Class: FileWatch::Watch

Inherits:
Object
  • Object
show all
Defined in:
lib/filewatch/watch.rb

Overview

TODO make a WatchedFilesDb class that holds the watched_files instead of a hash it should support an ‘identity’ of path + inode it should be serializable instead of the sincedb it should be deserializable to recreate the exact state all files were in as last seen some parts of the each method should be handled by it, e.g. wfs_db.<state>_iterator{|wf| }, trapping the Errno::ENOENT, auto_delete and yield wtached_file

Constant Summary collapse

MAX_FILES_WARN_INTERVAL = 
ENV.fetch("FILEWATCH_MAX_FILES_WARN_INTERVAL", 20).to_i

Instance Attribute Summary collapse

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(opts = {}) ⇒ Watch

Returns a new instance of Watch.



39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
 
# File 'lib/filewatch/watch.rb', line 39

def initialize(opts={})
  if opts[:logger]
    @logger = opts[:logger]
  else
    @logger = Logger.new(STDERR)
    @logger.level = Logger::INFO
  end
  @watching = []
  @exclude = []
  @files = Hash.new
  # we need to be threadsafe about the mutation
  # of the above 2 ivars because the public
  # methods each, discover and watch
  # can be called from different threads.
  @lock = Mutex.new
  # we need to be threadsafe about the quit mutation
  @quit = false
  @quit_lock = Mutex.new
  self.max_open_files = ENV["FILEWATCH_MAX_OPEN_FILES"].to_i
  @lastwarn_max_files = 0
end

Instance Attribute Details

#delimiter ⇒ Object

Returns the value of attribute delimiter.



36
37
38
 
# File 'lib/filewatch/watch.rb', line 36

def delimiter
  @delimiter
end

#logger ⇒ Object

Returns the value of attribute logger.



35
36
37
 
# File 'lib/filewatch/watch.rb', line 35

def logger
  @logger
end

#max_active ⇒ Object (readonly)

Returns the value of attribute max_active.



37
38
39
 
# File 'lib/filewatch/watch.rb', line 37

def max_active
  @max_active
end

Class Method Details

.inode(path, stat) ⇒ Object 



31
32
33
 
# File 'lib/filewatch/watch.rb', line 31

def self.inode(path, stat)
  send(FILEWATCH_INODE_METHOD, path, stat)
end

.nix_inode(path, stat) ⇒ Object 



27
28
29
 
# File 'lib/filewatch/watch.rb', line 27

def self.nix_inode(path, stat)
  [stat.ino.to_s, stat.dev_major, stat.dev_minor]
end

.win_inode(path, stat) ⇒ Object 



22
23
24
25
 
# File 'lib/filewatch/watch.rb', line 22

def self.win_inode(path, stat)
  fileId = Winhelper.GetWindowsUniqueFileIdentifier(path)
  [fileId, 0, 0] # dev_* doesn't make sense on Windows
end

Instance Method Details

#close_older=(value) ⇒ Object 



79
80
81
82
83
84
85
 
# File 'lib/filewatch/watch.rb', line 79

def close_older=(value)
  if !value.nil?
    val = value.to_f
    val = val <= 0 ? nil : val
  end
  @close_older = val
end

#discover ⇒ Object

def each



264
265
266
267
268
269
270
271
272
273
274
 
# File 'lib/filewatch/watch.rb', line 264

def discover
  synchronized do
    @watching.each do |path|
      _discover_file(path) do |filepath, stat|
        WatchedFile.new_ongoing(
            filepath, inode(filepath, stat), stat
          ).init_vars(@delimiter, @ignore_older, @close_older)
      end
    end
  end
end

#each(&block) ⇒ Object

Calls &block with params [event_type, path] event_type can be one of:

:create_initial - initially present file (so start at end for tail)
:create - file is created (new file after initial globs, start at 0)
:modify - file is modified (size increases)
:delete - file is deleted
:timeout - file is closable
:unignore - file was ignored, but since then it received new content



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
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
 
# File 'lib/filewatch/watch.rb', line 117

def each(&block)
  synchronized do
    return if @files.empty?
    begin
      file_deletable = []
      # creates this array just once
      watched_files = @files.values

      # look at the closed to see if its changed
      watched_files.select {|wf| wf.closed? }.each do |watched_file|
        path = watched_file.path
        break if quit?
        begin
          stat = watched_file.restat
          if watched_file.size_changed? || watched_file.inode_changed?(inode(path,stat))
            # if the closed file changed, move it to the watched state
            # not to active state because we want to use MAX_OPEN_FILES throttling.
            watched_file.watch
          end
        rescue Errno::ENOENT
          # file has gone away or we can't read it anymore.
          file_deletable << path
          @logger.debug? && @logger.debug("each: closed?: stat failed: #{path}: (#{$!}), deleting from @files")
        rescue => e
          @logger.error("each: closed?: #{path}: (#{e.inspect})")
        end
      end
      return if quit?

      # look at the ignored to see if its changed
      watched_files.select {|wf| wf.ignored? }.each do |watched_file|
        path = watched_file.path
        break if quit?
        begin
          stat = watched_file.restat
          if watched_file.size_changed? || watched_file.inode_changed?(inode(path,stat))
            # if the ignored file changed, move it to the watched state
            # not to active state because we want to use MAX_OPEN_FILES throttling.
            # this file has not been yielded to the block yet
            # but we must have the tail to start from the end, so when the file
            # was first ignored we updated the bytes_read to the stat.size at that time.
            # by adding this to the sincedb so that the subsequent modify
            # event can detect the change
            watched_file.watch
            yield(:unignore, watched_file)
          end
        rescue Errno::ENOENT
          # file has gone away or we can't read it anymore.
          file_deletable << path
          @logger.debug? && @logger.debug("each: ignored: stat failed: #{path}: (#{$!}), deleting from @files")
        rescue => e
          @logger.error("each: ignored?: #{path}: (#{e.inspect})")
        end
      end

      return if quit?

      # Send any creates.
      if (to_take = @max_active - watched_files.count{|wf| wf.active?}) > 0
        watched_files.select {|wf| wf.watched? }.take(to_take).each do |watched_file|
          break if quit?
          path = watched_file.path
          begin
            stat = watched_file.restat
            watched_file.activate
            # don't do create again
            next if watched_file.state_history_any?(:closed, :ignored)
            # if the file can't be opened during the yield
            # its state is set back to watched
            sym = watched_file.initial? ? :create_initial : :create
            yield(sym, watched_file)
          rescue Errno::ENOENT
            # file has gone away or we can't read it anymore.
            file_deletable << path
            watched_file.unwatch
            yield(:delete, watched_file)
            next
            @logger.debug? && @logger.debug("each: closed: stat failed: #{path}: (#{$!}), deleting from @files")
          rescue => e
            @logger.error("each: watched?: #{path}: (#{e.inspect})")
          end
        end
      else
        now = Time.now.to_i
        if (now - @lastwarn_max_files) > MAX_FILES_WARN_INTERVAL
          waiting = @files.size - @max_active
          specific = if @close_older.nil?
           ", try setting close_older. There are #{waiting} unopened files"
          else
            ", files yet to open: #{waiting}"
          end
          @logger.warn(@max_warn_msg + specific)
          @lastwarn_max_files = now
        end
      end

      return if quit?

      # wf.active means the actual files were opened
      # and have been read once - unless they were empty at the time
      watched_files.select {|wf| wf.active? }.each do |watched_file|
        path = watched_file.path
        break if quit?
        begin
          stat = watched_file.restat
        rescue Errno::ENOENT
          # file has gone away or we can't read it anymore.
          file_deletable << path
          @logger.debug? && @logger.debug("each: active: stat failed: #{path}: (#{$!}), deleting from @files")
          watched_file.unwatch
          yield(:delete, watched_file)
          next
        rescue => e
          @logger.error("each: active?: #{path}: (#{e.inspect})")
          next
        end

        if watched_file.file_closable?
          @logger.debug? && @logger.debug("each: active: file expired: #{path}")
          yield(:timeout, watched_file)
          watched_file.close
          next
        end

        _inode = inode(path,stat)
        read_thus_far = watched_file.bytes_read
        # we don't update the size here, its updated when we actually read
        if watched_file.inode_changed?(_inode)
          @logger.debug? && @logger.debug("each: new inode: #{path}: old inode was #{watched_file.inode.inspect}, new is #{_inode.inspect}")
          watched_file.update_inode(_inode)
          yield(:delete, watched_file)
          yield(:create, watched_file)
        elsif stat.size < read_thus_far
          @logger.debug? && @logger.debug("each: file rolled: #{path}: new size is #{stat.size}, old size #{read_thus_far}")
          yield(:delete, watched_file)
          yield(:create, watched_file)
        elsif stat.size > read_thus_far
          @logger.debug? && @logger.debug("each: file grew: #{path}: old size #{read_thus_far}, new size #{stat.size}")
          yield(:modify, watched_file)
        end
      end
    ensure
      file_deletable.each {|f| @files.delete(f)}
    end
  end
end

#exclude(path) ⇒ Object 



87
88
89
 
# File 'lib/filewatch/watch.rb', line 87

def exclude(path)
  path.to_a.each { |p| @exclude << p }
end

#ignore_older=(value) ⇒ Object 



70
71
72
73
74
75
76
77
 
# File 'lib/filewatch/watch.rb', line 70

def ignore_older=(value)
  #nil is allowed but 0 and negatives are made nil
  if !value.nil?
    val = value.to_f
    val = val <= 0 ? nil : val
  end
  @ignore_older = val
end

#inode(path, stat) ⇒ Object

def watch



105
106
107
 
# File 'lib/filewatch/watch.rb', line 105

def inode(path, stat)
  self.class.inode(path, stat)
end

#max_open_files=(value) ⇒ Object 



63
64
65
66
67
68
 
# File 'lib/filewatch/watch.rb', line 63

def max_open_files=(value)
  val = value.to_i
  val = 4095 if value.nil? || val <= 0
  @max_warn_msg = "Reached open files limit: #{val}, set by the 'max_open_files' option or default"
  @max_active = val
end

#quit ⇒ Object

def subscribe



293
294
295
 
# File 'lib/filewatch/watch.rb', line 293

def quit
  @quit_lock.synchronize { @quit = true }
end

#quit? ⇒ Boolean

def quit

Returns:

  • (Boolean) 


297
298
299
 
# File 'lib/filewatch/watch.rb', line 297

def quit?
  @quit_lock.synchronize { @quit }
end

#subscribe(stat_interval = 1, discover_interval = 5, &block) ⇒ Object 



276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
 
# File 'lib/filewatch/watch.rb', line 276

def subscribe(stat_interval = 1, discover_interval = 5, &block)
  glob = 0
  reset_quit
  while !quit?
    each(&block)
    break if quit?
    glob += 1
    if glob == discover_interval
      discover
      glob = 0
    end
    break if quit?
    sleep(stat_interval)
  end
  @files.values.each(&:file_close)
end

#watch(path) ⇒ Object 



91
92
93
94
95
96
97
98
99
100
101
102
103
 
# File 'lib/filewatch/watch.rb', line 91

def watch(path)
  synchronized do
    if !@watching.member?(path)
      @watching << path
      _discover_file(path) do |filepath, stat|
        WatchedFile.new_initial(
            filepath, inode(filepath, stat), stat
          ).init_vars(@delimiter, @ignore_older, @close_older)
      end
    end
  end
  return true
end