Class: Stamp::Translator

Inherits:
Object
  • Object
show all
Defined in:
lib/stamp/translator.rb

Constant Summary collapse

TIME_ZONE_ABBREVIATIONS =

Full list of time zone abbreviations from http://en.wikipedia.org/wiki/List_of_time_zone_abbreviations

%w{
  ACDT ACST ACT ADT AEDT AEST AFT AKDT AKST AMST AMT ART AST AWDT AWST AZOST AZT
  BDT BIOT BIT BOT BRT BST BTTCAT CCT CDT CEDT CEST CET CHADT CHAST CHOT ChST CHUT
  CIST CIT CKT CLST CLT COST COT CST CT CVT CWST CXT DAVT DDUT DFT EASST EAST EAT
  ECT EDT EEDT EEST EET EGST EGT EIT EST FET FJT FKST FKT FNT GALT GAMT GET GFT
  GILT GIT GMT GST GYT HADT HAEC HAST HKT HMT HOVT HST ICT IDT IOT IRDT IRKT IRST
  IST JST KGT KOST KRAT KST LHST LINT MAGT MART MAWT MDT MET MEST MHT MIST MIT MMT
  MSK MST MUT MVT MYT NCT NDT NFT NPT NST NT NUT NZDT NZST OMST ORAT PDT PET PETT
  PGT PHOT PHT PKT PMDT PMST PONT PST RET ROTT SAKT SAMT SAST SBT SCT SGT SLT SRT
  SST SYOT TAHT THA TFT TJT TKT TLT TMT TOT TVT UCT ULAT UTC UYST UYT UZT VET VLAT
  VOLT VOST VUT WAKT WAST WAT WEDT WEST WET WST YAKT YEKT
}
TIMEZONE_REGEXP =
/^(#{TIME_ZONE_ABBREVIATIONS.join('|')})$/
MONTHNAMES_REGEXP =
/^(#{Date::MONTHNAMES.compact.join('|')})$/i
ABBR_MONTHNAMES_REGEXP =
/^(#{Date::ABBR_MONTHNAMES.compact.join('|')})$/i
DAYNAMES_REGEXP =
/^(#{Date::DAYNAMES.join('|')})$/i
ABBR_DAYNAMES_REGEXP =
/^(#{Date::ABBR_DAYNAMES.join('|')})$/i
ONE_DIGIT_REGEXP =
/^\d{1}$/
TWO_DIGIT_REGEXP =
/^\d{2}$/
FOUR_DIGIT_REGEXP =
/^\d{4}$/
TIME_REGEXP =
/(\d{1,2})(:)(\d{2})(\s*)(:)?(\d{2})?(\s*)?([ap]m)?/i
MERIDIAN_LOWER_REGEXP =
/^(a|p)m$/
MERIDIAN_UPPER_REGEXP =
/^(A|P)M$/
ORDINAL_DAY_REGEXP =
/^(\d{1,2})(st|nd|rd|th)$/
OBVIOUS_24_HOUR =

Disambiguate based on value

13..23
OBVIOUS_DAY =
13..31
OBVIOUS_YEAR =
32..99
TWO_DIGIT_YEAR_EMITTER =
Emitters::TwoDigit.new(:year) { |year| year % 100 }
TWO_DIGIT_MONTH_EMITTER =
Emitters::TwoDigit.new(:month)
TWO_DIGIT_DAY_EMITTER =
Emitters::TwoDigit.new(:day)
HOUR_TO_12_HOUR =
lambda { |h| ((h - 1) % 12) + 1 }

Instance Method Summary collapse

Instance Method Details

#build_emitters(tokens) ⇒ Object

Transforms tokens that look like date/time parts to emitter objects.



70
71
72
73
74
# File 'lib/stamp/translator.rb', line 70

def build_emitters(tokens)
  tokens.map do |token|
    yield(token) || Emitters::String.new(token)
  end
end

#date_emitter(token) ⇒ Object



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
# File 'lib/stamp/translator.rb', line 107

def date_emitter(token)
  case token
  when MONTHNAMES_REGEXP
    Emitters::Lookup.new(:month, Date::MONTHNAMES)

  when ABBR_MONTHNAMES_REGEXP
    Emitters::Lookup.new(:month, Date::ABBR_MONTHNAMES)

  when DAYNAMES_REGEXP
    Emitters::Lookup.new(:wday, Date::DAYNAMES)

  when ABBR_DAYNAMES_REGEXP
    Emitters::Lookup.new(:wday, Date::ABBR_DAYNAMES)

  when TIMEZONE_REGEXP
    Emitters::Delegate.new(:zone)

  when FOUR_DIGIT_REGEXP
    Emitters::Delegate.new(:year)

  when ORDINAL_DAY_REGEXP
    Emitters::Ordinal.new(:day)

  when TWO_DIGIT_REGEXP
    value = token.to_i

    case value
    when OBVIOUS_DAY
      TWO_DIGIT_DAY_EMITTER
    when OBVIOUS_YEAR
      TWO_DIGIT_YEAR_EMITTER
    else
      Emitters::Ambiguous.new(
        TWO_DIGIT_MONTH_EMITTER,
        TWO_DIGIT_DAY_EMITTER,
        TWO_DIGIT_YEAR_EMITTER)
    end

  when ONE_DIGIT_REGEXP
    Emitters::Ambiguous.new(
      Emitters::Delegate.new(:month),
      Emitters::Delegate.new(:day))
  end
end

#time_emitter(token) ⇒ Object



76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
# File 'lib/stamp/translator.rb', line 76

def time_emitter(token)
  case token
  when MERIDIAN_LOWER_REGEXP
    Emitters::AmPm.new

  when MERIDIAN_UPPER_REGEXP
    Emitters::AmPm.new { |v| v.upcase }

  when TWO_DIGIT_REGEXP
    Emitters::Ambiguous.new(
      two_digit_hour_emitter(token),
      Emitters::TwoDigit.new(:min),
      Emitters::TwoDigit.new(:sec))

  when ONE_DIGIT_REGEXP
    # 12-hour clock without leading zero
    Emitters::Delegate.new(:hour, &HOUR_TO_12_HOUR)
  end
end

#translate(example) ⇒ Object



46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
# File 'lib/stamp/translator.rb', line 46

def translate(example)
  # extract any substrings that look like times, like "23:59" or "8:37 am"
  before, time_example, after = example.partition(TIME_REGEXP)

  # build emitters from the example date
  emitters = Emitters::Composite.new
  emitters << build_emitters(before.split(/\b/)) do |token|
    date_emitter(token)
  end

  # build emitters from the example time
  unless time_example.empty?
    time_parts = time_example.scan(TIME_REGEXP).first
    emitters << build_emitters(time_parts) do |token|
      time_emitter(token)
    end
  end

  # recursively process any remaining text
  emitters << translate(after) unless after.empty?
  emitters
end

#two_digit_hour_emitter(token) ⇒ Object



96
97
98
99
100
101
102
103
104
105
# File 'lib/stamp/translator.rb', line 96

def two_digit_hour_emitter(token)
  case token.to_i
  when OBVIOUS_24_HOUR
    # 24-hour clock
    Emitters::TwoDigit.new(:hour)
  else
    # 12-hour clock with leading zero
    Emitters::TwoDigit.new(:hour, &HOUR_TO_12_HOUR)
  end
end