update project

This commit is contained in:
2024-11-25 16:10:58 +01:00
parent a6f3325842
commit 958ef12e91
9 changed files with 108 additions and 91 deletions
+5 -3
View File
@@ -2,7 +2,6 @@ package scraper
import (
"fmt"
"github.com/PuerkitoBio/goquery"
"log/slog"
"regexp"
"sort"
@@ -10,6 +9,8 @@ import (
"strings"
"time"
"github.com/PuerkitoBio/goquery"
"github.com/gocolly/colly"
"github.com/jackc/pgx/v5/pgtype"
"github.com/zepyrshut/rating-orama/internal/sqlc"
@@ -47,14 +48,15 @@ const (
seasonsSelector = "ul.ipc-tabs a[data-testid='tab-season-entry']"
episodeCardSelector = "article.sc-f8507e90-1.cHtpvn.episode-item-wrapper"
seasonEpisodeAndTitleSelector = "div.ipc-title__text"
releasedDateSelector = "span.sc-ccd6e31b-10.dYquTu"
plotSelector = "div.sc-ccd6e31b-11.cVKeME"
releasedDateSelector = "span.sc-f2169d65-10.bYaARM"
plotSelector = "div.ipc-html-content-inner-div"
starRatingSelector = "span.ipc-rating-star--rating"
voteCountSelector = "span.ipc-rating-star--voteCount"
imdbEpisodesURL = "https://www.imdb.com/title/%s/episodes/?season=%d"
visitURL = "https://www.imdb.com/title/%s/episodes"
)
func ScrapeEpisodes(ttImdb string) (string, []Episode) {
c := colly.NewCollector(
colly.AllowedDomains("imdb.com", "www.imdb.com"),