Keep track of the current xml:base value

Tracks xml:base attributes in a stack of *url.URLs. Consumers of the
parser can access the top-level URL through `XMLPullParser.BaseStack.Top()`

This is useful for applications that need to resolve URLs in XML
documents relative to the xml:base attributes. To that end, a helper
method is provided which will resolve a relative string to an absolute
URL according to the current base:
`func (p *XMLPullParser) XmlBaseResolveUrl(u string) (*url.URL, error)`

Includes a single test. It is not comprehensive, but it checks for
xml:base to two levels, tests resolving a string against the current
base, as well as resolution of relative xml:base values.

Author: cristoper

xpp.go

@@ -5,12 +5,15 @@ import (
 	"errors"
 	"fmt"
 	"io"
+	"net/url"
 	"strings"
 )
 
 type XMLEventType int
 type CharsetReader func(charset string, input io.Reader) (io.Reader, error)
 
+const xmlNSURI = "http://www.w3.org/XML/1998/namespace"
+
 const (
 	StartDocument XMLEventType = iota
 	EndDocument
@@ -24,10 +27,33 @@ const (
 	// TODO: CDSECT ?
 )
 
+type urlStack []*url.URL
+
+func (s *urlStack) push(u *url.URL) {
+	*s = append([]*url.URL{u}, *s...)
+}
+
+func (s *urlStack) pop() *url.URL {
+	if s == nil || len(*s) == 0 {
+		return nil
+	}
+	var top *url.URL
+	top, *s = (*s)[0], (*s)[1:]
+	return top
+}
+
+func (s *urlStack) Top() *url.URL {
+	if s == nil || len(*s) == 0 {
+		return nil
+	}
+	return (*s)[0]
+}
+
 type XMLPullParser struct {
 	// Document State
 	Spaces      map[string]string
 	SpacesStack []map[string]string
+	BaseStack    urlStack
 
 	// Token State
 	Depth int
@@ -214,6 +240,7 @@ func (p *XMLPullParser) DecodeElement(v interface{}) error {
 	p.Depth--
 	p.Name = name
 	p.token = nil
+	p.popBase()
 	return nil
 }
 
@@ -263,6 +290,26 @@ func (p *XMLPullParser) EventType(t xml.Token) (event XMLEventType) {
 	return
 }
 
+// resolve the given string as a URL relative to current xml:base
+func (p *XMLPullParser) XmlBaseResolveUrl(u string) (*url.URL, error) {
+	curr := p.BaseStack.Top()
+	if curr == nil {
+		return nil, nil
+	}
+
+	relURL, err := url.Parse(u)
+	if err != nil {
+		return nil, err
+	}
+	if curr.Path != "" && u != "" && curr.Path[len(curr.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
+		curr.Path = curr.Path + "/"
+	}
+	absURL := curr.ResolveReference(relURL)
+	return absURL, nil
+}
+
 func (p *XMLPullParser) processToken(t xml.Token) {
 	switch tt := t.(type) {
 	case xml.StartElement:
@@ -286,6 +333,7 @@ func (p *XMLPullParser) processStartToken(t xml.StartElement) {
 	p.Name = t.Name.Local
 	p.Space = t.Name.Space
 	p.trackNamespaces(t)
+	p.pushBase()
 }
 
 func (p *XMLPullParser) processEndToken(t xml.EndElement) {
@@ -297,6 +345,7 @@ func (p *XMLPullParser) processEndToken(t xml.EndElement) {
 		p.Spaces = p.SpacesStack[len(p.SpacesStack)-1]
 	}
 	p.Name = t.Name.Local
+	p.popBase()
 }
 
 func (p *XMLPullParser) processCharDataToken(t xml.CharData) {
@@ -340,3 +389,40 @@ func (p *XMLPullParser) trackNamespaces(t xml.StartElement) {
 	p.Spaces = newSpace
 	p.SpacesStack = append(p.SpacesStack, newSpace)
 }
+
+// returns the popped base URL
+func (p *XMLPullParser) popBase() string {
+	url := p.BaseStack.pop()
+	if url != nil {
+		return url.String()
+	}
+	return ""
+}
+
+// Searches current attributes for xml:base and updates the urlStack
+func (p *XMLPullParser) pushBase() error {
+	var base string
+	// search list of attrs for "xml:base"
+	for _, attr := range p.Attrs {
+		if attr.Name.Local == "base" && attr.Name.Space == xmlNSURI {
+			base = attr.Value
+			break
+		}
+	}
+	if base == "" {
+		// no base attribute found
+		return nil
+	}
+
+	newURL, err := url.Parse(base)
+	if err != nil {
+		return err
+	}
+
+	topURL := p.BaseStack.Top()
+	if topURL != nil {
+		newURL = topURL.ResolveReference(newURL)
+	}
+	p.BaseStack.push(newURL)
+	return nil
+}

xpp_test.go

@@ -5,7 +5,7 @@ import (
 	"io"
 	"testing"
 
-	"github.com/mmcdole/goxpp"
+	xpp "github.com/mmcdole/goxpp"
 	"github.com/stretchr/testify/assert"
 )
 
@@ -85,6 +85,37 @@ func TestDecodeElementDepth(t *testing.T) {
 	p.DecodeElement(&v{})
 }
 
+func TestXMLBase(t *testing.T) {
+	crReader := func(charset string, input io.Reader) (io.Reader, error) {
+		return input, nil
+	}
+	r := bytes.NewBufferString(`<root xml:base="https://example.org/"><d2 xml:base="relative">foo</d2><d2>bar</d2></root>`)
+	p := xpp.NewXMLPullParser(r, false, crReader)
+
+	type v struct{}
+
+	// move to root
+	p.NextTag()
+	assert.Equal(t, "root", p.Name)
+	assert.Equal(t, "https://example.org/", p.BaseStack.Top().String())
+
+	// decode first <d2>
+	p.NextTag()
+	assert.Equal(t, "d2", p.Name)
+	assert.Equal(t, "https://example.org/relative", p.BaseStack.Top().String())
+	
+	resolved, err := p.XmlBaseResolveUrl("test")
+	assert.NoError(t, err)
+	assert.Equal(t, "https://example.org/relative/test", resolved.String())
+	p.DecodeElement(&v{})
+
+	// decode second <d2>
+	p.NextTag()
+	assert.Equal(t, "d2", p.Name)
+	assert.Equal(t, "https://example.org/", p.BaseStack.Top().String())
+	p.DecodeElement(&v{})
+}
+
 func toNextStart(t *testing.T, p *xpp.XMLPullParser) {
 	for {
 		tok, err := p.NextToken()