DataMuseum.dk

top - metrics - download

    Length: 4056 (0xfd8)
    Types: TextFile
    Names: »strfile.8«

└─⟦b20c6495f⟧ Bits:30007238 EUUGD18: Wien-båndet, efterår 1987
    └─⟦this⟧ »EUUGD18/General/Fortune/strfile.8« 

TextFile

.\" Copyright (c) 1980 Regents of the University of California.
.\" All rights reserved.  The Berkeley software License Agreement
.\" specifies the terms and conditions for redistribution.
.ie t \{\
.	ds lq ``
.	ds rq ''
.\}
.el \{\
.	ds lq "
.	ds rq "
.\}
.TH STRFILE 8 "31 Apr, 1987"
.UC 4
.SH NAME
strfile, unstr \- create a random access file for storing strings
.SH SYNOPSIS
.B strfile
[
.B \-
] [
.B \-svc\fIC\fP
] [
.B \-oir
]
sourcefile
[ datafile ]
.sp
.B unstr
[
.B \-o
]
datafile[.dat] [ outfile ]
.SH DESCRIPTION
.I strfile
is designed to take a file which contains a set of strings
and create a data file which contains those strings,
along with a seek pointer table
to the beginning of each.
This allows random access of the strings.
.PP
The source file contains strings
separated by lines starting with \*(lq%%\*(rq or \*(lq%\-\*(rq
Anything following these starting characters on the line
will be ignored,
so comments can be placed on these lines.
A \*(lq%%\*(rq simply separates strings;
a \*(lq%\-\*(rq separates not only strings but sections.
A file can have up to four sections,
.I i.e. ,
up to three delimiters.
This can be used in a program-defined way.
.PP
The data file,
if not specified on the command line,
is named
.IB sourcefile .out ,
and contains a header,
which describes the contents of the data file,
the seek pointers to the beginning of each string,
and the strings themselves,
terminated by null bytes.
.PP
The format of the header is:
.sp
.nf
.ta
.ta 8n +\w'unsigned long\ \ 'u +\w'str_delims[MAXDELIMS];\ 'u
# define	MAXDELIMS	3

# define	STR_RANDOM	0x1
# define	STR_ORDERED	0x2

typedef struct {
	unsigned long	str_numstr;	/* # of strings in the file */
	unsigned long	str_longlen;	/* length of longest string */
	unsigned long	str_shortlen;	/* length of shortest string */
	long	str_delims[MAXDELIMS];	/* delimiter markings */
	off_t	str_dpos[MAXDELIMS];	/* delimiter positions */
	short	str_flags;	/* bit field for flags */
} STRFILE;
.ev
.fi
.PP
The values in
.B str_delims
are the indices of the first string
which follow each \*(lq%\-\*(rq in the file.
The field
.B str_flags
fill have the bit
.SM STR_RANDOM
set if the
.B \-r
flag was specified,
or
.SM STR_ORDERED
if the
.B \-o
flag was specified.
.PP
The options are:
.TP
.B \-
Give a usage summary.
.TP
.B \-s
Run silently;
don't give the summary of data at the end.
.TP
.B \-v
Verbose (default).
.TP
.B \-o
Order the strings in alphabetical order.
The strings will be stored in the same order in the data file
as they were in the source,
but the seek pointer table will be sorted in alphabetical order
of the strings the point to.
Any
.I initial
non-alphanumeric characters are ignored.
This sets the
.SM STR_ORDERED
bit in the
.B str_flags
field of the header.
.TP
.B \-i
Ignore case when ordering.
.TP
.B \-r
Randomize the order of the seek pointers in the table.
The strings will be stored in the same order in the data file
as they were in the source,
but the seek pointer table will be randomized.
This sets the
.SM STR_RANDOM
bit in the
.B str_flags
field of the header.
.TP
.BI \-c C
Change the delimiting character from \*(lq%\*(rq to
.RI \\*(lq C \\*(rq,
making the delimiting lines start with
.RI \\*(lq CC \\*(rq
or
.RI \\*(lq C \-\\*(rq.
.PP
The purpose of
.I unstr
is to undo the work of
.I strfile .
It primarily exists as an emergency backup
in case you accidentally delete your source file,
but still have your data file around.
It reads data files and creates a corresponding output file.
If you don't want \*(lq%\*(rq as your delimiting character,
you can use the
.BI \-c C
option to change it.
.PP
.I unstr
will normally print out the strings
in the order they are in the data file.
If you give it the
.B \-o
option,
it will print them out in the header order,
which could be different if the file was
randomized or ordered when created.
Using this,
you can created sorted versions of your input file
by using
.B \-o
when you run
.IR strfile , 
and then using
.B "unstr \-o"
to dump them out in the table order.
.SH "SEE ALSO"
fortune(6)