From 36d773223481dd42d31499b3ea73e6999ff9f58e Mon Sep 17 00:00:00 2001 From: Frédéric Guillot Date: Wed, 18 Sep 2019 22:27:25 -0700 Subject: Disable strict XML parsing This change should improve parsing of broken XML feeds. See https://golang.org/pkg/encoding/xml/#Decoder --- reader/opml/parser.go | 1 + reader/opml/parser_test.go | 34 ++++++++++++++++++++++++++++++++++ 2 files changed, 35 insertions(+) (limited to 'reader/opml') diff --git a/reader/opml/parser.go b/reader/opml/parser.go index 1bdec20..97974bb 100644 --- a/reader/opml/parser.go +++ b/reader/opml/parser.go @@ -17,6 +17,7 @@ func Parse(data io.Reader) (SubcriptionList, *errors.LocalizedError) { feeds := new(opml) decoder := xml.NewDecoder(data) decoder.Entity = xml.HTMLEntity + decoder.Strict = false decoder.CharsetReader = encoding.CharsetReader err := decoder.Decode(feeds) diff --git a/reader/opml/parser_test.go b/reader/opml/parser_test.go index f6ce634..6c09db8 100644 --- a/reader/opml/parser_test.go +++ b/reader/opml/parser_test.go @@ -193,6 +193,40 @@ func TestParseOpmlVersion1WithoutOuterOutline(t *testing.T) { } } } + +func TestParseOpmlWithInvalidCharacterEntity(t *testing.T) { + data := ` + + + mySubscriptions.opml + + + + + + + + ` + + var expected SubcriptionList + expected = append(expected, &Subcription{Title: "Feed 1", FeedURL: "http://example.org/feed1/a&b", SiteURL: "http://example.org/c&d", CategoryName: ""}) + + subscriptions, err := Parse(bytes.NewBufferString(data)) + if err != nil { + t.Error(err) + } + + if len(subscriptions) != 1 { + t.Errorf("Wrong number of subscriptions: %d instead of %d", len(subscriptions), 1) + } + + for i := 0; i < len(subscriptions); i++ { + if !subscriptions[i].Equals(expected[i]) { + t.Errorf(`Subscription are different: "%v" vs "%v"`, subscriptions[i], expected[i]) + } + } +} + func TestParseInvalidXML(t *testing.T) { data := `garbage` _, err := Parse(bytes.NewBufferString(data)) -- cgit v1.2.3