Author Topic: New Search Engine  (Read 3766 times)

0 Members and 1 Guest are viewing this topic.

Offline rknize

  • * Administrator
  • Thread Starter
  • Posts: 1731
  • Location: Chicago
    • metaruss
New Search Engine
« on: Wed, 10 July 2013, 14:32:33 »
GeekHack is far too large for SMF's normal search indexer to cope with.  This problem has been on my back burner for a while.  We now have Sphinx up and running, which is not only much faster but should produce results with greater quality.  For example, you can use search terms with less than 3 characters.

Let me know if there are any issues.
Russ

Offline SpAmRaY

  • NOT a Moderator
  • * Certified Spammer
  • Posts: 14667
  • Location: ¯\(°_o)/¯
  • because reasons.......
Re: New Search Engine
« Reply #1 on: Wed, 10 July 2013, 14:33:32 »
Woot!! This is so awesome....now to try it out.... ;)
« Last Edit: Wed, 10 July 2013, 15:21:35 by SpAmRaY »

Offline Tym

  • [CTRL]ALT
  • * Maker
  • Posts: 1582
  • Location: England
Re: New Search Engine
« Reply #2 on: Wed, 10 July 2013, 14:38:38 »
Thank Christ, good work lads, thank you!
unless they have some unforeseeable downside (like they're actually made of cream cheese cunningly disguised as ABS)


Offline rowdy

  • HHKB Hapster
  • * Erudite Elder
  • Posts: 21175
  • Location: melbourne.vic.au
  • Missed another sale.
Re: New Search Engine
« Reply #3 on: Wed, 10 July 2013, 15:13:31 »
Assuming it is accessed from the search field in the upper right corner of the main page, I searched for "model m keycaps" (without quotes) and got this error:

Each word must be at least two characters long.
"Because keyboards are accessories to PC makers, they focus on minimizing the manufacturing costs. But that’s incorrect. It’s in HHKB’s slogan, but when America’s cowboys were in the middle of a trip and their horse died, they would leave the horse there. But even if they were in the middle of a desert, they would take their saddle with them. The horse was a consumable good, but the saddle was an interface that their bodies had gotten used to. In the same vein, PCs are consumable goods, while keyboards are important interfaces." - Eiiti Wada

NEC APC-H4100E | Ducky DK9008 Shine MX blue LED red | Ducky DK9008 Shine MX blue LED green | Link 900243-08 | CM QFR MX black | KeyCool 87 white MX reds | HHKB 2 Pro | Model M 02-Mar-1993 | Model M 29-Nov-1995 | CM Trigger (broken) | CM QFS MX green | Ducky DK9087 Shine 3 TKL Yellow Edition MX black | Lexmark SSK 21-Apr-1994 | IBM SSK 13-Oct-1987 | CODE TKL MX clear | Model M 122 01-Jun-1988

Ị̸͚̯̲́ͤ̃͑̇̑ͯ̊̂͟ͅs̞͚̩͉̝̪̲͗͊ͪ̽̚̚ ̭̦͖͕̑́͌ͬͩ͟t̷̻͔̙̑͟h̹̠̼͋ͤ͋i̤̜̣̦̱̫͈͔̞ͭ͑ͥ̌̔s̬͔͎̍̈ͥͫ̐̾ͣ̔̇͘ͅ ̩̘̼͆̐̕e̞̰͓̲̺̎͐̏ͬ̓̅̾͠͝ͅv̶̰͕̱̞̥̍ͣ̄̕e͕͙͖̬̜͓͎̤̊ͭ͐͝ṇ̰͎̱̤̟̭ͫ͌̌͢͠ͅ ̳̥̦ͮ̐ͤ̎̊ͣ͡͡n̤̜̙̺̪̒͜e̶̻̦̿ͮ̂̀c̝̘̝͖̠̖͐ͨͪ̈̐͌ͩ̀e̷̥͇̋ͦs̢̡̤ͤͤͯ͜s͈̠̉̑͘a̱͕̗͖̳̥̺ͬͦͧ͆̌̑͡r̶̟̖̈͘ỷ̮̦̩͙͔ͫ̾ͬ̔ͬͮ̌?̵̘͇͔͙ͥͪ͞ͅ

Offline alaricljs

  • I be WOT'ing all day...
  • ** Moderator Emeritus
  • Posts: 3715
  • Location: NE US
Re: New Search Engine
« Reply #4 on: Wed, 10 July 2013, 15:15:02 »
well, 2 is less than 3  :)
Filco w/ Imsto thick PBT
Ducky 1087XM PCB+Plate, w/ Matias "Quiet Click" spring-swapped w/ XM Greens

Offline rowdy

  • HHKB Hapster
  • * Erudite Elder
  • Posts: 21175
  • Location: melbourne.vic.au
  • Missed another sale.
Re: New Search Engine
« Reply #5 on: Wed, 10 July 2013, 17:19:27 »
1 is also less than three ;)
"Because keyboards are accessories to PC makers, they focus on minimizing the manufacturing costs. But that’s incorrect. It’s in HHKB’s slogan, but when America’s cowboys were in the middle of a trip and their horse died, they would leave the horse there. But even if they were in the middle of a desert, they would take their saddle with them. The horse was a consumable good, but the saddle was an interface that their bodies had gotten used to. In the same vein, PCs are consumable goods, while keyboards are important interfaces." - Eiiti Wada

NEC APC-H4100E | Ducky DK9008 Shine MX blue LED red | Ducky DK9008 Shine MX blue LED green | Link 900243-08 | CM QFR MX black | KeyCool 87 white MX reds | HHKB 2 Pro | Model M 02-Mar-1993 | Model M 29-Nov-1995 | CM Trigger (broken) | CM QFS MX green | Ducky DK9087 Shine 3 TKL Yellow Edition MX black | Lexmark SSK 21-Apr-1994 | IBM SSK 13-Oct-1987 | CODE TKL MX clear | Model M 122 01-Jun-1988

Ị̸͚̯̲́ͤ̃͑̇̑ͯ̊̂͟ͅs̞͚̩͉̝̪̲͗͊ͪ̽̚̚ ̭̦͖͕̑́͌ͬͩ͟t̷̻͔̙̑͟h̹̠̼͋ͤ͋i̤̜̣̦̱̫͈͔̞ͭ͑ͥ̌̔s̬͔͎̍̈ͥͫ̐̾ͣ̔̇͘ͅ ̩̘̼͆̐̕e̞̰͓̲̺̎͐̏ͬ̓̅̾͠͝ͅv̶̰͕̱̞̥̍ͣ̄̕e͕͙͖̬̜͓͎̤̊ͭ͐͝ṇ̰͎̱̤̟̭ͫ͌̌͢͠ͅ ̳̥̦ͮ̐ͤ̎̊ͣ͡͡n̤̜̙̺̪̒͜e̶̻̦̿ͮ̂̀c̝̘̝͖̠̖͐ͨͪ̈̐͌ͩ̀e̷̥͇̋ͦs̢̡̤ͤͤͯ͜s͈̠̉̑͘a̱͕̗͖̳̥̺ͬͦͧ͆̌̑͡r̶̟̖̈͘ỷ̮̦̩͙͔ͫ̾ͬ̔ͬͮ̌?̵̘͇͔͙ͥͪ͞ͅ

Offline CommunistWitchDr

  • Posts: 479
  • Location: St. Louis, MO
  • >implying keyboards
Re: New Search Engine
« Reply #6 on: Wed, 10 July 2013, 17:30:36 »
1 is also less than three ;)
Now searching for all posts containing phrases "a" or "e" or "i" or "o" or "u".

Offline rowdy

  • HHKB Hapster
  • * Erudite Elder
  • Posts: 21175
  • Location: melbourne.vic.au
  • Missed another sale.
Re: New Search Engine
« Reply #7 on: Wed, 10 July 2013, 17:35:28 »
Ok, so I searched for: "model m" keycaps

(quotes around "model m", no quote around keycaps, thinking that it might search for posts containing those two phrases, but ...

"Because keyboards are accessories to PC makers, they focus on minimizing the manufacturing costs. But that’s incorrect. It’s in HHKB’s slogan, but when America’s cowboys were in the middle of a trip and their horse died, they would leave the horse there. But even if they were in the middle of a desert, they would take their saddle with them. The horse was a consumable good, but the saddle was an interface that their bodies had gotten used to. In the same vein, PCs are consumable goods, while keyboards are important interfaces." - Eiiti Wada

NEC APC-H4100E | Ducky DK9008 Shine MX blue LED red | Ducky DK9008 Shine MX blue LED green | Link 900243-08 | CM QFR MX black | KeyCool 87 white MX reds | HHKB 2 Pro | Model M 02-Mar-1993 | Model M 29-Nov-1995 | CM Trigger (broken) | CM QFS MX green | Ducky DK9087 Shine 3 TKL Yellow Edition MX black | Lexmark SSK 21-Apr-1994 | IBM SSK 13-Oct-1987 | CODE TKL MX clear | Model M 122 01-Jun-1988

Ị̸͚̯̲́ͤ̃͑̇̑ͯ̊̂͟ͅs̞͚̩͉̝̪̲͗͊ͪ̽̚̚ ̭̦͖͕̑́͌ͬͩ͟t̷̻͔̙̑͟h̹̠̼͋ͤ͋i̤̜̣̦̱̫͈͔̞ͭ͑ͥ̌̔s̬͔͎̍̈ͥͫ̐̾ͣ̔̇͘ͅ ̩̘̼͆̐̕e̞̰͓̲̺̎͐̏ͬ̓̅̾͠͝ͅv̶̰͕̱̞̥̍ͣ̄̕e͕͙͖̬̜͓͎̤̊ͭ͐͝ṇ̰͎̱̤̟̭ͫ͌̌͢͠ͅ ̳̥̦ͮ̐ͤ̎̊ͣ͡͡n̤̜̙̺̪̒͜e̶̻̦̿ͮ̂̀c̝̘̝͖̠̖͐ͨͪ̈̐͌ͩ̀e̷̥͇̋ͦs̢̡̤ͤͤͯ͜s͈̠̉̑͘a̱͕̗͖̳̥̺ͬͦͧ͆̌̑͡r̶̟̖̈͘ỷ̮̦̩͙͔ͫ̾ͬ̔ͬͮ̌?̵̘͇͔͙ͥͪ͞ͅ

Offline alaricljs

  • I be WOT'ing all day...
  • ** Moderator Emeritus
  • Posts: 3715
  • Location: NE US
Re: New Search Engine
« Reply #8 on: Wed, 10 July 2013, 19:25:33 »
You tricked it!
Filco w/ Imsto thick PBT
Ducky 1087XM PCB+Plate, w/ Matias "Quiet Click" spring-swapped w/ XM Greens

Offline mkawa

  •  No Marketplace Access
  • Posts: 6562
  • (ツ)@@@. crankypants
Re: New Search Engine
« Reply #9 on: Wed, 10 July 2013, 23:44:59 »
in theory, the query "model m" AND keycaps should work.

the query language documentation is here: http://sphinxsearch.com/docs/current.html#searching

however, to get this working, rknize had to rip out the smf search engine and glue in the sphinx backend, so consider this still to be a beta feature. however, the changeover to sphinx has been planned for a long time (a year, basically), so this is not in any way experimental. report bugs to the massive bug report thread (rknize has a possible method for that madness in the works as well, by the way).

to all the brilliant friends who have left us, and all the students who climb on their shoulders.

Offline rknize

  • * Administrator
  • Thread Starter
  • Posts: 1731
  • Location: Chicago
    • metaruss
Re: New Search Engine
« Reply #10 on: Thu, 11 July 2013, 00:25:23 »
I have it setup for extended syntax:

http://sphinxsearch.com/docs/current.html#extended-syntax

It seems like the phrase operator (quotes) don't work as expected.  I'll look into this.  Perhaps SMF is stripping the quotes.
Russ

Offline Glod

  • * Elevated Elder
  • Posts: 1998
  • Location: Virginia, USA
  • Also Known As Ergonomech
    • YouTube Channel
Re: New Search Engine
« Reply #11 on: Thu, 11 July 2013, 17:30:23 »
Hmmm the syntax is similar to a combination of Concordance, dtSearch, and Lucene if anyone has worked in the legal industry like i do.

Offline hashbaz

  • Grand Ancient One
  • * Moderator Emeritus
  • Posts: 5057
  • Location: SF Bae Area
Re: New Search Engine
« Reply #12 on: Thu, 11 July 2013, 17:42:21 »
regex plz