format_test.go 3.9 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132
  1. // SPDX-FileCopyrightText: Copyright The Miniflux Authors. All rights reserved.
  2. // SPDX-License-Identifier: Apache-2.0
  3. package parser // import "miniflux.app/v2/internal/reader/parser"
  4. import (
  5. "strings"
  6. "testing"
  7. )
  8. func TestDetectRDF(t *testing.T) {
  9. data := `<?xml version="1.0"?><rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://my.netscape.com/rdf/simple/0.9/"></rdf:RDF>`
  10. format, _ := DetectFeedFormat(strings.NewReader(data))
  11. if format != FormatRDF {
  12. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatRDF)
  13. }
  14. }
  15. func TestDetectRSS(t *testing.T) {
  16. data := `<?xml version="1.0"?><rss version="2.0"><channel></channel></rss>`
  17. format, _ := DetectFeedFormat(strings.NewReader(data))
  18. if format != FormatRSS {
  19. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatRSS)
  20. }
  21. }
  22. func TestDetectAtom10(t *testing.T) {
  23. data := `<?xml version="1.0" encoding="utf-8"?><feed xmlns="http://www.w3.org/2005/Atom"></feed>`
  24. format, _ := DetectFeedFormat(strings.NewReader(data))
  25. if format != FormatAtom {
  26. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatAtom)
  27. }
  28. }
  29. func TestDetectAtom03(t *testing.T) {
  30. data := `<?xml version="1.0" encoding="utf-8"?><feed version="0.3" xmlns="http://purl.org/atom/ns#" xmlns:dc="http://purl.org/dc/elements/1.1/" xml:lang="en"></feed>`
  31. format, _ := DetectFeedFormat(strings.NewReader(data))
  32. if format != FormatAtom {
  33. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatAtom)
  34. }
  35. }
  36. func TestDetectAtomWithISOCharset(t *testing.T) {
  37. data := `<?xml version="1.0" encoding="ISO-8859-15"?><feed xmlns="http://www.w3.org/2005/Atom"></feed>`
  38. format, _ := DetectFeedFormat(strings.NewReader(data))
  39. if format != FormatAtom {
  40. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatAtom)
  41. }
  42. }
  43. func TestDetectJSON(t *testing.T) {
  44. data := `
  45. {
  46. "version" : "https://jsonfeed.org/version/1",
  47. "title" : "Example"
  48. }
  49. `
  50. format, _ := DetectFeedFormat(strings.NewReader(data))
  51. if format != FormatJSON {
  52. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatJSON)
  53. }
  54. }
  55. func TestDetectUnknown(t *testing.T) {
  56. data := `
  57. <!DOCTYPE html> <html> </html>
  58. `
  59. format, _ := DetectFeedFormat(strings.NewReader(data))
  60. if format != FormatUnknown {
  61. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatUnknown)
  62. }
  63. }
  64. func TestDetectJSONWithLargeLeadingWhitespace(t *testing.T) {
  65. leadingWhitespace := strings.Repeat(" ", 10000)
  66. data := leadingWhitespace + `{
  67. "version" : "https://jsonfeed.org/version/1",
  68. "title" : "Example with lots of leading whitespace"
  69. }`
  70. format, _ := DetectFeedFormat(strings.NewReader(data))
  71. if format != FormatJSON {
  72. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatJSON)
  73. }
  74. }
  75. func TestDetectJSONWithMixedWhitespace(t *testing.T) {
  76. leadingWhitespace := strings.Repeat("\n\t ", 10000)
  77. data := leadingWhitespace + `{
  78. "version" : "https://jsonfeed.org/version/1",
  79. "title" : "Example with mixed whitespace"
  80. }`
  81. format, _ := DetectFeedFormat(strings.NewReader(data))
  82. if format != FormatJSON {
  83. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatJSON)
  84. }
  85. }
  86. func TestDetectOnlyWhitespace(t *testing.T) {
  87. data := strings.Repeat(" \t\n\r", 10000)
  88. format, _ := DetectFeedFormat(strings.NewReader(data))
  89. if format != FormatUnknown {
  90. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatUnknown)
  91. }
  92. }
  93. func TestDetectJSONSmallerThanBuffer(t *testing.T) {
  94. data := `{"version":"1"}` // This is only 15 bytes, well below the 32-byte buffer
  95. format, _ := DetectFeedFormat(strings.NewReader(data))
  96. if format != FormatJSON {
  97. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatJSON)
  98. }
  99. }
  100. func TestDetectJSONWithWhitespaceSmallerThanBuffer(t *testing.T) {
  101. data := ` {"title":"test"} `
  102. format, _ := DetectFeedFormat(strings.NewReader(data))
  103. if format != FormatJSON {
  104. t.Errorf(`Wrong format detected: %q instead of %q`, format, FormatJSON)
  105. }
  106. }