BogoJoker

Stack Overflow - Edit Summary Quicklinks

joe — Thu, 04 Sep 2008 02:11:01 +0000

I requested a feature to make the few edit summary suggestions clickable. I, like many developers, don’t normally let my keys leave the keyboard. However, this was one case where I felt making those suggestions of “corrected spelling” and “fixed grammar” should automatically be inserted.

Well the suggestion was declined. I can’t blame the team. Nobody upvoted the suggestion. But I felt strongly enough about it, and knew that it was very simple to implement that I whipped up a GreaseMonkey script to do it myself. The script runs like a charm and even adds a few extra suggestions to the original three. It handles formatting and commas all automatically, so don’t worry about a thing, just click. Enjoy!

Click Here to Grab the GreaseMonkey Script

UPDATE: Fixed to use keng’s URL and added just plain old stackoverflow.com without the beta sub-domain in preparation for a launch. Thanks keng!

Spaces, Exposé, Hot Corners, Dashboard Broken?

joe — Sat, 16 Aug 2008 03:00:56 +0000

I ferociously make use of Spaces and Exposé. To utilize them I use QuickSilver, Cmd+Tab, the Dock, special Mouse Buttons, but most of all I use the Hot Corners! Occasionally, when I wake my Mac Book Pro these features won’t work. No reason given, just when I try I get the usual chime meaning “this can’t be done.” At the time the only sure solution was to reboot the computer. Not anymore. Just open up your terminal and run:

killall Dock

Since those major features (Spaces, Exposé, and the Dashboard) are tied to the Dock, this command will restart the Dock and, like magic, those features will work again. Killing the Dock and even Finder with `killall` are special cases. Those two will automatically restart on their own. However, for other programs, it will just quit. For instance if Firefox hangs you could type `killall Firefox` and Firefox would be force quit but would not reopen. Check the man page for more information.

Command Line svn:ignore a file

joe — Sun, 20 Jul 2008 17:59:07 +0000

This took me far too long to do, but its because no one explained it correctly to me. I’m frustrated enough that I’m not going to go back to those other websites and see if I just overlooked something. Instead I’m going to put it right here.

You don’t svn:ignore a file. You put an svn:ignore property on the directory to ignore that filename pattern.

It makes sense, but I didn’t immediately think of that, and no source sufficiently made that point to me. So if you were struggling, please grasp that concept and take a look at the commands below (which you no doubt have seen and did not think worked) and you’ll really understand what they do.

# ---------------------------------------------------------------------
#  Ignore all the .wdgtuser files in the /trunk/Blah.dcproj directory
# ---------------------------------------------------------------------

# Go to the directory
cd trunk/Blah.dcproj/     # The directory with the files

# Start editing the properties for the current directory
svn propedit svn:ignore .   # Opens an editor (SVN_EDITOR, EDITOR)

# Add the following value with a new line, save, and exit:
*.wdgtuser

# See that things worked
svn propget svn:ignore .    # So you can see the properties
svn status --no-ignore      # You should see an 'I' next to the ignored files

# Commit
svn commit -m "New Ignores" # You must commit the new property change


# ---------------------------------------------------------------------
#     Ignore a single file secret.txt in the /trunk/ directory
# ---------------------------------------------------------------------

# Go to the directory
cd trunk/

# Add just the single file to the current directories ignore list (like above)
# Note the dot at the end of the command is important
svn propset svn:ignore secret.txt .

# See that things worked
svn propget svn:ignore .    # Notice the single file was added to the list
svn status --no-ignore      # You should see an 'I' next to the ignored files

# Commit
svn commit -m "Its secret"  # You must commit the new property change

That also means in GUI programs if you can’t seem to ignore a single file that is unversioned you should instead go to the directory that file is in and (like the above) add the filename to the svn:ignore list.

Cheers.

My Blog Is

joe — Fri, 20 Jun 2008 01:55:54 +0000

That right there is the only tag cloud that I have ever found interesting. The cool part is that you can generate those tag clouds and with a lot more jazz at this really great service called Wordle.

How I Did It

For those interested on how I got all the tags for the above cloud (with weight) I decided to grab them myself. Here were the simple 4 steps I took.

The SQL grabs all of the tags used on blog posts (so not categories) at least once. It also grabs the number of times the tag appears on a post. Using a simple ruby script I take the tags (all single words) and multiply them by the number of times they appear. So a “ruby 3″ turns into “ruby ruby ruby.” Once that multiplication has taken place I can just plop them into Wordle and make the magic happen!

More clouds here!

SotD - Remove the First N Characters From a Line

joe — Wed, 18 Jun 2008 02:58:37 +0000

I was asked today to remove the first 6 characters from every line in a document. It was known that each line contained at least 6 characters. For me, that means a find a replace. Here is how I did it (in a number of different ways). Can you think of any other ways to do it?

cut -c7-
rr ^.{6}
colrm 1 6
rr s/^.{6}//
sed 's/^.\{6\}//'
perl -pe 's/^.{6}//'
ruby -pe 'sub /^.{6}/,""'
gawk 'BEGIN{FIELDWIDTHS="6 999"}{print$2}'

I was pretty happy with rr taking the prize for least characters. I do want to point out that since rr defaults to multi-line and global replacements the “^” is required. If you wanted to remove the “^” (like you could with sed, perl, and ruby) then you would have to use the “–line” or “-l” and “–notglobal” or “-ng” options. So a single character replaces 7! If you want to grab rr just do:

$ sudo gem install regex_replace

Also, here is a link to gawk - GNU’s awk. This has the very nice FIELDWIDTHS variable, which is extremely useful!

Also keep in mind that the regular expressions above ^.{6} only work because the lines were known to have at least 6 characters. If we didn’t know that we would have had to use something like ^.{0,6} to allow up to 6 characters (and even that could be ^.{1.6} ignoring blank lines which can’t change). So again the requirements for this challenge were important.

Update!

I came across a neat little unix utility I didn’t know about called “cut”. As you can see, the new command tops the list quite handily too. “cut -c 7-” is actually cutting from each line the 7th character onwards (counting starts from 1). In this case once it spits what it cuts out to stdout; so it leaves behind those first 6 characters and therefore accomplishes our goal. In the above list I removed the optional space after “-c” to make it just a tad shorter. Pretty neat.

Then I found “colrm” which is the most straightforward. This one wins in simplicity. No hacks, just straightforward does exactly what you think it does. Very cool.

Double Update!

I figured since quite a bit of my rr usage has an empty string for the second argument, I figured it would be okay to throw that in as the default case. So now there is a third usage for rr, which is great for pipes, that just strips something out. That brings rr back up to a tie for first place. Very cool.

Why CDATA Matters in XML

joe — Mon, 16 Jun 2008 03:31:51 +0000

You’ve seen it before, but you may not know what it means. Wikipedia describes CDATA as meaning “Character Data” which makes sense. w3school goes one step further and points out that this text should not be parsed by the XML Parser. The general idea is when you want to display straight textual data, without needing to encode characters or wanting them interpreted by the parser, you can just wrap that data inside of a CDATA tag.

Needless to say this is clearly the ugliest tag currently in existence (lets leave room for the future though):

I promised to tell you why it mattered

Yes, words in the title become a promise. You can hold me to that in the future. Why does CDATA matter? Well, I’ll actually side-step the question for a minute and show you what looks like a perfectly fine looking XHTML document (keep in mind that XHTML is a subset of XML):

DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
  "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
<head>
  <meta http-equiv="Content-type" content="text/html; charset=utf-8" />
  <title>Simple Exampletitle>
head>
<body>
  <h1>Welcomeh1>
  <script type="text/javascript">
    document.write('Hello, World
');
  script>
body>
html>

Looks simple enough. We can ignore the fact that its not really the best way of doing things, but who cares… its a Hello World example right? Well, technically this is not Valid XHTML. The validator shows we’ve got a single error:

Line 11, Column 20: document type does not allow element "p" here.
document.write('>Hello, World
');

The element named above was found in a context where it is not allowed. This could mean that you have incorrectly nested elements — such as a “style” element in the “body” section instead of inside “head” — or two elements that overlap (which is not allowed).

Well, the validator hints at the cause of the error, but it is hard to understand unless you really know your XML! We are inside a “