feed-fetcher/main.go
Marcus Noble 8cf51e1004
Added support for atom feeds
Signed-off-by: Marcus Noble <github@marcusnoble.co.uk>
2022-08-08 07:05:03 +01:00

122 lines
2.7 KiB
Go

package main
import (
"fmt"
"net/http"
"net/url"
"os"
"strings"
"embed"
"github.com/PuerkitoBio/goquery"
"github.com/gofiber/fiber/v2"
"github.com/mmcdole/gofeed"
)
//go:embed index.html
var content embed.FS
func main() {
port, ok := os.LookupEnv("PORT")
if !ok {
port = "8080"
}
app := fiber.New(fiber.Config{})
app.Get("/", func(c *fiber.Ctx) error {
feedUrl := c.Query("url")
if feedUrl == "" {
c.Type("html", "UTF8")
body, _ := content.ReadFile("index.html")
return c.Send(body)
}
feeds, statusCode := getFeeds(feedUrl)
if c.Is("json") {
if statusCode >= 400 || statusCode == 300 {
c.Status(statusCode)
}
return c.JSON(feeds)
} else {
c.Status(statusCode)
c.Location(feeds[0])
if len(feeds) > 1 {
responseBody := "Multiple Choices\n\n"
for _, feed := range feeds {
responseBody += feed + "\n"
}
return c.SendString(responseBody)
}
return c.Send(nil)
}
})
fmt.Println(app.Listen(fmt.Sprintf(":%s", port)))
}
func absoluteUrl(requestUrl, foundUrl string) string {
if !strings.HasPrefix(foundUrl, "http") {
parsedUrl, _ := url.Parse(requestUrl)
foundUrl = fmt.Sprintf("%s://%s%s", parsedUrl.Scheme, parsedUrl.Host, foundUrl)
}
return foundUrl
}
func getFeeds(requestURL string) ([]string, int) {
feeds := []string{}
fp := gofeed.NewParser()
_, err := fp.ParseURL(requestURL)
if err == nil {
feeds = []string{requestURL}
} else if err != nil && err == gofeed.ErrFeedTypeNotDetected {
res, err := http.Get(requestURL)
if err != nil {
fmt.Printf("Failed to fetch URL %s", requestURL)
return feeds, fiber.StatusInternalServerError
}
defer res.Body.Close()
if res.StatusCode >= 400 {
fmt.Println("Provided URL returned an error status code")
return feeds, res.StatusCode
}
doc, err := goquery.NewDocumentFromReader(res.Body)
if err != nil {
fmt.Println("Failed to parse response body")
return feeds, fiber.StatusInternalServerError
}
matches := doc.Find(`[rel="alternate"][type="application/rss+xml"],[rel="alternate"][type="application/atom+xml"]`)
if matches.Length() == 0 {
fmt.Printf("No RSS feeds found on page %s\n", requestURL)
return feeds, fiber.StatusNotFound
}
matches.Each(func(i int, s *goquery.Selection) {
feeds = append(feeds, absoluteUrl(requestURL, s.AttrOr("href", "")))
})
if matches.Length() > 1 {
fmt.Printf("Multiple RSS feed found on page %s\n", requestURL)
return feeds, fiber.StatusMultipleChoices
} else {
fmt.Printf("Single RSS feed found on page %s\n", requestURL)
return feeds, fiber.StatusTemporaryRedirect
}
} else if err != nil {
fmt.Println("Failed while attempting to parse feed")
return feeds, fiber.StatusInternalServerError
}
return feeds, 200
}