I should have mentioned in my previous blog post that my friend Jonathan originally told me about perl -00pe0. After I posted it, we exchanged a few more emails about various ways to squeeze multiple blank lines, and he pointed out that cat -s is not, in fact, exactly the same as perl -00pe0. The difference is how they deal with leading blank lines. cat -s will squeeze them down to a single blank line, but perl -00pe0 will remove all of the leading blank lines.

I learned a fun new perl one-liner today:

perl -00pe0 foo >bar

What does it do? Well, it’s nearly the same as cp foo bar, except that it squeezes multiple blank lines into foo into a single blank line in bar.

How does it do that? Well, let’s break it down piece by piece. Perl’s -0 flag lets you change perl’s “input record separator”. This lets you fiddle with perl’s idea of what a “line” is. Normally perl reads in text a line at a time, where a “line” is a string of text terminated by a newline character. But when you set the -0 flag to 0, perl will keep reading until it hits a string of 2 or more newlines in a row. In other words, perl will read in text a paragraph at a time, and stop when it hits a blank line. If there are multiple blank lines, perl will treat them as if there was just a single blank line and throw away the extras. (You can also do the same thing by setting perl’s special $/ variable to “”.)

The next flag is -p. That tells perl to read in text a “line” at a time, execute some code against each line, and then print the line to standard output. Of course, because of the -00 right before it, each “line” is actually a paragraph.

The final flag is -e. This tells perl that the following parameter is a program that it should execute after it reads in each line. Normally this is given in between single quotes, but here the entire program is simply “0”. This is a perl idiom for a no-op. This is because we don’t need any additional code — the combination of the -00 and -p flags are all we need.

Incidentally, blank lines can be squeezed much simply, and with far less obfuscation, by running cat(1) with the -s flag:

cat -s foo bar

Why did I pick “mawode.com” for my domain name? It goes back to when I visited my friends Jonathan and Chyun in Taipei back in 1998. While I was there they decided I needed to have a proper Chinese name. Most Chinese names are 3 syllables (think Mao Tse-tung, Sun Yat-Sen, and Yo-Yo Ma) and that’s what they were trying to come up with. After agonizing over this for several days, they finally came up with 馬沃德, which is written in Pinyin as Mǎ Wò-Dé. I strung them together to come up with “Mawode”.

In Chinese the usual practice is to say the surname first, so Mǎ is my “last name” and Wò-Dé is my “first name”. Mǎ means “horse”, and is a very common Chinese surname. It’s the same last name as Yo-Yo Ma (although he follows Western practice and puts it at the end). The character 馬 is supposed to be a stylized picture of a horse. If you look closely you can make out the 4 legs at the bottom.

They picked Mǎ because it was the first two letters of “Mankowski”. They tried to do the same thing with “Walt”, but couldn’t get it quite right. Although there are several different meanings for “wa” depending on the tone, none of them really worked for a name. As for the “lt”, well, Chinese doesn’t really have the same “l” sound we have in English, so that was never going to work out, either. They decided on “wò”, meaning “fertile” or “rich”, and “dé”, meaning “morality” or “virtue”. When you say the two syllables together, it sounds pretty close to “Walt”!