A little enhancement to the listing of stockpiled blog posts.

In previous post Stockpiling blog posts we took a quick look at, a small shell program that helps me write blog posts in advance and park them until I need them.

In time, in addition to keeping a few filler posts, I also found it useful to store other more serious ones, just to avoid scheduling too much stuff in advance. Who knows… I might find some very interesting topic that I want to anticipate, and decide to defer a post or two.

On the other hand, I like to at least guess the publishing date right, so I usually try to assign a reasonable date anyway. Which has become increasingly difficult lately, because I’m putting more posts and I have to always look at what the next date should be.

This is further made difficult by the Perl Weekly Challenge, by the way, because I try to reserve a couple of dates per week to those posts.

In summary… a mess.

In trying to make things a bit easier, I enhanced the list sub-command of to also include the date set for all stockpiled posts. This allows me to have the information quickly, and ease my routine.

For the curious, this is the new command implementation:

command_list() {
   local branch rest
   git branch \
      |  sed -ne '/^..stockpile\/item-/s/^..//p' \
      |  while read branch rest ; do
            git diff "$branch^..$branch" \
               | sed -ne '/^+---/,/^+---/{s/^+//;p}' \
               |  awk '
                     /^title:/ { $1 = ""; title = $0 }
                     /^date:/  { date = $2 }
                     END       { print date " " title }
                  ' \
               | sed -e "s#^#$branch  #"
         done \
      |  nl

The first two commands (git and sed) make sure to only isolate stockpile items and provide a list of branches related to a stockpiled post.

The big while loop that follows gets the branch name into variable branch and extracts the date and the title out of it.

These two pieces of information are put inside the initial front matter of the post, which is a YAML fragment delimited by the conventional --- lines. The sed command makes sure to only isolate those lines. And yes, if there’s anything like that in the following of the post… I’m pretty much in trouble. Something to do for future me, I suppose.

The awk part takes care to only get the date and the title, and print them in this order; it is followed by a last sed command that adds the branch name at the beginning.

Last, for good measure I threw a nl to all of this, because… it will come useful.

Stay safe, protect yourself!