The Internaut - Latest Comments

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Ben — Mon, 20 Apr 2009 14:42:32 -0000

sorry, that should be

aquarium.inhabitants << bob
aquarium.inhabitants << jaws

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Ben — Mon, 20 Apr 2009 14:40:13 -0000

Firstly, great library. Very handy. Has anyone tried to do something that would encapsulate an array of arbitrary objects ? Like this:

class Fish
xml_accessor :name, :text
end

class Shark < Fish
xml_accessor :teeth, :text
end

class Goldfish < Fish
xml_accessor :friendly, :text
end

class Fishtank
xml_accessor :inhabitants, [Fish]
end

bob = Goldfish.new
bob.friendly = true
bob.name = "bob"

jaws = Shark.new
jaws.name = "jaws"
jaws.teeth = "sharp"

aquarium = Fishtank.new
aquarium << bob
aquarium << jaws

the_xml = aquarium.to_xml.to_s

you should get:

"<fishtank><inhabitant><teeth>sharp</teeth><name>Jaws</name></inhabitant><inhabitant><friendly>true</friendly><name>Bob</name></inhabitant></fishtank>"

Which you notice contains all the user defined fields of all the objects. My expectation would be that since I declared the object array as type Fish, it should only contain Fish methods. How do I now get my array of different types out of this xml string now ? Simply using from_xml results in only Fish being returned. Using Object results in errors.

Any ideas ?

Re: Amtrak Cascades: Families & Discounts

Empact — Sun, 01 Mar 2009 12:04:31 -0000

Yes, if you read again the rules:

Child 2 - 15 50% Up to two children per paying adult. Children must travel with adult. 1, 2
Infant Under 2 Free One infant per paying adult. Infants must ride on adult’s lap.

It says that your 2 children and 1 infant could travel even with a single one of you at discouted rates, while a larger family, 4 children and 2 infants could travel at the discounted rate with both parents present.

Re: Amtrak Cascades: Families & Discounts

Colon Health — Sun, 01 Mar 2009 07:50:37 -0000

We are a couple with 2 children between 8 and 13 years and one infant - 1yr.
So this makes it 2 adults , 2 children and 1 infant.
Based on what you have written it can be 1 children/infant per adult but here we are 2 adults in 3 children/infant, how does this work out, do we get any consideration?

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Empact — Thu, 29 Jan 2009 01:32:24 -0000

So today I happened to run into my own problem with the extensions dependency, so I went ahead and removed it from the just-released ROXML 2.4.1. Hope this helps!

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Empact — Sat, 24 Jan 2009 06:40:28 -0000

Thanks Andy, the xml_name inheritance issue should be fixed in 2.4.0, thanks to this commit: http://github.com/Empact/ro.... You can install it from rubyforge or github.

As for extensions... I'd need more details to know what problem exactly it has with activescaffold, which I haven't used. Feel free to fork the code on github and I'll pick up whatever changes you make.

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Andy Nahapetian — Thu, 08 Jan 2009 07:41:26 -0000

ROXML is great, but i have a question. How can i define a superclass and subclass it to extend its xml attributes? The following does not work:

class MySuper
include ROXML
xml_name :sitt
end

class MySub < MySuper
xml_accessor :user_id
end

the to_xml method called on an instance of MySub will not have the xml name 'sitt' as defined in the superclass.
This would be a really useful feature to DRY up the code.

The second issue that i have are the dependencies on the extensions library, which causes problems when using activescaffold

Re: Don’t just sit there

austinindian — Wed, 12 Nov 2008 11:57:52 -0000

And to think we could of had such a piece of art sitting on MLK Jr. across from the Texas History Museum. Absolute shame.

One of their latest efforts:
http://www.56leonardtribeca...

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Empact — Fri, 24 Oct 2008 05:33:42 -0000

They're equivalent. The former is used internally when we translate the the arguments into something more manageable, and can be used without problem. The latter is the preferred syntax, and I think the only documented one.

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

James — Fri, 24 Oct 2008 03:21:06 -0000

Ben,

Are you able to explain the difference between these 2 attribute declarations? They seem to do more or less the same thing in my testing, but I'm probably missing something.

xml_accessor :contributors, Contributor, :as => :array
xml_accessor :contributors, [Contributor]

Re: Diamonds aren’t Forever

Empact — Mon, 13 Oct 2008 17:26:56 -0000

You are wrong. De Beers in 1980 owned 80% of world diamond production. Now in 2008 De Beers owns only 40% of diamond production. The Russians, Canadians, Australians and numerous other smaller countries have taken over. De Beers does take a pivotal role

Sorry, but without any citation, I'll take the word of Smithsonian magazine over that of a random commentator, who presumably has a vested interest in undermining such narratives.

More rubbish... most major miners pay special attention to the environment. It's part of the terms and conditions when mining leases are granted.

Whether or not they "pay special attention," there are unavoidable environmental impacts to mining. If the mining is unnecessary and can be avoided, why put those costs on society?

Its not diamonds that hurt people. It's corrupt governments.

Again, the fact of the matter is, much of diamond mining occurs in countries with corrupt governments. Whether you like it or not, the diamond mining funds regimes and supports conflict.

Let you in ona little secret.... nobody cares anyway. They still want to wear real diamonds

What people want to wear is primarily a product of social conditioning. For example, compare the clothing styles of one era over another, or between different cultures. Gems have been valued for their beauty and scarcity, but when that scarcity is taken away, and the beauty can be had without the aforementioned social costs, society will discourage people from wearing actual earth diamonds just as it now discourages them from driving SUVs.

Re: Diamonds aren’t Forever

Daniel Katz — Mon, 13 Oct 2008 08:36:55 -0000

In reply to : " diamond production is owned by a single cartel (De Beers) which constrains the output to prop up prices and create the appearance of scarcity "

You are wrong. De Beers in 1980 owned 80% of world diamond production. Now in 2008 De Beers owns only 40% of diamond production. The Russians, Canadians, Australians and numerous other smaller countries have taken over. De Beers does take a pivotal role

In reply to : " Diamonds will soon be easily producible in a lab, which will be excellent for industry " Nothing new. More than 50% of world production of industrial diamond is synthetically manufactured.

The fact is natural diamond is becoming rarer for gem quality material

The largest synthetic diamond manufacturer is Element 6 owned by De Beers

In reply to : "Unlike diamonds that come from the dirt, man-made diamonds aren’t socially nor environmentally destructive"

More rubbish... most major miners pay special attention to the environment. It's part of the terms and conditions when mining leases are granted.

Its not diamonds that hurt people. It's corrupt governments.
Let you in ona little secret.... nobody cares anyway. They still want to wear real diamonds

Daniel Katz
www.DiamondImports.com.au

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

James — Sun, 12 Oct 2008 09:04:22 -0000

I too worship at the alter of git, so that'll be no problem :)

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Empact — Sun, 12 Oct 2008 09:02:15 -0000

Just a reminder: if you have any trouble or feel a need to extend the library, please modify it publicly on github so I can pick up the changes. Anyway, thanks!

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

James — Sun, 12 Oct 2008 08:51:23 -0000

Hey, thanks for this overhaul.

We've been using ROXML 1.X for an internal gem for working with the (rather large) ONIX standard. I started hacking on a new branch last week that replaces ROXML with raw libxml for performance reasons, but it looks like I won't need to continue with it any more :)

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Zak Mandhro — Thu, 09 Oct 2008 19:03:47 -0000

Great work Ben! Thanks for picking up ROXML and giving it some much needed TLC.

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Empact — Tue, 30 Sep 2008 19:03:32 -0000

Update: This is not a libxml bug, and the problem is fixed in ROXML 2.1

Yeah Jimmy, unfortunately this is a LibXml bug. I made a first stab at forking their rubyforge svn into github, to fix this, but ran into a git import error, which may be fixed in a more current version of git-svn.

Until this is fixed, you can use REXML. I've added fallback to REXML in my own development branch on github, and that will be in 2.1. Rexml is used if libxml is not present or if you precede your require with the following (admittedly ugly) declaration. This may be cleaned up for final release:

module ROXML XML_PARSER = 'rexml' end require 'roxml'
Until the libxml bug is fixed, you also might consider forking ROXML (ideally from my development branch) and having it manually remove the namespace declaration in this case (if libxml is being used). That would go early in the parse function. Otherwise, I'll get around to fixing it myself soon, and include the workaround in 2.1.

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Jimmy Zimmerman — Tue, 30 Sep 2008 18:50:14 -0000

The only problem I'm running into is namespace support. It's not finding the elements and attributes if there is a namespace present in the root node.

Does this ROXML have a way of dealing with namespaces?

Re: ROXML 2.0: An open-source takeover (now with defaults, blocks, hash mapping & better syntax)

Jimmy Zimmerman — Tue, 30 Sep 2008 13:44:39 -0000

Awesome! I was just getting ready to fork the ROXML project myself to switch the XML library over to libxml-ruby. REXML was one of the reasons I wasn't going to use ROXML, despite it having the best Ruby/XML binding API currently available.

I like the changes you made to the syntax. It adds even more power and cleanliness to the API.

The only suggestion I would like to make is to change the name of the project enough to distinguish it from the old ROXML. I almost missed this because I didn't realize that it was actually a new project. I just thought the ROXML project had moved over to GitHub.

I'd like to sometime build a code generator for creating your ROXML classes based on an XSD schema (kind of like JAXB does for Java).

Anyways, this all looks really cool. I'll be sure to report any bugs I might find.

Very cool.

Re: Syntax Matters

Samuel A. Falvo II — Wed, 30 Jul 2008 19:06:46 -0000

Use the name "my" instead of "self" for Python methods, and it usually makes far more sense.

class usa_telephone_number(object):
# Partial class implementation
def _get_areacode(my):
return my._areacode

def _get_exchange(my):
return my._exchange

def _get_number(my):
return my._number

areacode = property(fget=_get_areacode, fset=None)
exchange = property(fget=_get_exchange, fset=None)
number = property(fget=_get_number, fset=None)

def fromString(my, s):
(a, e, n) = somehowParseNumberFrom(s)
my._areacode = a
my._exchange = e
my._number = n

def __equals__(my, another):
return (my._areacode == another._areacode) and ...etc...

Re: The Ex-motorist

Tatiana — Mon, 17 Mar 2008 22:02:39 -0000

Hey, you should get into cycling. If not for the mode of transportation, then for the crowd.

It's kind of fun like upwards skiing, if you know what I mean, and way more social if you fall in with the right people, of which there are plentier than in any other "subculture."

Here, if you find this one funny, then... Well, enjoy.
http://bikesnobnyc.blogspot...

Re: Syntax Matters

Heewa — Mon, 04 Feb 2008 08:46:21 -0000

Hey, Ben. I just started reading your blog, and came across this.

I only recently started oging in Python, but I like it a lot. I've tried Ruby, too, but I don't like its syntax. I agree that it's annoying to have to type self.var, or self.__class__.var, but I think it's better than Ruby (or Perl) style syntax. The @var or @@var syntax is obscure, and doesn't really tell you what's going on, unless you already know. I know it's fine if you're already a fluent Ruby coder, but it's just one step closer to the syntax disaster that is Perl.

In fact, I think according to the quote you started with, Python is the clear winner. I think the Naggum didn't mean optimizing for number of chars, but for clarity of expression. In that sense, the Python syntax is far better than Ruby's.

Re: Urban Shenanigans

Elizabeth Grigg — Tue, 03 Jul 2007 20:28:10 -0000

Let me guess, was that the 49?

Re: Creationist Revisionism

Empact — Wed, 30 May 2007 20:03:56 -0000

This is a test comment. hooray!