[lnkForumImage]
TotalShareware - Download Free Software

Confronta i prezzi di migliaia di prodotti.
Asp Forum
 Home | Login | Register | Search 


 

Forums >

comp.lang.ruby

Re: Large File Reading and Writing

Joel VanderWerf

10/13/2006 6:29:00 AM

Daniel N wrote:
> Hi,
>
> A project that I've currently got at work consists of reading large files
> and writing another larger file. Unfortunately the machine I'm on is
> pretty
> lame and it runs out of RAM pretty quick and so takes a long time.
> (This is
> currently in perl and I'm not keen to play with it in perl.)
>
> I have had a look around the docs and searched the mailing list ( a quick
> search ) for how to do buffered reading and writing but I've come up with
> nothing.
>
> I have a bunch of text files. I want to read some lines, and then write a
> line, the move on to the next line etc but I only want to have a few lines
> in memory at once. Any clues as to how to achive this?
>
> Cheers
> Daniel
>

Something like this maybe...

history = []
ARGF.each do |line|
history << line
if enough_history_to_generate_some_output
write_output
history.clear
end
end

The ARGF thingy is explained in 'ri IO':

The global constant ARGF (also accessible as $<) provides an
IO-like stream which allows access to all files mentioned on the
command line (or STDIN if no files are mentioned). ARGF provides
the methods #path and #filename to access the name of the file
currently being read.

You can of course open a file by name from your code.

--
vjoel : Joel VanderWerf : path berkeley edu : 510 665 3407

1 Answer

Joel VanderWerf

10/13/2006 7:04:00 AM

0

Daniel N wrote:
> Thanx. I'm not sure that my head is really around it though...
>
> Pls see inline
>
> On 10/13/06, Joel VanderWerf <vjoel@path.berkeley.edu> wrote:
>
>>
>> Something like this maybe...
>>
>> history = []
>> ARGF.each do |line|
>> history << line
>> if enough_history_to_generate_some_output
>> write_output
>> history.clear
>> end
>> end
>
>
> Does the output file not stay in memory? In my case the output file is
> almost a concatenation of large files so once I've written a line I don't
> really want to keep that line in memory.

Instead of the line

write_output

let's say you have something like

output_line = ...
puts output_line

Each time these lines are executed, you have a variable that refers to
the _current_ line of output, but there is no reference to the string
that was printed last time around. This means that the garbage collector
can reclaim that space if it needs to. So the whole output file need
not be kept in memory.

> The ARGF thingy is explained in 'ri IO':
>>
>> The global constant ARGF (also accessible as $<) provides an
>> IO-like stream which allows access to all files mentioned on the
>> command line (or STDIN if no files are mentioned). ARGF provides
>> the methods #path and #filename to access the name of the file
>> currently being read.
>>
>> You can of course open a file by name from your code.
>
>
>
> If I have a named file and I open it. I think it will be easier if I open
> specify the files from withing ruby by feeding it only a directory. Once I
> have my list of files if I open a file
>
> File.open( "my_file", "r" )
>
> Is there a way to buffer this input so that the entire file is not read?

Sure. If you use IO.gets (or IO.each, as above), then only one line at a
time is read.

File.open( "my_file", "r" ) do |f|
f.each do |line|
... # do something with line
end
end

--
vjoel : Joel VanderWerf : path berkeley edu : 510 665 3407