mbox_merge 0.1.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: e39d77dd7c066dd650439c52f9ca3e6cd122749fa76c9822b8b2b10f9a37c606
+  data.tar.gz: a7e4809e4675e7fc01222f96d81c5a979b2db0896cbd7fad6ee9107b55b14637
+SHA512:
+  metadata.gz: 7aaa43cb6e3d643fb3b2850c04cbc3f9a00b20f68cc19bd855c196b6def23d1e81e10d9b80cafc34b9cf94efaeaa8b30b25cb6eb30f0260526724a79d59ba025
+  data.tar.gz: 0a8551999f6e85f39c1165153905cc6e2a5f4c12589b82c5f700d52504b814875d15376383c21181f2adc6684adf74f77510942f6c07c09ca6570bdd534732b8

data/lib/mbox_merge.rb

@@ -0,0 +1,86 @@
+require 'mail'
+require 'ruby-progressbar'
+
+module MboxMerge
+    class Merger
+        # Merges multiple mbox files into a single mbox file, sorting emails by date
+        # and presenting statistics of merged emails.
+        #
+        # @param output_mbox [String] Path to the output mbox file
+        # @param mbox_files [Array<String>] List of mbox files to be merged
+        # @return [Statistics] Statistics object containing information about the merging
+        def self.merge_mboxes(output_mbox, *mbox_files)
+            emails = []
+
+            # Load emails from each mbox file
+            mbox_files.each do |mbox_file|
+                File.open(mbox_file, 'r') do |f|
+                    # Split the file into chunks using the standard mbox 'From ' separator
+                    f.read.split(/^From /).each do |raw_message|
+                        begin
+                            # Only try to read valid messages (non-empty chunks)
+                            unless raw_message.strip.empty?
+                                # Prepend the 'From ' as it's removed by the split
+                                email = Mail.read_from_string("From #{raw_message}")
+                                emails << email
+                            end
+                        rescue => e
+                            puts "Failed to parse email: #{e.message}"
+                        end
+                    end
+                end
+            end
+
+            # Calculate the total emails
+            total_emails = emails.size
+
+            # Sort emails by date (convert dates to Time objects for consistent comparison)
+            sorted_emails = emails.sort_by do |email|
+                begin
+                    email.date.to_time
+                rescue
+                    Time.now  # emails without a valid date (TODO)
+                end
+            end
+
+            # Progress bar for merging
+            progress_bar = ProgressBar.create(title: "Merging", total: sorted_emails.size, format: '%t: |%B| %p%%')
+
+            # Write the sorted emails to the output mbox file
+            File.open(output_mbox, 'w') do |out|
+                sorted_emails.each do |email|
+                    # Ensure email date is present; fallback to Time.now if nil
+                    email_date = email.date ? email.date.to_time.strftime("%a, %d %b %Y %H:%M:%S %z") : Time.now.strftime("%a, %d %b %Y %H:%M:%S %z")
+                    out.puts "From #{email.envelope_from} #{email_date}"
+                    out.puts email.to_s
+                    progress_bar.increment
+                end
+            end
+
+            # Return statistics
+            Statistics.new(mbox_files.size, total_emails)
+        end
+    end
+
+    class Statistics
+        attr_reader :file_count, :email_count
+
+        # Initializes the statistics for the merge process
+        #
+        # @param file_count [Integer] The number of mbox files merged
+        # @param email_count [Integer] The total number of emails merged
+        def initialize(file_count, email_count)
+            @file_count = file_count
+            @email_count = email_count
+        end
+
+        # Prints the statistics to the console
+        #
+        # @return [void]
+        def print_statistics
+            puts "Merging mbox files completed!"
+            puts "Total Mbox Files: #{@file_count}"
+            puts "Total Emails Merged: #{@email_count}"
+        end
+    end
+end

metadata

@@ -0,0 +1,73 @@
+--- !ruby/object:Gem::Specification
+name: mbox_merge
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+platform: ruby
+authors:
+- firefly-cpp
+autorequire:
+bindir: bin
+cert_chain: []
+date: 2024-11-12 00:00:00.000000000 Z
+dependencies:
+- !ruby/object:Gem::Dependency
+  name: mail
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.7'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '2.7'
+- !ruby/object:Gem::Dependency
+  name: ruby-progressbar
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.11'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '1.11'
+description: Merges multiple mbox files, orders emails by date, and provides statistics
+  on the merged emails.
+email:
+- iztok@iztok-jr-fister.eu
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/mbox_merge.rb
+homepage: https://codeberg.org/firefly-cpp/mbox_merge
+licenses:
+- MIT
+metadata: {}
+post_install_message:
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - ">="
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubygems_version: 3.5.16
+signing_key:
+specification_version: 4
+summary: A tool to merge multiple mbox files and order emails by date.
+test_files: []