From: "Jon Smirl" <jonsmirl@gmail.com>
To: "Johannes Schindelin" <Johannes.Schindelin@gmx.de>
Cc: "Shawn Pearce" <spearce@spearce.org>,
"John Rigby" <jcrigby@gmail.com>,
"linux@horizon.com" <linux@horizon.com>,
git@vger.kernel.org
Subject: Re: Compression and dictionaries
Date: Wed, 16 Aug 2006 10:43:30 -0400 [thread overview]
Message-ID: <9e4733910608160743k22362dc4r4fa22c791707c73c@mail.gmail.com> (raw)
In-Reply-To: <Pine.LNX.4.63.0608160908070.28360@wbgn013.biozentrum.uni-wuerzburg.de>
On 8/16/06, Johannes Schindelin <Johannes.Schindelin@gmx.de> wrote:
> Hi,
>
> On Wed, 16 Aug 2006, Shawn Pearce wrote:
>
> > Johannes Schindelin <Johannes.Schindelin@gmx.de> wrote:
> > > Hi,
> > >
> > > On Wed, 16 Aug 2006, Jon Smirl wrote:
> > >
> > > > On 8/16/06, John Rigby <jcrigby@gmail.com> wrote:
> > > > > Sorry if this is off topic, but could the dictionary be used to make
> > > > > git-grep alot faster?
> > > >
> > > > It would be almost instant.
> > >
> > > But only if you are not using a regular expression, but a single word.
> >
> > Yes and no. If the inverted index contains terms broken by some
> > known pattern (e.g. break on word-type boundaries) and the regex
> > in question has constant sections (it should, otherwise it might
> > as well just be '.') then you can reduce your search space to a
> > fraction of the overall data by looking at the inverted index to
> > select likely terms, select the related revisions containing those
> > possible terms, then run the regex only on those revisions.
> >
> > Sure you would be possibly pulling out a number of false positives
> > but if the constant sequence(s) in the regex reduce your search
> > space to below 1/2 of the overall data that's probably a lot less
> > I/O and CPU required to complete the query, even if you have to
> > read the entire dictionary and apply each term in the dictionary
> > to the regex to look for those possible matches.
>
> So it would speed up the search, but no, in case of regular expressions,
> particularly any interesting one, the result would not be instantaneous.
Instant is a relative term. Google is instant compared to running grep
over 10TB of data. How long would that take, a month?
Shawn is correct, the inverted indexes are used to eliminate as many
files as possible. So the response time is a more of a function of how
many hits you have instead of how big the data set is. Of course if
you give it a pattern that matches everything it will just as slow as
grep. Give it a pattern that is only in one file and detectable by the
index and it will be very fast. If you are going to give it a bunch of
patterns that aren't in the index, then we need to adjust how the
index is built.
--
Jon Smirl
jonsmirl@gmail.com
next prev parent reply other threads:[~2006-08-16 14:43 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-08-15 8:33 Compression and dictionaries linux
2006-08-15 13:29 ` Jon Smirl
2006-08-15 14:55 ` Jon Smirl
2006-08-16 0:37 ` linux
[not found] ` <4b73d43f0608152243i15b37036x7aa50aa3afc2b02f@mail.gmail.com>
2006-08-16 5:50 ` Jon Smirl
2006-08-16 6:33 ` Johannes Schindelin
2006-08-16 6:55 ` Shawn Pearce
2006-08-16 7:09 ` Johannes Schindelin
2006-08-16 14:43 ` Jon Smirl [this message]
2006-08-17 22:33 ` linux
-- strict thread matches above, loose matches on Subject: below --
2006-08-14 3:37 Jon Smirl
2006-08-14 3:56 ` Shawn Pearce
2006-08-14 4:07 ` Jon Smirl
2006-08-14 4:17 ` Shawn Pearce
2006-08-14 7:48 ` Alex Riesen
2006-08-14 10:06 ` Erik Mouw
2006-08-14 12:33 ` Johannes Schindelin
2006-08-14 14:08 ` Jon Smirl
2006-08-14 14:45 ` Johannes Schindelin
2006-08-14 16:15 ` Jon Smirl
2006-08-14 16:32 ` David Lang
2006-08-14 16:55 ` Jakub Narebski
2006-08-14 17:15 ` Jeff Garzik
2006-08-14 17:34 ` David Lang
2006-08-14 17:50 ` Jeff Garzik
2006-08-14 18:48 ` Jon Smirl
2006-08-14 19:08 ` David Lang
2006-08-14 19:38 ` Johannes Schindelin
2006-08-14 15:14 ` Alex Riesen
2006-08-14 15:26 ` Johannes Schindelin
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: http://vger.kernel.org/majordomo-info.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=9e4733910608160743k22362dc4r4fa22c791707c73c@mail.gmail.com \
--to=jonsmirl@gmail.com \
--cc=Johannes.Schindelin@gmx.de \
--cc=git@vger.kernel.org \
--cc=jcrigby@gmail.com \
--cc=linux@horizon.com \
--cc=spearce@spearce.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
Code repositories for project(s) associated with this public inbox
https://80x24.org/mirrors/git.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).