Resolve xml:base URLs without switching out the BaseStack

This provides an equivalent fix that doesn't do any inelegant swapping
out of the BaseStack. It also doesn't change `goxpp`'s public API by
essentially copying `XmlBaseResolveUrl` to `gofeed`.

atom/parser.go

@@ -658,20 +658,12 @@ func (ap *Parser) parseAtomText(p *xpp.XMLPullParser) (string, error) {
 		InnerXML string `xml:",innerxml"`
 	}
 
-    // DecodeElement pops the base stack if the element contains an xml:base
-    // attribute, so we need to save and restore it before resolving any
-    // relative URLs below
-	oldBase := p.BaseStack
+	// get current base URL before it is clobbered by DecodeElement
+	base := p.BaseStack.Top()
 	err := p.DecodeElement(&text)
 	if err != nil {
 		return "", err
 	}
-	newBase := p.BaseStack
-	p.BaseStack = oldBase
-	defer func() {
-		// pop base when we're done with the decoded element
-		p.BaseStack = newBase
-	}()
 
 	result := text.InnerXML
 	result = strings.TrimSpace(result)
@@ -682,7 +674,7 @@ func (ap *Parser) parseAtomText(p *xpp.XMLPullParser) (string, error) {
 	if strings.Contains(result, "<![CDATA[") {
 		result = shared.StripCDATA(result)
 		if lowerType == "html" || strings.Contains(lowerType, "xhtml") {
-			result, _ = shared.ResolveHTML(p, result)
+			result, _ = shared.ResolveHTML(base, result)
 		}
 	} else {
 		// decode non-CDATA contents depending on type
@@ -693,12 +685,12 @@ func (ap *Parser) parseAtomText(p *xpp.XMLPullParser) (string, error) {
 			result, err = shared.DecodeEntities(result)
 		} else if strings.Contains(lowerType, "xhtml") {
 			result = ap.stripWrappingDiv(result)
-			result, _ = shared.ResolveHTML(p, result)
+			result, _ = shared.ResolveHTML(base, result)
 		} else if lowerType == "html" {
 			result = ap.stripWrappingDiv(result)
 			result, err = shared.DecodeEntities(result)
 			if err == nil {
-				result, _ = shared.ResolveHTML(p, result)
+				result, _ = shared.ResolveHTML(base, result)
 			}
 		} else {
 			decodedStr, err := base64.StdEncoding.DecodeString(result)
@@ -711,7 +703,7 @@ func (ap *Parser) parseAtomText(p *xpp.XMLPullParser) (string, error) {
 	// resolve relative URIs in URI-containing elements according to xml:base
 	name := strings.ToLower(p.Name)
 	if atomUriElements[name] {
-		resolved, err := p.XmlBaseResolveUrl(result)
+		resolved, err := shared.XmlBaseResolveUrl(base, result)
 		if resolved != nil && err == nil {
 			result = resolved.String()
 		}

internal/shared/xmlbase.go

@@ -3,6 +3,7 @@ package shared
 import (
 	"bytes"
 	"fmt"
+	"net/url"
 	"strings"
 
 	xpp "github.com/mmcdole/goxpp"
@@ -95,11 +96,31 @@ func resolveAttrs(p *xpp.XMLPullParser) error {
 	return nil
 }
 
+// resolve u relative to b
+func XmlBaseResolveUrl(b *url.URL, u string) (*url.URL, error) {
+	relURL, err := url.Parse(u)
+	if err != nil {
+		return nil, err
+	}
+
+	if b == nil {
+		return relURL, nil
+	}
+
+	if b.Path != "" && u != "" && b.Path[len(b.Path)-1] != '/' {
+		// There's no reason someone would use a path in xml:base if they
+		// didn't mean for it to be a directory
+		b.Path = b.Path + "/"
+	}
+	absURL := b.ResolveReference(relURL)
+	return absURL, nil
+}
+
 // Transforms html by resolving any relative URIs in attributes
 // if an error occurs during parsing or serialization, then the original string
 // is returned along with the error.
-func ResolveHTML(p *xpp.XMLPullParser, relHTML string) (string, error) {
-	if p.BaseStack.Top() == nil {
+func ResolveHTML(base *url.URL, relHTML string) (string, error) {
+	if base == nil {
 		return relHTML, nil
 	}
 
@@ -117,7 +138,7 @@ func ResolveHTML(p *xpp.XMLPullParser, relHTML string) (string, error) {
 		if n.Type == html.ElementNode {
 			for i, a := range n.Attr {
 				if htmlURIAttrs[a.Key] {
-					absVal, err := p.XmlBaseResolveUrl(a.Val)
+					absVal, err := XmlBaseResolveUrl(base, a.Val)
 					if absVal != nil && err == nil {
 						n.Attr[i].Val = absVal.String()
 					}