Lazy SAX Machine
Description
Tired of writing parsers for XML files? Now you can parse XML into Ruby objects with ease! SAXMachine provides a simple declarative language for describing how to convert XML into Ruby objects.
This library can parse gigantic XML files without blowing up your computer. (Unlike the original SAXMachine gem.)
As long as you declare the root object as "lazy", the parser will parse the XML file as a stream and return an enumerable, which allows you to pluck one object out of the XML file at a time, storing as little of the XML file in memory as possible.
Installation
Add these lines to your application's Gemfile:
gem 'lazy-sax-machine'
gem 'nokogiri'
And then execute:
$ bundle
Usage
SAX Machine currently only supports nokogiri
, but eventually it will try to find the best XML library automatically. (First looking for ox
, then oga
, then nokogiri
.)
Examples
Lazy parsing/Enumerator - Constant memory usage
The Nokogiri Sax parser already operates in constant memory with respect to the file contents. However, Sax Machine fills up memory as it creates your Ruby objects. Using the laziness option will parse top-level elements one at a time- allowing for constant memory usage with respect to the Ruby objects. The lazy option only allows one set of elements to be declared as lazy.
class Atom
include SAXMachine
element :title
elements :entry, :lazy => true, :as => :entries, :class => AtomEntry
end
feed = Atom.parse(xml_file_handle, :lazy => true)
feed.entries # => #<Enumerator: #<Enumerator::Generator:0x00000004c41ea0>:each>
feed.entries.each do |entry|
# every time the block is called the next entry is parsed- no memory blow up!
#
# This is probably where you save the entry to a database
end
Warning !!!
- If you use the lazy option twice, or if the class using this laziness is referenced from multiple other classes, the behavior will be incorrect.
- Does not work with Ruby version 1.8. Uses ruby fibers and Enumerator capabilities, both of which were added in Ruby version 1.9.
Caution
- The overall speed with laziness may be slower if you can already fit everything into memory.
- Once you traverse an enumerator (with #each), it will be empty.
Eager Parsing
Include SAXMachine
in any class and define properties to parse:
class AtomContent
include SAXMachine
attribute :type
value :text
end
class AtomEntry
include SAXMachine
element :title
def title=(val)
@title = val.titlecase
end
# The :as argument makes this available through entry.author instead of .name
element :name, as: :author
element "feedburner:origLink", as: :url
# The :default argument specifies default value for element when it's missing
element :summary, class: String, default: "No summary available"
element :content, class: AtomContent
element :published
ancestor :ancestor
end
class Atom
include SAXMachine
# Use block to modify the returned value
# Blocks are working with pretty much everything,
# except for `elements` with `class` attribute
element :title do |title|
title.strip
end
# The :with argument means that you only match a link tag
# that has an attribute of type: "text/html"
element :link, value: :href, as: :url, with: {
type: "text/html"
}
# The :value argument means that instead of setting the value
# to the text between the tag, it sets it to the attribute value of :href
element :link, value: :href, as: :feed_url, with: {
type: "application/atom+xml"
}
elements :entry, as: :entries, class: AtomEntry
end
Then parse any XML with your class:
feed = Atom.parse(xml_text)
feed.title # Whatever the title of the blog is
feed.url # The main URL of the blog
feed.feed_url # The URL of the blog feed
feed.entries.first.title # Title of the first entry
feed.entries.first.author # The author of the first entry
feed.entries.first.url # Permalink on the blog for this entry
feed.entries.first.summary # Returns "No summary available" if summary is missing
feed.entries.first.ancestor # The Atom ancestor
feed.entries.first.content # Instance of AtomContent
feed.entries.first.content.text # Entry content text
You can also use the elements method without specifying a class:
class ServiceResponse
include SAXMachine
elements :message, as: :messages
end
response = ServiceResponse.parse("
<response>
<message>hi</message>
<message>world</message>
</response>
")
response.messages.first # hi
response.messages.last # world
To limit conflicts in the class used for mappping, you can use the alternate
SAXMachine.configure
syntax:
class X < ActiveRecord::Base
# This way no element, elements or ancestor method will be added to X
SAXMachine.configure(X) do |c|
c.element :title
end
end
Multiple elements can be mapped to the same alias:
class RSSEntry
include SAXMachine
# ...
element :pubDate, as: :published
element :pubdate, as: :published
element :"dc:date", as: :published
element :"dc:Date", as: :published
element :"dcterms:created", as: :published
end
If more than one of these elements exists in the source, the value from the last one is used. The order of
the element
declarations in the code is unimportant. The order they are encountered while parsing the
document determines the value assigned to the alias.
If an element is defined in the source but is blank (e.g., <pubDate></pubDate>
), it is ignored, and non-empty one is picked.
Contributing
- Fork it
- Create your feature branch (
git checkout -b my-new-feature
) - Commit your changes (
git commit -am 'Add some feature'
) - Push to the branch (
git push origin my-new-feature
) - Create new Pull Request
LICENSE
The MIT License
Copyright (c) 2009-2015:
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the 'Software'), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.