I LIKE stamping things with the date, whether it’s a status report email or filename. (Like say I’ve downloaded todays news in Spanish, it really seems to make sense to tag it as spanish_news_12_8_08.mp3 instead of just spanish_news.mp3 doesn’t it?)

Okay, so this is something I do in a lot of my shell scripts. Here’s some code to play with.

Thankfully, linux has a great (and flexible) date command…

In the top part of your script just assign the variable date as such:

date=`date +%b%d%Y`

this uses the date command with the following format: Dec082008

Of course, there are other ways to format it – like date +%F which gives you 2008-12-08

As you’re seeing the +and then %letter formats the date. There are MANY different ways to format the date, in fact, if you just want the month, or day of week, that’s possible too. you might want to consult the manual page for date to see all the choices.

So… from the console type man date

now, you can have the date or time stamp you want in any of your shell scripts.

Recently I’ve run across a realmedia stream that I really want to convert to mp3 (mainly because my portable player can’t read rm files. I’ll talk about capturing rm streams in another post, but this one is going to talk about converting from rm to mp3.

For starters, it’s possible to use mencoder for this as follows:

mencoder input_file.rm -ovc frameno -oac mp3lame -of rawaudio -lameopts cbr:br=128 -o output_file.mp3

This failed for me with an error that video was not optional. Rather than trying to figure out how to make mencoder deal with an audio only rm… I looked for another way around it.
So, …if you want another way to approach it, try this…

This makes use of mplayer to dump the audio to wav and then lame to encode to mp3

mplayer -ao pcm:file=audiodump.wav myrealmediafile.rm
lame -h -v -b 128 audiodump.wav finalemp3file.mp3

The -b rate in the lame command sets the minimum bitrate – if you’re converting audio and you know your device can deal with variable bit rates you could drop that to 32. (That’s what I did…)

If you have a linux machine and haven’t ever made use of sed (stream editor) you’re missing out on a great automation utility. I’ve saved myself probably 20 hours of manual editing with about an hour of work TWICE today. Here’s how…. over on the North Carolina Genealogy site I was opening forums for each county in the state of North Carolina (100 counties.) Now, I could have gone through and typed out a description, slug (address) and name for each one, but that looked too tedious. So…

I copied and pasted the list of counties to a text file (there was a carriage return after each) so I had one per line. Then I made a second file which I used tr to convert from upper case to lower case (since I wanted lower case “slugs”) I then went through and used sed to convert a space to an underscore. Now, I could have done this by hand because the only county with a space in the name is New Hanover, but…. it was easy as the following…

cat county_lowercase.txt | sed 's/ /_/' > county_lowercase_nospace.txt


So, the next step was crafting the slugs…

cat county_lowercase_nospace.txt | sed 's/$/-county-nc-genealogy-queries/' > slugs.txt


I checked the forum list table for my forum software through phpmyadmin and it was a pretty simple structure, about 8 fields…. a unique id which I set to NULL in my database, forum name, description, slug, forum order, number of posts, number of topics…. I set the last two to 0 throughout (100 lines), the forum order I started a bit past the current forum order numbers that were listed in the current database and I copied and pasted from the textfile the slugs.

So, next I crafted the forum titles….

cat county_list.txt | sed 's/$/ County, NC Genealogy Query Forum/' > forum_titles.txt


And of course, those get copied and pasted into the correct column. By the way, the $ above tells sed to put the text at the end of each line so that the county name comes first.

Then the descriptions, again I cut and paste and things look good. From the spreadsheet now I export to csv and import via phpmyadmin and check, lo and behold I now have 100 forums correctly added.

The other big thing though was a script I use to retrieve rss feeds. Each page was to have one uniquely named that would be included via a php include call in the page.

So… again sed to the rescue.

#!/bin/bash
for name in `cat county_lower_nospace.txt`;
do
cat template_gen.php | sed "s/template_search/$name+county%22+nc/" > "$name"_gen.php
done

VERY easy and it worked like a charm…. 100 php scripts created, one for each county in …. 5 seconds? Nice.

Finally to top things off, I used sed to make the links to insert in to the pages… again all 100. So, it turned into a simple copy paste for each one.

Thanks sed.. I don’t know how else I could have done this that didn’t involve hours of menial/tedious editing.

For more information on some of the neat textfile tricks you can do with sed… take a look at this sed guide.