stream_parser 0.1 → 0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 0e47f922c40712cf85591d6fe3f5e39b8fd974964a9c8604bc1d91dcaebedd36
-  data.tar.gz: fed08e64284d9b798822d620d06d7618932ff3ac3546e5033186aba807befc54
+  metadata.gz: 3cc1101615985f944d1a8f88cd37509d985860be8950fed026bd75399184ce2d
+  data.tar.gz: c5088acc9378bb08a988a4d53c6fc3d9ec0d92ddfb20a4484a8c5b83e5ce6031
 SHA512:
-  metadata.gz: d25b559fb053168d5637ad0f774eb291b98ee34cd7baf30858699b1584e533f8322fdcd677685f4647bc101df997251cea1cb843b9d3d5d511fbd234b75e1ce7
-  data.tar.gz: e2fa502f87ed3285044086780c0e615641edccdb1e3680418a004e247f19dbd389a2f6465068bfb1081b44a2599e159ede05101f5f1e691d1f6f89802d967b4c
+  metadata.gz: 8519f3bbcd0dc82e16fb52d366f893fc2465720ac626aa7822604a294c55baed11567bf62ae4b2f6971de0df8f67c2b5722413f1fdd8a86466b894810ef105ec
+  data.tar.gz: 7c5888f18e4dd6f65d86d117826294bfa4d3aac80a9c14c529764ac93243beb06fbef751107374818b5497fd9e571cfeab5f585c4012cf6763b440d5d23c8885

data/README.md

@@ -55,5 +55,7 @@ QuotedStringFinder.parse(%q{Here "ar})
 # => SyntaxError "Unbalanced Quotes in string"
 ```
 
+Although we grab quoted values ourselfs in this example there is a `quoted_value` helper as well as a
+`StreamParser::HTML` which provides additional helpers such as `next_tag`, `scan_for_tag`, `next_end_tag` and others.
 
 

data/lib/stream_parser/html/tag.rb

@@ -0,0 +1,39 @@
+class StreamParser::HTML::Tag
+  attr_reader :name, :attributes
+  attr_writer :self_closing
+  
+  def initialize(name, closing=false)
+    @name = name
+    @attributes = {}
+    @closing = closing
+    @self_closing = false
+  end
+
+  def [](key)
+    @attributes[key.to_sym]
+  end
+
+  def []=(key, value)
+    @attributes[key.to_sym] = value
+  end
+  
+  def self_closing?
+    @self_closing
+  end
+  
+  def closing?
+    @closing
+  end
+  
+  def opening?
+    !@closing
+  end
+  
+  def match(name: nil, closing: nil, attributes: nil)
+    return false if name && @name != name
+    return false if !closing.nil? && @closing != closing
+    return false if attributes && !attributes.all? { |k,v| @attributes[k] == v }
+    
+    true
+  end
+end

data/lib/stream_parser/html.rb

@@ -0,0 +1,112 @@
+module StreamParser::HTML
+  
+  autoload :Tag, File.expand_path('../html/tag', __FILE__)
+  
+  def self.included(base)
+    base.include(StreamParser)
+  end
+  
+  def next_tag(old_index: nil)
+    old_index ||= @index
+    return unless scan_until(/<\s*/)
+    start_index = @index-1
+
+    while peek(3) == '!--'
+      forward(3)
+      scan_until(/-->\s*/)
+      scan_until(/<\s*/)
+    end
+    
+    # HTMLComment.new(pre_match)
+    if peek(1) == '/'
+      scan_until(/[^>\s\/]+/)
+      scan_tag(match, old_index: old_index, start_index: start_index, closing: true)
+    else
+      scan_until(/[^>\s\/]+/)
+      scan_tag(match, old_index: old_index, start_index: start_index) 
+    end
+  end
+  
+  def scan_for_tag(name, closing: nil, **attributes)
+    old_index ||= @index
+    tag = next_tag
+    while tag && !tag.match(name: name, closing: closing, attributes: attributes)
+      tag = next_tag(old_index: old_index)
+    end
+    tag
+  end
+  
+  def scan_for_closing_tag
+    old_index = @index
+    heap = []
+    
+    tag = next_tag
+    puts tag.inspect
+    while tag && !(tag.closing? && heap.empty?)
+      if !tag.closing? && !tag.self_closing?
+        heap << tag
+      elsif !tag.self_closing?
+        heap.pop
+      end
+      tag = next_tag(old_index: old_index)
+    end
+    @old_index = old_index
+    tag
+  end
+  
+  def scan_tag(name, closing: false, old_index:, start_index:)
+    tag = Tag.new(name, closing)
+    
+    while !eos?
+      gobble(/\s+/)
+      key = case peek(1)
+      when '>'
+        forward(1)
+        @old_index = old_index
+        @match = @source[start_index...@index]
+        return tag
+      when '/'
+        forward(1)
+        gobble(/\s*\>/)
+        @old_index = old_index
+        @match = @source[start_index...@index]
+        tag.self_closing = true
+        return tag
+      when '"', "'"
+        quote_char = next_char
+        forward(1)
+        quoted_value(quote_char)
+      else
+        scan_until(/[^>\s\/=]+/)[0]
+      end
+      
+      tag[key] = if next?(/\s*=/)
+        gobble(/\s*=/)
+        html_tag_value
+      else
+        true
+      end
+    end
+
+    @old_index = old_index
+    @match = @source[start_index...@index]
+    tag
+  end
+  
+  def html_tag_value
+    gobble(/\s+/)
+    case peek(1)
+    when '"', "'"
+      quote_char = next_char
+      forward(1)
+      quoted_value(quote_char)
+    else
+      scan_until(/[^>\s\/=]+/)[0]
+    end
+  end
+  
+  def next_end_tag(name)
+    scan_until(/<\/\s*li>/)
+  end
+  
+end

data/lib/stream_parser/version.rb

@@ -1,3 +1,3 @@
 module StreamParser
-  VERSION = '0.1'
+  VERSION = '0.3'
 end

data/lib/stream_parser.rb

@@ -1,5 +1,7 @@
 module StreamParser
 
+  autoload :HTML, File.expand_path('../stream_parser/html', __FILE__)
+  
   def self.included(base)
     base.extend ClassMethods
   end
@@ -24,13 +26,13 @@ module StreamParser
   # end
 
   def eos?
-    @index >= (@source.size - 1)
+    @index >= (@source.size - 0)
   end
 
   def scan_until(r)
     r = Regexp.new(Regexp.escape(r)) if r.is_a?(String)
-    index = @source.index(r, @index)
     match = @source.match(r, @index)
+    index = match&.begin(0)
     
     if match
       @match = match.to_s
@@ -39,13 +41,20 @@ module StreamParser
     else
       @match = nil
       @old_index = @index
-      @index = @source.size - 1
+      @index = @source.size
     end
     match
   end
+  
+  def gobble(r)
+    m = @source.match(r, @index)
+    if m&.begin(0) == @index
+      scan_until(r)
+    end
+  end
 
   def pre_match
-    @source[@old_index...(@index-@match.size)]
+    @source[@old_index...(@index-(@match&.size || 0))]
   end
 
   def rewind(by=1)
@@ -63,13 +72,25 @@ module StreamParser
   end
 
   def next_char
-    @source[@index+1]
+    @source[@index]
   end
   
   def prev_char
     @source[@index-1]
   end
   
+  def next?(r)
+    @source.match(r, @index)&.begin(0) == @index
+  end
+
+  def peek(n=1)
+    if n.is_a?(Regexp)
+      @source.match(n, @index)
+    else
+      @source.slice(@index, n)
+    end
+  end
+
   def next_word
     nw = @source.match(/\s*(\S+)/, @index)
     nw.nil? ? nil : nw[1]
@@ -103,4 +124,18 @@ module StreamParser
     output
   end
 
+  def quoted_value(quote_char = '"', escape_chars = ["\\"])
+    ret_value = ""
+    while scan_until(/(#{quote_char}|\Z)/)
+      if match != quote_char
+        raise Net::HTTPHeaderSyntaxError.new("Invalid Set-Cookie header format: unbalanced quotes (#{quote_char})")
+      elsif !escape_chars.include?(pre_match[-1])
+        ret_value << pre_match
+        return ret_value
+      else
+        ret_value << pre_match[0...-1] << match
+      end
+    end
+  end
+  
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: stream_parser
 version: !ruby/object:Gem::Version
-  version: '0.1'
+  version: '0.3'
 platform: ruby
 authors:
 - Jon Bracy
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2021-05-06 00:00:00.000000000 Z
+date: 2023-08-09 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake
@@ -94,6 +94,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: activesupport
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 description:
 email:
 - jonbracy@gmail.com
@@ -104,6 +118,8 @@ extra_rdoc_files:
 files:
 - README.md
 - lib/stream_parser.rb
+- lib/stream_parser/html.rb
+- lib/stream_parser/html/tag.rb
 - lib/stream_parser/version.rb
 homepage: https://github.com/malomalo/stream_parser
 licenses:
@@ -126,7 +142,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.2.3
+rubygems_version: 3.4.13
 signing_key:
 specification_version: 4
 summary: SAX/Stream style parse helpers