/html-scan.go
Created Jan 4, 2019
| package main | |
| import ( | |
| "bufio" | |
| "bytes" | |
| "fmt" | |
| ) | |
| const html = ` | |
| <html> | |
| <head> | |
| <title>text</title> | |
| </head> | |
| <body> | |
| <h1>more text</h1> | |
| </body> | |
| </html>` | |
| func main() { | |
| src := bytes.NewReader([]byte(html)) | |
| scanner := bufio.NewScanner(src) | |
| scanner.Split(splitHTML) | |
| for scanner.Scan() { | |
| fmt.Printf("token: %q\n", scanner.Text()) | |
| } | |
| fmt.Println(scanner.Err()) | |
| } | |
| func splitHTML(data []byte, atEOF bool) (advance int, token []byte, err error) { | |
| tagStart := bytes.Index(data, []byte("<")) | |
| if tagStart > 0 { | |
| trimmed := bytes.TrimSpace(data[:tagStart]) | |
| if len(trimmed) > 0 { | |
| return tagStart, trimmed, nil | |
| } | |
| } | |
| if tagStart == -1 { | |
| if atEOF { | |
| return 0, bytes.TrimSpace(data), bufio.ErrFinalToken | |
| } else { | |
| return 0, nil, nil | |
| } | |
| } | |
| tagEnd := bytes.Index(data, []byte(">")) | |
| if tagEnd == -1 { | |
| if atEOF { | |
| return 0, nil, fmt.Errorf("malformed tag: %q", string(data)) | |
| } else { | |
| return 0, nil, nil | |
| } | |
| } | |
| return tagEnd + 1, data[tagStart : tagEnd+1], nil | |
| } |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment