add charsetreader to xmlreader

This commit is contained in:
Nazar Kanaev
2021-03-23 11:31:46 +00:00
parent 454eff0155
commit 646519e074
7 changed files with 19 additions and 6 deletions

View File

@@ -59,7 +59,7 @@ func (links atomLinks) First(rel string) string {
func ParseAtom(r io.Reader) (*Feed, error) {
srcfeed := atomFeed{}
decoder := xml.NewDecoder(r)
decoder := xmlDecoder(r)
if err := decoder.Decode(&srcfeed); err != nil {
return nil, err
}

View File

@@ -19,7 +19,7 @@ func sniff(lookup string) (string, processor) {
lookup = strings.TrimSpace(lookup)
switch lookup[0] {
case '<':
decoder := xml.NewDecoder(strings.NewReader(lookup))
decoder := xmlDecoder(strings.NewReader(lookup))
for {
token, _ := decoder.Token()
if token == nil {

View File

@@ -12,6 +12,10 @@ func TestSniff(t *testing.T) {
`<?xml version="1.0"?><rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"></rdf:RDF>`,
"rdf",
},
{
`<?xml version="1.0" encoding="ISO-8859-1"?><rss version="2.0"><channel></channel></rss>`,
"rss",
},
{
`<?xml version="1.0"?><rss version="2.0"><channel></channel></rss>`,
"rss",

View File

@@ -27,7 +27,7 @@ type rdfItem struct {
func ParseRDF(r io.Reader) (*Feed, error) {
srcfeed := rdfFeed{}
decoder := xml.NewDecoder(r)
decoder := xmlDecoder(r)
if err := decoder.Decode(&srcfeed); err != nil {
return nil, err
}

View File

@@ -6,7 +6,6 @@ package feed
import (
"encoding/xml"
"fmt"
"io"
)
@@ -59,10 +58,9 @@ type rssEnclosure struct {
func ParseRSS(r io.Reader) (*Feed, error) {
srcfeed := rssFeed{}
decoder := xml.NewDecoder(r)
decoder := xmlDecoder(r)
decoder.DefaultSpace = "rss"
if err := decoder.Decode(&srcfeed); err != nil {
fmt.Println(err)
return nil, err
}

View File

@@ -1,6 +1,9 @@
package feed
import (
"encoding/xml"
"io"
"golang.org/x/net/html/charset"
"time"
)
@@ -13,6 +16,13 @@ func firstNonEmpty(vals ...string) string {
return ""
}
func xmlDecoder(r io.Reader) *xml.Decoder {
decoder := xml.NewDecoder(r)
decoder.Strict = false
decoder.CharsetReader = charset.NewReaderLabel
return decoder
}
// taken from github.com/mjibson/goread
var dateFormats = []string{
time.RFC822, // RSS