A Text Retrieval Package for the Unix Operating System

Liam R. E. Quin

SoftQuad Inc. (lee at sq.com)

Note: the author of this paper has moved to liamquin at interlog dot com


Appendix II - Sample lq session

This listing shows part of a session using lq, a shell-script that uses lq-text. The faq command invokes lq after setting up environment variables and options to use the news.answers database.

: sqrex!lee; faq
Using database in /usr/spool/news/faq.db...
| Type words or phrases to find, one per line, followed by a blank line.
| Use control-D to quit.  Type ? for more information.
> text retrieval
> 
Computer Science Technical Report Archive Sites == news/answers/17480
  1:v.edu> comments: research reports on text retrieval and OCR orgcode: ISRI 
Interleaf FAQ -- Frequently Asked Questions for comp.text.interleaf == news/answers/17607
  2:g with hypertext navigation and full-text retrieval. 1.2. What platform
alt.cd-rom FAQ == news/answers/17643
  3:and Mac. 56. Where can I find a full text retrieval engine for a CDROM I a
  4:======== 56. Where can I find a full text retrieval engine for a CDROM I a
  5: I am making? Here is a list of Full-Text Retrieval Engines from the CD-PU
OPEN LOOK GUI FAQ 02/04: Sun OpenWindows DeskSet Questions == news/answers/18078
  6:OOK/XView/mf-fonts FAQs;lq-text unix text retrieval who is my neighbour? 
OPEN LOOK GUI FAQ 04/04: List of programs with an OPEN LOOK UI == news/answers/18079
  7: Description: Networked, distributed text-retrieval system. OLIT-based fr
  8:OOK/XView/mf-fonts FAQs;lq-text unix text retrieval who is my neighbour? 
[comp.text.tex] Metafont: All fonts available in .mf format == news/answers/18080
  9:OOK/XView/mf-fonts FAQs;lq-text unix text retrieval who is my neighbour? 
OPEN LOOK GUI FAQ 03/04: the XView Toolkit == news/answers/18082
 10:OOK/XView/mf-fonts FAQs;lq-text unix text retrieval who is my neighbour? 
Catalog of free database systems == news/answers/18236
 11:------------------ name: Liam Quin's text retrieval package (lq-text) vers
 12: are bugs. description: lq-text is a text retrieval package. That means 
| Type words or phrases to find, one per line, followed by a blank line.
| Use control-D to quit.  Type ? for more information.
> :less 9 (brings up the 9th match in a pager)
> :help
| Commands
|
| :help
|    Shows you this message.
| :view [n,n1-n2,n...]
|    Use :view to see the text surrounding matches.
|    The number n is from the left-hand edge of the index;
|    :set maxhits explains ranges in more detail.
| :page [n,n1-n2,n...]
|    Uses less -c (which you can set
|    in the $PAGER Unix environment variable) to show the files matched.
| :less [n,n1-n2,n...]
|    The same as :page except that it starts on the line number
|    containing the match, firing up the pager separately on
|    each file.
| :index [n,n1-n2,n...]
|    This shows the indexfor the phrases you've typed.
| :files [n,n1-n2,n...]
|    This simply lists all of the files that were matched,
|    in ranked order.
| :prefix prefix-string
|    Shows all of the words in the database that begin with that prefix,
|    together with the number of times they occur.
| :grep egrep-pattern
|    Shows all of the words in the database that match that egrep pattern,
|    together with the number of times they occur (not the number of files
|    in which they appear, though).
| :set option ...
|    Type :set to see more information about setting options.
| :shell [command ...]
|    Use /home/sqrex/lee/bin/sun4os4/msh to run commands.
|
| Commands that take a list of matches (n,n1-n2,n...)
| only work when you have generated a list of matches.  If you don't give
| any arguments, you get the whole list.
|
> :set
| :set match [precise|heuristic|rough]
|	to set the phrase matching level,
|	currently heuristic matching (the default)
| :set index
|	to see a Key Word In Context (KWIC) index of the matches
| :set text
|	to browse the actual text that was matched
| :set database directory
|	to use the database in the named directory.
|	Current value: /usr/spool/news/faq.db
| :set rank [all|most|any]
|	all presents only documents containing all the phrases you typed;
|	most shows those first, and then documents with all but one of
|	the phrases, and so on.
|	any doesn't bother sorting the matches; this is the default,
|	because it's the fastest.
|	Currently: find documents containing all of the phrases
| :set tags string
|	determine whether SGML tags are shown or hidden in the KWIK index
| :set tagchar string
|	set the character used as a replacement for hidden SGML tags
| :set maxhits n|all
|	show only the first n matches (currently 200)
| :set prompt string
|	set the prompt for typing in phrases to string
> ^C
Interrupted, bye.
: sqrex!lee; 


Top