I have a server which parses through a text file and creates a XML page.
The XML creation can involve a lot of loops (up to ~8000).
I am confused regarding the asynchronous nature of NodeJs.
Would the below function benefit from a non-blocking call? Or does the
createServer function maintain its
Hi!
Ummm... makeXmlFromString, beginning at line 83, could directly return the
value xmlFile instead of calling a callback. No I/O in its body, so you
don't leverage the power of node.js to manage async I/O
Maybe, at line 61, when req.on('data') you could start to create the
xml, incrementall
So, how do I get the function to be asynchronous if I added other
functionality? Is there a module to import? I just do not get the threading
aspect of it all I guess.
On Sun, Jun 16, 2013 at 4:37 PM, Angel Java Lopez wrote:
> Hi!
>
> Ummm... makeXmlFromString, beginning at line 83, could direct
> So, how do I get the function to be asynchronous
You can't. You could put in setTimeouts with zero delay if you are worried
about giving other tasks a chance of running more often.
> I just do not get the threading aspect of it all
It is actually simple, there is not threading. You can launc
Do you have any numbers to proof that you need to optimise here? How much
time does it usually take for makeXmlFromString to complete? Could you post
first lines of processed v8.log after running with --profile?
If you are sure that this function does block io loop for some significant
amount o
Wow, you have many solutions to that one. I was under the impression that a
consistent asynchronous patterned was expected in NodeJs programming. I
think I understand how it works now. It just serves users with data when it
is time/ready to. I was thinking the wrong way about the whole thing. I did
"consistent asynchronous pattern" is expected for IO-bound tasks.
CPU-bound solutions are less consistent as 1) it's a little bit
outside of node.js scope 2) there are more solutions 3) it depend a
lot on domain area (not all people crunch fibonacci numbers) and
hardware ( number of CPUs/GPU/vector