Time object to String Object to XML and Back Again in Ruby - ruby

Question

I am storing Time objects in XML as strings. I am having trouble figuring out the best way to reinitialize them back. From strings to Time objects in order to perform a subtraction on them.
here is how they are stored in xml
<time>
<category> batin </category>
<in>2014-10-29 18:20:47 -0400</in>
<out>2014-10-29 18:20:55 -0400</out>
</time>
using
t = Time.now
i am accessing them from xml with
doc = Nokogiri::XML(File.open("time.xml"))
nodes = doc.xpath("//time").each do |node|
temp = TimeClock.new
temp.category = node.xpath('category').inner_text
temp.in = node.xpath('in').inner_text.
temp.out = node.xpath('out').inner_text.
#times << temp
end
what is the best way to reconvert them back to Time objects? i do not see a method of Time object that does this. I found that it is possible to convert to a Date object. but that seems to only give me a format of mm/dd/yyyy which is partly what i want.
in need to be able to subtract
<out>2014-10-29 18:20:55 -0400</out>
from
<in>2014-10-29 18:20:47 -0400</in>
the XML will at some point be stored based on dates but i also need the exact time "hh/mm/ss" as well to perform calculations.
any sugguestions?

Answer 1

The time stdlib extends the class with parsing/conversion methods.
require 'time'
Time.parse('2014-10-29 18:20:47 -0400')

Answer 2

I'd do something like:
require 'nokogiri'
require 'time'
doc = Nokogiri::XML(<<EOT)
<xml>
<time>
<category>
<in>2014-10-29 18:20:47 -0400</in>
<out>2014-10-29 18:20:55 -0400</out>
</category>
</time>
</xml>
EOT
times = doc.search('time category').map{ |category|
in_time, out_time = %w[in out].map{ |n| Time.parse(category.at(n).text) }
{
in: in_time,
out: out_time
}
}
times # => [{:in=>2014-10-29 15:20:47 -0700, :out=>2014-10-29 15:20:55 -0700}]
Both the DateTime and Time classes allow parsing of a small variety of date/time formats. Some formats can cause explosions but this one is safe. Use DateTime if the date could be before the Unix epoch.
in_time, out_time = %w[in out].map{ |n| Time.parse(category.at(n).text) }
Looking at that in IRB:
>> doc.search('time category').to_html
"<category>\n <in>2014-10-29 18:20:47 -0400</in>\n <out>2014-10-29 18:20:55 -0400</out>\n</category>"
doc.search('time category') returns a NodeSet of all <category> nodes.
>> %w[in out]
[
[0] "in",
[1] "out"
]
returns an array of strings.
Time.parse(category.at(n).text)
returns the n node under the <category> node, where n is first 'in', then 'out'.