rss parser

src/feed/rss.go

@@ -6,21 +6,23 @@ package feed
 
 import (
 	"encoding/xml"
+	"fmt"
+	"io"
 )
 
 type rssFeed struct {
 	XMLName xml.Name  `xml:"rss"`
 	Version string    `xml:"version,attr"`
 	Title   string    `xml:"channel>title"`
-	Links   []rssLink `xml:"channel>link"`
+	Link    string    `xml:"channel>link"`
 	Items   []rssItem `xml:"channel>item"`
 }
 
 type rssItem struct {
 	GUID           string         `xml:"guid"`
-	Title          []rssTitle     `xml:"title"`
-	Links          []rssLink      `xml:"link"`
-	Description    string         `xml:"description"`
+	Title          string         `xml:"title"`
+	Link           string         `xml:"link"`
+	Description    string         `xml:"rss description"`
 	PubDate        string         `xml:"pubDate"`
 	EnclosureLinks []rssEnclosure `xml:"enclosure"`
 
@@ -53,3 +55,31 @@ type rssEnclosure struct {
 	Type   string `xml:"type,attr"`
 	Length string `xml:"length,attr"`
 }
+
+func ParseRSS(r io.Reader) (*Feed, error) {
+	f := rssFeed{}
+
+	decoder := xml.NewDecoder(r)
+	decoder.DefaultSpace = "rss"
+	if err := decoder.Decode(&f); err != nil {
+		fmt.Println(err)
+		return nil, err
+	}
+
+	feed := &Feed{
+		Title:   f.Title,
+		SiteURL: f.Link,
+	}
+	for _, e := range f.Items {
+		date, _ := dateParse(first(e.DublinCoreDate, e.PubDate))
+
+		feed.Items = append(feed.Items, Item{
+			GUID:       first(e.GUID, e.Link),
+			Date:       date,
+			URL:        e.Link,
+			Title:      e.Title,
+			Content:    e.Description,
+		})
+	}
+	return feed, nil
+}

src/feed/rss_test.go

@@ -0,0 +1,56 @@
+package feed
+
+import (
+	"reflect"
+	"strings"
+	"testing"
+)
+
+func TestRSSFeed(t *testing.T) {
+	have, _ := ParseRSS(strings.NewReader(`
+		<?xml version="1.0"?>
+		<!DOCTYPE rss SYSTEM "http://my.netscape.com/publish/formats/rss-0.91.dtd">
+		<rss version="0.91">
+		<channel>
+			<language>en</language>
+			<description>???</description>
+			<link>http://www.scripting.com/</link>
+			<title>Scripting News</title>
+			<item>
+				<title>Title 1</title>
+				<link>http://www.scripting.com/one/</link>
+				<description>Description 1</description>
+			</item>
+			<item>
+				<title>Title 2</title>
+				<link>http://www.scripting.com/two/</link>
+				<description>Description 2</description>
+			</item>
+		</channel>
+		</rss>
+	`))
+	want := &Feed{
+		Title: "Scripting News",
+		SiteURL: "http://www.scripting.com/",
+		Items: []Item{
+			{
+				GUID: "http://www.scripting.com/one/",
+				URL: "http://www.scripting.com/one/",
+				Title: "Title 1",
+				Content: "Description 1",
+			},
+			{
+				GUID: "http://www.scripting.com/two/",
+				URL: "http://www.scripting.com/two/",
+				Title: "Title 2",
+				Content: "Description 2",
+			},
+		},
+	}
+
+	if !reflect.DeepEqual(want, have) {
+		t.Logf("want: %#v", want)
+		t.Logf("have: %#v", have)
+		t.Fatal("invalid rss")
+	}
+}