Class: FilesHunter::Decoders::Text

Inherits:
FilesHunter::Decoder show all
Defined in:
lib/fileshunter/Decoders/Text.rb

Constant Summary collapse

UTF_16BE_BOM = 
"\xFE\xFF".force_encoding(Encoding::ASCII_8BIT)
UTF_16LE_BOM = 
"\xFF\xFE".force_encoding(Encoding::ASCII_8BIT)
NULL_CHAR = 
"\x00".force_encoding(Encoding::ASCII_8BIT)
NL_CHAR = 
"\n".force_encoding(Encoding::ASCII_8BIT)

Instance Method Summary collapse

Methods inherited from FilesHunter::Decoder

#segments_found, #setup

Instance Method Details

#find_segments ⇒ Object

Find segments from a given data



13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
 
# File 'lib/fileshunter/Decoders/Text.rb', line 13

def find_segments
  current_offset = @begin_offset
  while (current_offset < @end_offset)
    # First find a new line character from current_offset
    newline_offset = @data.index(NL_CHAR, current_offset)
    if ((newline_offset == nil) or
        (newline_offset >= @end_offset))
      # No text
      current_offset = @end_offset
      log_debug "Contains no more Text."
    else
      # We have a candidate
      # Get back to see the beginning of Text
      text_begin_offset = nil
      text_header_size = 0
      # Detect if it might be UTF-16 encoded
      if (((newline_offset > @begin_offset) and
           (@data[newline_offset-1] == NULL_CHAR)) or
          ((newline_offset < @end_offset-1) and
           (@data[newline_offset+1] == NULL_CHAR)))
        # Cursor should always be on a \x00 unless it arrived at the end
        cursor = newline_offset - 1
        while ((cursor >= @begin_offset+1) and
               (@data[cursor] == NULL_CHAR) and
               ((((c = @data[cursor-1].ord) >= 32) and
                 (c != 127)) or
                (c == 9) or
                (c == 13)))
          cursor -= 2
        end
        # Here we several possibilities:
        # * cursor is on @begin_offset-1 and data begins with \x00: UTF-16 string starts at @begin_offset and is big endian,
        # * else cursor is on @begin_offset and @data[@begin_offset+1] is \x00 but we did not check @data[@begin_offset] (if @data[@begin_offset] is ASCII then it means UTF-16 begins at @begin_offset and is little endian ; otherwise it starts at @begin_offset+1 and is big endian),
        # * else cursor is at least on @begin_offset+1 (we have at least 2 bytes before it), and
        #   * @data[cursor] is not \x00 and @data[cursor+2] is \x00: we could be on the endianness marker, or out of the string already ; if not endianness marker, if @data[cursor+1] is valid ASCI then UTF-16 string starts at cursor+1 and is little endian, otherwise it starts at cursor+2 and is big endian,
        #   * else @data[cursor] is \x00 but preceding character is not ASCII (meaning it can't be the endianness marker either): UTF-16 string starts at cursor end is big endian
        # UTF_16BE = "\xFE\xFF\x00\x??"
        # UTF_16LE = "\xFF\xFE\x??\x00"
        # In following comments, here are the conventions:
        # * \xAA means valid ASCII character
        # * \xBB means not a valid ASCII character
        # * \x11 means a non zero character
        # * \x?? means unknown character
        # * other values represent their corresponding character
        if (cursor == @begin_offset-1)
          # @data[@begin_offset..@begin_offset+1] == \xAA\x00
          text_begin_offset = @begin_offset
          encoding = Encoding::UTF_16LE
        elsif (cursor == @begin_offset)
          # @data[@begin_offset..@begin_offset+2] == \x??\xAA\x00
          if ((c = @data[@begin_offset].ord) == 0)
            # @data[@begin_offset..@begin_offset+2] == \x00\xAA\x00
            text_begin_offset = @begin_offset
            encoding = Encoding::UTF_16BE
          elsif (((c >= 32) and
                  (c != 127)) or
                 (c == 9) or
                 (c == 13))
            # @data[@begin_offset..@begin_offset+2] == \xAA\xAA\x00
            if (@data[@begin_offset..@begin_offset+1] == UTF_16BE_BOM)
              # @data[@begin_offset..@begin_offset+2] == \xFE\xFF\x00
              text_begin_offset = @begin_offset
              encoding = Encoding::UTF_16BE
              text_header_size = 2
            else
              text_begin_offset = @begin_offset + 1
              encoding = Encoding::UTF_16LE
            end
          else
            # @data[@begin_offset..@begin_offset+2] == \xBB\xAA\x00
            text_begin_offset = @begin_offset + 1
            encoding = Encoding::UTF_16LE
          end
        elsif (@data[cursor] == NULL_CHAR)
          # @data[cursor-1..cursor+2] == \xBB\x00\xAA\x00
          text_begin_offset = cursor
          encoding = Encoding::UTF_16BE
        elsif (@data[cursor-1..cursor] == UTF_16LE_BOM)
          # @data[cursor-1..cursor+2] == \xFF\xFE\xAA\x00
          text_begin_offset = cursor - 1
          encoding = Encoding::UTF_16LE
          text_header_size = 2
        elsif (@data[cursor..cursor+1] == UTF_16BE_BOM)
          # @data[cursor-1..cursor+2] == \x??\xFE\xFF\x00
          text_begin_offset = cursor
          encoding = Encoding::UTF_16BE
          text_header_size = 2
        else
          # @data[cursor-1..cursor+2] == \x??\x11\xAA\x00
          text_begin_offset = cursor + 1
          encoding = Encoding::UTF_16LE
        end
      else
        encoding = Encoding::ASCII_8BIT
        cursor = newline_offset - 1
        while ((cursor >= @begin_offset) and
               ((((c = @data[cursor].ord) >= 32) and
                 (c != 127)) or
                (c == 9) or
                (c == 13)))
          cursor -= 1
        end
        text_begin_offset = cursor + 1
      end
      # Now find forward
      keep_alive
      text_end_offset = nil
      truncated = false
      case encoding
      when Encoding::ASCII_8BIT
        cursor = newline_offset + 1
        while ((cursor < @end_offset) and
               ((((c = @data[cursor].ord) >= 32) and
                 (c != 127)) or
                (c == 9) or
                (c == 10) or
                (c == 13)))
          cursor += 1
        end
        text_end_offset = cursor
      when Encoding::UTF_16BE
        # cursor points on \x00
        cursor = newline_offset + 1
        while ((cursor < @end_offset-1) and
               (@data[cursor] == NULL_CHAR) and
               ((((c = @data[cursor+1].ord) >= 32) and
                 (c != 127)) or
                (c == 9) or
                (c == 10) or
                (c == 13)))
          cursor += 2
        end
        # Several possibilities:
        # * cursor is at @end_offset, meaning the string ends at @end_offset,
        # * else cursor is at @end_offset-1, meaning the string ends at @end_offset-1 or at @end_offset and is truncated if @data[@end_offset-1] is "\x00",
        # * else the string ends at cursor
        if (cursor == @end_offset-1)
          if (@data[cursor] == NULL_CHAR)
            truncated = true
            text_end_offset = @end_offset
          else
            text_end_offset = @end_offset - 1
          end
        else
          text_end_offset = cursor
        end
      when Encoding::UTF_16LE
        # cursor points on the ASCII value
        cursor = newline_offset
        while ((cursor < @end_offset-1) and
               (@data[cursor+1] == NULL_CHAR) and
               ((((c = @data[cursor].ord) >= 32) and
                 (c != 127)) or
                (c == 9) or
                (c == 10) or
                (c == 13)))
          cursor += 2
        end
        # Several possibilities:
        # * cursor is at @end_offset, meaning the string ends at @end_offset,
        # * else cursor is at @end_offset-1, meaning the string ends at @end_offset-1 or at @end_offset and is truncated if @data[@end_offset-1] is a valid ASCII,
        # * else the string ends at cursor
        if (cursor == @end_offset-1)
          if ((((c = @data[cursor].ord) >= 32) and
               (c != 127)) or
              (c == 9) or
              (c == 10) or
              (c == 13))
            truncated = true
            text_end_offset = @end_offset
          else
            text_end_offset = @end_offset - 1
          end
        else
          text_end_offset = cursor
        end
      end
      # Consider text files longer than a certain size only
      if (text_end_offset - text_begin_offset < 512*((encoding == Encoding::ASCII_8BIT) ? 1 : 2))
        #log_debug "@#{text_begin_offset} - Text section is too short (#{text_end_offset - text_begin_offset}) to be identified as text"
      else
        log_debug "@#{text_begin_offset} - Found text up to #{text_end_offset} with encoding #{encoding} and header of size #{text_header_size}"
        # Now check some formats
        text = @data[text_begin_offset+text_header_size..text_end_offset-1].clone.force_encoding(encoding)
        lines = text.split("\r\n".encode(encoding))
        lines = text.split("\n".encode(encoding)) if (lines.size == 1)
        extension = [ :txt, :log ] # By default
        if is_text_srt?(lines, encoding)
          extension = :srt
        elsif is_text_rtf?(lines, encoding)
          extension = :rtf
        elsif is_text_html?(lines, encoding)
          extension = :html
        elsif is_text_xml?(lines, encoding)
          extension = :xml
        end
        found_segment(text_begin_offset, text_end_offset, extension, false, false, :encoding => encoding)
      end
      current_offset = text_end_offset + 1
    end
  end
end