extras_de_cont 1.1.0 → 1.3.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: '029f7b904cb76faa331a686ec02f81a0fad9ffce39e0bea1c05e0ed95a6bb34c'
-  data.tar.gz: c98d9dc06a00db1cface2c7ea7b58f563502c2fa58da4481d2a928e99fe08766
+  metadata.gz: df78cdd18c6c97610fd739d7154922d9d5cafe9d8460c54104429698d93d0479
+  data.tar.gz: 8ea76ba23c3c32f5a4d39c7610abdb3dbef95a8c96cb831592316c81165bd610
 SHA512:
-  metadata.gz: cd20be09910a3ca8ad2a13bc07ce619d76950e080e3bb927a2f3dee0f24441f173ad842437dfc9f8196b24579d5976519c5558031aa8cd33e93cd5263e229705
-  data.tar.gz: 27c5cda6efa0043af432ac596c7d2ff736dd3f26c520f3cd2ea79a45506eb41a48d5c9ca236432d1f9999e73f9d8c11dabc201e4ee23cab82f08e9bc7a351b6b
+  metadata.gz: ccfbbd9f393fe5de67d966edbdfae2b50e2cc23a40d8ec5def3e532176648e5243b08e22a7387423ecce688c81e0e0f9dd78a5b493678f620b079ee0e63b5ab5
+  data.tar.gz: 44b5c55785d2b93c0c85797826a90fc56733c9fa2dc60e494e61ee2e3041857498d32603cf410ac9c70998f463971b72134f4a9d8cd986374249713c3fe8a124

data/README.md

@@ -35,3 +35,78 @@ Run the Revolut parser test with:
 ```bash
 ruby -Ilib:test test/revolut_rule_test.rb
 ```
+
+## Supported Banks
+
+| Bank | Symbol | Currencies | Features |
+|---|---|---|---|
+| Revolut | `:revolut` | RON, EUR, USD, GBP, PLN, CZK, HUF, BGN, TRY, UAH | Personal & Business, multi-section, symbol currencies |
+| UniCredit | `:unicredit` | RON, EUR | Romanian month names, page breaks, transaction markers |
+| BRD | `:brd` | RON, EUR | Below-line amounts, Romanian number format |
+
+## Development
+
+```bash
+bundle install          # Install dependencies
+bundle exec rake test   # Run all tests
+bundle exec rake standard  # Run linter
+bundle exec rake build  # Build gem
+```
+
+## Contributing
+
+Contributions are welcome. Here is how to add a new bank:
+
+### 1. Gather information
+
+Obtain a sample PDF statement from the bank and extract its text:
+
+```bash
+ruby -Ilib -e 'require "extras_de_cont"; puts ExtrasDeCont::Parser.new(ARGV[0]).text' /path/to/statement.pdf
+```
+
+### 2. Create a rule class
+
+Add `lib/extras_de_cont/rules/<bank>.rb` inheriting from `Rules::Base`. See existing rules in `lib/extras_de_cont/rules/` for patterns.
+
+### 3. Register the bank
+
+Add the require and `BANK_RULES` entry in `lib/extras_de_cont.rb`:
+
+```ruby
+require "extras_de_cont/rules/<bank>"
+# ...
+BANK_RULES = {
+  brd: Rules::Brd,
+  <bank>: Rules::<BankName>,
+  revolut: Rules::Revolut,
+  unicredit: Rules::UniCredit
+}.freeze
+```
+
+### 4. Write tests
+
+Add `test/extras_de_cont/rules/<bank>_rule_test.rb`. Use sanitized fixtures — never commit real financial data. Run with:
+
+```bash
+bundle exec ruby -Ilib:test test/extras_de_cont/rules/<bank>_rule_test.rb
+bundle exec rake standard
+```
+
+### Code style
+
+This project uses [Standard Ruby](https://github.com/standardrb/standard). Run `bundle exec rake standard` before submitting changes.
+
+## AI-Assisted Development
+
+This project supports AI-assisted development via [OpenCode](https://github.com/anomalyco/opencode).
+
+A skill is provided for adding new bank statement parsers. When using an AI coding assistant, it can load the skill 
+at `.agents/skills/add-bank-statement/SKILL.md` to guide the process. 
+The skill covers rule class creation, registration, RBS signatures, test patterns, and fixture anonymization.
+
+The project maintains a [MemPalace](https://github.com/anomalyco/mempalace) knowledge graph (`mempalace.yaml`) 
+that organizes the codebase into wings and rooms. AI agents can query this to learn about the project structure, 
+existing parsers, test patterns, and conventions without reading every file.
+
+For more context, see `AGENTS.md`.

data/extras_de_cont.gemspec

@@ -2,7 +2,7 @@
 
 Gem::Specification.new do |s|
   s.name = "extras_de_cont"
-  s.version = "1.1.0"
+  s.version = "1.3.0"
   s.licenses = ["GPLv3"]
   s.summary = "A simple library which helps you extract transactions from a PDF bank statement."
   s.description = <<~TEXT
@@ -14,8 +14,8 @@ Gem::Specification.new do |s|
   s.authors = ["Denis Nutiu"]
   s.email = "dnutiu@nuculabs.dev"
   s.homepage = "https://nuculabs.dev"
-  s.metadata = { "source_code_uri" => "https://gitlab.nuculabs.dev/dnutiu/extras-de-cont",
-                 "rubygems_mfa_required" => "true" }
+  s.metadata = {"source_code_uri" => "https://gitlab.nuculabs.dev/dnutiu/extras-de-cont",
+                 "rubygems_mfa_required" => "true"}
   s.required_ruby_version = ">= 3.0.0"
 
   # Files to include in the gem

data/lib/extras_de_cont/rules/brd.rb

@@ -0,0 +1,185 @@
+# frozen_string_literal: true
+
+require "date"
+require "extras_de_cont/transaction"
+
+module ExtrasDeCont
+  module Rules
+    class Brd < Rules::Base
+      DATE_PREFIX = /\A(?<date>\d{2}\/\d{2}\/\d{4})\b/
+      AMOUNT_PATTERN = /\d{1,3}(?:\.\d{3})*,\d{2}/
+      TABLE_HEADER_PATTERN = /Data oper\.\s+Descriere operatiune\s+Debit\s+Credit\s+Data val\./
+
+      NOISE_PATTERNS = [
+        /\APag\./,
+        /\ABRD-Groupe/,
+        /\ACAPITAL SOCIAL/,
+        /\AMihalache/,
+        /\ATel:/,
+        /\ARO361579/,
+        /\A255\/06/,
+        /\APJR01INCR/,
+        /\ACIFRE CHEIE/,
+        /\ACONTURI DETINUTE/,
+        /\AFonduri proprii/,
+        /\ALimita de credit/,
+        /\ACredit neutilizat/,
+        /\ADescoperit/,
+        /\ANr\. Zile/,
+        /\ATotal disponibil/,
+        /\ATotal sume/,
+        /\ADomicilierea contului/,
+        /\AReferinte bancare/,
+        /\ATitular \/ Account/,
+        /\AIBAN /,
+        /\ANumar cont/,
+        /\AExtras de cont/,
+        /\ADe la \/ From/,
+        /\ADocumentul este/,
+        /\ACNP\/CUI:/,
+        /\ASWIFT/,
+        /\AAg\. /,
+        /\AStr\. /,
+        /\A•/,
+        /\Ahttp/,
+        /\ASold/,
+        /\ATotal debit/,
+        /\ACard:MBS/,
+        /\ATrans\.Date/
+      ].freeze
+
+      def parse(text)
+        transactions = []
+        current_table = nil
+        current_currency = nil
+        above_lines = []
+        below_lines = []
+        date_line = nil
+
+        each_normalized_line(text) do |line|
+          if line.start_with?("Valuta / Currency")
+            current_currency = line.split.last
+            next
+          end
+
+          if table_header?(line)
+            try_flush(date_line, above_lines, below_lines, current_table, current_currency, transactions)
+            current_table = extract_columns(line)
+            above_lines = []
+            below_lines = []
+            date_line = nil
+            next
+          end
+
+          if noise?(line)
+            try_flush(date_line, above_lines, below_lines, current_table, current_currency, transactions)
+            above_lines = []
+            below_lines = []
+            date_line = nil
+            next
+          end
+
+          next if current_table.nil?
+
+          if date_line?(line)
+            try_flush(date_line, above_lines, below_lines, current_table, current_currency, transactions)
+            date_line = line
+            below_lines = []
+            next
+          end
+
+          if date_line
+            below_lines << line
+          else
+            above_lines << line
+          end
+        end
+
+        try_flush(date_line, above_lines, below_lines, current_table, current_currency, transactions)
+        transactions
+      end
+
+      private
+
+      def each_normalized_line(text)
+        text.each_line do |line|
+          normalized = line.tr("\u00A0", " ").strip
+          next if normalized.empty?
+          yield normalized
+        end
+      end
+
+      def table_header?(line)
+        line.match?(TABLE_HEADER_PATTERN)
+      end
+
+      def date_line?(line)
+        line.match?(DATE_PREFIX)
+      end
+
+      def noise?(line)
+        NOISE_PATTERNS.any? { |pattern| line.match?(pattern) }
+      end
+
+      def extract_columns(line)
+        {
+          debit: line.index("Debit"),
+          credit: line.index("Credit")
+        }
+      end
+
+      def try_flush(date_line, above_lines, below_lines, table, currency, transactions)
+        return if date_line.nil? || table.nil?
+
+        transaction = build_transaction(date_line, above_lines, below_lines, table, currency)
+        transactions << transaction if transaction
+      end
+
+      def build_transaction(date_line, above_lines, below_lines, table, currency)
+        date_match = date_line.match(DATE_PREFIX)
+        return if date_match.nil?
+
+        date = parse_date(date_match[:date])
+
+        amounts_on_date = date_line.to_enum(:scan, AMOUNT_PATTERN).map { Regexp.last_match }
+
+        if amounts_on_date.any?
+          transaction_amount_match = amounts_on_date.first
+          main_description = date_line[date_match.end(0)...transaction_amount_match.begin(0)].to_s.strip
+        else
+          amount_line = below_lines.find { |l| l.match?(AMOUNT_PATTERN) }
+          return unless amount_line
+
+          amount_match = amount_line.match(AMOUNT_PATTERN)
+          transaction_amount_match = amount_match
+          main_description = date_line[date_match.end(0)..].to_s.strip
+          main_description = main_description.sub(/\s*\d{2}\/\d{2}\/\d{4}\s*\z/, "")
+        end
+
+        midpoint = (table[:debit] + 2 * table[:credit]) / 3
+        debit = transaction_amount_match.begin(0) < midpoint
+        amount = parse_amount(transaction_amount_match[0])
+        amount = -amount if debit
+
+        description_lines = below_lines.map(&:strip).reject(&:empty?)
+        if !amounts_on_date.any? && description_lines.first
+          stripped = description_lines.first.sub(AMOUNT_PATTERN, "").strip
+          description_lines[0] = stripped unless stripped.empty?
+        end
+        description_parts = [*above_lines.map(&:strip), main_description, *description_lines]
+        description = description_parts.reject(&:empty?).join(" | ")
+
+        Transaction.new(date, description, amount, currency)
+      end
+
+      def parse_date(value)
+        day, month, year = value.split("/").map(&:to_i)
+        Date.new(year, month, day)
+      end
+
+      def parse_amount(value)
+        value.delete(".").sub(",", ".").to_f
+      end
+    end
+  end
+end

data/lib/extras_de_cont/rules/ing.rb

@@ -0,0 +1,174 @@
+# frozen_string_literal: true
+
+require "date"
+require "extras_de_cont/transaction"
+
+module ExtrasDeCont
+  module Rules
+    class Ing < Rules::Base
+      ROMANIAN_MONTHS = {
+        "ianuarie" => 1, "februarie" => 2, "martie" => 3, "aprilie" => 4,
+        "mai" => 5, "iunie" => 6, "iulie" => 7, "august" => 8,
+        "septembrie" => 9, "octombrie" => 10, "noiembrie" => 11, "decembrie" => 12
+      }.freeze
+
+      RO_MONTH_NAMES = ROMANIAN_MONTHS.keys.freeze
+      DATE_PATTERN = /\b(\d{1,2})\s+(#{RO_MONTH_NAMES.join("|")})\s+(\d{4})\b/i
+      DATE_PREFIX = /\A\s*#{DATE_PATTERN}/
+
+      TABLE_HEADER_PATTERN = /Data\s+Detalii tranzactie\s+Debit\s+Credit/
+
+      AMOUNT_PATTERN = /\d{1,3}(?:\.\d{3})*,\d{2}/
+
+      NOISE_PATTERNS = [
+        /\AExtras de cont\z/,
+        /\APentru perioada:/,
+        /\AValabil fara semnatura/,
+        /\AING Bank/,
+        /\ASediul:/,
+        /\ANr\. inregistrare/,
+        /\ACIF:/,
+        /\ATitular cont:/,
+        /\ACNP:/,
+        /\AStr\. /,
+        /\ATip cont:/,
+        /\ANumar cont:/,
+        /\AMoneda:/,
+        /\A\d{6},/,
+        /\ARoxana Petria/,
+        /\AAlexandra Ilie/,
+        /\AȘef Serviciu/,
+        /\ASef Serviciu/,
+        /\ASucursala/,
+        /\AÎN/,
+        /\AInformatii despre/,
+        /\Ape www\./,
+        /\d+\/\d+$/
+      ].freeze
+
+      def parse(text)
+        transactions = []
+        current_currency = nil
+        current_table = nil
+        above_lines = []
+        below_lines = []
+        date_line = nil
+
+        each_normalized_line(text) do |line|
+          if line.start_with?("Moneda:")
+            current_currency = line.split.last
+            next
+          end
+
+          if table_header?(line)
+            try_flush(date_line, above_lines, below_lines, current_table, current_currency, transactions)
+            current_table = extract_column_positions(line)
+            above_lines, below_lines, date_line = [], [], nil
+            next
+          end
+
+          if noise?(line)
+            try_flush(date_line, above_lines, below_lines, current_table, current_currency, transactions)
+            above_lines, below_lines, date_line = [], [], nil
+            next
+          end
+
+          next if current_table.nil?
+
+          if date_line?(line)
+            try_flush(date_line, above_lines, below_lines, current_table, current_currency, transactions)
+            date_line = line
+            below_lines = []
+            next
+          end
+
+          if date_line
+            below_lines << line
+          else
+            above_lines << line
+          end
+        end
+
+        try_flush(date_line, above_lines, below_lines, current_table, current_currency, transactions)
+        transactions
+      end
+
+      private
+
+      def each_normalized_line(text)
+        text.each_line do |line|
+          normalized = line.tr("\u00A0", " ").strip
+          next if normalized.empty?
+          yield normalized
+        end
+      end
+
+      def table_header?(line)
+        line.match?(TABLE_HEADER_PATTERN)
+      end
+
+      def date_line?(line)
+        line.match?(DATE_PREFIX)
+      end
+
+      def noise?(line)
+        NOISE_PATTERNS.any? { |pattern| line.match?(pattern) }
+      end
+
+      def extract_column_positions(line)
+        {
+          debit: line.index("Debit"),
+          credit: line.index("Credit")
+        }
+      end
+
+      def try_flush(date_line, above_lines, below_lines, table, currency, transactions)
+        return if date_line.nil? || table.nil?
+
+        transaction = build_transaction(date_line, above_lines, below_lines, table, currency)
+        transactions << transaction if transaction
+      end
+
+      def build_transaction(date_line, above_lines, below_lines, table, currency)
+        date_match = date_line.match(DATE_PREFIX)
+        return if date_match.nil?
+
+        amounts = date_line.to_enum(:scan, AMOUNT_PATTERN).map { Regexp.last_match }
+        return if amounts.empty?
+
+        transaction_amount_match = amounts.last
+        description_start = date_match.end(0)
+        description_end = transaction_amount_match.begin(0)
+        main_description = date_line[description_start...description_end].to_s.strip
+
+        amount_string = transaction_amount_match[0]
+        amount = parse_amount(amount_string)
+        midpoint = (table[:debit] + table[:credit]) / 2
+        amount = -amount if transaction_amount_match.begin(0) < midpoint
+
+        description = build_description(main_description, above_lines, below_lines)
+
+        Transaction.new(
+          parse_date(date_match[1].to_i, date_match[2], date_match[3].to_i),
+          description,
+          amount,
+          currency
+        )
+      end
+
+      def parse_date(day, month_name, year)
+        month = ROMANIAN_MONTHS[month_name.downcase]
+        Date.new(year, month, day)
+      end
+
+      def parse_amount(value)
+        value.delete(".").sub(",", ".").to_f
+      end
+
+      def build_description(main_desc, above_lines, below_lines)
+        parts = [*above_lines.map(&:strip), main_desc, *below_lines.map(&:strip)]
+        parts.reject(&:empty?).join(" | ")
+      end
+    end
+  end
+end

data/lib/extras_de_cont.rb

@@ -2,6 +2,8 @@
 
 require "extras_de_cont/parser"
 require "extras_de_cont/rules/base"
+require "extras_de_cont/rules/brd"
+require "extras_de_cont/rules/ing"
 require "extras_de_cont/rules/revolut"
 require "extras_de_cont/rules/unicredit"
 
@@ -9,8 +11,10 @@ require "extras_de_cont/rules/unicredit"
 module ExtrasDeCont
   # Map of supported banks (symbol → rule class)
   BANK_RULES = {
-    unicredit: Rules::UniCredit,
-    revolut: Rules::Revolut
+    brd: Rules::Brd,
+    ing: Rules::Ing,
+    revolut: Rules::Revolut,
+    unicredit: Rules::UniCredit
   }.freeze
 
   class << self

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: extras_de_cont
 version: !ruby/object:Gem::Version
-  version: 1.1.0
+  version: 1.3.0
 platform: ruby
 authors:
 - Denis Nutiu
@@ -39,6 +39,8 @@ files:
 - lib/extras_de_cont.rb
 - lib/extras_de_cont/parser.rb
 - lib/extras_de_cont/rules/base.rb
+- lib/extras_de_cont/rules/brd.rb
+- lib/extras_de_cont/rules/ing.rb
 - lib/extras_de_cont/rules/revolut.rb
 - lib/extras_de_cont/rules/unicredit.rb
 - lib/extras_de_cont/transaction.rb