core: harden watch-order relations

This commit is contained in:
2026-04-11 22:06:26 +02:00
parent d07ba90892
commit 6edf47c553
6 changed files with 505 additions and 210 deletions

View File

@@ -0,0 +1,184 @@
package watchorder
import (
"context"
"errors"
"fmt"
"net/http"
"regexp"
"strconv"
"strings"
"github.com/PuerkitoBio/goquery"
)
const defaultUserAgent = "anime-relations-scraper/1.0 (+https://github.com/mkelvers/anime-relations)"
var idPattern = regexp.MustCompile(`/id/(\d+)`)
var ErrInvalidWatchOrderURL = errors.New("invalid watch order url")
type WatchOrderEntry struct {
ID int `json:"id"`
Type string `json:"type"`
Title string `json:"title"`
TitleAlt string `json:"title_alt,omitempty"`
}
type WatchOrderResult struct {
ID int `json:"id"`
WatchOrder []WatchOrderEntry `json:"watch_order"`
}
type watchOrderRow struct {
id int
typeID int
title string
alternativeTitle string
}
func parseRootID(url string) (int, error) {
match := idPattern.FindStringSubmatch(url)
if len(match) != 2 {
return 0, ErrInvalidWatchOrderURL
}
id, err := strconv.Atoi(match[1])
if err != nil {
return 0, ErrInvalidWatchOrderURL
}
return id, nil
}
func fetchDocument(ctx context.Context, httpClient *http.Client, url string) (*goquery.Document, error) {
client := httpClient
if client == nil {
client = http.DefaultClient
}
request, err := http.NewRequestWithContext(ctx, http.MethodGet, url, nil)
if err != nil {
return nil, fmt.Errorf("failed to create request: %w", err)
}
request.Header.Set("User-Agent", defaultUserAgent)
response, err := client.Do(request)
if err != nil {
return nil, fmt.Errorf("request failed: %w", err)
}
defer response.Body.Close()
if response.StatusCode != http.StatusOK {
return nil, fmt.Errorf("unexpected status code: %d", response.StatusCode)
}
document, err := goquery.NewDocumentFromReader(response.Body)
if err != nil {
return nil, fmt.Errorf("failed to parse html: %w", err)
}
return document, nil
}
func extractTypeLabelsByID(doc *goquery.Document) map[int]string {
typeLabels := make(map[int]string)
doc.Find("#wo_type_filter label").Each(func(_ int, selection *goquery.Selection) {
input := selection.Find("input[type='checkbox']")
rawID, exists := input.Attr("value")
if !exists {
return
}
typeID, err := strconv.Atoi(strings.TrimSpace(rawID))
if err != nil {
return
}
label := strings.TrimSpace(selection.Text())
if label == "" {
return
}
typeLabels[typeID] = label
})
return typeLabels
}
func parseAttrInt(selection *goquery.Selection, attrName string) (int, bool) {
rawValue, exists := selection.Attr(attrName)
if !exists {
return 0, false
}
value, err := strconv.Atoi(strings.TrimSpace(rawValue))
if err != nil {
return 0, false
}
return value, true
}
func extractRows(doc *goquery.Document) []watchOrderRow {
rows := make([]watchOrderRow, 0)
doc.Find("tr[data-id]").Each(func(_ int, selection *goquery.Selection) {
id, ok := parseAttrInt(selection, "data-id")
if !ok {
return
}
typeID, ok := parseAttrInt(selection, "data-type")
if !ok {
return
}
title := strings.TrimSpace(selection.Find(".wo_title").First().Text())
alternativeTitle := strings.TrimSpace(selection.Find(".uk-text-small").First().Text())
rows = append(rows, watchOrderRow{
id: id,
typeID: typeID,
title: title,
alternativeTitle: alternativeTitle,
})
})
return rows
}
func FetchWatchOrder(ctx context.Context, httpClient *http.Client, url string) (WatchOrderResult, error) {
rootID, err := parseRootID(url)
if err != nil {
return WatchOrderResult{}, err
}
doc, err := fetchDocument(ctx, httpClient, url)
if err != nil {
return WatchOrderResult{}, err
}
rows := extractRows(doc)
if len(rows) == 0 {
return WatchOrderResult{ID: rootID, WatchOrder: []WatchOrderEntry{}}, nil
}
typeByID := extractTypeLabelsByID(doc)
entries := make([]WatchOrderEntry, 0, len(rows))
for _, row := range rows {
typeName := strings.TrimSpace(typeByID[row.typeID])
entries = append(entries, WatchOrderEntry{
ID: row.id,
Type: typeName,
Title: row.title,
TitleAlt: row.alternativeTitle,
})
}
return WatchOrderResult{ID: rootID, WatchOrder: entries}, nil
}

View File

@@ -0,0 +1,105 @@
package watchorder
import (
"context"
"net/http"
"net/http/httptest"
"testing"
"time"
)
func testServer(body string) *httptest.Server {
handler := http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
w.Header().Set("Content-Type", "text/html; charset=utf-8")
_, _ = w.Write([]byte(body))
})
return httptest.NewServer(handler)
}
func testHTMLWithMetadata() string {
return `
<!doctype html>
<html>
<body>
<div id="wo_type_filter">
<label><input type="checkbox" value="1" checked> TV</label>
<label><input type="checkbox" value="3" checked> Movie</label>
</div>
<table id="wo_list">
<tr data-id="442" data-anilist-id="442" data-type="3">
<td>
<span class="wo_title">Naruto Movie 1</span>
<span class="uk-text-small">Naruto the Movie 1</span>
</td>
</tr>
</table>
</body>
</html>`
}
func testHTMLEmptyRows() string {
return `
<!doctype html>
<html>
<body>
<div id="wo_type_filter">
<label><input type="checkbox" value="1" checked> TV</label>
<label><input type="checkbox" value="3" checked> Movie</label>
</div>
<table id="wo_list"></table>
</body>
</html>`
}
func TestFetchWatchOrder_OutputShape(t *testing.T) {
server := testServer(testHTMLWithMetadata())
defer server.Close()
url := server.URL + "/?/tools/watch_order/id/442"
result, err := FetchWatchOrder(context.Background(), &http.Client{Timeout: time.Second}, url)
if err != nil {
t.Fatalf("expected no error, got %v", err)
}
if result.ID != 442 {
t.Fatalf("expected root id 442, got %d", result.ID)
}
if len(result.WatchOrder) != 1 {
t.Fatalf("expected 1 watch_order entry, got %d", len(result.WatchOrder))
}
entry := result.WatchOrder[0]
if entry.ID != 442 {
t.Fatalf("expected entry id 442, got %d", entry.ID)
}
if entry.Type != "Movie" {
t.Fatalf("expected type Movie, got %q", entry.Type)
}
if entry.Title != "Naruto Movie 1" {
t.Fatalf("expected title Naruto Movie 1, got %q", entry.Title)
}
if entry.TitleAlt != "Naruto the Movie 1" {
t.Fatalf("expected title_alt Naruto the Movie 1, got %q", entry.TitleAlt)
}
}
func TestFetchWatchOrder_NoRowsReturnsEmpty(t *testing.T) {
server := testServer(testHTMLEmptyRows())
defer server.Close()
url := server.URL + "/?/tools/watch_order/id/1535"
result, err := FetchWatchOrder(context.Background(), &http.Client{Timeout: time.Second}, url)
if err != nil {
t.Fatalf("expected no error, got %v", err)
}
if result.ID != 1535 {
t.Fatalf("expected root id 1535, got %d", result.ID)
}
if len(result.WatchOrder) != 0 {
t.Fatalf("expected no entries, got %d", len(result.WatchOrder))
}
}