palkovic@linac.fnal.gov (John Palkovic) (03/01/91)
I came up with a shell script to remove new news articles with old
Date: headers. Neil Rickert made some useful modifications and this is
the result. It needs header.c, so that is included. I got header.c
from Chip Salzenberg's deliver package. You will have to compile and
link header and place it in the PATH in order for the script to work.
-John
#!/bin/sh
# This is a shell archive (shar 3.46)
# made 02/28/1991 23:22 UTC by palkovic@linac
# Source directory /home/palkovic/news
#
# existing files will NOT be overwritten unless -c is specified
#
# This shar contains:
# length mode name
# ------ ---------- ------------------------------------------
# 502 -rwxr-xr-x old.rm
# 4796 -rw-r--r-- header.c
#
# ============= old.rm ==============
if test -f 'old.rm' -a X"$1" != X"-c"; then
echo 'x - skipping old.rm (File already exists)'
else
echo 'x - extracting old.rm (Text)'
sed 's/^X//' << 'SHAR_EOF' > 'old.rm' &&
#!/bin/sh
#
# Old.rm. Trash new news articles with old Date: headers.
# J. Palkovic 2/28/91
#
# You may need to change this next line:
PATH=/usr/lib/newsbin:/usr/local/bin:/usr/bin
TODAY="`date`"
TODAY=`getdate "$TODAY"`
LIMIT=`expr $TODAY - 1209600`
cd /usr/spool/news
find `ls |egrep -v '\.'` -mtime -7 -name '[0-9]*' -type f -print |(
X while read f
X do
X DATE=`header -f Date $f`
X THEN=`getdate "$DATE" 2>/dev/null`
X if test "$THEN"
X then
X test $THEN -lt $LIMIT && rm -f $f
X fi
X done
)
SHAR_EOF
chmod 0755 old.rm ||
echo 'restore of old.rm failed'
Wc_c="`wc -c < 'old.rm'`"
test 502 -eq "$Wc_c" ||
echo 'old.rm: original size 502, current size' "$Wc_c"
fi
# ============= header.c ==============
if test -f 'header.c' -a X"$1" != X"-c"; then
echo 'x - skipping header.c (File already exists)'
else
echo 'x - extracting header.c (Text)'
sed 's/^X//' << 'SHAR_EOF' > 'header.c' &&
/* $Header: header.c,v 2.2 89/06/09 13:08:07 network Exp $
X *
X * A program to parse RFC 822 mail/news headers.
X *
X * usage: header [-c] [-n] [-f field] ... files
X *
X * Default action is to print entire header. If one or more -f options
X * are given, only the specified fields are printed. The field names are
X * not printed unless -n is specified. Field name comparisons are case
X * insensitive unless -c is specified.
X *
X * Output lines are preceeded by the filename if more than one file is
X * specified.
X *
X * This program is intended for use in delivery files, to extract multi-
X * line header fields.
X *
X * $Log: header.c,v $
X * Revision 2.2 89/06/09 13:08:07 network
X * Adapt to BSD quirks.
X *
X * Revision 2.1 89/06/09 12:25:29 network
X * Update RCS revisions.
X *
X * Revision 1.5 89/06/09 12:23:51 network
X * Baseline for 2.0 release.
X *
X */
X
#include <stdio.h>
#include <ctype.h>
X
/*
X * Manifest constants
X */
X
#define TRUE 1
#define FALSE 0
X
/*
X * Other useful macros.
X */
X
#define GETSIZE(buf) (sizeof(buf) - 1)
X
#define ISFROM(p) ((p)[0] == 'F' && (p)[1] == 'r' && (p)[2] == 'o' \
X && (p)[3] == 'm' && (p)[4] == ' ')
X
/*
X * External data.
X */
X
/* Variables set by getopt() [blech] */
X
extern int optind, opterr;
extern char *optarg;
X
/*
X * Library functions.
X */
X
extern char *malloc();
extern char *realloc();
extern void free();
X
/*
X * Global data
X */
X
int field_count = 0;
int field_alloc = 0;
char **field_names = NULL;
X
int nocasematch = TRUE; /* ignore case in header matches */
int printnames = FALSE; /* print field names with data */
X
/*----------------------------------------------------------------------
X * The Program.
X */
X
main(argc, argv)
int argc;
char **argv;
{
X int c, errors;
X
X field_alloc = 8;
X field_names = (char **) malloc(field_alloc * sizeof(char **));
X if (field_names == NULL)
X nomem();
X
X errors = FALSE;
X while ((c = getopt(argc, argv, "cnf:")) != EOF)
X {
X switch (c)
X {
X case 'c':
X nocasematch = FALSE;
X break;
X case 'n':
X printnames = TRUE;
X break;
X case 'f':
X if (field_count >= field_alloc)
X {
X field_alloc *= 2;
X field_names =
X (char **) realloc((char *)field_names,
X field_alloc * sizeof(char **));
X if (field_names == NULL)
X nomem();
X }
X field_names[field_count++] = optarg;
X break;
X default:
X errors = TRUE;
X break;
X }
X }
X
X if (errors)
X usage();
X
X if (optind == argc)
X header(stdin, (char *)NULL);
X else
X {
X FILE *fp;
X int a, filenames;
X
X filenames = ((argc - optind) > 1);
X for (a = optind; a < argc; ++a)
X {
X if ((fp = fopen(argv[a], "r")) == NULL)
X {
X errors = TRUE;
X perror(argv[a]);
X continue;
X }
X
X header(fp, (filenames ? argv[a] : (char *)NULL));
X fclose(fp);
X }
X }
X
X exit(errors ? 1 : 0);
}
X
usage()
{
X fprintf(stderr, "usage: header [-c] [-f fieldname] ... files\n");
X exit(1);
}
X
nomem()
{
X fprintf(stderr, "header: out of memory\n");
X exit(1);
}
X
header(fp, filename)
FILE *fp;
char *filename;
{
X char buf[1024];
X
X if (fgets(buf, GETSIZE(buf), fp) == NULL)
X return;
X
X /* Ignore From_ line(s). */
X
X while (ISFROM(buf) || buf[0] == '>')
X {
X if (fgets(buf, GETSIZE(buf), fp) == NULL)
X return;
X }
X
X while (buf[0] != '\n')
X {
X char *p;
X int print_this;
X
X p = buf;
X while (isupper(*p) || islower(*p) || isdigit(*p) || *p == '-')
X ++p;
X if (p == buf || *p != ':')
X break;
X print_this = field(buf, p - buf);
X if (print_this)
X {
X if (filename)
X {
X fputs(filename, stdout);
X putc(':', stdout);
X }
X ++p;
X if (*p == ' ' || *p == '\t')
X ++p;
X if (field_count == 0 || printnames)
X fputs(buf, stdout);
X else
X fputs(p, stdout);
X }
X
X /* get the next input line */
X if (fgets(buf, GETSIZE(buf), fp) == NULL)
X break;
X
X /* deal with continuation lines */
X while (buf[0] == ' ' || buf[0] == '\t')
X {
X if (print_this)
X {
X if (filename)
X {
X fputs(filename, stdout);
X putc(':', stdout);
X }
X fputs(buf, stdout);
X }
X
X if (fgets(buf, GETSIZE(buf), fp) == NULL)
X {
X buf[0] = '\n';
X break;
X }
X }
X }
}
X
int
field(s, n)
char *s;
int n;
{
X int i;
X
X if (field_count == 0)
X return TRUE;
X
X for (i = 0; i < field_count; ++i)
X {
X char *f = field_names[i];
X
X if (strlen(f) == n)
X {
X if (nocasematch)
X {
X if (ci_strncmp(f, s, n) == 0)
X return TRUE;
X }
X else
X {
X if (strncmp(f, s, n) == 0)
X return TRUE;
X }
X }
X }
X
X return FALSE;
}
X
int
ci_strncmp(s, t, n)
char *s, *t;
int n;
{
X char c, d;
X
X while (n-- > 0)
X {
X c = *s++;
X d = *t++;
X if ((c == 0) && (d == 0))
X break;
X if (isupper(c))
X c = tolower(c);
X if (isupper(d))
X d = tolower(d);
X if (c > d)
X return 1;
X if (c < d)
X return -1;
X }
X
X return 0;
}
SHAR_EOF
chmod 0644 header.c ||
echo 'restore of header.c failed'
Wc_c="`wc -c < 'header.c'`"
test 4796 -eq "$Wc_c" ||
echo 'header.c: original size 4796, current size' "$Wc_c"
fi
exit 0
--
palkovic@linac.fnal.gov || {royko,tellab5,simon}!linac!palkovicsow@cad.luth.se (Sven-Ove Westberg) (03/02/91)
In article <9Y`#-6-@linac.fnal.gov> palkovic@linac.fnal.gov (John Palkovic) writes: | |I came up with a shell script to remove new news articles with old |Date: headers. Neil Rickert made some useful modifications and this is |the result. It needs header.c, so that is included. I got header.c |from Chip Salzenberg's deliver package. You will have to compile and |link header and place it in the PATH in order for the script to work. Hmmmmm.... This is from the manual page for expire. -p causes expire to use the date the article was posted, rather than the date it arrived at your machine, as the basis for expiration. Every now and then there is a time warp that causes a batch of very very old news to be dumped onto the network; judicious use of the -p option can eradicate it. Sven-Ove