unofficial mirror of guix-devel@gnu.org 
 help / color / mirror / code / Atom feed
From: Julien Lepiller <julien@lepiller.eu>
To: Pierre Neidhardt <mail@ambrevar.xyz>,
	guix-devel@gnu.org, Ricardo Wurmus <rekado@elephly.net>
Subject: Re: File search progress: database review and question on triggers
Date: Wed, 12 Aug 2020 17:29:39 -0400	[thread overview]
Message-ID: <72845E8B-35E1-4A27-95E6-452D1D1F626B@lepiller.eu> (raw)
In-Reply-To: <87o8nfy4qu.fsf@ambrevar.xyz>

[-- Attachment #1: Type: text/plain, Size: 1880 bytes --]

Why wouldn't it help? Can't you make it a trie from basename -> complete name? If I'm looking for "libcord.so" (which is a key in the trie), I don't think I need to look for every path. I only need to follow the trie until I find a pointer to some structure that contains the data I look for (ex: a list of complete filenames).

On 2020年8月12日 16:43:37 GMT-04:00, Pierre Neidhardt <mail@ambrevar.xyz> wrote:
>Julien Lepiller <julien@lepiller.eu> writes:
>
>> Have you tried something more structured? I have some code for
>creating a binary search tree and even compressing/decompressing
>strings with huffman, as well as code to serialize all that (my
>deserialization is in Java though, so not very useful to you):
>https://framagit.org/nani-project/nani-website
>>
>> See modules/nani/trie.scm for instance.
>>
>> The idea is to have a binary search tree whose keys are filenames and
>value is a pointer in the file to a structure that holds data for this
>filerame (packages and versions of guix for instance). It's super fast
>to look it up, because you don't load the whole file, you only seek to
>the right position. It's a lot longer to build, and not so easy to
>update though.
>
>Assuming we'd have only 1 Guix generation per file, I'm not sure a trie
>would help because we _always_ need to search _all_ file paths, so in
>all cases we've got some 10,000+ entries to load in memory and loop
>over
>them.
>
>The total number of entries is the bottleneck here, both for the
>database load and the individual search queries.
>
>An obvious optimization is to memoize the database load.  This has a
>significant memory cost though.
>
>The trie could be helpful for multiple Guix generations in the same
>database, but I'm not sure it warrant the increased complexity.
>
>Thoughts?
>
>-- 
>Pierre Neidhardt
>https://ambrevar.xyz/

[-- Attachment #2: Type: text/html, Size: 2275 bytes --]

  reply	other threads:[~2020-08-12 21:30 UTC|newest]

Thread overview: 73+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-10 14:32 File search progress: database review and question on triggers Pierre Neidhardt
2020-08-11  9:43 ` Mathieu Othacehe
2020-08-11 12:35   ` Pierre Neidhardt
2020-08-15 12:48     ` Hartmut Goebel
2020-08-11 15:43 ` Ricardo Wurmus
2020-08-11 17:54   ` Pierre Neidhardt
2020-08-11 17:58     ` Pierre Neidhardt
2020-08-11 20:08       ` Ricardo Wurmus
2020-08-12 19:10         ` Pierre Neidhardt
2020-08-12 20:13           ` Julien Lepiller
2020-08-12 20:43             ` Pierre Neidhardt
2020-08-12 21:29               ` Julien Lepiller [this message]
2020-08-12 22:29                 ` Ricardo Wurmus
2020-08-13  6:55                   ` Pierre Neidhardt
2020-08-13  6:52                 ` Pierre Neidhardt
2020-08-13  9:34                   ` Ricardo Wurmus
2020-08-13 10:04                     ` Pierre Neidhardt
2020-08-15 12:47                       ` Hartmut Goebel
2020-08-15 21:20                         ` Bengt Richter
2020-08-16  8:18                           ` Hartmut Goebel
2020-08-12 20:32           ` Pierre Neidhardt
2020-08-13  0:17           ` Arun Isaac
2020-08-13  6:58             ` Pierre Neidhardt
2020-08-13  9:40             ` Pierre Neidhardt
2020-08-13 10:08               ` Pierre Neidhardt
2020-08-13 11:47               ` Ricardo Wurmus
2020-08-13 13:44                 ` Pierre Neidhardt
2020-08-13 12:20               ` Arun Isaac
2020-08-13 13:53                 ` Pierre Neidhardt
2020-08-13 15:14                   ` Arun Isaac
2020-08-13 15:36                     ` Pierre Neidhardt
2020-08-13 15:56                       ` Pierre Neidhardt
2020-08-15 19:33                         ` Arun Isaac
2020-08-24  8:29                           ` Pierre Neidhardt
2020-08-24 10:53                             ` Pierre Neidhardt
2020-09-04 19:15                               ` Arun Isaac
2020-09-05  7:48                                 ` Pierre Neidhardt
2020-09-06  9:25                                   ` Arun Isaac
2020-09-06 10:05                                     ` Pierre Neidhardt
2020-09-06 10:33                                       ` Arun Isaac
2020-08-18 14:58 ` File search progress: database review and question on triggers OFF TOPIC PRAISE Joshua Branson
2020-08-27 10:00 ` File search progress: database review and question on triggers zimoun
2020-08-27 11:15   ` Pierre Neidhardt
2020-08-27 12:56     ` zimoun
2020-08-27 13:19       ` Pierre Neidhardt
2020-09-26 14:04         ` Pierre Neidhardt
2020-09-26 14:12           ` Pierre Neidhardt
2020-10-05 12:35           ` Ludovic Courtès
2020-10-05 18:53             ` Pierre Neidhardt
2020-10-09 21:16               ` zimoun
2020-10-10  8:57                 ` Pierre Neidhardt
2020-10-10 14:58                   ` zimoun
2020-10-12 10:16                   ` Ludovic Courtès
2020-10-12 11:18                     ` Pierre Neidhardt
2020-10-13 13:48                       ` Ludovic Courtès
2020-10-13 13:59                         ` Pierre Neidhardt
2020-10-10 16:03               ` zimoun
2020-10-11 11:19                 ` Pierre Neidhardt
2020-10-11 13:02                   ` zimoun
2020-10-11 14:25                     ` Pierre Neidhardt
2020-10-11 16:05                       ` zimoun
2020-10-12 10:20               ` Ludovic Courtès
2020-10-12 11:21                 ` Pierre Neidhardt
2020-10-13 13:45                   ` Ludovic Courtès
2020-10-13 13:56                     ` Pierre Neidhardt
2020-10-13 21:22                       ` Ludovic Courtès
2020-10-14  7:50                         ` Pierre Neidhardt
2020-10-16 10:30                           ` Ludovic Courtès
2020-10-17  9:14                             ` Pierre Neidhardt
2020-10-17 19:17                               ` Pierre Neidhardt
2020-10-21  9:53                               ` Ludovic Courtès
2020-10-21  9:58                                 ` Pierre Neidhardt
2020-10-12 11:23                 ` zimoun

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: https://guix.gnu.org/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=72845E8B-35E1-4A27-95E6-452D1D1F626B@lepiller.eu \
    --to=julien@lepiller.eu \
    --cc=guix-devel@gnu.org \
    --cc=mail@ambrevar.xyz \
    --cc=rekado@elephly.net \
    /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://git.savannah.gnu.org/cgit/guix.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).