Initial commit

1 files changed, 70 insertions, 0 deletions

diff --git a/unxml.go b/unxml.go
new file mode 100644
index 0000000..d8d588e
--- /dev/null
+++ b/unxml.go
@@ -0,0 +1,70 @@
+/* See LICENSE file for copyright and license details. */
+
+package unxml
+
+import (
+	"bytes"
+	"fmt"
+	"io"
+	"os"
+
+	"golang.org/x/net/html"
+)
+
+type Reader struct {
+	reader io.Reader
+	tagmap map[string]bool
+}
+
+func NewReaderKeepTags(r io.Reader, tagstokeep []string) *Reader {
+	var tagmap map[string]bool
+
+	if len(tagstokeep) > 0 {
+		tagmap = make(map[string]bool, 10)
+		for _, tag := range tagstokeep {
+			tagmap[tag] = true
+		}
+	} else {
+		tagmap = nil
+	}
+
+	return &Reader{reader: r,
+		tagmap: tagmap,
+	}
+}
+
+func (r *Reader) Read(out []byte) (n int, err error) {
+	hr := html.NewTokenizer(r.reader)
+	buf := bytes.NewBuffer(make([]byte, len(out)))
+	depth := 0
+
+	for {
+		tt := hr.Next()
+
+		switch tt {
+		case html.ErrorToken:
+			fmt.Fprintf(os.Stderr, "There was an error when reading from the underlying os.Reader: %s", tt)
+			return 0, hr.Err()
+
+		case html.TextToken:
+			if depth > 0 {
+				// emitBytes should copy the []byte it receives,
+				// if it doesn't process it immediately.
+				n, err = buf.Write(hr.Text())
+			}
+
+		case html.StartTagToken, html.EndTagToken:
+			tn, _ := hr.TagName()
+
+			if len(tn) == 1 && tn[0] == 'a' {
+				if tt == html.StartTagToken {
+					depth++
+				} else {
+					depth--
+				}
+			}
+		}
+	}
+
+	return
+}