From: James Gray <james@...>
Date: 2008-09-19T11:47:11+09:00
Subject: [ruby-core:18717] Re: Request: File.binread (Or File.read_binary)

On Sep 18, 2008, at 6:56 PM, Eric Hodel wrote:

> On Sep 17, 2008, at 09:48 AM, Gregory Brown wrote:
>> On Wed, Sep 17, 2008 at 12:35 PM, Yukihiro Matsumoto <matz@ruby-lang.org 
>> > wrote:
>>> |class File
>>> |  def self.read_binary(file_name)
>>> |    File.open(file_name, "rb:BINARY") { |f| f.read }
>>> |  end
>>> |end
>>>
>>> Defines a exact behavior.  Does it take length or offset option, as
>>> File.read does?
>>
>> Oh, sorry for not being specific: So long as it makes sense to do so,
>> this should behave as close to how File.read does as possible, so it
>> should be able to take those options.
>
> I have a method in RubyGems to do the same (Gem.read_binary) and I  
> only use it for reading whole files (no length, offset options).

That's a good idea.  Reading by a byte count less than the full amount  
is pretty dangerous if you have any kind of m17n support.

I'm still struggling with this issue in the CSV code.  I read ahead to  
find your line endings, but when I do it can cause error.  The reason  
is that, with some encodings, I may read a partial character.  Then if  
I later hit that String with a Regexp, Ruby blows up on the malformed  
data.

James Edward Gray II