Class: CapitalOneStatement

Inherits:
Object
  • Object
show all
Defined in:
lib/parse_capitalone_pdf_statement.rb,
lib/parse_capitalone_pdf_statement/version.rb

Overview

CapitalOneStatement object, with data parsed from a PDF monthly statement.

Defined Under Namespace

Classes: Transaction

Constant Summary collapse

CARD_NUMBER_REGEX = 
/(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
DATE_REGEX = 
/(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
AMOUNT_REGEX = 
/\(?\$[\d,]+\.\d\d\)?/
AMOUNT_ONLY_REGEX = 
/^ *#{AMOUNT_REGEX.source} *$/
FEES_REGEX = 
/Total Fees This Period +(#{AMOUNT_REGEX.source})/
INTEREST_REGEX = 
/Total Interest This Period +(#{AMOUNT_REGEX.source})/
TRANSACTION_REGEX = 
/^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
  AMOUNT_REGEX.source
}) *$/
VERSION = 
"1.1.0"

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(pdf_path) ⇒ CapitalOneStatement

Returns a new instance of CapitalOneStatement.



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
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 59

def initialize(pdf_path)
  @card_number        = nil
  @dec_from_prev_year = nil
  @year               = nil
  @start_date         = nil
  @end_date           = nil
  @previous_balance   = nil
  @new_balance        = nil
  @total_payments     = nil
  @total_transactions = nil
  @total_fees         = nil
  @payments           = []
  @transactions       = []
  @fees               = []

  parse_from_pdf pdf_path

  %w{payments transactions fees}.each do |type|
    trxs  = "@#{type}"
    total = "@total_#{type}"

    instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

    check_total(
      type,
      instance_variable_get(total),
      instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
    )
  end
end

Instance Attribute Details

#end_date ⇒ Date (readonly)

Returns the final day of the monthly statement.

Returns:

  • (Date) —

    the final day of the monthly statement



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

#fees ⇒ Array<CapitalOneStatement::Transaction> (readonly)

Returns array of fee transactions.

Returns:



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

#new_balance ⇒ Float (readonly)

Returns the New Balance listed in the statement.

Returns:

  • (Float) —

    the New Balance listed in the statement



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

#payments ⇒ Array<CapitalOneStatement::Transaction> (readonly)

Returns array of payment transactions.

Returns:



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

#previous_balance ⇒ Float (readonly)

Returns the “Previous Balance” field listed in the statement.

Returns:

  • (Float) —

    the “Previous Balance” field listed in the statement



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

#start_date ⇒ Date (readonly)

Returns the first day of the monthly statement.

Returns:

  • (Date) —

    the first day of the monthly statement



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

#total_fees ⇒ Float (readonly)

Returns the “Fees and Interest Charged” field listed in the statement.

Returns:

  • (Float) —

    the “Fees and Interest Charged” field listed in the statement



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

#total_payments ⇒ Float (readonly)

Returns the “Payments and Credits” field listed in the statement.

Returns:

  • (Float) —

    the “Payments and Credits” field listed in the statement



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

#total_transactions ⇒ Float (readonly)

Returns the “Transactions” field listed in the statement.

Returns:

  • (Float) —

    the “Transactions” field listed in the statement



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

#transactions ⇒ Array<CapitalOneStatement::Transaction> (readonly)

Returns array of charge transactions.

Returns:



38
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
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
267
268
269
270
271
272
273
274
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 38

class CapitalOneStatement
  CARD_NUMBER_REGEX = /(?:TRANSACTIONS|ADJUSTMENTS) FOR [^#]+ #(\d{4})/
  DATE_REGEX        = /(\w{3})\. (\d\d) - (\w{3})\. (\d\d), (\d{4})/
  AMOUNT_REGEX      = /\(?\$[\d,]+\.\d\d\)?/
  AMOUNT_ONLY_REGEX = /^ *#{AMOUNT_REGEX.source} *$/
  FEES_REGEX        = /Total Fees This Period +(#{AMOUNT_REGEX.source})/
  INTEREST_REGEX    = /Total Interest This Period +(#{AMOUNT_REGEX.source})/
  TRANSACTION_REGEX = /^ *(\d+) +(\d\d) ([A-Z][A-Z][A-Z]) (.+[^ ]) +(#{
                        AMOUNT_REGEX.source
                      }) *$/

  attr_reader :start_date,
              :end_date,
              :previous_balance,
              :total_payments,
              :total_fees,
              :total_transactions,
              :new_balance,
              :payments,
              :transactions

  def initialize(pdf_path)
    @card_number        = nil
    @dec_from_prev_year = nil
    @year               = nil
    @start_date         = nil
    @end_date           = nil
    @previous_balance   = nil
    @new_balance        = nil
    @total_payments     = nil
    @total_transactions = nil
    @total_fees         = nil
    @payments           = []
    @transactions       = []
    @fees               = []

    parse_from_pdf pdf_path

    %w{payments transactions fees}.each do |type|
      trxs  = "@#{type}"
      total = "@total_#{type}"

      instance_variable_set(trxs, instance_variable_get(trxs).sort_by {|trx| trx[:id]})

      check_total(
        type,
        instance_variable_get(total),
        instance_variable_get(trxs).inject(0) {|sum, trx| sum += trx[:amount]}
      )
    end
  end

  def to_json(*args)
    {
      :start_date         => @start_date,
      :end_date           => @end_date,
      :previous_balance   => @previous_balance,
      :total_payments     => @total_payments,
      :total_fees         => @total_fees,
      :total_transactions => @total_transactions,
      :new_balance        => @new_balance,
      :payments           => @payments,
      :transactions       => @transactions,
      :fees               => @fees
    }.to_json(*args)
  end

  private

  def parse_from_pdf(pdf_path)
    PDF::Reader.new(pdf_path).pages.each_with_index do |page, page_num|
      if @year.nil?
        walker = Struct.new(:year, :offset, :start_date, :end_date) do
          def respond_to?(_)
            true
          end

          def method_missing(name, *args)
            return unless name =~ /show_text/

            if args.any? {|str| str.to_s =~ DATE_REGEX}
              self.offset     = ($1.upcase == 'DEC' && $3.upcase == 'JAN') ? 1 : 0
              self.year       = $5.to_i
              self.start_date = Date.parse('%s-%s-%s' % [year - offset, $1, $2])
              self.end_date   = Date.parse('%s-%s-%s' % [year, $3, $4])
            end
          end
        end.new

        page.walk walker

        @dec_from_prev_year = walker.offset == 1
        @year               = walker.year
        @start_date         = walker.start_date
        @end_date           = walker.end_date
      end

      enum = page.text.split("\n").each

      loop do
        current_line = enum.next
        enum.next until (enum.peek rescue nil) != ''
        next_line    = (enum.peek rescue nil)

        parse_pdf_line page_num, current_line, next_line
      end
    end
  end

  def parse_pdf_line(page_num, line, next_line)
    if line =~ CARD_NUMBER_REGEX
      @card_number = $1.to_i
    end

    if @previous_balance.nil?
      amount_strs = line.scan AMOUNT_REGEX
      if amount_strs.size == 5
        @previous_balance,
        @total_payments,
        @total_fees,
        @total_transactions,
        @new_balance = amount_strs.map {|amount| parse_amount(amount)}

        @total_payments = -@total_payments
      end
    end

    if line =~ FEES_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "CAPITAL ONE MEMBER FEE",
        $1,
        parse_amount($1)
      )
    end

    if line =~ INTEREST_REGEX && $1 != '$0.00'
      check_parse_state
      @fees << Transaction.new(
        @fees.size + 1,
        @card_number,
        @end_date, "INTEREST CHARGE:PURCHASES",
        $1,
        parse_amount($1)
      )
    end

    transactions, payments = [(0..78), (80..-1)].map do |index|
      str      = line     .to_s[index].to_s
      next_str = next_line.to_s[index].to_s

      repair_transaction_line str, next_str
    end.map do |str|
      parse_transaction(str)
    end.compact.partition do |trx|
      trx[:amount] >= 0
    end

    @transactions += transactions
    @payments     += payments
  end

  def repair_transaction_line(line, next_line)
    if next_line =~ AMOUNT_ONLY_REGEX && !(line =~ AMOUNT_REGEX)
      line += " #{next_line.strip}"
    else
      line
    end
  end

  def parse_transaction(line)
    return nil unless line =~ TRANSACTION_REGEX &&
                      $4   != "CAPITAL ONE MEMBER FEE"

    check_parse_state

    year = ($3.upcase == 'DEC' && @dec_from_prev_year) ? @year - 1 : @year
    date = Date.parse('%s-%s-%s' % [year, $3, $2])

    Transaction.new($1.to_i, @card_number, date, $4, $5, parse_amount($5))
  end

  def parse_amount(amount)
    num = amount.gsub(/[^\d.]/, '').to_f
    amount.start_with?(?() ? -num : num
  end

  def check_parse_state
    raise "Failed to determine billing cycle dates" if @year.nil?
    raise "Failed to determine card number"         if @card_number.nil?
  end

  def check_total(type, expected, actual)
    return if actual.round(2) == expected.round(2)

    raise "Calculated %s mismatch %.2f != %.2f" % [
      type,
      actual,
      expected
    ]
  end

  # CapitalOneStatement::Transaction represents a single credit transaction
  class CapitalOneStatement::Transaction < Struct.new(
                                             :id,
                                             :card_number,
                                             :date,
                                             :description,
                                             :amount_str,
                                             :amount
                                           )
    # @!attribute id
    #    @return [Fixnum] transaction id
    #
    # @!attribute card_number
    #    @return [Fixnum] the last four digits of the account number
    #
    # @!attribute date
    #    @return [Date] the date of the transaction
    #
    # @!attribute description
    #    @return [String] the description of the transaction
    #
    # @!attribute amount_str
    #    @return [String] the dollar amount string of the transaction
    #
    # @!attribute amount
    #    @return [Float] the dollar amount parsed into a Float, negative for payments

    # @return [String] JSON representation of Transaction
    def to_json(*args)
      to_h.to_json(*args)
    end
  end
end

Instance Method Details

#to_json(*args) ⇒ Object 



90
91
92
93
94
95
96
97
98
99
100
101
102
103
 
# File 'lib/parse_capitalone_pdf_statement.rb', line 90

def to_json(*args)
  {
    :start_date         => @start_date,
    :end_date           => @end_date,
    :previous_balance   => @previous_balance,
    :total_payments     => @total_payments,
    :total_fees         => @total_fees,
    :total_transactions => @total_transactions,
    :new_balance        => @new_balance,
    :payments           => @payments,
    :transactions       => @transactions,
    :fees               => @fees
  }.to_json(*args)
end