Use composition

1 files changed, 50 insertions, 22 deletions

diff --git a/unxml.go b/unxml.go
index c34f29f..cf3f238 100644
--- a/unxml.go
+++ b/unxml.go
@@ -18,6 +18,21 @@ type Reader struct {
 	tokenizer *html.Tokenizer
 }
 
+type ElementReader struct {
+	r           Reader
+	tagsinstack map[string]int
+}
+
+//type stack []string
+//
+//func (s stack) Empty() bool  { return len(s) == 0 }
+//func (s stack) Peek() string { return s[len(s)-1] }
+//func (s *stack) Pop() string {
+//	d := (*s)[len(*s)-1]
+//	(*s) = (*s)[:len(*s)-1]
+//	return d
+//}
+
 func NewReaderKeepTags(r io.Reader, tagstokeep []string) *Reader {
 	var tagmap map[string]bool
 
@@ -36,52 +51,65 @@ func NewReaderKeepTags(r io.Reader, tagstokeep []string) *Reader {
 	}
 }
 
-func (r *Reader) Read(out []byte) (int, error) {
+func (r *ElementReader) Read(out []byte) (int, error) {
 	fmt.Fprintf(os.Stderr, "Read has been called.\n")
 	var err error
 
-	r.count = 0
+	r.r.count = 0
 	n := 0
 	lenout := len(out)
 	if lenout == 0 {
-		return r.count, nil
+		return r.r.count, nil
 	}
 
-	lenlr := len(r.lastread)
+	lenlr := len(r.r.lastread)
 	if lenlr > 0 {
-		n = copy(out[0:], r.lastread)
-		r.count += n
-		r.lastread = make([]byte, len(out))
+		n = copy(out[0:], r.r.lastread)
+		r.r.count += n
+		r.r.lastread = make([]byte, len(out))
 		lenout -= n
 	}
 
 	for {
-		tt := r.tokenizer.Next()
+		tt := r.r.tokenizer.Next()
 
 		switch tt {
 		case html.ErrorToken:
-			fmt.Fprintf(os.Stderr, "There was an error when parsing the html: %s, %s\n", tt, r.tokenizer.Err())
-			return r.count, io.EOF
+			//fmt.Fprintf(os.Stderr, "There was an error when parsing the html: %s, %s\n", tt, r.r.tokenizer.Err())
+			return r.r.count, io.EOF
 
 		case html.TextToken:
-			text := r.tokenizer.Text()
+			text := r.r.tokenizer.Text()
 			lentext := len(text)
 			if lentext <= lenout {
-				n = copy(out[r.count:], text)
-				r.count += n
+				n = copy(out[r.r.count:], text)
+				r.r.count += n
 				lenout -= n
-				//fmt.Printf("HAD SPACE: %q, count: %d, err: %s\n", text, r.count, err)
+				//fmt.Printf("HAD SPACE: %q, count: %d, err: %s\n", text, r.r.count, err)
 			} else {
-				n = copy(out[r.count:], text[:lenout-1])
-				r.count += n
-				r.lastread = text[lenout-1:]
-				//fmt.Printf("HAD NO SPACE: count: %d, err: %s\n", r.count, err)
-				return r.count, err
+				n = copy(out[r.r.count:], text[:lenout-1])
+				r.r.count += n
+				r.r.lastread = text[lenout-1:]
+				//fmt.Printf("HAD NO SPACE: count: %d, err: %s\n", r.r.count, err)
+				return r.r.count, err
 			}
 
-		case html.StartTagToken, html.EndTagToken:
-			_, _ = r.tokenizer.TagName()
-			//fmt.Printf("TagName: %s\n", tn)
+		case html.StartTagToken:
+			tn, _ := r.r.tokenizer.TagName()
+			if _, ok := r.r.tagmap[string(tn)]; ok {
+			}
+			fmt.Printf("TagName: %s\n", tn)
+
+		case html.EndTagToken:
+			tn, _ := r.r.tokenizer.TagName()
+			fmt.Printf("TagEndName: %s\n", tn)
+			if count, ok := r.tagsinstack[string(tn)]; ok {
+				if count == 1 {
+					delete(r.tagsinstack, string(tn))
+				} else {
+					r.tagsinstack[string(tn)]--
+				}
+			}
 		}
 	}
 }