bank_statement_parser 0.0.10 → 0.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 5c93eae163c9caceb5c2f638c676fbab045be046
-  data.tar.gz: bee74ae67ecf727306124b56d560b0d9182b1127
+  metadata.gz: 28cc31e88b6633ad90a259383e0fb6b4df8954f5
+  data.tar.gz: 703b94437f058e944a1ec40bc0f301b084518ff9
 SHA512:
-  metadata.gz: e7d81ebe279828dee500512950e43d901e7775128a9754749337db1b47258d3423e4858d4f97f142593b3c24557c64d6dde895a675b22d11baf0e3431bf7f4f2
-  data.tar.gz: aa13cc9f1f3a13945852ea4fe0633b42f1cdb18997b9f141c3806f8d08d927b1f8cb505d6a535520c1adf4108d2b308c65a1e6a890e8b3693ae1133a41e00431
+  metadata.gz: ccfcb6d0cdb6a92691d3ea9a257a88020e835b1c1ce272de80c302873ccccdc13bdf612498f2dc5f23918d84660be4c24f144c3f7d424595f9418233bb165ae6
+  data.tar.gz: 84378266b04c539eda026af368e81541b81f582b77b4aec32b356e3cbac80c845720941abe58109b560af97272277d15cee0fd3bc7bfed537040f3bcb0b07519

data/bin/bank_statement_to_text.sh

@@ -0,0 +1,22 @@
+#!/bin/sh -
+#
+# Convert the specified bank statement PDF file to text
+
+# Copyright 2015 Simon Dawson <spdawson@gmail.com>
+
+# This file is part of bank_statement_parser.
+#
+# bank_statement_parser is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# bank_statement_parser is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with bank_statement_parser. If not, see <http://www.gnu.org/licenses/>.
+
+exec pdftotext -layout -eol unix -enc UTF-8 "$1"

data/bin/bank_statement_to_yaml.rb

@@ -19,6 +19,8 @@
 # You should have received a copy of the GNU General Public License
 # along with bank_statement_parser. If not, see <http://www.gnu.org/licenses/>.
 
+require 'yaml'
+
 require 'bank_statement_parser'
 
 parser = BankStatementParser::HSBC.new
@@ -26,23 +28,4 @@ parser = BankStatementParser::HSBC.new
 # Attempt to parse the specified file
 parser.parse ARGV[0]
 
-# Statement metadata
-puts <<METADATA
-bank_statement:
-  account_number: #{parser.account_number}
-  sort_code: #{parser.sort_code}
-  statement_date: #{parser.statement_date}
-  records:
-METADATA
-
-# Statement records
-parser.records.each do |record|
-  puts <<RECORD
-    - date: #{record.date}
-      type: #{record.type}
-      credit: #{record.credit}
-      amount: #{record.amount || ''}
-      detail: #{record.detail}
-      balance: #{record.balance || ''}
-RECORD
-end
+puts YAML.dump(parser.bank_statement)

data/lib/bank_statement_parser/bank_statement.rb

@@ -0,0 +1,48 @@
+# Copyright 2015 Simon Dawson <spdawson@gmail.com>
+
+# This file is part of bank_statement_parser.
+#
+# bank_statement_parser is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# bank_statement_parser is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with bank_statement_parser. If not, see <http://www.gnu.org/licenses/>.
+
+require 'yaml'
+
+module BankStatementParser
+
+  # A bank statement
+  class BankStatement
+    attr_accessor :sort_code, :account_number, :statement_date,
+      :opening_balance, :closing_balance, :records
+
+    # Constructor
+    def initialize
+      @records = []
+    end
+
+    # Stringify
+    def to_s
+      to_yaml
+    end
+
+    # Equality test
+    def ==(other)
+      super || (sort_code == other.sort_code &&
+                account_number == other.account_number &&
+                statement_date == other.statement_date &&
+                opening_balance == other.opening_balance &&
+                closing_balance == other.closing_balance &&
+                records == other.records)
+    end
+  end
+
+end

data/lib/bank_statement_parser/base.rb

@@ -15,6 +15,8 @@
 # You should have received a copy of the GNU General Public License
 # along with bank_statement_parser. If not, see <http://www.gnu.org/licenses/>.
 
+require 'bank_statement_parser/bank_statement'
+
 module BankStatementParser
 
   # Base class for statement parsers
@@ -31,7 +33,7 @@ module BankStatementParser
 
     require 'fileutils.rb'
 
-    attr_accessor :sort_code, :account_number, :statement_date, :records
+    attr_accessor :bank_statement
 
     # Constructor
     def initialize
@@ -53,9 +55,9 @@ module BankStatementParser
       end
 
       # Sanity checking
-      raise "Failed to find sort code" if @sort_code.nil?
-      raise "Failed to find account number" if @account_number.nil?
-      raise "Failed to find statement date" if @statement_date.nil?
+      raise "Failed to find sort code" if sort_code.nil?
+      raise "Failed to find account number" if account_number.nil?
+      raise "Failed to find statement date" if statement_date.nil?
     end
 
     protected
@@ -67,10 +69,55 @@ module BankStatementParser
 
     # Reset the parser
     def reset
-      @sort_code = nil
-      @account_number = nil
-      @statement_date = nil
-      @records = []
+      @bank_statement = BankStatement.new
+    end
+
+    # @todo FIXME: Why can't we use Forwardable for these methods?
+    #
+    # Partially works, but doesn't seem to be accessible from subclasses...
+
+    def sort_code
+      @bank_statement.sort_code
+    end
+
+    def sort_code= sort_code
+      @bank_statement.sort_code = sort_code
+    end
+
+    def account_number
+      @bank_statement.account_number
+    end
+
+    def account_number= account_number
+      @bank_statement.account_number = account_number
+    end
+
+    def statement_date
+      @bank_statement.statement_date
+    end
+
+    def statement_date= statement_date
+      @bank_statement.statement_date = statement_date
+    end
+
+    def opening_balance
+      @bank_statement.opening_balance
+    end
+
+    def opening_balance= opening_balance
+      @bank_statement.opening_balance = opening_balance
+    end
+
+    def closing_balance
+      @bank_statement.closing_balance
+    end
+
+    def closing_balance= closing_balance
+      @bank_statement.closing_balance = closing_balance
+    end
+
+    def add_record record
+      @bank_statement.records << record
     end
 
   end

data/lib/bank_statement_parser/hsbc.rb

@@ -17,6 +17,7 @@
 
 require 'date'
 require 'bank_statement_parser/base'
+require 'bank_statement_parser/statement_record'
 require 'bank_statement_parser/utils'
 module BankStatementParser
 
@@ -49,23 +50,23 @@ module BankStatementParser
 
       # Look for sort code and account number lines, if we haven't found
       # one yet
-      if @sort_code.nil? && @account_number.nil?
+      if sort_code.nil? && account_number.nil?
         if line =~ /(?:\A[A-Z][\w\s]+|,)\s+(?<sort_code>\d{2}-\d{2}-\d{2})\s+(?<account_number>\d{8})(?:\s*|\s+\d+)\z/
           logger.debug { "Found sort code and account number" }
-          @sort_code = Regexp.last_match(:sort_code)
-          @account_number = Regexp.last_match(:account_number)
+          self.sort_code = Regexp.last_match(:sort_code)
+          self.account_number = Regexp.last_match(:account_number)
         end
       end
 
       # Look for statement date lines, if we haven't found one yet
-      if @statement_date.nil?
+      if statement_date.nil?
         if line =~ /\A\s*(?<statement_date>\d{2} (?:#{MONTHS.map{|m| m[0,3]}.join('|')}) \d{4})\s*\z/
           logger.debug { "Found statement date (1st form)" }
           @statement_format = StatementFormat::FORMAT_1ST
 
           # Parse statement date
           date_string = Regexp.last_match(:statement_date)
-          @statement_date = Date.parse(date_string)
+          self.statement_date = Date.parse(date_string)
         elsif line =~ /\A(?<date_range_start>\d+\s+(?:#{MONTHS.join('|')})(?:\s+\d{4})?)\s+to\s+(?<date_range_end>\d+\s+(?:#{MONTHS.join('|')})\s+\d{4})\b/
           logger.debug { "Found statement date (2nd form)" }
           @statement_format = StatementFormat::FORMAT_2ND
@@ -75,11 +76,11 @@ module BankStatementParser
           logger.debug { "Found statement date range #{date_range_start}-#{date_range_end}" }
 
           # Parse range end date
-          @statement_date = Date.parse(date_range_end)
+          self.statement_date = Date.parse(date_range_end)
         end
       end
 
-      if !@sort_code.nil? && !@account_number.nil? && !@statement_date.nil?
+      if !sort_code.nil? && !account_number.nil? && !statement_date.nil?
 
         # Look for statement records proper
         headings = nil
@@ -161,14 +162,14 @@ module BankStatementParser
       # The date we have parsed will have the year set to the current year.
       #
       # We need to figure out the correct year, from the statement date.
-      raise "No statement date" unless @statement_date
-      record_date = Date.new(@statement_date.year,
+      raise "No statement date" unless statement_date
+      record_date = Date.new(statement_date.year,
                              record_date.month,
                              record_date.day)
       logger.debug { "record date #{record_date}" }
-      if @statement_date.month != record_date.month
+      if statement_date.month != record_date.month
         logger.debug { "record month differs from statement month" }
-        if 1 == @statement_date.month
+        if 1 == statement_date.month
           # Assume that the statement crosses a year boundary: the record
           # must be from the end of the previous year
           raise "Expected a record from December" unless
@@ -371,7 +372,7 @@ module BankStatementParser
                                      detail: full_details,
                                      balance: balance)
         logger.debug { "Created statement record: #{record}" }
-        @records << record
+        add_record record
 
         @cached_payment_type = nil
         @cached_details = []

data/lib/bank_statement_parser/statement_record.rb

@@ -15,6 +15,8 @@
 # You should have received a copy of the GNU General Public License
 # along with bank_statement_parser. If not, see <http://www.gnu.org/licenses/>.
 
+require 'yaml'
+
 module BankStatementParser
 
   # A bank statement record
@@ -35,12 +37,7 @@ module BankStatementParser
 
     # Stringify
     def to_s
-      "%s:%s:%s:%s:%s:%s" % [date,
-                             type,
-                             credit.to_s,
-                             (amount || ''),
-                             detail,
-                             (balance || '')]
+      to_yaml
     end
 
     # Equality test

data/lib/bank_statement_parser.rb

@@ -17,7 +17,6 @@
 
 require 'logger'
 require 'bank_statement_parser/hsbc'
-require 'bank_statement_parser/statement_record'
 module BankStatementParser
 
   @@logger = Logger.new(STDERR)

metadata

@@ -1,24 +1,27 @@
 --- !ruby/object:Gem::Specification
 name: bank_statement_parser
 version: !ruby/object:Gem::Version
-  version: 0.0.10
+  version: 0.1.0
 platform: ruby
 authors:
 - Simon Dawson
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-03-26 00:00:00.000000000 Z
+date: 2015-03-28 00:00:00.000000000 Z
 dependencies: []
 description: A gem for parsing bank statements
 email: spdawson@gmail.com
 executables:
 - bank_statement_to_yaml.rb
+- bank_statement_to_text.sh
 extensions: []
 extra_rdoc_files: []
 files:
+- bin/bank_statement_to_text.sh
 - bin/bank_statement_to_yaml.rb
 - lib/bank_statement_parser.rb
+- lib/bank_statement_parser/bank_statement.rb
 - lib/bank_statement_parser/base.rb
 - lib/bank_statement_parser/hsbc.rb
 - lib/bank_statement_parser/statement_record.rb