Fetching the status of blog entries using grep and cut
I write my blog entries in some pseudo-POD/HTML that grew organically. In other words, it is a mess.
I wanted to easily see which articles are ready to be published and which one might need some more writing.
Each article is a text file with .tt extension (don't ask! probably something to do with Template::Toolkit), and each file has a line staring with =status followed by a value.
Originally each value could be either show or draft. The former for when the article is published and it is to be shown on the web site. The latter when it is still only a draft.
Lately, going with the organic evolution, I started to use all kinds of other status values. For example hide for articles I decided to pull from the web site, draft sent to Smylers for an article I sent to Smylers who is kind enough to proofread some of my articles posted on the Perl Maven web site. I use ready for proofreading for articles that are, well, ready for proofreading. I don't want to overwhelm Smylers any more, but sometime I have a few article written ahead of time.
(Gabor, please get to the point already!)
So I wanted to see what status values are there.
grep =status *
gives me a list like this:
how-to-read-a-csv-file-using-perl.tt:=status show some-other-article.tt:=status draft ready for proofreading fetching-the-status-of-blog-entries.tt:=status draft fetching-the-status-of-blog-entries.tt: has a line starting with =status
It is quite unreadable and there is a line where the =status was inside the text. (probably this article)
Slight improvement in the regex of grep to only match =status at the beginning of the line:
grep ^=status *
This will print:
how-to-read-a-csv-file-using-perl.tt:=status show some-other-article.tt:=status draft ready for proofreading fetching-the-status-of-blog-entries.tt:=status draft
It is still unreadable so I'd better cut it up.
grep ^=status * | cut -d ':' -f 2
Will take each line, split it at the chose field-delimiter character (:) and show the 2nd field:
=status show =status draft ready for proofreading =status draft
That's better, but I'd also like to get rid of the =status. Another cut splitting the lines on space and showing the second field:
grep ^=status drafts/* | cut -d ':' -f 2 | cut -d ' ' -f 2
This is almost good, but because the words in the value of the =status were also separated by spaces, they were considered as separate fields and are not displayed.
show draft draft
Adding a dash at the end will tell cut to include a range of fields:
grep ^=status drafts/* | cut -d ':' -f 2 | cut -d ' ' -f 2-
show draft ready for proofreading draft
Of course my list is much longer and includes lots of duplicates. To get the unique list of used values I could sort the output and use the uniq function to eliminate duplicates.
grep ^=status drafts/* | cut -d ':' -f 2 | cut -d ' ' -f 2- | sort | uniq
Processing with Perl
Another way I processed the file also involved using a Perl one-liner:
grep ^=status drafts/* | perl -nE 'chomp; ($file, $status) = (split /[: ]/, $_, 3)[0, 2]; printf "%-25s %s\n", $status,$file' | sort
(I broke up the one-line code to several lines to make it more readable in the post.)
This allowed me to keep both the file name and the status with one call to split and then to print them in reverse order before sorting the lines. Using the unix sort again. This helps me locate articles that I should work on.
BTW for more Perl-related articles check out the Perl Maven site.