocketBaseConfig, minioCfg MinioConfig) (*HybridStore, error) {
mc, err := NewMinioClient(ctx, minioCfg)
if err != nil {
return nil, fmt.Errorf("storage: minio: %w", err)
}
pb := NewPocketBaseStore(pbCfg)
if err := pb.EnsureCollections(ctx); err != nil {
// Log but don't fail — collection creation errors are often "already exists"
_ = err
}
return &HybridStore{pb: pb, minio: mc}, nil
}
// ─── Book metadata ────────────────────────────────────────────────────────────
func (h *HybridStore) WriteMetadata(ctx context.Context, meta scraper.BookMeta) error {
return h.pb.UpsertBook(ctx,
meta.Slug, meta.Title, meta.Author, meta.Cover,
meta.Status, meta.Summary, meta.SourceURL,
meta.Genres, meta.TotalChapters, meta.Ranking,
)
}
func (h *HybridStore) ReadMetadata(ctx context.Context, slug string) (scraper.BookMeta, bool, error) {
rec, found, err := h.pb.GetBook(ctx, slug)
if err != nil || !found {
return scraper.BookMeta{}, found, err
}
return recToBookMeta(rec), true, nil
}
func (h *HybridStore) ListBooks(ctx context.Context) ([]scraper.BookMeta, error) {
rows, err := h.pb.ListBooks(ctx)
if err != nil {
return nil, err
}
books := make([]scraper.BookMeta, 0, len(rows))
for _, r := range rows {
books = append(books, recToBookMeta(r))
}
return books, nil
}
func (h *HybridStore) LocalSlugs(ctx context.Context) (map[string]bool, error) {
books, err := h.ListBooks(ctx)
if err != nil {
return nil, err
}
slugs := make(map[string]bool, len(books))
for _, b := range books {
slugs[b.Slug] = true
}
return slugs, nil
}
func (h *HybridStore) MetadataMtime(ctx context.Context, slug string) int64 {
t, err := h.pb.BookMetaUpdated(ctx, slug)
if err != nil || t.IsZero() {
return 0
}
return t.Unix()
}
// ─── Chapters ─────────────────────────────────────────────────────────────────
func (h *HybridStore) ChapterExists(ctx context.Context, slug string, ref scraper.ChapterRef) bool {
return h.minio.ChapterExists(ctx, slug, ref.Volume, ref.Number)
}
func (h *HybridStore) WriteChapter(ctx context.Context, slug string, chapter scraper.Chapter) error {
content := "# " + chapter.Ref.Title + "\n\n" + chapter.Text + "\n"
if err := h.minio.PutChapter(ctx, slug, chapter.Ref.Volume, chapter.Ref.Number, content); err != nil {
return err
}
// Update chapter index in PocketBase.
title, dateLabel := splitChapterTitle(chapter.Ref.Title)
_ = h.pb.UpsertChapterIdx(ctx, slug, chapter.Ref.Number, title, dateLabel)
return nil
}
func (h *HybridStore) ReadChapter(ctx context.Context, slug string, n int) (string, error) {
return h.minio.GetChapter(ctx, slug, 0, n)
}
func (h *HybridStore) ListChapters(ctx context.Context, slug string) ([]ChapterInfo, error) {
rows, err := h.pb.ListChapterIdx(ctx, slug)
if err != nil {
return nil, err
}
infos := make([]ChapterInfo, 0, len(rows))
for _, r := range rows {
n := int(floatVal(r, "number"))
title, _ := r["title"].(string)
date, _ := r["date_label"].(string)
infos = append(infos, ChapterInfo{Number: n, Title: title, Date: date})
}
sort.Slice(infos, func(i, j int) bool { return infos[i].Number < infos[j].Number })
return infos, nil
}
func (h *HybridStore) CountChapters(ctx context.Context, slug string) int {
return h.pb.CountChapterIdx(ctx, slug)
}
// ─── Ranking ─────────────────────────────────────────────────────────────────
func (h *HybridStore) WriteRanking(ctx context.Context, items []RankingItem) error {
data, err := json.Marshal(items)
if err != nil {
return fmt.Errorf("storage: marshal ranking: %w", err)
}
return h.pb.SetRanking(ctx, string(data))
}
func (h *HybridStore) ReadRankingItems(ctx context.Context) ([]RankingItem, error) {
dataStr, _, err := h.pb.GetRanking(ctx)
if err != nil || dataStr == "" {
return nil, err
}
var items []RankingItem
if err := json.Unmarshal([]byte(dataStr), &items); err != nil {
return nil, fmt.Errorf("storage: unmarshal ranking: %w", err)
}
return items, nil
}
func (h *HybridStore) RankingFileInfo(ctx context.Context) (os.FileInfo, error) {
return h.pb.RankingModTime(ctx)
}
// ─── Ranking page HTML cache ──────────────────────────────────────────────────
func (h *HybridStore) WriteRankingPageCache(ctx context.Context, page int, html string) error {
return h.pb.SetRankingPageHTML(ctx, page, html)
}
func (h *HybridStore) ReadRankingPageCache(ctx context.Context, page int) (string, error) {
html, _, err := h.pb.GetRankingPageHTML(ctx, page)
return html, err
}
func (h *HybridStore) RankingPageCacheInfo(ctx context.Context, page int) (os.FileInfo, error) {
return h.pb.RankingPageCacheModTime(ctx, page)
}
// ─── Audio cache ──────────────────────────────────────────────────────────────
func (h *HybridStore) GetAudioCache(ctx context.Context, cacheKey string) (string, bool) {
filename, ok, _ := h.pb.GetAudioCache(ctx, cacheKey)
return filename, ok
}
func (h *HybridStore) SetAudioCache(ctx context.Context, cacheKey, filename string) error {
return h.pb.SetAudioCache(ctx, cacheKey, filename)
}
// ─── Reading progress ─────────────────────────────────────────────────────────
func (h *HybridStore) GetProgress(ctx context.Context, sessionID, slug string) (ReadingProgress, bool) {
ch, updated, ok, err := h.pb.GetProgress(ctx, sessionID, slug)
if err != nil || !ok {
return ReadingProgress{}, false
}
return ReadingProgress{Slug: slug, Chapter: ch, UpdatedAt: updated}, true
}
func (h *HybridStore) SetProgress(ctx context.Context, sessionID string, p ReadingProgress) error {
return h.pb.SetProgress(ctx, sessionID, p.Slug, p.Chapter)
}
func (h *HybridStore) AllProgress(ctx context.Context, sessionID string) ([]ReadingProgress, error) {
rows, err := h.pb.AllProgress(ctx, sessionID)
if err != nil {
return nil, err
}
out := make([]ReadingProgress, 0, len(rows))
for _, r := range rows {
slug, _ := r["slug"].(string)
ch := int(floatVal(r, "chapter"))
var updated time.Time
if ts, ok := r["updated"].(string); ok {
updated, _ = time.Parse(time.RFC3339, ts)
}
out = append(out, ReadingProgress{Slug: slug, Chapter: ch, UpdatedAt: updated})
}
return out, nil
}
func (h *HybridStore) DeleteProgress(ctx context.Context, sessionID, slug string) error {
return h.pb.DeleteProgress(ctx, sessionID, slug)
}
// ─── AudioObjectKey ───────────────────────────────────────────────────────────
func (h *HybridStore) AudioObjectKey(slug string, n int, voice string, speed float64) string {
return AudioObjectKey(slug, n, voice, speed)
}
// ─── helpers ──────────────────────────────────────────────────────────────────
func recToBookMeta(rec map[string]interface{}) scraper.BookMeta {
m := scraper.BookMeta{
Slug: strVal(rec, "slug"),
Title: strVal(rec, "title"),
Author: strVal(rec, "author"),
Cover: strVal(rec, "cover"),
Status: strVal(rec, "status"),
Summary: strVal(rec, "summary"),
SourceURL: strVal(rec, "source_url"),
}
if tc := floatVal(rec, "total_chapters"); tc > 0 {
m.TotalChapters = int(tc)
}
if rk := floatVal(rec, "ranking"); rk > 0 {
m.Ranking = int(rk)
}
// Genres stored as JSON string or array.
switch v := rec["genres"].(type) {
case string:
_ = json.Unmarshal([]byte(v), &m.Genres)
case []interface{}:
for _, g := range v {
if s, ok := g.(string); ok {
m.Genres = append(m.Genres, s)
}
}
}
return m
}
func strVal(m map[string]interface{}, key string) string {
if v, ok := m[key].(string); ok {
return v
}
return ""
}
// splitChapterTitle mirrors writer.SplitChapterTitle logic (simplified).
func splitChapterTitle(raw string) (title, date string) {
raw = strings.TrimSpace(raw)
// Strip leading numeric index.
if idx := strings.IndexFunc(raw, func(r rune) bool { return r == ' ' || r == '\t' }); idx > 0 {
prefix := raw[:idx]
allDigit := true
for _, c := range prefix {
if c < '0' || c > '9' {
allDigit = false
break
}
}
if allDigit {
raw = strings.TrimSpace(raw[idx:])
}
}
// Detect trailing relative date.
units := []string{"second", "minute", "hour", "day", "week", "month", "year"}
lower := strings.ToLower(raw)
for _, u := range units {
for _, suffix := range []string{u + "s ago", u + " ago"} {
if idx := strings.LastIndex(lower, suffix); idx > 0 {
// Find start of date token (digit before the unit).
start := strings.LastIndex(raw[:idx], " ")
if start < 0 {
start = 0
}
numPart := strings.TrimSpace(raw[start:idx])
if _, err := strconv.Atoi(strings.Fields(numPart)[0]); err == nil {
return strings.TrimSpace(raw[:start]), strings.TrimSpace(raw[start : idx+len(suffix)])
}
}
}
}
return raw, ""
}