io-segmenter 0.6.0

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA256:
+  metadata.gz: 794c025371c0f2a36e5648199101b20294bfb4ffb160c9d1c8d55d016115d660
+  data.tar.gz: 821db4894a29799ab3f2a84a0574839470d9080bebeedfee8e368ef09f73b2ba
+SHA512:
+  metadata.gz: 5e4a7a3bfedac02138e053800bb619a5e38e5e19a4d3b623f7cb82abb04fddb9b3458d5c88bc0a07ffc8d1fa81350e5e8db227cb8cab8dd736d3dc585c127789
+  data.tar.gz: 7a12973857ed46aaef093870ff5a008649500404bb7911966eb9258d5d7e96ac5d765988d721ec50ad9e71e3a4870fa48180a3f6bf7cdfb32eef5742291841d0

data/lib/io-segmenter.rb

@@ -0,0 +1,88 @@
+# frozen_string_literal: true
+
+# IO iterator for parsing segments of streams
+class IOSegmenter
+  include Enumerable
+
+  def initialize(io, starting_char = '{', ending_char = '}', quote_char = '"', escape_char = '\\', max_read_size = 8192)
+    @io = io
+    @starting_char = starting_char
+    @ending_char = ending_char
+    @quote_char = quote_char
+    @escape_char = escape_char
+    @max_read_size = max_read_size
+
+    terms = [
+      @starting_char,
+      @ending_char,
+      @quote_char,
+      @escape_char
+    ]
+    terms.compact!
+    terms.map! { |str| Regexp.escape(str) }
+
+    @search = Regexp.new('(:?' + terms.join('|') + ')')
+  end
+
+  def each
+    buffer = String.new
+
+    until @io.eof?
+      buffer << @io.read(@max_read_size)
+      each_segment(buffer) do |segment|
+        yield segment
+      end
+    end
+  end
+
+  def self.foreach(*args, &block)
+    new(*args).each(&block)
+  end
+
+  private
+
+  def each_segment(buffer)
+    return unless (start_offset = buffer.index(@starting_char))
+
+    brackets = 1
+    offset = start_offset
+
+    opened_quote = false
+
+    while (offset = buffer.index(@search, offset + 1))
+      case buffer[offset]
+      when @ending_char
+        next if opened_quote
+        brackets -= 1
+      when @starting_char
+        next if opened_quote
+        brackets += 1
+      when @quote_char
+        opened_quote = !opened_quote
+        next
+      when @escape_char
+        offset += @escape_char.size
+        next
+      else
+        if @ending_char == buffer[offset, @ending_char.size]
+          next if opened_quote
+          brackets -= 1
+        elsif @starting_char == buffer[offset, @starting_char.size]
+          next if opened_quote
+          brackets += 1
+        else
+          raise("unhandled offset #{offset}, at #{buffer[offset, 20]}...")
+        end
+      end
+
+      next unless brackets.zero?
+
+      len = (offset + @ending_char.size) - start_offset
+      yield buffer[start_offset, len]
+      buffer.slice!(0, offset + @ending_char.size)
+      return unless (start_offset = buffer.index(@starting_char))
+      offset = start_offset
+      brackets = 1
+    end
+  end
+end

metadata

@@ -0,0 +1,43 @@
+--- !ruby/object:Gem::Specification
+name: io-segmenter
+version: !ruby/object:Gem::Version
+  version: 0.6.0
+platform: ruby
+authors:
+- Doug Youch
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2018-09-09 00:00:00.000000000 Z
+dependencies: []
+description: Used to iterate over segments of data in IO objects
+email: dyouch5@yahoo.com
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/io-segmenter.rb
+homepage: https://github.com/dougyouch/io-segemnter
+licenses: []
+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: []
+rubyforge_project: 
+rubygems_version: 2.7.6
+signing_key: 
+specification_version: 4
+summary: IO Segemnter
+test_files: []