fix: ranking refresh deadlock, switch to direct HTTP, add CI pipeline
- Fix channel drain goroutine deadlock in handleRankingRefresh: replace
'for { ... continue ... if nil break }' with 'for A != nil || B != nil'
so the select never blocks on two nil channels
- Switch ScrapeRanking from urlClient (Browserless) to client (direct HTTP)
since novelfire.net/ranking is fully server-rendered — no JS needed
- Add ranking unit tests (single page, multi-page, empty page, write round-trip)
- Add .gitea/workflows/ci.yaml: lint, test, build jobs with commented-out
Docker image push step for when runner has Docker available
This commit is contained in:
Admin
106
.gitea/workflows/ci.yaml
Normal file
106
.gitea/workflows/ci.yaml
Normal file
@@ -0,0 +1,106 @@
|
||||
name: CI
|
||||
|
||||
on:
|
||||
push:
|
||||
branches: ["main", "master"]
|
||||
paths:
|
||||
- "scraper/**"
|
||||
- ".gitea/workflows/**"
|
||||
pull_request:
|
||||
branches: ["main", "master"]
|
||||
paths:
|
||||
- "scraper/**"
|
||||
- ".gitea/workflows/**"
|
||||
|
||||
defaults:
|
||||
run:
|
||||
working-directory: scraper
|
||||
|
||||
jobs:
|
||||
# ── lint & vet ───────────────────────────────────────────────────────────────
|
||||
lint:
|
||||
name: Lint
|
||||
runs-on: ubuntu-latest
|
||||
steps:
|
||||
- uses: actions/checkout@v4
|
||||
|
||||
- uses: actions/setup-go@v5
|
||||
with:
|
||||
go-version-file: scraper/go.mod
|
||||
cache-dependency-path: scraper/go.sum
|
||||
|
||||
- name: go vet
|
||||
run: go vet ./...
|
||||
|
||||
- name: staticcheck
|
||||
run: |
|
||||
go install honnef.co/go/tools/cmd/staticcheck@latest
|
||||
staticcheck ./...
|
||||
|
||||
# ── tests ────────────────────────────────────────────────────────────────────
|
||||
test:
|
||||
name: Test
|
||||
runs-on: ubuntu-latest
|
||||
steps:
|
||||
- uses: actions/checkout@v4
|
||||
|
||||
- uses: actions/setup-go@v5
|
||||
with:
|
||||
go-version-file: scraper/go.mod
|
||||
cache-dependency-path: scraper/go.sum
|
||||
|
||||
- name: Run tests
|
||||
run: go test -race -count=1 -timeout=60s ./...
|
||||
|
||||
# ── build binary ─────────────────────────────────────────────────────────────
|
||||
build:
|
||||
name: Build
|
||||
runs-on: ubuntu-latest
|
||||
needs: [lint, test]
|
||||
steps:
|
||||
- uses: actions/checkout@v4
|
||||
|
||||
- uses: actions/setup-go@v5
|
||||
with:
|
||||
go-version-file: scraper/go.mod
|
||||
cache-dependency-path: scraper/go.sum
|
||||
|
||||
- name: Build binary
|
||||
run: |
|
||||
CGO_ENABLED=0 GOOS=linux GOARCH=amd64 \
|
||||
go build -ldflags="-s -w" -o bin/scraper ./cmd/scraper
|
||||
|
||||
- name: Upload binary artifact
|
||||
uses: actions/upload-artifact@v4
|
||||
with:
|
||||
name: scraper-linux-amd64
|
||||
path: scraper/bin/scraper
|
||||
retention-days: 7
|
||||
|
||||
# ── docker build (& push) ────────────────────────────────────────────────────
|
||||
# Uncomment once the runner has Docker available and a registry is configured.
|
||||
#
|
||||
# docker:
|
||||
# name: Docker
|
||||
# runs-on: ubuntu-latest
|
||||
# needs: [lint, test]
|
||||
# # Only push images on commits to the default branch, not on PRs.
|
||||
# # if: github.event_name == 'push'
|
||||
# steps:
|
||||
# - uses: actions/checkout@v4
|
||||
#
|
||||
# - name: Log in to Gitea registry
|
||||
# uses: docker/login-action@v3
|
||||
# with:
|
||||
# registry: gitea.kalekber.cc
|
||||
# username: ${{ secrets.REGISTRY_USER }}
|
||||
# password: ${{ secrets.REGISTRY_TOKEN }}
|
||||
#
|
||||
# - name: Build and push
|
||||
# uses: docker/build-push-action@v5
|
||||
# with:
|
||||
# context: ./scraper
|
||||
# push: true
|
||||
# tags: |
|
||||
# gitea.kalekber.cc/kamil/libnovel:latest
|
||||
# gitea.kalekber.cc/kamil/libnovel:${{ gitea.sha }}
|
||||
187
scraper/internal/novelfire/ranking_test.go
Normal file
187
scraper/internal/novelfire/ranking_test.go
Normal file
@@ -0,0 +1,187 @@
|
||||
package novelfire
|
||||
|
||||
import (
|
||||
"context"
|
||||
"os"
|
||||
"path/filepath"
|
||||
"testing"
|
||||
|
||||
"github.com/libnovel/scraper/internal/scraper"
|
||||
"github.com/libnovel/scraper/internal/writer"
|
||||
)
|
||||
|
||||
// rankingPage1HTML is a realistic mock of novelfire.net/ranking?page=1.
|
||||
// It contains two novel-item entries and a "next" link for pagination tests.
|
||||
func rankingPage1HTML() string {
|
||||
return `<!DOCTYPE html>
|
||||
<html><body>
|
||||
<ul class="rank-novels">
|
||||
<li class="novel-item">
|
||||
<figure class="cover"><a href="/book/the-iron-throne"><img data-src="/covers/iron-throne.jpg"></a></figure>
|
||||
<div class="item-body">
|
||||
<h2 class="title"><a href="/book/the-iron-throne">The Iron Throne</a></h2>
|
||||
<span class="status">Ongoing</span>
|
||||
<div class="categories"><div class="scroll"><span>Fantasy</span><span>Action</span></div></div>
|
||||
</div>
|
||||
</li>
|
||||
<li class="novel-item">
|
||||
<figure class="cover"><a href="/book/shadow-mage"><img data-src="/covers/shadow-mage.jpg"></a></figure>
|
||||
<div class="item-body">
|
||||
<h2 class="title"><a href="/book/shadow-mage">Shadow Mage</a></h2>
|
||||
<span class="status">Completed</span>
|
||||
<div class="categories"><div class="scroll"><span>Magic</span></div></div>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
<a class="next" href="/ranking?page=2">Next</a>
|
||||
</body></html>`
|
||||
}
|
||||
|
||||
func rankingPage2HTML() string {
|
||||
return `<!DOCTYPE html>
|
||||
<html><body>
|
||||
<ul class="rank-novels">
|
||||
<li class="novel-item">
|
||||
<figure class="cover"><a href="/book/void-hunter"><img data-src="/covers/void-hunter.jpg"></a></figure>
|
||||
<div class="item-body">
|
||||
<h2 class="title"><a href="/book/void-hunter">Void Hunter</a></h2>
|
||||
<span class="status">Ongoing</span>
|
||||
<div class="categories"><div class="scroll"><span>Sci-Fi</span></div></div>
|
||||
</div>
|
||||
</li>
|
||||
</ul>
|
||||
<!-- no .next link → last page -->
|
||||
</body></html>`
|
||||
}
|
||||
|
||||
// drainRanking collects all entries from a ScrapeRanking call without
|
||||
// deadlocking. It uses "for A != nil || B != nil" — nil channels are never
|
||||
// selected, so setting one to nil effectively removes it from the select.
|
||||
func drainRanking(t *testing.T, entryCh <-chan scraper.BookMeta, errCh <-chan error) []scraper.BookMeta {
|
||||
t.Helper()
|
||||
var entries []scraper.BookMeta
|
||||
for entryCh != nil || errCh != nil {
|
||||
select {
|
||||
case meta, ok := <-entryCh:
|
||||
if !ok {
|
||||
entryCh = nil
|
||||
} else {
|
||||
entries = append(entries, meta)
|
||||
}
|
||||
case err, ok := <-errCh:
|
||||
if !ok {
|
||||
errCh = nil
|
||||
} else if err != nil {
|
||||
t.Fatalf("unexpected scrape error: %v", err)
|
||||
}
|
||||
}
|
||||
}
|
||||
return entries
|
||||
}
|
||||
|
||||
// TestScrapeRanking_SinglePage verifies a single page is parsed into entries
|
||||
// with sequential Ranking numbers using a stub client.
|
||||
// ScrapeRanking uses s.client (the main client, not urlClient) because the
|
||||
// ranking page is fully server-rendered.
|
||||
func TestScrapeRanking_SinglePage(t *testing.T) {
|
||||
// newScraper passes the stub as s.client — exactly what ScrapeRanking uses.
|
||||
s := newScraper(rankingPage1HTML())
|
||||
entryCh, errCh := s.ScrapeRanking(context.Background(), 1)
|
||||
entries := drainRanking(t, entryCh, errCh)
|
||||
|
||||
if len(entries) != 2 {
|
||||
t.Fatalf("expected 2 entries, got %d", len(entries))
|
||||
}
|
||||
if entries[0].Ranking != 1 || entries[0].Title != "The Iron Throne" {
|
||||
t.Errorf("entry[0]: got rank=%d title=%q, want rank=1 title=%q",
|
||||
entries[0].Ranking, entries[0].Title, "The Iron Throne")
|
||||
}
|
||||
if entries[1].Ranking != 2 || entries[1].Title != "Shadow Mage" {
|
||||
t.Errorf("entry[1]: got rank=%d title=%q, want rank=2 title=%q",
|
||||
entries[1].Ranking, entries[1].Title, "Shadow Mage")
|
||||
}
|
||||
}
|
||||
|
||||
// TestScrapeRanking_MultiPage verifies pagination across two pages yields
|
||||
// contiguous rank numbers (1, 2, 3).
|
||||
func TestScrapeRanking_MultiPage(t *testing.T) {
|
||||
// Use pagedStubClient for s.client so each GetContent call returns the
|
||||
// next page. ScrapeRanking now calls s.client directly.
|
||||
urlClient := &pagedStubClient{pages: []string{rankingPage1HTML(), rankingPage2HTML()}}
|
||||
s := New(urlClient, nil, nil) // urlClient == nil → falls back to client
|
||||
|
||||
entryCh, errCh := s.ScrapeRanking(context.Background(), 0) // 0 = all pages
|
||||
entries := drainRanking(t, entryCh, errCh)
|
||||
|
||||
if len(entries) != 3 {
|
||||
t.Fatalf("expected 3 entries across 2 pages, got %d", len(entries))
|
||||
}
|
||||
want := []struct {
|
||||
rank int
|
||||
title string
|
||||
}{
|
||||
{1, "The Iron Throne"},
|
||||
{2, "Shadow Mage"},
|
||||
{3, "Void Hunter"},
|
||||
}
|
||||
for i, w := range want {
|
||||
if entries[i].Ranking != w.rank || entries[i].Title != w.title {
|
||||
t.Errorf("entry[%d]: got rank=%d title=%q, want rank=%d title=%q",
|
||||
i, entries[i].Ranking, entries[i].Title, w.rank, w.title)
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// TestScrapeRanking_EmptyPage verifies that a page with no .rank-novels
|
||||
// container produces zero entries and closes channels cleanly (no deadlock).
|
||||
func TestScrapeRanking_EmptyPage(t *testing.T) {
|
||||
s := newScraper(`<!DOCTYPE html><html><body><div class="no-rankings"></div></body></html>`)
|
||||
entryCh, errCh := s.ScrapeRanking(context.Background(), 1)
|
||||
entries := drainRanking(t, entryCh, errCh)
|
||||
|
||||
if len(entries) != 0 {
|
||||
t.Errorf("expected 0 entries for empty page, got %d", len(entries))
|
||||
}
|
||||
}
|
||||
|
||||
// TestWriteRanking_RoundTrip verifies WriteRanking → ReadRankingItems
|
||||
// faithfully reconstructs the original slice.
|
||||
func TestWriteRanking_RoundTrip(t *testing.T) {
|
||||
dir := t.TempDir()
|
||||
w := writer.New(dir)
|
||||
|
||||
items := []writer.RankingItem{
|
||||
{Rank: 1, Slug: "the-iron-throne", Title: "The Iron Throne", Status: "Ongoing",
|
||||
Genres: []string{"Fantasy", "Action"}, SourceURL: "https://novelfire.net/book/the-iron-throne"},
|
||||
{Rank: 2, Slug: "shadow-mage", Title: "Shadow Mage", Status: "Completed",
|
||||
Genres: []string{"Magic"}, SourceURL: "https://novelfire.net/book/shadow-mage"},
|
||||
}
|
||||
|
||||
if err := w.WriteRanking(items); err != nil {
|
||||
t.Fatalf("WriteRanking failed: %v", err)
|
||||
}
|
||||
|
||||
rankingFile := filepath.Join(dir, "ranking.md")
|
||||
if _, err := os.Stat(rankingFile); err != nil {
|
||||
t.Fatalf("ranking.md not created: %v", err)
|
||||
}
|
||||
|
||||
got, err := w.ReadRankingItems()
|
||||
if err != nil {
|
||||
t.Fatalf("ReadRankingItems failed: %v", err)
|
||||
}
|
||||
if len(got) != len(items) {
|
||||
t.Fatalf("expected %d items, got %d", len(items), len(got))
|
||||
}
|
||||
for i, want := range items {
|
||||
if got[i].Rank != want.Rank {
|
||||
t.Errorf("item[%d].Rank = %d, want %d", i, got[i].Rank, want.Rank)
|
||||
}
|
||||
if got[i].Title != want.Title {
|
||||
t.Errorf("item[%d].Title = %q, want %q", i, got[i].Title, want.Title)
|
||||
}
|
||||
if got[i].Status != want.Status {
|
||||
t.Errorf("item[%d].Status = %q, want %q", i, got[i].Status, want.Status)
|
||||
}
|
||||
}
|
||||
}
|
||||
@@ -389,13 +389,10 @@ func (s *Scraper) ScrapeRanking(ctx context.Context, maxPages int) (<-chan scrap
|
||||
pageURL := fmt.Sprintf("%s%s?page=%d", baseURL, rankingPath, page)
|
||||
s.log.Info("scraping ranking page", "page", page, "url", pageURL)
|
||||
|
||||
// Always use the urlClient for the ranking page — it requires JS rendering.
|
||||
raw, err := s.urlClient.GetContent(ctx, browser.ContentRequest{
|
||||
// The ranking page is fully server-rendered; a direct HTTP GET is
|
||||
// sufficient and avoids the Browserless round-trip overhead.
|
||||
raw, err := s.client.GetContent(ctx, browser.ContentRequest{
|
||||
URL: pageURL,
|
||||
WaitFor: &browser.WaitForSelector{Selector: ".rank-novels", Timeout: 30000},
|
||||
RejectResourceTypes: rejectResourceTypes,
|
||||
GotoOptions: &browser.GotoOptions{Timeout: 60000},
|
||||
BestAttempt: true,
|
||||
})
|
||||
if err != nil {
|
||||
s.log.Debug("ranking page fetch failed", "page", page, "url", pageURL, "err", err)
|
||||
|
||||
@@ -455,13 +455,12 @@ func (s *Server) handleRankingRefresh(w http.ResponseWriter, r *http.Request) {
|
||||
rankingCh, errCh := s.novel.ScrapeRanking(ctx, maxPages)
|
||||
|
||||
var rankingItems []writer.RankingItem
|
||||
for {
|
||||
for rankingCh != nil || errCh != nil {
|
||||
select {
|
||||
case meta, ok := <-rankingCh:
|
||||
if !ok {
|
||||
rankingCh = nil
|
||||
continue
|
||||
}
|
||||
} else {
|
||||
rankingItems = append(rankingItems, writer.RankingItem{
|
||||
Rank: meta.Ranking,
|
||||
Slug: meta.Slug,
|
||||
@@ -472,18 +471,14 @@ func (s *Server) handleRankingRefresh(w http.ResponseWriter, r *http.Request) {
|
||||
Genres: meta.Genres,
|
||||
SourceURL: meta.SourceURL,
|
||||
})
|
||||
}
|
||||
case err, ok := <-errCh:
|
||||
if !ok {
|
||||
errCh = nil
|
||||
continue
|
||||
}
|
||||
if err != nil {
|
||||
} else if err != nil {
|
||||
s.log.Error("ranking scrape error", "err", err)
|
||||
}
|
||||
}
|
||||
if rankingCh == nil && errCh == nil {
|
||||
break
|
||||
}
|
||||
}
|
||||
|
||||
if len(rankingItems) > 0 {
|
||||
|
||||
Reference in New Issue
Block a user