You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

escape.go 1.6KB

1234567891011121314151617181920212223242526272829303132333435363738394041424344
  1. // Copyright 2022 The Gitea Authors. All rights reserved.
  2. // SPDX-License-Identifier: MIT
  3. //go:generate go run invisible/generate.go -v -o ./invisible_gen.go
  4. //go:generate go run ambiguous/generate.go -v -o ./ambiguous_gen.go ambiguous/ambiguous.json
  5. package charset
  6. import (
  7. "html/template"
  8. "io"
  9. "strings"
  10. "code.gitea.io/gitea/modules/log"
  11. "code.gitea.io/gitea/modules/setting"
  12. "code.gitea.io/gitea/modules/translation"
  13. )
  14. // RuneNBSP is the codepoint for NBSP
  15. const RuneNBSP = 0xa0
  16. // EscapeControlHTML escapes the unicode control sequences in a provided html document
  17. func EscapeControlHTML(html template.HTML, locale translation.Locale, allowed ...rune) (escaped *EscapeStatus, output template.HTML) {
  18. sb := &strings.Builder{}
  19. escaped, _ = EscapeControlReader(strings.NewReader(string(html)), sb, locale, allowed...) // err has been handled in EscapeControlReader
  20. return escaped, template.HTML(sb.String())
  21. }
  22. // EscapeControlReader escapes the unicode control sequences in a provided reader of HTML content and writer in a locale and returns the findings as an EscapeStatus
  23. func EscapeControlReader(reader io.Reader, writer io.Writer, locale translation.Locale, allowed ...rune) (escaped *EscapeStatus, err error) {
  24. if !setting.UI.AmbiguousUnicodeDetection {
  25. _, err = io.Copy(writer, reader)
  26. return &EscapeStatus{}, err
  27. }
  28. outputStream := &HTMLStreamerWriter{Writer: writer}
  29. streamer := NewEscapeStreamer(locale, outputStream, allowed...).(*escapeStreamer)
  30. if err = StreamHTML(reader, streamer); err != nil {
  31. streamer.escaped.HasError = true
  32. log.Error("Error whilst escaping: %v", err)
  33. }
  34. return streamer.escaped, err
  35. }