Stripping CVS keywords
[project/porterstemmer.git] / README.txt
CommitLineData
7e5dbb93 1
a4ce718f
JH
2GENERAL INFORMATION
3-------------------
7e5dbb93 4
fae24be6
JH
5This module implements the Porter-Stemmer algorithm, version 2, to improve
6English-language searching with the Drupal built-in Search module. Information
7about the algorithm can be found at
a4ce718f 8http://snowball.tartarus.org/algorithms/english/stemmer.html
7e5dbb93 9
fae24be6
JH
10Stemming reduces a word to its basic root or stem (e.g. 'blogging' to 'blog') so
11that variations on a word ('blogs', 'blogger', 'blogging', 'blog') are
12considered equivalent when searching. This generally results in more relevant
13results.
a4ce718f 14
9e4f7505
JH
15Note that a few parts of the Porter Stemmer algorithm work better for American
16English than British English, so some British spellings will not be stemmed
17correctly.
a4ce718f 18
73ea7032 19This module will use the PECL "stem" library's implementation of the Porter
fae24be6
JH
20Stemmer algorithm, if it is installed on your server. If the PECL "stem" library
21is not available, the module uses its own PHP implementation of the
73ea7032
JH
22algorithm. The output is the same in either case. More information about the
23PECL "stem" library: http://pecl.php.net/package/stem
24
a4ce718f
JH
25
26INSTALLATION
27------------
fae24be6 28
a4ce718f
JH
29See the INSTALL.txt file for installation instructions.
30
31
32TESTING
33-------
34
fae24be6
JH
35The Porter Stemmer module includes tests for the stemming algorithm and
36functionality. If you would like to run the tests, enable the core Testing
37module, and then navigate to Administer > Configuration / Development / Testing.
a4ce718f 38
fae24be6
JH
39Each "Stemming output" test for the Porter Stemmer module includes approximately
402000 individual word stemming tests (which test the module against a standard
41word list downloaded from the site above). Due to the way output is displayed
42in SimpleTest, you may run into browser timeout or memory issues if you try to
43run all 16 of the "Stemming output" tests during the same test run.
73ea7032
JH
44
45Tests are provided both for the internal algorithm and the PECL library.
fae24be6
JH
46
47There are also functional tests and tests for some of the internal steps of the
48stemming algorithm.