sary: a suffix array library and tools

What is sary?

sary is a suffix array library and tools. It provides fast full-text search facilities for text files on the order of 10 to 100 MB using a data structure called a suffix array. It can also search specific fields in a text file by assigning index points to those fields.

Table of Contents

• What's New
• Characteristics
• Brief Introduction to Suffix Array
• libsary Reference Manual
• Using the Included Tools
• FAQ
• Download
• TODO
• Links

What's New

• 2005-03-30: sary 1.2.0 Released!
  
  • Changed ABI.
  • Fixed some minor bugs.
• 2002-09-18: sary 1.0.4 Released!
  
  • Improve perfonrmance of display of search results.
  • Modify help messages.
• 2000-11-06: sary 0.1.0 Released!
  It's the first version.

Characteristics

• Fast full-text search for huge text files.
  With suffix array. `mmap' are used for performance.
• Flexible construction of suffix array.
  Indexer can be easily extended.
• Useful included tools.
  
  • mksary: constructs suffix array.
  • sary: performs full-text search using suffix array.
• Easy-to-understand source codes. (I hope so.)
  Written in C but OO fashion. Simple and compact.
• The minimum functionality is implemented. Maybe new functions will be added in the future.
• GLib is required.
  Reference Manual

Using the Included Tools

mksary

mksary constructs a suffix array. Try -b option to use mksary with a machine which has limited memories. Try --help option for finding out other options.

  # Creating a suffix array for HUGE-TEXT.

  % mksary HUGE-TEXT
  % ls HUGE-TEXT*
  HUGE-TEXT HUGE-TEXT.ary  # .ary file is created.

sary

sary searchs a text file with a suffix array created by mksary. -i -A -B -C -c options can be used like GNU grep. Try --help option for finding out other options.

  # Searching HUGE-TEXT for PATTERN.

  % sary PATTERN HUGE-TEXT
  (search results follow...)

Download

sary is a free software under the terms of the GNU Lesser General Public License.

Stable

• sary-1.2.0.tar.gz
• sary-ruby-1.2.0.tar.gz
• CVS

TODO

• Implement limited approximate searchs.
• Implement limited regular expression searchs.
• Prepare more language bindings.

Links

• SUFARY
  Another suffix array library.
• McIlroy's suffix sort
  highly-developed algorithm. sary doesn't use this codes.