pedro/rating-orama
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

feat: ✨ add title to tvshow

Browse Source
This commit is contained in:
Pedro Pérez 2024-11-06 02:51:04 +01:00
parent e332660f24
commit bcbb02a663
2 changed files with 24 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
core/internal/handlers/tvshow.go
View File

@ -12,11 +12,12 @@ func (hq *Handlers) GetTVShow(c *gin.Context) {
ttShowID := c.Query("ttid") ttShowID := c.Query("ttid")
slog.Info("GetTVShow", "ttid", ttShowID) slog.Info("GetTVShow", "ttid", ttShowID)
seasons := scraper.ScrapeSeasons(ttShowID) title, seasons := scraper.ScrapeSeasons(ttShowID)
slog.Info("scraped seasons", "ttid", ttShowID) slog.Info("scraped seasons", "ttid", ttShowID, "title", title)
c.JSON(http.StatusOK, gin.H{ c.JSON(http.StatusOK, gin.H{
"title": title,
"seasons": seasons, "seasons": seasons,
}) })
} }

29
core/internal/scraper/tvshow.go
View File

@ -24,20 +24,29 @@ type Episode struct {
type Season []Episode type Season []Episode
const seasonsSelector = "ul.ipc-tabs a[data-testid='tab-season-entry']" const (
const episodesSelector = "section.sc-1e7f96be-0.ZaQIL" titleSelector = "h2.sc-b8cc654b-9.dmvgRY"
const nextSeasonButtonSelector = "#next-season-btn" seasonsSelector = "ul.ipc-tabs a[data-testid='tab-season-entry']"
const imdbEpisodesURL = "https://www.imdb.com/title/%s/episodes?season=%d" episodesSelector = "section.sc-1e7f96be-0.ZaQIL"
nextSeasonButtonSelector = "#next-season-btn"
imdbEpisodesURL = "https://www.imdb.com/title/%s/episodes?season=%d"
visitURL = "https://www.imdb.com/title/%s/episodes"
)
func ScrapeSeasons(ttImdb string) []Season { func ScrapeSeasons(ttImdb string) (string, []Season) {
c := colly.NewCollector( c := colly.NewCollector(
colly.AllowedDomains("imdb.com", "www.imdb.com"), colly.AllowedDomains("imdb.com", "www.imdb.com"),
) )
c.OnRequest(func(r *colly.Request) {
r.Headers.Set("Accept-Language", "en-US")
})
var allSeasons []Season var allSeasons []Season
var seasons []int var seasons []int
var title string
c.OnHTML("ul.ipc-tabs a[data-testid='tab-season-entry']", func(e *colly.HTMLElement) { c.OnHTML(seasonsSelector, func(e *colly.HTMLElement) {
seasonText := strings.TrimSpace(e.Text) seasonText := strings.TrimSpace(e.Text)
seasonNum, err := strconv.Atoi(seasonText) seasonNum, err := strconv.Atoi(seasonText)
if err == nil { if err == nil {
@ -45,6 +54,10 @@ func ScrapeSeasons(ttImdb string) []Season {
} }
}) })
c.OnHTML(titleSelector, func(e *colly.HTMLElement) {
title = e.Text
})
c.OnScraped(func(r *colly.Response) { c.OnScraped(func(r *colly.Response) {
seasonMap := make(map[int]bool) seasonMap := make(map[int]bool)
uniqueSeasons := []int{} uniqueSeasons := []int{}
@ -75,10 +88,10 @@ func ScrapeSeasons(ttImdb string) []Season {
episodeCollector.Wait() episodeCollector.Wait()
}) })
c.Visit(fmt.Sprintf("https://www.imdb.com/title/%s/episodes", ttImdb)) c.Visit(fmt.Sprintf(visitURL, ttImdb))
c.Wait() c.Wait()
return allSeasons return title, allSeasons
} }
func extractEpisodesFromSeason(data string) Season { func extractEpisodesFromSeason(data string) Season {