On the $& fix, all I can say is *thank you*. I had investigated this previously, but I was completely misreading the warning in NYTProf as complaining that line 32 itself was doing this, which I couldn't figure out at all. You actually found the offending lines and even offered a fix! After making your change, the warning does indeed go away. From the docs this seems like a safe change to make.

Sadly, this seems to have a very small effect in this particular environment / workload. A 38-second run was virtually unaffected... +/- one second. On a 15-minute repo, it fell by about 4-5 seconds. There are bigger repos that might show a bigger benefit, but I suspect we'll only shave at most a few minutes off over a full day's work. Still, every little bit helps and fixing the warning is nice in and of itself. :)

I'm working on some parallelization for this, which should help. I'm slightly concerned this might overload the system it runs on, but that can be fixed with the proper application of a wallet.

For example, we have a set of repos that have this script run on them every 4 hours. With a very rudimentary parallelization algorithm in the calling script, the runtime fell from about 1h50m to about 1h15m. This was just moving from serial execution to 2 parallel jobs... with some inefficiencies due to very simple shell backgrounding and "wait" semantics. A better implementation (say, using GNU parallel) may get it down to an hour flat.

At the same time there's a much bigger set of repos that get processed daily. I think it takes over 24 hours to run (separate issue: it stopped reporting its status regularly). This runs concurrently, and obviously can overlap the 4-hour jobs. If I parallelize this daily job as well (2 processes), I run the risk of simultaneously maxxing out all 4 cores in this particular system... which would (possibly) be bad, because it also runs the web app that these scripts populate the data for. So I kinda need to be careful how I do this.

Still, it's clearly a very good approach to reducing wall-clock time. I was primarily hoping someone would spot something egregiously wrong with the regex's that I could fix, but that seems not to be the case. Oh well... guess we'll do it the hard way. :)

In reply to Re^2: tight loop regex optimization by superawesome
in thread tight loop regex optimization by superawesome

Title:
Use:  <p> text here (a paragraph) </p>
and:  <code> code here </code>
to format your post, it's "PerlMonks-approved HTML":


  • Posts are HTML formatted. Put <p> </p> tags around your paragraphs. Put <code> </code> tags around your code and data!
  • Titles consisting of a single word are discouraged, and in most cases are disallowed outright.
  • Read Where should I post X? if you're not absolutely sure you're posting in the right place.
  • Please read these before you post! —
  • Posts may use any of the Perl Monks Approved HTML tags:
    a, abbr, b, big, blockquote, br, caption, center, col, colgroup, dd, del, details, div, dl, dt, em, font, h1, h2, h3, h4, h5, h6, hr, i, ins, li, ol, p, pre, readmore, small, span, spoiler, strike, strong, sub, summary, sup, table, tbody, td, tfoot, th, thead, tr, tt, u, ul, wbr
  • You may need to use entities for some characters, as follows. (Exception: Within code tags, you can put the characters literally.)
            For:     Use:
    & &amp;
    < &lt;
    > &gt;
    [ &#91;
    ] &#93;
  • Link using PerlMonks shortcuts! What shortcuts can I use for linking?
  • See Writeup Formatting Tips and other pages linked from there for more info.