Work the Shell - More Fun with Days and Dates

Figuring out how to calculate the year for a given date and day of week is a task that's not as easy as it sounds.

I received a very interesting note from a reader—a note that referred to a very interesting problem:

Many UNIX commands (for example, last) and log files show brain-dead date strings, such as “Thu Feb 24”. Does anybody out there have a script that will convert that to a year, given a five-year interval and defaulting to the present?

Given a day of the week, a month and a day, is it possible to calculate quickly the most recent year in the past when that particular date occurred on that day of the week? Of course it is!

Various formulas exist for calculating this sort of thing, but I realized pretty quickly that the handy cal utility can do the work for us. If you haven't experimented with it, you'll be surprised at what it can do. Here are two quick, relevant examples:

$ cal
     March 2011
Su Mo Tu We Th Fr Sa
      1  2  3  4  5
6  7  8  9 10 11 12
13 14 15 16 17 18 19
20 21 22 23 24 25 26
27 28 29 30 31 

$ cal mar 2007
     March 2007
Su Mo Tu We Th Fr Sa
            1  2  3
4  5  6  7  8  9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31

Any light bulb starting to glow for you? If you know the month and day, you simply can go backward looking at that month's day-of-week layout until finally you find a match.

In a rudimentary fashion, the basic idea can be illustrated with a loop, like this:

  cal $month $year | grep $day
  if day-of-week matches
     echo date $month $day most recently occurred in $year
     year=$(( $year - 1 ))
end repeat

Of course, the problem is a bit more complicated (as they always are), partially because of the complexity of calculating what day a specific date occurs in the cal output. There's another complication too, however; the requested date actually might have occurred in the current year, so it's not as simple as starting with the year 2010 and going backward.

Normalizing Data

The first task is to figure out how to get the information from the user. We'll have only three input parameters and do relatively little testing for misspelled day names and so on:

if [ $# -ne 3 ] ; then
  echo "Usage: $(basename $0) weekday month day"
  echo "  (example: $(basename $0) wed aug 3  )"
  exit 1

That's straightforward and pretty typical, offering a nice usage tip if you forget how to use the script. As is typical of scripts, we return a nonzero result upon error too.

We can't work with completely arbitrary data, however, so when we grab the first few parameters, we'll transliterate them into lowercase and chop off all but the first three letters:

weekday=$(echo $1 | tr '[[:upper:]]' '[[:lower:]]'; | cut -c1-3)
  month=$(echo $2 | tr '[[:upper:]]' '[[:lower:]]'; | cut -c1-3)

Given “Monday, February 8”, it'd be converted automatically to “mon” and “feb” for subsequent testing.

The Current Date

We also need the current date fields for testing, and to do this, I'll demonstrate a very neat trick of date that makes this incredibly efficient:

eval $(date "+thismonth=%m; thisday=%d; thisyear=%Y")

The eval function interprets its argument as if it were a direct script command. More interesting, date can output arbitrary formats (as documented in strftime if you want to read the man page) by using the + prefix, with %m the month number, %d the day of the month and %Y the year. The result is that date creates the string:

thismonth=03; thisday=01; thisyear=2011

which then is interpreted by the shell to create and instantiate the three named variables. Neat, eh?

It turns out that users can specify a month by name or number on the command line. If it's already a number, it'll survive the transforms intact. If it's a name though, we also need the number, so we can figure out whether the date specified could be earlier this year. There are several ways to do this, including a case statement, but that's a lot of work. Instead, I'll lean on sed as I quite frequently do:

monthnum=$(echo $month | sed

Here's where a misspelled month name is a problem, but that's a challenge beyond the scope of this script. For now, however, we'll just roll with it.


Dave Taylor has been hacking shell scripts for over thirty years. Really. He's the author of the popular "Wicked Cool Shell Scripts" and can be found on Twitter as @DaveTaylor and more generally at


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

GNU `date' is enough...

zeroxia's picture

As we are talking about Linux, the GNU `date' utility could be far more clever than you might thought, so there is no need to mess with the output of `cal'...

Quick hint:
date -d "Sep 25 2008" +%A
date -d "Sep 25 2008" +%u

But this probably is GNU only, AFAIK, at least the BSD `date' does not have such magic.

Here is an `sh' script listing I just crafted:


export LC_TIME=C

    cat <<!
    ${0##*/} WEEKDAY MONTH_NAME Day
    ${0##*/} WEEKDAY MONTH-DAY
    ${0##*/} WEEKDAY MONTH/DAY

if [ $# -ne 3 -a $# -ne 2 ]; then
    exit 1
if [ $# -eq 3 ]; then
    # GNU `date' accepts "Sep 25 2008"
    fmt="$2 $3 %d"
    # And also accepts "2011-9-25" or "9/25/2011"
    case "$2" in
            echo "Uknown date: $2"
            exit 1

case $(echo $1 | tr '[:upper:]' '[:lower:]' | cut -c1-3) in
    mon) weekday=1 ;;
    tue) weekday=2 ;;
    wed) weekday=3 ;;
    thu) weekday=4 ;;
    fri) weekday=5 ;;
    sat) weekday=6 ;;
    sun) weekday=7 ;;
        echo "$1: Unknown weekday"
        exit 1



y0=$(date +%Y)
while [ $i -lt $MAX_TRY ]; do
    y=$((y0 - i))
    str=$(printf "$fmt" $y)
    if ! j=$(date -d "$str" +%u); then
        # `date' will complain, so I keep quiet
        exit 1
    if [ $j -eq $weekday ]; then
        echo $(date -d "$str" +$MY_DATE_FMT) is \
            $(date -d "$str" +$MY_WDAY_FMT)
    i=$((i + 1))

test $found -eq 0 && exit 1
exit 0
White Paper
Linux Management with Red Hat Satellite: Measuring Business Impact and ROI

Linux has become a key foundation for supporting today's rapidly growing IT environments. Linux is being used to deploy business applications and databases, trading on its reputation as a low-cost operating environment. For many IT organizations, Linux is a mainstay for deploying Web servers and has evolved from handling basic file, print, and utility workloads to running mission-critical applications and databases, physically, virtually, and in the cloud. As Linux grows in importance in terms of value to the business, managing Linux environments to high standards of service quality — availability, security, and performance — becomes an essential requirement for business success.

Learn More

Sponsored by Red Hat

White Paper
Private PaaS for the Agile Enterprise

If you already use virtualized infrastructure, you are well on your way to leveraging the power of the cloud. Virtualization offers the promise of limitless resources, but how do you manage that scalability when your DevOps team doesn’t scale? In today’s hypercompetitive markets, fast results can make a difference between leading the pack vs. obsolescence. Organizations need more benefits from cloud computing than just raw resources. They need agility, flexibility, convenience, ROI, and control.

Stackato private Platform-as-a-Service technology from ActiveState extends your private cloud infrastructure by creating a private PaaS to provide on-demand availability, flexibility, control, and ultimately, faster time-to-market for your enterprise.

Learn More

Sponsored by ActiveState