tools/fix_date.sh

   1 #!/bin/sh
   2
   3 #  * Copyright 2004 Tristan Chabredier <wwp@claws-mail.org>
   4 #  *
   5 #  * This file is free software; you can redistribute it and/or modify it
   6 #  * under the terms of the GNU General Public License as published by
   7 #  * the Free Software Foundation; either version 3 of the License, or
   8 #  * (at your option) any later version.
   9 #  *
  10 #  * This program is distributed in the hope that it will be useful, but
  11 #  * WITHOUT ANY WARRANTY; without even the implied warranty of
  12 #  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  13 #  * General Public License for more details.
  14 #  *
  15 #  * You should have received a copy of the GNU General Public License
  16 #  * along with this program; if not, write to the Free Software
  17 #  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18 #
  19 # fix_date.sh           helper script to fix non-standard date or add missing
  20 #                                       date header to emails
  21
  22 # usage: fix_date.sh <filename> [<filename> ..]
  23 # It will replace the Date: value w/ the one picked up from more recent
  24 # Fetchinfo time header, Received: field.. Otherwise, it will take the file
  25 #  modification time (using a RFC 2822-compliant form).
  26 # Any already existing X-Original-Date is kept, if missing we're adding it
  27 # if the Date: was set (even if set w/ non conform value)
  28
  29 # TODO: fallback to X-OriginalArrivalTime: ?
  30
  31 VERSION="0.1.3"
  32
  33
  34 version()
  35 {
  36         echo "$VERSION"
  37         exit 0
  38 }
  39
  40 usage()
  41 {
  42         echo "usage:"
  43         echo "  ${0##*/} [<switches>] <filename> [<filename> ..]"
  44         echo "switches:"
  45         echo "  -h --help     display this help then exit"
  46         echo "  -v --version  display version information then exit"
  47         echo "  -d --debug    turn on debug information (be more verbose)"
  48         echo "  -f --force    always force (re-)writing of Date: header"
  49         echo "  -r --rfc      force re-writing of Date: header when it's not RFC-compliant"
  50         echo "  -s --strict   use RFC-strict matching patterns for dates"
  51         echo "  --            end of switches (in case a filename starts with a -)"
  52         exit $1
  53 }
  54
  55 date_valid()
  56 {
  57         test $STRICT -eq 1 && \
  58                 REGEXP="$DATE_REGEXP_STRICT" || \
  59                 REGEXP="$DATE_REGEXP"
  60
  61         echo "$1" | grep -qEim 1 "$REGEXP"
  62         DATE_VALID=$?
  63 }
  64
  65 dump_date_fields()
  66 {
  67         test -z "$X_ORIGINAL_DATE" -a -n "$DATE" && \
  68                 echo "X-Original-Date:$DATE" >> "$TMP"
  69         echo "Date:$REPLACEMENT_DATE" >> "$TMP"
  70 }
  71
  72
  73 # use --force to always (re-)write the Date header
  74 # otherwise, the Date header will be written if only it doesn't exist
  75 FORCE=0
  76 # use --rfc to (re-)write the Date header when it's not RFC-compliant
  77 # otherwise, the Date header will be written if only it doesn't exist
  78 RFC=0
  79 # use --debug to display more information about what's performed
  80 DEBUG=0
  81 # use --strict to use strict matching patterns for date validation
  82 STRICT=0
  83 # 0 = valid, always valid until --strict is used, then date_valid overrides this value
  84 DATE_VALID=0
  85 # max header lines (300 is a reasonable minimum value but 600 has already been encountered, set to 1000 by security)
  86 MAX_HEADER_LINES=1000
  87
  88 while [ -n "$1" ]
  89 do
  90         case "$1" in
  91         -h|--help)              usage 0;;
  92         -v|--version)   version;;
  93         -f|--force)             FORCE=1;;
  94         -d|--debug)             DEBUG=1;;
  95         -r|--rfc)               RFC=1;;
  96         -s|--strict)    STRICT=1;;
  97         --)                             shift
  98                                         break;;
  99         -*)                             echo "error: unrecognized switch '$1'"
 100                                         usage 1;;
 101         *)                              break;;
 102         esac
 103         shift
 104 done
 105
 106 if [ $FORCE -eq 1 -a $RFC -eq 1 ]
 107 then
 108         echo "error: use either --force or --rfc, but not both at the same time"
 109         usage 1
 110 fi
 111
 112 test $# -lt 1 && \
 113         usage 1
 114
 115 TMP="/tmp/${0##*/}.$$.tmp"
 116 HEADERS="/tmp/${0##*/}.$$.headers.tmp"
 117 BODY="/tmp/${0##*/}.$$.body.tmp"
 118
 119 DATE_REGEXP='( (Mon|Tue|Wed|Thu|Fri|Sat|Sun),)? [0-9]+ (Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec) [0-9]+ [0-9]+:[0-9]+:[0-9]+ [+-]?[0-9]+'
 120 DATE_REGEXP_STRICT='(Mon|Tue|Wed|Thu|Fri|Sat|Sun), [0-9]+ (Jan|Feb|Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov|Dec) [0-9]+ [0-9]+:[0-9]+:[0-9]+ [+-]?[0-9]+'
 121
 122 while [ -n "$1" ]
 123 do
 124         # skip if file is empty or doesn't exist
 125         if [ ! -s "$1" ]
 126         then
 127                 test $DEBUG -eq 1 && \
 128                         echo "$1: no found or empty, skipping"
 129                 shift
 130                 continue
 131         fi
 132         SKIP=0
 133
 134         # split headers and body
 135         # find the empty line that separates body (if any) from headers,
 136         # work on a temporary dos2unix'ed copy because body might
 137         # contain DOS CRLF and grep '^$' won't work
 138         head -$MAX_HEADER_LINES "$1" | dos2unix > "$TMP"
 139         SEP=`grep -nEm1 "^$" "$TMP" 2>/dev/null | cut -d ':' -f 1`
 140         rm -f "$TMP"
 141         if [ -z "$SEP" -o "$SEP" = "0" -o $? -ne 0 ]
 142         then
 143                 cp -f "$1" "$HEADERS"
 144                 :> "$BODY"
 145                 test $DEBUG -eq 1 && \
 146                         echo "$1: no body part could be found before line $MAX_HEADER_LINES"
 147         else
 148                 sed -n '1,'`expr $SEP - 1`'p' "$1" > "$HEADERS"
 149                 sed '1,'`expr $SEP - 1`'d' "$1" > "$BODY"
 150         fi
 151
 152         # work on headers only
 153
 154         # get the Date and X-Original-Date
 155         X_ORIGINAL_DATE=`sed -n '/^X-Original-Date:/,/^[^\t]/p' "$HEADERS" | head -n -1 | cut -d ':' -f 2-`
 156         DATE=`sed -n '/^Date:/,/^[^\t]/p' "$HEADERS" | head -n -1 | cut -d ':' -f 2-`
 157
 158         # work on headers, minus Date and X-Original-Date
 159         test -n "$X_ORIGINAL_DATE" && \
 160                 sed -i '/^X-Original-Date:/,/^[^\t]/d' "$HEADERS"
 161         test -n "$DATE" && \
 162                 sed -i '/^Date:/,/^[^\t]/d' "$HEADERS"
 163
 164         # find a replacement date in Fetchinfo: header
 165         FETCH_DATE=`grep -im1 'X-FETCH-TIME: ' "$HEADERS" | cut -d ' ' -f 2-`
 166
 167         # or in Received: headers ..
 168         test $STRICT -eq 1 && \
 169                 REGEXP="$DATE_REGEXP" || \
 170                 REGEXP="$DATE_REGEXP_STRICT"
 171         RECEIVED_DATE=`sed -n '/^Received:/,/^[^\t]/p' "$HEADERS" | head -n -1 | grep -Eoim 1 "$REGEXP"`
 172
 173         # .. or from file properties
 174         FILE_DATE=`LC_ALL=POSIX LANG=POSIX ls -l --time-style="+%a, %d %b %Y %X %z" "$1" | tr -s ' ' | cut -d ' ' -f 6-11`
 175         # we could also use the system date as a possible replacement
 176         SYSTEM_DATE="`date -R`"
 177
 178         # determine which replacement date to use
 179         if [ -z "$FETCH_DATE" ]
 180         then
 181                 if [ -z "$RECEIVED_DATE" ]
 182                 then
 183                         # don't forget the leading whitespace here
 184                         REPLACEMENT_DATE=" $FILE_DATE"
 185                         REPLACEMENT="file date"
 186 #                       REPLACEMENT_DATE=" $SYSTEM_DATE"
 187 #                       REPLACEMENT="system date"
 188                 else
 189                         REPLACEMENT_DATE="$RECEIVED_DATE"
 190                         REPLACEMENT="received date"
 191                 fi
 192         else
 193                 # don't forget the leading whitespace here
 194                 REPLACEMENT_DATE=" $FETCH_DATE"
 195                 REPLACEMENT="Fetchinfo time header"
 196         fi
 197
 198         # ensure that the original X-Original-Date is kept
 199         :> "$TMP"
 200         if [ -n "$X_ORIGINAL_DATE" ]
 201         then
 202                 echo "X-Original-Date:$X_ORIGINAL_DATE" >> "$TMP"
 203         fi
 204
 205         # replace/set the date and write all lines
 206         test $RFC -eq 1 && \
 207                 date_valid "$DATE"
 208         if [ -z "$DATE" ]
 209         then
 210                 test $DEBUG -eq 1 && \
 211                         echo "$1: date not found, using $REPLACEMENT now"
 212                 dump_date_fields
 213         else
 214                 if [ $FORCE -eq 1 ]
 215                 then
 216                         test $DEBUG -eq 1 && \
 217                                 echo "$1: date already found, replacing with $REPLACEMENT"
 218                         dump_date_fields
 219                 else
 220                         if [ $RFC -eq 1 ]
 221                         then
 222                                 if [ $DATE_VALID -ne 0 ]
 223                                 then
 224                                         test $DEBUG -eq 1 && \
 225                                                 echo "$1: date already found but not RFC-compliant, replacing with $REPLACEMENT"
 226                                         dump_date_fields
 227                                 else
 228                                         test $DEBUG -eq 1 && \
 229                                                 echo "$1: date already found and RFC-compliant, skipping"
 230                                         SKIP=1
 231                                 fi
 232                         else
 233                                 test $DEBUG -eq 1 && \
 234                                         echo "$1: date already found, skipping"
 235                                 SKIP=1
 236                         fi
 237                 fi
 238         fi
 239
 240         if [ $SKIP -eq 0 ]
 241         then
 242                 # uncomment the following line to backup the original file
 243                 #mv -f "$1" "$1.bak"
 244
 245                 cat "$HEADERS" >> "$TMP"
 246                 cat "$BODY" >> "$TMP"
 247                 mv -f "$TMP" "$1"
 248                 if [ $? -ne 0 ]
 249                 then
 250                         echo "error while moving '$TMP' to '$1'"
 251                         exit 1
 252                 fi
 253         fi
 254         rm -f "$HEADERS" "$BODY" "$TMP" >/dev/null 2>&1
 255
 256         shift
 257 done
 258 exit 0