Class: Gel::TailFile

Inherits:
Object
  • Object
show all
Defined in:
lib/gel/tail_file.rb

Constant Summary collapse

MAXIMUM_CHAIN =

The number of redirects etc we’ll follow before giving up

8
CONTENT_OVERLAP =

When attempting a partial file download, we’ll back up by this many bytes to ensure the preceding content matches our local file

100
PARTIAL_MINIMUM =

Only bother trying for a partial download if we have at least this many bytes of local content. If we *don’t* make a partial request, 1) we’re guaranteed not to need a second request due to bad overlap, and 2) the response can be gzipped.

65536

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(uri, pinboard, httpool: Gel::Httpool.new) ⇒ TailFile

Returns a new instance of TailFile.



24
25
26
27
28
29
30
31
32
33
 
# File 'lib/gel/tail_file.rb', line 24

def initialize(uri, pinboard, httpool: Gel::Httpool.new)
  @uri = uri
  @pinboard = pinboard
  @httpool = httpool

  @filename = pinboard.filename(uri)
  @etag = pinboard.etag(uri)

  @done = false
end

Instance Attribute Details

#filename ⇒ Object

Returns the value of attribute filename.



23
24
25
 
# File 'lib/gel/tail_file.rb', line 23

def filename
  @filename
end

#pinboard ⇒ Object

Returns the value of attribute pinboard.



23
24
25
 
# File 'lib/gel/tail_file.rb', line 23

def pinboard
  @pinboard
end

#uri ⇒ Object

Returns the value of attribute uri.



23
24
25
 
# File 'lib/gel/tail_file.rb', line 23

def uri
  @uri
end

Instance Method Details

#debug(message) ⇒ Object 



202
203
204
 
# File 'lib/gel/tail_file.rb', line 202

def debug(message)
  #$stderr.puts message if $DEBUG
end

#done? ⇒ Boolean

Returns:

  • (Boolean) 


35
36
37
 
# File 'lib/gel/tail_file.rb', line 35

def done?
  @done
end

#update(force_reset: false) ⇒ Object 



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
 
# File 'lib/gel/tail_file.rb', line 39

def update(force_reset: false)
  uri = self.uri

  File.open(@filename, "a+b") do |f|
    f.seek(0, IO::SEEK_END)

    MAXIMUM_CHAIN.times do
      f.seek(0, IO::SEEK_SET) if force_reset

      request = Net::HTTP::Get.new(uri)

      requested_from = 0

      if f.pos > PARTIAL_MINIMUM
        requested_from = f.pos - CONTENT_OVERLAP
        request["Range"] = "bytes=#{requested_from}-"
        request["Accept-Encoding"] = "identity"
      else
        f.pos = 0
      end
      request["If-None-Match"] = @etag if @etag

      response = @httpool.request(uri, request)

      case response
      when Net::HTTPNotModified
        @done = true
        pinboard.updated(uri, response["ETag"], false)

        return # no-op

      when Net::HTTPRequestedRangeNotSatisfiable
        # This should never happen, but sometimes does: either the
        # file has been rebuilt, and we should do a full fetch, or
        # we're ahead of the server (because an intermediate is
        # caching too aggressively, say), and we should do nothing.

        if response["Content-Range"] =~ /\Abytes \*\/(\d+)\z/
          current_length = $1.to_i

          debug "File is smaller than we expected: only #{current_length} bytes"

          # Back up a bit, and check whether the end matches what we
          # already have

          f.pos = current_length
          next
        else
          force_reset = true
          next
        end

      when Net::HTTPOK
        f.pos = 0
        f.truncate(0)
        f.write response.body
        f.flush

        @done = true
        pinboard.updated(uri, response["ETag"], true)

        return # Done

      when Net::HTTPPartialContent
        if response["Content-Range"] =~ /\Abytes (\d+)-(\d+)\/(\d+)\z/
          from, to, size = $1.to_i, $2.to_i, $3.to_i
        else
          # Not what we asked for
          debug "Bad response range"

          force_reset = true
          next
        end

        if from != requested_from
          # Server didn't give us what we asked for
          debug "Incorrect response range"

          force_reset = true
          next
        end

        if to - from + 1 != response.body.size
          # Server didn't give us what it claimed to
          debug "Bad response length"

          force_reset = true
          next
        end

        debug "Current file size is #{File.size(@filename)}"
        debug "Remote size is #{size}"

        f.pos = from

        if to < f.size
          # No new content, but check the overlap in case the file's
          # been reset

          overlap = f.read(to - from + 1)
          if response.body == overlap
            # Good overlap, but nothing new
            debug "Overlap is good, but no new content"

            @done = true
            pinboard.updated(uri, @etag, false) # keep old etag

            return # Done
          else
            # Bad overlap
            debug "Bad overlap on short response"

            force_reset = true
            next
          end
        else
          overlap = f.read
          if response.body[0, overlap.size] == overlap
            # Good overlap; use rest
            rest = response.body[overlap.size..-1]

            debug "#{overlap.size} byte overlap is okay"
            debug "Using remaining #{rest.size} bytes"

            f.write(rest)
            f.flush

            @done = true
            pinboard.updated(uri, response["ETag"], true)

            return # Done
          else
            # Bad overlap
            debug "Bad overlap on long response"

            force_reset = true
            next
          end
        end

      when Net::HTTPRedirection
        uri = URI(response["Location"])
        next

      when Net::HTTPTooManyRequests
        # https://github.com/rubygems/rubygems-infrastructure/blob/adc0668c1f1539f281ffe86c6e0ad12743c5c8bd/cookbooks/rubygems-balancer/templates/default/site.conf.erb#L12
        debug "Rate limited"

        sleep 1 + rand
        next

      else
        # splat
        response.value

        raise "Unexpected HTTP success code"
      end
    end

    raise "Giving up after #{MAXIMUM_CHAIN} requests"
  end
end