XPATFFW

Section: User Commands (1)
Updated: November 2000
Index Return to Main Contents

NAME

xpatffw - XPAT Fast-Find Word List builder

SYNOPSIS

xpatffw [ -v ] [ -b block_size ] [ -I index_name ] [ -m memory k|m ] [ -o output_name ] [ [ -w word_wheel_prefix ] [ -r region_name ] [ -f region_names_list_file_name ] ] -D data_dictionary

DESCRIPTION

xpatffw builds a Fast-Find Word List for a Main Index of the database specified by data_dictionary. The Fast-Find Word List, combined with the Fast-Find Index, greatly increases the speed of text searching in xpat.

xpatffw builds the Fast-Find Word List using the memory that is specified. If the index cannot be built in the given amount of memory, it builds the index in pieces, making one pass over the text for each piece.

xpatffw can also be used to build a Word Wheel database at the same time as it builds the Fast-Find Word List. See the -w and -r options, below, for more details.

OPTIONS

The following options are available:

-v: Specify verbose mode. This option tells xpatffw to print progress messages while it builds the Fast-Find Word List. By default, xpatffw works quietly, only printing a message if an error occurs.
-b block_size: Specify compression block size (in bytes). Must be the same as that specified to xpatffi (refer to the xpatffi(1) man page for more detail on this parameter).
-I index_name: Specify the index name over which to build the Fast-Find index. The default is the first index in the data_dictionary.
-m Nk
-m Nm: memory size - use N kilobytes or N megabytes of physical memory for building the Fast-Find index. A larger memory specification results in faster indexing. Note that the amount of memory specified is limited by the amount of available physical memory. Also note that the amount of memory that is specified is usually the same as that specified to the xpatbld program. Refer to the xpatbld(1) man page for further details.
-o output_name: Specify file name prefix for the Fast-Find index. This may be a complete path. xpatffw produces one file, with the suffix `.ffw'. By default xpatffw takes the name for that file from the `FastFind' section of the data_dictionary, or uses the prefix of the data_dictionary file itself if that file does not contain any Fast-Find specifications.
-w word_wheel_prefix: Specify Word Wheel database generation. This option tells xpatffw to generate the Word Wheel database files for this database while it generates the Fast-Find Word List. xpatffw creates three Word Wheel files: the tagged Word Wheel text file (with a `.ww' suffix), the Word Wheel regions intermediate file (with a `.ptr' suffix), and the Word Wheel tagnames file (with a `.tag' suffix). These files are all named with the prefix, `word_wheel_prefix'.
: The -w option may be used in conjunction with one or more -r options, or with the -f option, to produce region-specific Word Wheels, in addition to the main Word Wheel. See the -r option and the -f option, below, for more details.
: After you have run xpatffw to produce the Word Wheel files, you must build a separate database on those files. A Word Wheel database is built by first running xpatbld to produce a Data Dictionary and a `.idx' file. Then, you must run multirgn, giving it the -sw option and passing it the `.ptr' file and the `.tag' file. multirgn will produce a `.rgn' file and will update the Word Wheel database's Data Dictionary. Note that if you are using dbbuild to build the main database, all the Word Wheel generation and database building operations will be performed automatically.
-r region_name: Specify a region to build a region-specific Word Wheel on. This option can only be used in conjunction with the -w option. A region-specific Word Wheel is a Word Wheel on the words that occur within members of the region, region_name. The -r option can be specified multiple times to build region-specific Word Wheels on several regions (see the examples, below). xpatffw merges all the region-specific Word Wheels into the main tagged Word Wheel (`.ww') file. The format of that file, including the region-specific Word Wheels, is detailed in the word_wheel(5) man page.
-f region_names_list_file_name: Specify a file containing a list of regions to build a region-specific Word Wheel on. The specified file must list region names only, with each region name separated by a newline. This option can only be used in conjunction with the -w option. This option may be used in conjunction with one or more -r options.

RESTRICTIONS

The Main Index must already have been built using xpatbld before xpatffw is run.

If any region-specific Word Wheels are to be built, both the region indexes and the Fast-Region indexes for those regions must have already been built before xpatffw is run.

EXAMPLES

EXAMPLE

xpatffw -m 8m -o /data/text -I second -D text.dd

This example builds a Word List on the database specified by the Data Dictionary, `text.dd' (`-D text.dd'). It uses 8 MB of memory (`-m 8m') (presumably that was the same amount given to xpatbld). It puts the output Word List file in the `/data' directory and calls it `text.ffw' (`-o /data/text'). It uses the index called `second' (`-I second') in the `text.dd' Data Dictionary (`-D text.dd').

xpatffw -v -m 45m -w dbww -r Headline -r Author -D db.dd

This example builds an Fast-Find Word List on the database specified by the Data Dictionary, `db.dd' (`-D db.dd'). It prints out progress messages as it builds the index (`-v') and it uses 45 MB of memory (`-m 45m'). The Word List is placed in a file called `db.ffw' (the default action). This example also creates a Word Wheel which it places in the files, `dbww.ww', `dbww.ptr' and `dbww.tag' (`-w dbww'). This Word Wheel database also includes region-specific Word Wheels for the `Headline' (`-r Headline') and `Author' (`-r Author') regions.

This document was created by man2html, using the manual pages.
Time: 18:03:38 GMT, March 26, 2001