.travis.yml

@@ -1,5 +0,0 @@
-language: go
-
-go:
-  - 1.6.2
-  - tip

README.md

@@ -1,7 +1,5 @@
 # MediaLink IRC Bot
 
-[![Build Status](https://travis-ci.org/icedream/irc-medialink.svg?branch=master)](https://travis-ci.org/icedream/irc-medialink)
-
 This IRC bot automatically parses links posted in chat rooms and prints information about them.
 
 Currently explicit support has been built in for:
@@ -13,7 +11,7 @@ Currently explicit support has been built in for:
 
 Generally, for websites that are not directly supported the bot will print the page title.
 
-## How to run the bot
+# How to run the bot
 
 In order to properly run the bot, you need to [register a SoundCloud application](http://soundcloud.com/you/apps/new) and [get a YouTube Data API key](https://console.developers.google.com/apis/api/youtube/overview) for it and then feed the API data to the bot through the command line arguments.
 
@@ -27,7 +25,7 @@ Then you can find out which options you can pass to the bot directly by running
 
 You need to at least pass the `--server`, `--youtube-key`, `--soundcloud-id` and `--soundcloud-secret` parameters.
 
-### ...with Docker
+# ...with Docker
 
 You can use the `icedream/irc-medialink` image in order to run this bot in Docker. You can pull it using this command:
 
@@ -56,12 +54,12 @@ services:
     restart: always
 ```
 
-## Support
+# Support
 
 This bot is officially tested and running on the LibraIRC IRC network (irc.librairc.net) though also being able to run on other IRC networks.
 
 For support on LibraIRC please use the channel #MediaLink there to get in contact with Icedream.
 
-## License
+# License
 
 This project is licensed under the **GNU General Public License Version 2 or later**. For more information check the [LICENSE](LICENSE) file.

main.go

@@ -32,10 +32,7 @@ func main() {
 	var soundcloudClientId string
 	var soundcloudClientSecret string
 
-	var webEnableImages bool
-
 	var debug bool
-	var noInvite bool
 	var useTLS bool
 	var server string
 	var password string
@@ -51,7 +48,6 @@ func main() {
 	kingpin.Flag("nick", "The nickname.").Short('n').StringVar(&nickname)
 	kingpin.Flag("ident", "The ident.").Short('i').StringVar(&ident)
 	kingpin.Flag("debug", "Enables debug mode.").Short('d').BoolVar(&debug)
-	kingpin.Flag("no-invite", "Disables auto-join on invite.").BoolVar(&noInvite)
 	kingpin.Flag("tls", "Use TLS.").BoolVar(&useTLS)
 	kingpin.Flag("server", "The server to connect to.").Short('s').StringVar(&server)
 	kingpin.Flag("password", "The password to use for logging into the IRC server.").Short('p').StringVar(&password)
@@ -62,14 +58,9 @@ func main() {
 
 	// Youtube config
 	kingpin.Flag("youtube-key", "The API key to use to access the YouTube API.").StringVar(&youtubeApiKey)
-
-	// SoundCloud config
 	kingpin.Flag("soundcloud-id", "The SoundCloud ID.").StringVar(&soundcloudClientId)
 	kingpin.Flag("soundcloud-secret", "The SoundCloud secret.").StringVar(&soundcloudClientSecret)
 
-	// Web parser config
-	kingpin.Flag("images", "Enables parsing links of images. Disabled by default for legal reasons.").BoolVar(&webEnableImages)
-
 	kingpin.Parse()
 
 	if len(nickname) == 0 {
@@ -83,17 +74,12 @@ func main() {
 	m := manager.NewManager()
 
 	// Load youtube parser
-	if len(youtubeApiKey) > 0 {
 	youtubeParser := &youtube.Parser{
 		Config: &youtube.Config{ApiKey: youtubeApiKey},
 	}
 	must(m.RegisterParser(youtubeParser))
-	} else {
-		log.Println("No YouTube API key provided, YouTube parsing via API is disabled.")
-	}
 
 	// Load soundcloud parser
-	if len(soundcloudClientId) > 0 && len(soundcloudClientSecret) > 0 {
 	soundcloudParser := &soundcloud.Parser{
 		Config: &soundcloud.Config{
 			ClientId:     soundcloudClientId,
@@ -101,18 +87,12 @@ func main() {
 		},
 	}
 	must(m.RegisterParser(soundcloudParser))
-	} else {
-		log.Println("No SoundCloud client ID or secret provided, SoundCloud parsing via API is disabled.")
-	}
 
 	// Load wikipedia parser
 	must(m.RegisterParser(new(wikipedia.Parser)))
 
 	// Load web parser
-	webParser := &web.Parser{
+	must(m.RegisterParser(new(web.Parser)))
-		EnableImages: webEnableImages,
-	}
-	must(m.RegisterParser(webParser))
 
 	// IRC
 	conn := m.AntifloodIrcConn(irc.IRC(nickname, ident))
@@ -149,8 +129,6 @@ func main() {
 	conn.AddCallback("JOIN", func(e *irc.Event) {
 		// Is this JOIN not about us?
 		if !strings.EqualFold(e.Nick, conn.GetNick()) {
-			// Save this user's details for a temporary ignore
-			m.NotifyUserJoined(e.Arguments[0], e.Source)
 			return
 		}
 
@@ -160,7 +138,6 @@ func main() {
 		default:
 		}
 	})
-	if !noInvite {
 	conn.AddCallback("INVITE", func(e *irc.Event) {
 		// Is this INVITE not for us?
 		if !strings.EqualFold(e.Arguments[0], conn.GetNick()) {
@@ -184,7 +161,7 @@ func main() {
 						time.Sleep(1 * time.Second)
 						conn.Privmsgf(targetChannel, "Thanks for inviting me, %s! I am %s, the friendly bot that shows information about links posted in this channel. I hope I can be of great help for everyone here in %s! :)", sourceNick, conn.GetNick(), targetChannel)
 						time.Sleep(2 * time.Second)
-							conn.Privmsg(targetChannel, "If you ever run into trouble with me (or find any bugs), please use the channel #MediaLink for contact on this IRC.")
+						conn.Privmsg(targetChannel, "If you ever run into trouble with me (or find any bugs), please us the channel #MediaLink for contact on this IRC.")
 						break joinWaitLoop
 					}
 				case channel := <-inviteChan:
@@ -199,7 +176,6 @@ func main() {
 		}(e.Nick, e.Arguments[1])
 		conn.Join(e.Arguments[1])
 	})
-	}
 	conn.AddCallback("PRIVMSG", func(e *irc.Event) {
 		go func(event *irc.Event) {
 			//sender := event.Nick
@@ -220,12 +196,6 @@ func main() {
 
 			log.Printf("<%s @ %s> %s", event.Nick, target, msg)
 
-			// Ignore user if they just joined
-			if shouldIgnore := m.TrackUser(target, event.Source); shouldIgnore {
-				log.Print("This message will be ignored since the user just joined.")
-				return
-			}
-
 			urlStr := xurls.Relaxed.FindString(msg)
 
 			switch {

main.tpl

@@ -21,17 +21,10 @@
 	{{- else -}}
 		Link info
 	{{- end -}}
-	{{- reset }}
+	{{- bold }}
 
 	»
 	
-	{{- if index . "AgeRestriction" }}
-		{{ color 4 -}}
-		{{ bold -}}
-		[{{- index . "AgeRestriction" }}]
-		{{- reset }}
-	{{- end }}
-	
 	{{ if index . "IsProfile" }}
 		{{- if index . "Title" }}
 			{{ bold -}}
@@ -67,19 +60,15 @@
 				({{ . }})
 			{{ end }}
 		{{ else }}
-			{{ with index . "Description" }}
+			{{ if index . "Description" }}
-				{{ excerpt 384 . }}
+				{{ excerpt 384 (index . "Description") }}
+			{{ else }}
+				{{ with index . "ImageType" }}
+					{{ . }} image,
 				{{ end }}
-		{{ end }}
-		
-		{{ if index . "ImageType" }}
-			{{ if index . "Title" }}
-				·
-			{{ end }}
-			{{ .ImageType }} image,
 				{{ if (index . "ImageSize") (index . "Size") }}
 					{{ with index . "ImageSize" }}
-					{{ .X }}×{{ .Y }}
+						{{ .X }}x{{ .Y }}
 					{{ end }}
 					{{ with index . "Size" }}
 						({{ size . }})
@@ -87,6 +76,7 @@
 				{{ end }}
 			{{ end }}
 		{{ end }}
+	{{ end }}
 
 	{{ if or (index . "Author") }}
 		{{ if index . "Author" }}

manager/antiflood.go

@@ -17,28 +17,6 @@ func (m *Manager) initAntiflood() {
 	m.cache = cache.New(1*time.Minute, 5*time.Second)
 }
 
-func (m *Manager) TrackUser(target string, source string) (shouldIgnore bool) {
-	key := normalizeUserAntiflood(target, source)
-
-	if _, ok := m.cache.Get(key); ok {
-		// User just joined here recently, ignore them
-		shouldIgnore = true
-	}
-
-	return
-}
-
-func (m *Manager) NotifyUserJoined(target string, source string) {
-	key := normalizeUserAntiflood(target, source)
-
-	// When a user joins, he will be ignored for the first 30 seconds,
-	// enough to prevent parsing links from people who only join to spam their
-	// links immediately
-	if _, exists := m.cache.Get(key); !exists {
-		m.cache.Add(key, nil, 30*time.Second)
-	}
-}
-
 func (m *Manager) TrackUrl(target string, u *url.URL) (shouldIgnore bool) {
 	key := normalizeUrlAntiflood(target, u)
 
@@ -92,17 +70,6 @@ func normalizeTextAntiflood(target, text string) string {
 	return fmt.Sprintf("TEXT/%s/%X", strings.ToUpper(target), s.Sum([]byte{}))
 }
 
-func normalizeUserAntiflood(target, source string) string {
-	sourceSplitHost := strings.SplitN(source, "@", 2)
-	sourceSplitHostname := strings.Split(sourceSplitHost[1], ".")
-	if len(sourceSplitHostname) > 1 &&
-		strings.EqualFold(sourceSplitHostname[len(sourceSplitHostname)-1], "IP") {
-		sourceSplitHostname[0] = "*"
-	}
-	source = fmt.Sprintf("%s!%s@%s", "*", "*", strings.Join(sourceSplitHostname, "."))
-	return fmt.Sprintf("USER/%s/%s", strings.ToUpper(target), source)
-}
-
 // Proxies several methods of the IRC connection in order to drop repeated messages
 type ircConnectionProxy struct {
 	*irc.Connection

manager/parser.go

@@ -52,40 +52,33 @@ func (m *Manager) RegisterParser(parser Parser) error {
 	return nil
 }
 
-func (m *Manager) Parse(currentUrl *url.URL) (string, parsers.ParseResult) {
+func (m *Manager) Parse(u *url.URL) (string, parsers.ParseResult) {
-	var referer *url.URL
+	var oldU *url.URL
 	attempt := 0
 followLoop:
-	for currentUrl != nil {
+	for u != nil {
+		log.Printf("Parsing %s (referer %s)...", u, oldU)
 		attempt++
 		if attempt > 15 {
-			log.Printf("WARNING: Potential infinite loop for url %s, abort parsing", currentUrl)
+			log.Printf("WARNING: Potential infinite loop for url %s, abort parsing", u)
 			break
 		}
 		for _, p := range m.GetParsers() {
-			var refererCopy *url.URL
+			r := p.Parse(u, oldU)
-			if referer != nil {
-				refererCopy = &url.URL{}
-				*refererCopy = *referer
-			}
-			currentUrlCopy := &url.URL{}
-			*currentUrlCopy = *currentUrl
-			r := p.Parse(currentUrlCopy, refererCopy)
 			if r.Ignored {
 				continue
 			}
 			if r.FollowUrl != nil {
-				if *currentUrl == *r.FollowUrl {
+				if *u == *r.FollowUrl {
 					log.Printf("WARNING: Ignoring request to follow to same URL, ignoring.")
 					break followLoop
 				}
-				referer = currentUrl
+				oldU, u = u, r.FollowUrl
-				currentUrl = r.FollowUrl
 				continue followLoop
 			}
 			return p.Name(), r
 		}
-		currentUrl = nil
+		u = nil
 	}
 
 	// No parser matches, link ignored

parsers/web/parser.go

@@ -2,9 +2,9 @@ package web
 
 import (
 	"errors"
+	"log"
 	"net/http"
 	"net/url"
-	"regexp"
 	"strings"
 
 	"golang.org/x/net/html"
@@ -16,25 +16,14 @@ import (
 	_ "image/png"
 
 	"github.com/icedream/irc-medialink/parsers"
-	"github.com/icedream/irc-medialink/util/limitedio"
 	"github.com/yhat/scrape"
 )
 
 var (
 	ErrCorruptedImage = errors.New("Corrupted image.")
-
-	rxNewlines = regexp.MustCompile(`(?:\r?\n)+`)
 )
 
-const (
+type Parser struct{}
-	runeHash    = '#'
-	noTitleStr  = "(no title)"
-	maxHtmlSize = 8 * 1024
-)
-
-type Parser struct {
-	EnableImages bool
-}
 
 func (p *Parser) Init() error {
 	return nil
@@ -51,12 +40,6 @@ func (p *Parser) Parse(u *url.URL, referer *url.URL) (result parsers.ParseResult
 		return
 	}
 
-	// Remove hash reference from URL since that's not meant to be in the request
-	if strings.Contains(u.Path, string(runeHash)) {
-		u = &(*u) // avoid modifying original URL object
-		u.Path = u.Path[0:strings.IndexRune(u.Path, runeHash)]
-	}
-
 	// Make request
 	req, err := http.NewRequest("GET", u.String(), nil)
 	if err != nil {
@@ -66,11 +49,12 @@ func (p *Parser) Parse(u *url.URL, referer *url.URL) (result parsers.ParseResult
 	if referer != nil {
 		req.Header.Set("Referer", referer.String())
 	}
-	req.Header.Set("User-Agent", "MediaLink IRC Bot")
 	if resp, err := http.DefaultTransport.RoundTrip(req); err != nil {
+		log.Print("HTTP Get failed")
 		result.Error = err
 		return
 	} else {
+		log.Printf("Web parser result: %+v", resp)
 		defer resp.Body.Close()
 		if 300 <= resp.StatusCode && resp.StatusCode < 400 {
 			if u2, err := resp.Location(); err == nil && u2 != nil && *u2 != *u {
@@ -91,40 +75,30 @@ func (p *Parser) Parse(u *url.URL, referer *url.URL) (result parsers.ParseResult
 		if sep < 0 {
 			sep = len(contentType)
 		}
+		log.Print(contentType[0:sep])
 		switch strings.ToLower(contentType[0:sep]) {
 		case "text/html":
 			// Parse the page
-			var contentLength int
+			root, err := html.Parse(resp.Body)
-			if resp.ContentLength < 0 || resp.ContentLength > maxHtmlSize {
-				contentLength = maxHtmlSize
-			} else {
-				contentLength = int(resp.ContentLength)
-			}
-			limitedBody := limitedio.NewLimitedReader(resp.Body, contentLength)
-			root, err := html.Parse(limitedBody)
 			if err != nil {
 				result.Error = err
 				return
 			}
 			// Search for the title
-			result.Information = []map[string]interface{}{
-				map[string]interface{}{
-					"IsUpload": false,
-				},
-			}
 			title, ok := scrape.Find(root, scrape.ByTag(atom.Title))
 			if ok {
 				// Got it!
-				result.Information[0]["Title"] = rxNewlines.ReplaceAllString(scrape.Text(title), " ")
+				result.Information = []map[string]interface{}{
+					map[string]interface{}{
+						"IsUpload": false,
+						"Title":    scrape.Text(title),
+					},
+				}
 			} else {
-				// No title found
+				result.Ignored = true
-				result.Information[0]["Title"] = noTitleStr
 			}
 		case "image/png", "image/jpeg", "image/gif":
-			if p.EnableImages {
+			log.Print("Parsing image...")
-
-				// No need to limit the reader to a specific size here as
-				// image.DecodeConfig only reads as much as needed anyways.
 			if m, imgType, err := image.DecodeConfig(resp.Body); err != nil {
 				result.UserError = ErrCorruptedImage
 			} else {
@@ -132,19 +106,16 @@ func (p *Parser) Parse(u *url.URL, referer *url.URL) (result parsers.ParseResult
 					"IsUpload":  true,
 					"ImageSize": image.Point{X: m.Width, Y: m.Height},
 					"ImageType": strings.ToUpper(imgType),
-						"Title":     u.Path[strings.LastIndex(u.Path, "/")+1:],
 				}
 				if resp.ContentLength > 0 {
 					info["Size"] = uint64(resp.ContentLength)
 				}
 				result.Information = []map[string]interface{}{info}
+				log.Printf("Got through: %+v!", info)
 			}
-				break
-			}
-
-			fallthrough
 		default:
 			// TODO - Implement generic head info?
+			log.Printf("web parser: Ignoring content of type %s", resp.Header.Get("content-type"))
 			result.Ignored = true
 		}
 	}

parsers/web/parser_test.go

@@ -1,133 +0,0 @@
-package web
-
-import (
-	"net/url"
-	"testing"
-	"time"
-
-	"github.com/icedream/irc-medialink/parsers"
-	"github.com/stretchr/testify/assert"
-)
-
-func mustNewParser(t *testing.T) *Parser {
-	p := new(Parser)
-	if !assert.Nil(t, p.Init(), "Parser.Init must throw no errors") {
-		panic("Can't run test without a proper parser")
-	}
-	return p
-}
-
-func parseWithTimeout(p *Parser, t *testing.T, timeout time.Duration, u *url.URL, ref *url.URL) (retval parsers.ParseResult) {
-	resultChan := make(chan parsers.ParseResult)
-	go func(resultChan chan<- parsers.ParseResult, p *Parser, u *url.URL, ref *url.URL) {
-		resultChan <- p.Parse(u, ref)
-	}(resultChan, p, u, ref)
-
-	select {
-	case r := <-resultChan:
-		retval = r
-		return
-	case <-time.After(timeout):
-		t.Fatal("Didn't succeed parsing URL in time")
-		return
-	}
-}
-
-func Test_Parser_Parse_IRCBotScience_NoTitle(t *testing.T) {
-	p := mustNewParser(t)
-	result := p.Parse(&url.URL{
-		Scheme: "https",
-		Host:   "irc-bot-science.clsr.net",
-		Path:   "notitle",
-	}, nil)
-
-	t.Logf("Result: %+v", result)
-	assert.False(t, result.Ignored)
-	assert.Nil(t, result.Error)
-	assert.Nil(t, result.UserError)
-	assert.Len(t, result.Information, 1)
-	assert.Equal(t, noTitleStr, result.Information[0]["Title"])
-}
-
-func Test_Parser_Parse_IRCBotScience_LongHeaders(t *testing.T) {
-	p := mustNewParser(t)
-	result := parseWithTimeout(p, t, 5*time.Second, &url.URL{
-		Scheme: "https",
-		Host:   "irc-bot-science.clsr.net",
-		Path:   "longheaders",
-	}, nil)
-	for result.FollowUrl != nil {
-		result = parseWithTimeout(p, t, 5*time.Second, result.FollowUrl, nil)
-	}
-
-	t.Logf("Result: %+v", result)
-	assert.True(t, result.Ignored)
-}
-
-func Test_Parser_Parse_IRCBotScience_BigHeader(t *testing.T) {
-	p := mustNewParser(t)
-	result := parseWithTimeout(p, t, 5*time.Second, &url.URL{
-		Scheme: "https",
-		Host:   "irc-bot-science.clsr.net",
-		Path:   "bigheader",
-	}, nil)
-	for result.FollowUrl != nil {
-		result = parseWithTimeout(p, t, 5*time.Second, result.FollowUrl, nil)
-	}
-
-	t.Logf("Result: %+v", result)
-	assert.True(t, result.Ignored)
-}
-
-func Test_Parser_Parse_IRCBotScience_Large(t *testing.T) {
-	p := mustNewParser(t)
-
-	result := parseWithTimeout(p, t, 5*time.Second, &url.URL{
-		Scheme: "https",
-		Host:   "irc-bot-science.clsr.net",
-		Path:   "large",
-	}, nil)
-	for result.FollowUrl != nil {
-		result = parseWithTimeout(p, t, 5*time.Second, result.FollowUrl, nil)
-	}
-
-	t.Logf("Result: %+v", result)
-	assert.False(t, result.Ignored)
-	assert.Nil(t, result.Error)
-	assert.Nil(t, result.UserError)
-	assert.Len(t, result.Information, 1)
-	assert.Equal(t, "If this title is printed, it works correctly.", result.Information[0]["Title"])
-
-}
-
-func Test_Parser_Parse_IRCBotScience_Redirect(t *testing.T) {
-	p := mustNewParser(t)
-	originalUrl := &url.URL{
-		Scheme: "https",
-		Host:   "irc-bot-science.clsr.net",
-		Path:   "redirect",
-	}
-	result := p.Parse(originalUrl, nil)
-
-	t.Logf("Result: %+v", result)
-	assert.False(t, result.Ignored)
-	assert.Nil(t, result.Error)
-	assert.Nil(t, result.UserError)
-	assert.NotNil(t, result.FollowUrl)
-	assert.Equal(t, originalUrl.String(), result.FollowUrl.String())
-}
-
-func Test_Parser_Parse_Hash(t *testing.T) {
-	p := mustNewParser(t)
-	originalUrl := &url.URL{
-		Scheme: "https",
-		Host:   "www.google.com",
-		Path:   "/#invalid",
-	}
-	result := p.Parse(originalUrl, nil)
-
-	t.Logf("Result: %+v", result)
-	assert.False(t, result.Ignored)
-	assert.Nil(t, result.Error)
-	assert.Nil(t, result.UserError)
-}

parsers/wikipedia/parser.go

@@ -23,10 +23,7 @@ func (p *Parser) Init() error {
 }
 
 func (p *Parser) Parse(u *url.URL, referer *url.URL) (result parsers.ParseResult) {
-	if !(strings.EqualFold(u.Scheme, "http") ||
+	if !strings.HasSuffix(strings.ToLower(u.Host), ".wikipedia.org") {
-		strings.EqualFold(u.Scheme, "https")) ||
-		(!strings.HasSuffix(strings.ToLower(u.Host), ".wikipedia.org") &&
-			!strings.EqualFold(u.Host, "wikipedia.org")) {
 		result.Ignored = true
 		return
 	}
@@ -40,11 +37,6 @@ func (p *Parser) Parse(u *url.URL, referer *url.URL) (result parsers.ParseResult
 		}
 
 		// We're using the original host for link localization
-		// or en.wikipedia.org for (www.)wikipedia.org
-		if strings.EqualFold(u.Host, "wikipedia.org") ||
-			strings.EqualFold(u.Host, "www.wikipedia.org") {
-			u.Host = "en.wikipedia.org"
-		}
 		r, err := http.Get("https://" + u.Host + "/api/rest_v1/page/summary/" + titleEscaped)
 		if err != nil {
 			result.Error = err

parsers/youtube/parser.go

@@ -22,7 +22,9 @@ const (
 	youtubeIdType_ChannelId
 	youtubeIdType_Playlist
 
-	header = "\x0301,00You\x0300,04Tube"
+	header = "\x031,0You\x030,4Tube\x03" +
+		"99,99" + /* Fix for KiwiIRC not flushing background color on empty color tag */
+		"\x03" /* Fix for Mibbit interpreting 99 as green instead of transparent */
 )
 
 var (
@@ -166,12 +168,6 @@ func (p *Parser) Parse(u *url.URL, referer *url.URL) (result parsers.ParseResult
 				} else {
 					log.Print(err)
 				}
-
-				if item.ContentDetails.ContentRating != nil {
-					if item.ContentDetails.ContentRating.YtRating == "ytAgeRestricted" {
-						r["AgeRestriction"] = "NSFW"
-					}
-				}
 			}
 			if item.Statistics != nil {
 				r["Views"] = item.Statistics.ViewCount
@@ -212,7 +208,8 @@ func (p *Parser) Parse(u *url.URL, referer *url.URL) (result parsers.ParseResult
 			r := map[string]interface{}{
 				"Header":      header,
 				"IsProfile":   true,
-				"Name":        item.Snippet.Title,
+				"Title":       "Channel",
+				"Author":      item.Snippet.Title,
 				"CountryCode": item.Snippet.Country,
 				"Description": item.Snippet.Description,
 				"ShortUrl":    item.Snippet.CustomUrl,

util.go

@@ -1,6 +1,7 @@
 package main
 
 import (
+	"net/url"
 	"regexp"
 	"strings"
 )
@@ -27,3 +28,34 @@ func stripIrcFormatting(text string) string {
 	text = rxIrcColor.ReplaceAllLiteralString(text, "")
 	return text
 }
+
+func getYouTubeId(uri *url.URL) string {
+	u := &(*uri)
+	u.Scheme = strings.ToLower(u.Scheme)
+	u.Host = strings.ToLower(u.Host)
+
+	// Must be an HTTP URL
+	if u.Scheme != "http" && u.Scheme != "https" {
+		return ""
+	}
+
+	// Remove www. prefix from hostname
+	if strings.HasPrefix(u.Host, "www.") {
+		u.Host = u.Host[4:]
+	}
+
+	switch strings.ToLower(u.Host) {
+	case "youtu.be":
+		// http://youtu.be/{id}
+		if s, err := url.QueryUnescape(strings.TrimLeft(u.Path, "/")); err == nil {
+			return s
+		} else {
+			return ""
+		}
+	case "youtube.com":
+		// http://youtube.com/watch?v={id}
+		return u.Query().Get("v")
+	}
+
+	return ""
+}

util/limitedio/limited_reader.go

@@ -1,48 +0,0 @@
-package limitedio
-
-import "io"
-
-type limitedReader struct {
-	io.Reader
-	rest int
-}
-
-func NewLimitedReader(r io.Reader, limit int) io.Reader {
-	return &limitedReader{r, limit}
-}
-
-func (r *limitedReader) Read(data []byte) (n int, err error) {
-	if r.rest <= 0 {
-		err = io.EOF
-		return
-	}
-
-	var dataSize int
-	if len(data) < r.rest {
-		dataSize = len(data)
-	} else {
-		dataSize = r.rest
-	}
-
-	actualData := make([]byte, dataSize)
-	n, err = r.Reader.Read(actualData)
-	if n > 0 {
-		copy(data, actualData)
-	}
-	r.rest -= (n)
-
-	return
-}
-
-type limitedReadCloser struct {
-	*limitedReader
-	closeMethod func() error
-}
-
-func NewLimitedReadCloser(r io.ReadCloser, limit int) io.Reader {
-	return &limitedReadCloser{&limitedReader{r, limit}, r.Close}
-}
-
-func (rc *limitedReadCloser) Close() error {
-	return rc.closeMethod()
-}

util_test.go

@@ -1,3 +1,33 @@
 package main
 
-// TODO - unit test stripIrcFormatting
+import (
+	"net/url"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func mustParseUrl(u string) *url.URL {
+	if uri, err := url.Parse(u); err == nil {
+		return uri
+	} else {
+		panic(err)
+	}
+}
+
+func Test_GetYouTubeId(t *testing.T) {
+	assert.Equal(t, "aYz-9jUlav-", getYouTubeId(mustParseUrl("http://youtube.com/watch?v=aYz-9jUlav-")))
+	assert.Equal(t, "aYz-9jUlav-", getYouTubeId(mustParseUrl("https://youtube.com/watch?v=aYz-9jUlav-")))
+	assert.Equal(t, "aYz-9jUlav-", getYouTubeId(mustParseUrl("http://www.youtube.com/watch?v=aYz-9jUlav-")))
+	assert.Equal(t, "aYz-9jUlav-", getYouTubeId(mustParseUrl("https://www.youtube.com/watch?v=aYz-9jUlav-")))
+	assert.Equal(t, "aYz-9jUlav-", getYouTubeId(mustParseUrl("http://youtu.be/aYz-9jUlav-")))
+	assert.Equal(t, "aYz-9jUlav-", getYouTubeId(mustParseUrl("https://youtu.be/aYz-9jUlav-")))
+	assert.Equal(t, "aYz-9jUlav-", getYouTubeId(mustParseUrl("http://www.youtu.be/aYz-9jUlav-")))
+	assert.Equal(t, "aYz-9jUlav-", getYouTubeId(mustParseUrl("https://www.youtu.be/aYz-9jUlav-")))
+}
+
+func Benchmark_GetYouTubeId(b *testing.B) {
+	for n := 0; n < b.N; n++ {
+		getYouTubeId(mustParseUrl("http://youtube.com/watch?v=aYz-9jUlav-"))
+	}
+}