Module: Sprockets::EncodingUtils

Extended by:
EncodingUtils
Included in:
EncodingUtils
Defined in:
lib/sprockets/encoding_utils.rb

Overview

Internal: HTTP transport encoding and charset detecting related functions. Mixed into Environment.

Constant Summary collapse

CHARSET_DETECT =

Internal: Shorthand aliases for detecter functions.

{}
BOM =

Internal: Mapping unicode encodings to byte order markers.

{
  Encoding::UTF_32LE => [0xFF, 0xFE, 0x00, 0x00],
  Encoding::UTF_32BE => [0x00, 0x00, 0xFE, 0xFF],
  Encoding::UTF_8    => [0xEF, 0xBB, 0xBF],
  Encoding::UTF_16LE => [0xFF, 0xFE],
  Encoding::UTF_16BE => [0xFE, 0xFF]
}
CHARSET_START =

Internal: @charset bytes

[0x40, 0x63, 0x68, 0x61, 0x72, 0x73, 0x65, 0x74, 0x20, 0x22]
CHARSET_SIZE =
CHARSET_START.size

Instance Method Summary collapse

Instance Method Details

#base64(str) ⇒ Object

Public: Use base64 to encode data.

str - String data

Returns a encoded String


73
74
75
# File 'lib/sprockets/encoding_utils.rb', line 73

def base64(str)
  Base64.strict_encode64(str)
end

#charlock_detect(str) ⇒ Object

Internal: Use Charlock Holmes to detect encoding.

To enable this code path, require 'charlock_holmes'

Returns encoded String.


122
123
124
125
126
127
128
129
130
# File 'lib/sprockets/encoding_utils.rb', line 122

def charlock_detect(str)
  if defined? CharlockHolmes::EncodingDetector
    if detected = CharlockHolmes::EncodingDetector.detect(str)
      str.force_encoding(detected[:encoding]) if detected[:encoding]
    end
  end

  str
end

#deflate(str) ⇒ Object

Public: Use deflate to compress data.

str - String data

Returns a compressed String


19
20
21
22
23
24
25
26
27
28
# File 'lib/sprockets/encoding_utils.rb', line 19

def deflate(str)
  deflater = Zlib::Deflate.new(
    Zlib::BEST_COMPRESSION,
    -Zlib::MAX_WBITS,
    Zlib::MAX_MEM_LEVEL,
    Zlib::DEFAULT_STRATEGY
  )
  deflater << str
  deflater.finish
end

#detect(str) ⇒ Object

Public: Basic string detecter.

Attempts to parse any Unicode BOM otherwise falls back to the environment's external encoding.

str - ASCII-8BIT encoded String

Returns encoded String.


100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
# File 'lib/sprockets/encoding_utils.rb', line 100

def detect(str)
  str = detect_unicode_bom(str)

  # Attempt Charlock detection
  if str.encoding == Encoding::BINARY
    charlock_detect(str)
  end

  # Fallback to environment's external encoding
  if str.encoding == Encoding::BINARY
    str.force_encoding(Encoding.default_external)
  end

  str
end

#detect_css(str) ⇒ Object

Public: Detect and strip @charset from CSS style sheet.

str - String.

Returns a encoded String.


178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
# File 'lib/sprockets/encoding_utils.rb', line 178

def detect_css(str)
  str = detect_unicode_bom(str)

  if name = scan_css_charset(str)
    encoding = Encoding.find(name)
    str = str.dup
    str.force_encoding(encoding)
    len = "@charset \"#{name}\";".encode(encoding).size
    str.slice!(0, len)
    str
  end

  # Fallback to UTF-8
  if str.encoding == Encoding::BINARY
    str.force_encoding(Encoding::UTF_8)
  end

  str
end

#detect_html(str) ⇒ Object

Public: Detect charset from HTML document.

Attempts to parse any Unicode BOM otherwise attempt Charlock detection and finally falls back to the environment's external encoding.

str - String.

Returns a encoded String.


245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
# File 'lib/sprockets/encoding_utils.rb', line 245

def detect_html(str)
  str = detect_unicode_bom(str)

  # Attempt Charlock detection
  if str.encoding == Encoding::BINARY
    charlock_detect(str)
  end

  # Fallback to environment's external encoding
  if str.encoding == Encoding::BINARY
    str.force_encoding(Encoding.default_external)
  end

  str
end

#detect_unicode(str) ⇒ Object

Public: Detect Unicode string.

Attempts to parse Unicode BOM and falls back to UTF-8.

str - ASCII-8BIT encoded String

Returns encoded String.


139
140
141
142
143
144
145
146
147
148
# File 'lib/sprockets/encoding_utils.rb', line 139

def detect_unicode(str)
  str = detect_unicode_bom(str)

  # Fallback to UTF-8
  if str.encoding == Encoding::BINARY
    str.force_encoding(Encoding::UTF_8)
  end

  str
end

#detect_unicode_bom(str) ⇒ Object

Public: Detect and strip BOM from possible unicode string.

str - ASCII-8BIT encoded String

Returns UTF 8/16/32 encoded String without BOM or the original String if no BOM was present.


157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
# File 'lib/sprockets/encoding_utils.rb', line 157

def detect_unicode_bom(str)
  bom_bytes = str.byteslice(0, 4).bytes.to_a

  BOM.each do |encoding, bytes|
    if bom_bytes[0, bytes.size] == bytes
      str = str.dup
      str.force_encoding(Encoding::BINARY)
      str.slice!(0, bytes.size)
      str.force_encoding(encoding)
      return str
    end
  end

  return str
end

#gzip(str) ⇒ Object

Public: Use gzip to compress data.

str - String data

Returns a compressed String


59
60
61
62
63
64
65
66
# File 'lib/sprockets/encoding_utils.rb', line 59

def gzip(str)
  io = StringIO.new
  gz = Zlib::GzipWriter.new(io, Zlib::BEST_COMPRESSION)
  gz.mtime = 1
  gz << str
  gz.finish
  io.string
end

#scan_css_charset(str) ⇒ Object

Internal: Scan binary CSS string for @charset encoding name.

str - ASCII-8BIT encoded String

Returns encoding String name or nil.


208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
# File 'lib/sprockets/encoding_utils.rb', line 208

def scan_css_charset(str)
  buf = []
  i = 0

  str.each_byte.each do |byte|
    # Halt on line breaks
    break if byte == 0x0A || byte == 0x0D

    # Only ascii bytes
    next unless 0x0 < byte && byte <= 0xFF

    if i < CHARSET_SIZE
    elsif i == CHARSET_SIZE
      if buf == CHARSET_START
        buf = []
      else
        break
      end
    elsif byte == 0x22
      return buf.pack('C*')
    end

    buf << byte
    i += 1
  end

  nil
end

#unmarshaled_deflated(str, window_bits = -Zlib::MAX_WBITS)) ⇒ Object

Internal: Unmarshal optionally deflated data.

Checks leading marshal header to see if the bytes are uncompressed otherwise inflate the data an unmarshal.

str - Marshaled String window_bits - Integer deflate window size. See ZLib::Inflate.new()

Returns unmarshaled Object or raises an Exception.


39
40
41
42
43
44
45
46
47
48
49
50
51
52
# File 'lib/sprockets/encoding_utils.rb', line 39

def unmarshaled_deflated(str, window_bits = -Zlib::MAX_WBITS)
  major, minor = str[0], str[1]
  if major && major.ord == Marshal::MAJOR_VERSION &&
      minor && minor.ord <= Marshal::MINOR_VERSION
    marshaled = str
  else
    begin
      marshaled = Zlib::Inflate.new(window_bits).inflate(str)
    rescue Zlib::DataError
      marshaled = str
    end
  end
  Marshal.load(marshaled)
end