Module: NotesStructuredTextJsonMessages

Defined in:
lib/notes_structured_text_json_messages.rb

Constant Summary collapse

MESSAGE_ID = 
"$MessageID"
POSTED_DATE = 
"PostedDate"
IN_REPLY_TO = 
"in_reply_to"
REFERENCES = 
"references"
FROM = 
"From"
TO = 
"SendTo"
CC = 
"CopyTo"
BCC = 
"BlindCopyTo"
INET_FROM = 
"InetFrom"
INET_TO = 
"InetSendTo"
INET_CC = 
"InetCopyTo"
INET_BCC = 
"InetBlindCopyTo"
NOTES_US_DATE_FORMAT = 
"%m/%d/%Y %I:%M:%S %p"

Class Attribute Summary collapse

Class Method Summary collapse

Class Attribute Details

.logger ⇒ Object

Returns the value of attribute logger.



6
7
8
 
# File 'lib/notes_structured_text_json_messages.rb', line 6

def logger
  @logger
end

.mappings ⇒ Object

Returns the value of attribute mappings.



8
9
10
 
# File 'lib/notes_structured_text_json_messages.rb', line 8

def mappings
  @mappings
end

.stats ⇒ Object

Returns the value of attribute stats.



7
8
9
 
# File 'lib/notes_structured_text_json_messages.rb', line 7

def stats
  @stats
end

Class Method Details

.collect_mapping(addr1, addr2) ⇒ Object 



21
22
23
 
# File 'lib/notes_structured_text_json_messages.rb', line 21

def collect_mapping(addr1, addr2)
  self.mappings << [addr1, addr2] if self.mappings
end

.extract_json_message(block, options = {}) ⇒ Object 



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/notes_structured_text_json_messages.rb', line 221

def extract_json_message(block, options={})
  message_id_h = header_value(block, MESSAGE_ID) 
  if !message_id_h
    increment_stats(:failure_no_message_id)
    raise "no #{MESSAGE_ID}"
  end
  message_id = strip_angles(message_id_h)

  posted_date_h = header_value(block, POSTED_DATE)
  if !posted_date_h
    increment_stats(:failure_no_posted_date)
    raise "no #{POSTED_DATE}"
  end
  posted_date = parse_date(posted_date_h, options)

  in_reply_to_h = header_value(block, IN_REPLY_TO)
  in_reply_to = strip_angles(in_reply_to_h) if in_reply_to_h

  references_h = header_values(block, REFERENCES, " ")
  references = references_h.map{|r| strip_angles(r)} if references_h

  froms = process_addresses(block, INET_FROM, FROM, options)
  if !froms || froms.size>1
    increment_stats(:failure_from)
    raise "no From:, or more than one From:"
  end
  from = froms[0]
  to = process_addresses(block, INET_TO, TO, options)
  cc = process_addresses(block, INET_CC, CC, options)
  bcc = process_addresses(block, INET_BCC, BCC, options)

  if (to||[]).size + (cc||[]).size + (bcc||[]).size == 0
    increment_stats(:failure_no_recipients)
    raise "no recipients" 
  end
  
  { :message_type=>"email",
    :message_id=>message_id,
    :sent_at=>posted_date,
    :in_reply_to=>in_reply_to,
    :references=>references,
    :from=>from,
    :to=>to,
    :cc=>cc,
    :bcc=>bcc}
end

.header_value(block, header) ⇒ Object 



84
85
86
87
88
 
# File 'lib/notes_structured_text_json_messages.rb', line 84

def header_value(block, header)
  patt = /^#{Regexp.quote(header)}: /i
  h = block.find{|l| l =~ patt}
  h.gsub(patt, '').strip if h
end

.header_values(block, header, split_on = ",") ⇒ Object 



90
91
92
93
94
95
96
97
98
99
 
# File 'lib/notes_structured_text_json_messages.rb', line 90

def header_values(block, header, split_on=",")
  h = header_value(block, header)
  if h
    if split_on.is_a?(Symbol)
      self.send(split_on, h)
    else
      h.split(split_on)
    end.map(&:strip)
  end
end

.increment_stats(key) ⇒ Object 



17
18
19
 
# File 'lib/notes_structured_text_json_messages.rb', line 17

def increment_stats(key)
  self.stats[key] = (self.stats[key]||0) + 1
end

.is_distinguished_name?(addr) ⇒ Boolean

Returns:

  • (Boolean) 


80
81
82
 
# File 'lib/notes_structured_text_json_messages.rb', line 80

def is_distinguished_name?(addr)
  !!(addr =~ /CN=/)
end

.is_message_block?(block) ⇒ Boolean

Returns:

  • (Boolean) 


76
77
78
 
# File 'lib/notes_structured_text_json_messages.rb', line 76

def is_message_block?(block)
  !!header_value(block, MESSAGE_ID)
end

.json_messages(output_dir, input_files, options = {}) ⇒ Object 



49
50
51
52
53
54
55
56
57
58
59
 
# File 'lib/notes_structured_text_json_messages.rb', line 49

def json_messages(output_dir, input_files, options={})
  with_stats do
    with_mappings(options) do
      [*input_files].each do |input_file|
        File.open(input_file, "r") do |input|
          json_messages_from_stream(output_dir, input, options)
        end
      end
    end
  end
end

.json_messages_from_stream(output_dir, input, options = {}) ⇒ Object 



61
62
63
64
 
# File 'lib/notes_structured_text_json_messages.rb', line 61

def json_messages_from_stream(output_dir, input, options={})
  block = nil
  process_block(output_dir, block, options) while block=read_block(input)
end

.log {|logger| ... } ⇒ Object

Yields:



13
14
15
 
# File 'lib/notes_structured_text_json_messages.rb', line 13

def log
  yield logger if logger
end

.output_json_message(output_dir, json_message) ⇒ Object 



268
269
270
271
 
# File 'lib/notes_structured_text_json_messages.rb', line 268

def output_json_message(output_dir, json_message)
  fname = File.join(output_dir, "#{MD5.hexdigest(json_message[:message_id])}.json")
  File.open(fname, "w"){|out| out << json_message.to_json}
end

.parse_date(date, options = {}) ⇒ Object 



217
218
219
 
# File 'lib/notes_structured_text_json_messages.rb', line 217

def parse_date(date, options={})
  DateTime.strptime(date, NOTES_US_DATE_FORMAT)
end

.process_address(addr) ⇒ Object 



163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
 
# File 'lib/notes_structured_text_json_messages.rb', line 163

def process_address(addr)
  if is_distinguished_name?(addr)
    name = addr[/CN=([^\/]*)/, 1]
    h = {:notes_dn=>addr}
    h[:name] = name if name
    h
  else
    ta = TMail::Address.parse(addr)
    if ta.is_a?(TMail::Address)
      h = {:email_address=>ta.address.downcase}
      h[:name] = ta.name if ta.name
      h
    else
      log{|logger| logger.warn("addr does not parse to a TMail::Address: #{addr}")}
    end
  end
end

.process_address_pair(inet_addr, notes_addr, options) ⇒ Object 



181
182
183
184
185
186
187
188
189
190
191
192
 
# File 'lib/notes_structured_text_json_messages.rb', line 181

def process_address_pair(inet_addr, notes_addr, options)
  if inet_addr == "."
    process_address(notes_addr)
  else
    inet_addr = process_address(inet_addr)
    notes_addr = process_address(notes_addr)

    collect_mapping(notes_addr, inet_addr)

    inet_addr
  end
end

.process_addresses(block, inet_field, notes_field, options) ⇒ Object 



194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
 
# File 'lib/notes_structured_text_json_messages.rb', line 194

def process_addresses(block, inet_field, notes_field, options)
  inet_h = header_values(block, inet_field, :split_rfc822_addresses)
  notes_h = header_values(block, notes_field, :split_rfc822_addresses)

  if inet_h && notes_h
    if inet_h.length == notes_h.length
      inet_h.zip(notes_h).map do |inet_addr, notes_addr|
        process_address_pair(inet_addr, notes_addr, options)
      end
    else
      raise "#{inet_field}: does not match #{notes_field}:"
    end
  elsif inet_h
    inet_h.map{|addr| process_address(addr)}
  elsif notes_h
    notes_h.map{|addr| process_address(addr)}
  else
    nil
  end
end

.process_block(output_dir, block, options = {}) ⇒ Object 



134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
 
# File 'lib/notes_structured_text_json_messages.rb', line 134

def process_block(output_dir, block, options={})
  if is_message_block?(block)
    json_message = extract_json_message(block, options)
    output_json_message(output_dir, json_message)
    increment_stats(:message)
  else
    increment_stats(:non_message)
  end
rescue Exception=>e
  increment_stats(:failed_message)
  log do |logger| 
    logger.error(e)
    logger.error(block.join("\n"))
  end
end

.read_block(input) ⇒ Object 



66
67
68
69
70
71
72
73
74
 
# File 'lib/notes_structured_text_json_messages.rb', line 66

def read_block(input)
  return nil if input.eof?
  block = []
  begin
    l = input.readline.chomp
    block << l if l.length>0
  end while !input.eof? && l != ""
  block
end

.split_rfc822_addresses(header) ⇒ Object 



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
 
# File 'lib/notes_structured_text_json_messages.rb', line 101

def split_rfc822_addresses(header)
  addresses = []
  quoted_pair = false
  quoted_string = false
  buf = ""
  header.each_char do |c|
    if quoted_pair
      buf << c
      quoted_pair = false
    elsif quoted_string && c=='\\'
      buf << c
      quoted_pair = true
    elsif !quoted_string && c==','
      addresses << buf
      buf = ""
    elsif !quoted_string && c=='"'
      buf << c
      quoted_string = true
    elsif quoted_string && c=='"'
      buf << c
      quoted_string = false
    else
      buf << c
    end
  end
  addresses << buf if buf.length>0
  addresses
end

.strip_angles(value) ⇒ Object 



130
131
132
 
# File 'lib/notes_structured_text_json_messages.rb', line 130

def strip_angles(value)
  value.gsub(/<([^>]*)>/, '\1')
end

.with_mappings(options) ⇒ Object 



25
26
27
28
29
30
31
32
33
34
35
36
37
38
 
# File 'lib/notes_structured_text_json_messages.rb', line 25

def with_mappings(options)
  mapping_file = options[:mapping_file]
  log{|logger| logger.info "using mapping_file: #{options[:mapping_file]}"} if mapping_file
  self.mappings=[]
  yield
ensure
  if mapping_file
    File.open(mapping_file, "w") do |output|
      output << "[\n"
      output << self.mappings.map(&:to_json).join(",\n")
      output << "\n]"
    end
  end
end

.with_stats ⇒ Object 



40
41
42
43
44
45
46
47
 
# File 'lib/notes_structured_text_json_messages.rb', line 40

def with_stats
  self.stats={}
  yield
ensure
  self.stats.each do |k,v|
    log{|logger| logger.info("#{k}: #{v}")}
  end
end