Class: Argos::Ds

Inherits:
Array
  • Object
show all
Includes:
Ascii
Defined in:
lib/argos/ds.rb

Overview

Argos DS|DAT file parser

Usage

ds = Argos::Ds.new
puts ds.parse(filename).to_json

Author:

  • Espen Egeland

  • Conrad Helgeland

Constant Summary collapse

START_REGEX = 
/^\d{5} \d{5,6} +\d+ +\d+/
START_REGEX_LEGACY = 
/\s+\d\.\d{3}\s\d{9}\s+\w{4}$/
LOCATION_CLASS = 
[nil, "0","1","2","3","A","B","G","Z"]

Instance Attribute Summary collapse

Instance Method Summary collapse

Methods included from Ascii

argos?, factory, #latitudes, #longitudes, #platforms, #programs, source, type

Constructor Details

#initialize ⇒ Ds

Returns a new instance of Ds.



26
27
28
29
 
# File 'lib/argos/ds.rb', line 26

def initialize
  @errors = []
  @log = Logger.new(STDERR)
end

Instance Attribute Details

#bundle ⇒ Object

Returns the value of attribute bundle.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def bundle
  @bundle
end

#errors ⇒ Object (readonly)

Returns the value of attribute errors.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def errors
  @errors
end

#filename ⇒ Object

Returns the value of attribute filename.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def filename
  @filename
end

#filesize ⇒ Object (readonly)

Returns the value of attribute filesize.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def filesize
  @filesize
end

#filter ⇒ Object

Returns the value of attribute filter.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def filter
  @filter
end

#filtername ⇒ Object (readonly)

Returns the value of attribute filtername.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def filtername
  @filtername
end

#log ⇒ Object 



44
45
46
47
48
49
 
# File 'lib/argos/ds.rb', line 44

def log
  if @log.nil?
    @log = Logger.new(STDERR)
  end
  @log
end

#messages ⇒ Object (readonly)

Returns the value of attribute messages.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def messages
  @messages
end

#multiplicates ⇒ Object (readonly)

Returns the value of attribute multiplicates.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def multiplicates
  @multiplicates
end

#sha1 ⇒ Object (readonly)

Returns the value of attribute sha1.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def sha1
  @sha1
end

#updated ⇒ Object (readonly)

Returns the value of attribute updated.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def updated
  @updated
end

#valid ⇒ Object (readonly)

Returns the value of attribute valid.



18
19
20
 
# File 'lib/argos/ds.rb', line 18

def valid
  @valid
end

Instance Method Details

#combine_header_with_transmission(measurements, header) ⇒ Object

Header is is a space-separated string containing 

 [0] Program number
 [1] Platform number
 [2] Number of lines of data per satellite pass
 [3] Number of sensors
 [4] Satellite identifier
 [5] Location class (lc)
 [6] Location date 2007-03-02
 [7] Location UTC time
 [8] Latitude (decimal degrees)
 [9] Longitude, may be > 180 like 255.452°, equivalent to 255.452 - 360 = -104.548 (°E)
[10] Altitude (km)
[11] Frequency (calculated)

The header varies in information elemenet, often either 0..4|5 or 0..11. Header examples (plit on “ ”):

["09660", "10788", "4", "3", "D", "0"]
["09660", "10788", "5", "3", "H", "2", "1992-04-06", "22:12:16", "78.248", "15.505", "0.000", "401649604"]
["09660", "10788", "2", "3", "D"]

www.argos-system.org/files/pmedia/public/r363_9_argos_users_manual-v1.5.pdf page 42

Warning, the parser does not support this header format from 1989 [AUO89.DAT] 19890800-19891000: [“09660”, “14653”, “10”, “41”, “14”, “1”, “-.42155E+1”, “00”, “112”, “17DD”]

Parameters:

  • header (String) 


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
263
264
265
266
 
# File 'lib/argos/ds.rb', line 195

def combine_header_with_transmission(measurements, header)
  unless header.is_a? Array
    header = header.split(" ")
  end
  latitude = longitude = positioned = nil
  warn = []
  errors = []

  lc = header[5]

  if not header[6].nil? and not header[7].nil?
    positioned = convert_datetime(header[6]+" "+header[7])
  end

  if header[8] != nil && valid_float?(header[8])
    latitude = header[8].to_f
  end
  
  if header[9] != nil && valid_float?(header[9])
    longitude = header[9].to_f
    if (180..360).include? longitude
      longitude = (longitude - 360)
    end
  end

  altitude = header[10]
  if not altitude.nil?
     altitude = altitude.to_f*1000
  end
    
  if positioned.nil? and measurements.nil?
    warn << "missing-time"
  end
    
  if latitude.nil? or longitude.nil?
    warn << "missing-position"
  else
  
    unless latitude.between?(-90, 90) and longitude.between?(-180, 180)
      errors << "invalid-position"
    end
  end

  unless LOCATION_CLASS.include? lc
    errors << "invalid-lc"
  end

  # Satellites 
  #  ["A", "B", "K", "L", "M", "N", "P", "R"] 

  document = { program:  header[0].to_i,
    platform: header[1].to_i,
    lines: header[2].to_i,
    sensors: header[3].to_i,
    satellite: header[4],
    lc: lc,
    positioned: positioned,
    latitude: latitude,
    longitude: longitude,
    altitude: altitude,        
    measurements: measurements,
    headers: header.size
  }
  if warn.any?
    document[:warn]=warn
  end
  if errors.any?
    document[:errors]=errors
  end
  
  document
end

#filter? ⇒ Boolean

Returns:

  • (Boolean) 


31
32
33
 
# File 'lib/argos/ds.rb', line 31

def filter?
  not @filter.nil?
end

#merge(ds, measurement) ⇒ Array

Merges a DS header hash into each measurement

Returns:

  • (Array) —

    Measurements with header and static metadata merged in



307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
 
# File 'lib/argos/ds.rb', line 307

def merge(ds, measurement)
  m = ds.select {|k,v| k != :measurements and k != :errors and k != :warn }     
  m = m.merge(measurement)          
  m = m.merge ({ technology: "argos",
    type: type,
    file: "file://"+filename,
    source: sha1
  })

  if not ds[:errors].nil? and ds[:errors].any?
    m[:errors] = ds[:errors].clone
  end

  if not ds[:warn].nil? and ds[:warn].any?
    m[:warn] = ds[:warn].clone
  end

  if not m[:sensor_data].nil? and m[:sensor_data].size != ds[:sensors]
    if m[:warn].nil?
      m[:warn] = []
    end
    m[:warn] << "sensors-count-mismatch"
  end

  # Create id as SHA1 hash of measurement minus stuff that may vary (like filename)
  # 
  # Possible improvement for is to base id on a static list of keys
  # :program,
  # :platform,
  # :lines,
  # :sensors,
  # :satellite,
  # :lc,
  # :positioned,
  # :latitude,
  # :longitude,
  # :altitude,
  # :headers,
  # :measured,
  # :identical,
  # :sensor_data,
  # :technology,
  # :type,
  # :source

  idbase = m.clone
  idbase.delete :errors
  idbase.delete :file
  idbase.delete :warn
  
  id = Digest::SHA1.hexdigest(idbase.to_json)

  m[:parser] = Argos.library_version
  m[:id] = id
  m[:bundle] = bundle
  m
end

#parse(filename = nil) ⇒ Argos::Ds

Parses Argos DS file and returns Argos::Ds -> Array

The parser loops all messages (stored in @messages), before #unfold creates a sorted Array of measurements

Parameters:

  • filename (String) (defaults to: nil) —

    Filename of Argos DS file

Returns:



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
 
# File 'lib/argos/ds.rb', line 58

def parse(filename=nil)

  self.clear # Needed if you parse multiple times
  @messages = []
  @valid = false

  if filename.nil?
    filename = @filename
  end
  

  filename = File.realpath(filename)
  @filename = filename
  if filename.nil? or not File.exists? filename
    raise ArgumentError, "Missing ARGOS DS file: \"#{filename}\""
  end
  @sha1 = Digest::SHA1.file(filename).hexdigest 

  contact = []
  file = File.open(filename)
  @filesize = file.size
  @updated = file.mtime.utc
  
  log.debug "Parsing ARGOS DS file #{filename} source:#{sha1} (#{filesize} bytes)"
  if filter?
    log.debug "Using filter: #{@filtername.nil? ? filter : @filtername }"
  end

  firstline = file.readline
  file.rewind

  if firstline =~ START_REGEX_LEGACY
    return parse_legacy(file)
  end

  file.each_with_index do |line, c|
    line = line.strip

    #if (c+1) % 1000 == 0
    #  log.debug "Line: #{c+1}"
    #end
    
    if line =~ START_REGEX

      @valid = true

      if contact.any?
        item = parse_message(contact)

        if self.class.valid_item? item  
          
          if not filter? or filter.call(item)
            @messages << item
          end
          
        else
          raise "Argos DS message #{filename}:#{c} lacks required program and/or platform"
        end
      end
  
      contact = [line]

    else 
      # 2010-12-14 15:11:34  1         00           37           01           52
      if contact.any? and line != ""
        contact << line
      end
    end
  end
  
  if false == @valid
    #log.debug file.read
    message = "Cannot parse file: #{filename}"
    raise ArgumentError, message 
  end

  last = parse_message(contact)

  # The last message
  if last
    if not filter? or filter.call(last)
      @messages << last
    end
  end
  
  log.debug "Parsed #{@messages.size} Argos DS messages into #{self.class.name} Array"  
  @segments = @messages.size
  unfold.each do |d|
    self << d
  end

  @multiplicates = group_by { |e| e }.select { |k, v| v.size > 1 }.map(&:first)
  if multiplicates.any?
    log.warn "#{multiplicates.size} multiplicates in source sha1 #{sha1} #{filename}): #{multiplicates.map {|a|a[:id]} }"
    self.uniq!
    log.info "Unique DS messages: #{self.size} sha1: #{sha1} #{filename}"
  end
  self.sort_by! {|ds| ds[:measured]}
  self
end

#parse_message(contact) ⇒ Object

Pare one DS segment



160
161
162
163
164
165
 
# File 'lib/argos/ds.rb', line 160

def parse_message(contact)
  header = contact[0]
  body = contact[1,contact.count]
  items = process_item_body(body)
  combine_header_with_transmission(items, header)
end

#process_item_body(body_arr) ⇒ Object 



365
366
367
368
369
 
# File 'lib/argos/ds.rb', line 365

def process_item_body(body_arr)
  @buf =""
  @transmission_arr = []
  @transmission_arr = recursive_transmission_parse(body_arr)
end

#recursive_transmission_parse(body_arr) ⇒ Aray

 @param [Array] body_arr

Returns:

  • (Aray) 


374
375
376
377
378
379
380
381
382
383
384
385
386
 
# File 'lib/argos/ds.rb', line 374

def recursive_transmission_parse(body_arr)
  if  body_arr.nil? or body_arr.empty?
    return
  end
  @buf =@buf + " " + body_arr[0]
  
  if body_arr[1] =~ /\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2}/ or body_arr[1]==nil
    @transmission_arr << transmission_package(@buf)
    @buf=""
  end
  recursive_transmission_parse(body_arr[1,body_arr.length])
  @transmission_arr
end

#source ⇒ Object 



412
413
414
 
# File 'lib/argos/ds.rb', line 412

def source
  @sha1
end

#start ⇒ Object 



404
405
406
 
# File 'lib/argos/ds.rb', line 404

def start
  positioned.map {|ds| ds [:positioned] }.first
end

#stop ⇒ Object 



408
409
410
 
# File 'lib/argos/ds.rb', line 408

def stop
  positioned.map {|ds| ds [:positioned] }.last
end

#transmission_package(data) ⇒ Object 



388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
 
# File 'lib/argos/ds.rb', line 388

def transmission_package(data)
  transmission_time = data[/(\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2})/,1]
  transmission_time = convert_datetime(transmission_time)

  identical = data.split(" ")[2].to_i    
  data = data.strip[23,data.length]

  if not data.nil?
    sensor_data = data.split(" ")
  end
  { measured: transmission_time,
    identical: identical,
    sensor_data: sensor_data
  }
end

#type ⇒ Object 



167
168
169
 
# File 'lib/argos/ds.rb', line 167

def type
  "ds"
end

#unfold ⇒ Object

Merge position and all other top-level DS fields with each measurement line (containing sensor data) The 3 lines below will unfold to 2 documents, each with “positioned”:2010-03-05T14:19:06Z, “platform”: “23695”, “latitude”:“79.989”, etc. 23695 074772 3 4 M B 2010-03-05 14:19:06 79.989 12.644 0.036 401639707

2010-03-05 14:17:35  1         01           25        37630           36
2010-03-05 14:20:38  1         00           28           00           65



275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
 
# File 'lib/argos/ds.rb', line 275

def unfold

  # First, grab all segments *without* measurements (if any)
  unfolded = messages.reject {|ds| ds.key?(:measurements) or ds[:measurements].nil? }
  log.debug "#{messages.size - unfolded.size} / #{messages.size} messages contained measurements"

  messages.select {|ds|
    ds.key?(:measurements) and not ds[:measurements].nil?
  }.each do |ds|
    
    ds[:measurements].each do |measurement|
      unfolded << merge(ds,measurement)
    end
  end
 
  unfolded = unfolded.sort_by {|ds|
    if not ds[:measured].nil?
      DateTime.parse(ds[:measured]) 
    elsif not ds[:positioned].nil?
      DateTime.parse(ds[:positioned])
    else
      ds[:program]
    end
  }

  log.info "Unfolded #{messages.size} ARGOS DS position and sensor messages into #{unfolded.size} new documents source:#{sha1} #{filename}"

  unfolded
end