Thursday, June 20, 2013

jQuery's extend() Method and Prototypes - A Deadly Combination

I recently have been working with some code where an object hierarchy was useful to me, and I implemented it using this sort of pattern:

function ObjectParent () {
}
ObjectParent.prototype = {
    constructor: ObjectParent,
    method: function () { ... },
    property: true,
    objectProperty: {
        subMethod: function () { ... },
        subProperty: true
    }
};
function ObjectChild () {
}
ObjectChild.prototype = new ObjectParent();
    $.extend(true, ObjectChild.prototype, {
        childMethod: function () { ... },
        childProperty: false
    });
function ObjectGrandChild () {
}
ObjectGrandChild.prototype = new ObjectGrandChild();
    $.extend(true, ObjectGrandChild.prototype, {
        grandChildMethod: function () { ... },
        childProperty: 'Now this is just getting weird.',
        objectProperty: {
            grandchildSubProperty: true
        }
    });
And it worked ... kind of.   Mostly?  Sort of?
I was seeing stuff from the grand child in the children and in the parent.  Why?  Well, first I took off the deep copy flag $.extend(true, ... ) becomes $.extend(...) and that made everything work.... sort of.

I had those sub-objects that I wanted to be extending and they weren't making it through properly.  So I made an explicit shallow copy for them, which looks like this:

$.extend(ObjectGrandChild.prototype.objectProperty, { grandchildsubProperty: true });

The question becomes ... where is ObjectGrandChild.prototype.objectProperty located?  If that property is actually coming from some previous level of prototypal inheritance, when we extend it in this way, we extend the carrying prototype and tend to pollute our prototype chain.

What we're actually doing here is saying:  take the item referenced by X (ObjectGrandChild.prototype.objectProperty -> ObjectParent.prototype.objectProperty) and update it with values from Y

I like the pattern I wrote, but the consequences of using correct prototypal inheritance are not always obvious.  Remember that object properties coming from the prototype are references and so changes pass through.  There is no way to overlay string / primitive data type properties in the fashion I was attempting.  You will always have to make a 'clone' of the prototype in order to extend it.

Which is easy enough:

ObjectGrandChild.prototype.objectProperty = $.extend({},
    ObjectGrandChild.prototype.objectProperty,
    { grandchildsubProperty: true });

Note that what we're doing is this:  redefine the reference A (ObjectGrandChild.prototype.objectProperty) to point to the new object B ({}) which has been created by taking all the direct properties of X (ObjectGrandChild.prototype.objectProperty -> ObjectParent.prototype.objectProperty) and adding / replacing them with the direct properties of Y.

Monday, October 22, 2012

Compass / SASS for Placing Elements on a Circle

Need to put some elements on a circle?  Already using Compass / SASS to compile your CSS?

Woot!  Compass has you covered with trigonometric sine and cosine functions and of course the SASSy @for iterator helps, too.

This jsfiddle demonstrates the compiled result.

Your HTML document is composed of a container for the circle, an element for the circle, a container for items placed on the circle, and the items which will appear on the circle.

Your SASS file is composed of the following.  Modify the $positions, $ringSize, and $itemSize variables according to your needs:

// Use either Compass' border-radius mix-in or prefixfree plugin
@import "compass/css3/border-radius";

/**
 * on-circle takes a radius, position on a circle, number of possible positions and returns top and left properties
 *
 * $radius {Number}    radius of the circle in pixels
 * $ordinalPosition {Number} the position of the item on the circle, counting from 1 (North) through N
 * $positions {Number}   the number of positions on the circle
 * $originX {Number}   optional X origin point for the circle, defaults to the radius
 * $originY {Number}   optional Y origin point for the circle, defaults to the radius
 * $offsetX {Number}   optional X offset for the item, you might use 1/2 of the width of an item, default is 0
 * $offsetY {Number}   optional Y offset for the item, you might use 1/2 of the height of an item, default is 0
 */
@mixin on-circle ($radius, $ordinalPosition, $positions, $originX: $radius, $originY: $radius, $offsetX: 0, $offsetY: 0) {

 /*
 Determine the angle for the position:
  Multiply the adjusted zero-based index of the position by the degrees-per-position (360 degrees divided by the 
  number of positions) and subtract 90 degrees (adjusting to begin at North)
  */
 $positionAngleDegrees: ($ordinalPosition - 1)*360/$positions - 90;
 /*
 Convert the angle to radians:
  Multiply the angle by pi and then divide by 180 degrees.

 NOTE: This step is necessary because of a bug in handling of the degrees unit when doing iterations, AFAICT
  */
 $positionAngleRadians: $positionAngleDegrees * pi() / 180;

 /*
 Apply the parametric equation of the circle,
  http://en.wikipedia.org/wiki/Circle#Equations
 via:
  http://stackoverflow.com/questions/839899/how-do-i-calculate-a-point-on-a-circles-circumference

  x = [origin x] + (r * cos angle)
  y = [origin y] + (r * sin angle)
  */
 top: #{$originY + $offsetY + $radius * sin($positionAngleRadians)}px;
 left: #{$originX + $offsetX + $radius * cos($positionAngleRadians)}px;

}

// How many positions will there be on the circle?
$positions: 13;

// How big is the circle (diameter)?
$ringSize: 180;

// How big is an item?  Offsets and margins are based on this value
// such that the center of an item is located on the circle, rather than
// the top left point
$itemSize: 28;

body {
 margin: 50px;
}

.ring-container {
 border: solid 1px black;
 display: inline-block;
}

.ring {
 position: relative;
 // Use border-box so that margin and border are not included and the size of the element containing the
 // rendered circle is the circle size + border + ...
 box-sizing: border-box;
 width: #{$ringSize}px;
 height: #{$ringSize}px;
 margin: #{$itemSize/2}px;
 border: solid 1px black;

 // Make any square into a circle by setting the border radius to it's full width/height
 @include border-radius(#{$ringSize}px);
}

.ring-positions {
 position: relative;
 width: #{$ringSize}px;
 height: #{$ringSize}px;
 top: #{$itemSize/-2}px;
 left: #{$itemSize/-2}px;
}

.ring-position {
 position: absolute;
 width: #{$itemSize}px;
 height: #{$itemSize}px;
 @include border-radius(#{$itemSize}px);

 // Use box-sizing: border-box if you'll have a border
 box-sizing: border-box;
 border-color: black;
 border-style: solid;
}

@for $i from 1 through $positions {
 .ring-position-#{$i} {
  @include on-circle($radius: $ringSize / 2, $ordinalPosition: $i, $positions: $positions);

  // Just for demonstration of placement
  border-width: #{$i/2+3}px;
 }
}

Thursday, August 16, 2012

html5 boilerplate build Woes

On Friday I spent a few hours trying to debug doing a build on my current project using the html5 boilerplate build script (ant based).  Didn't happen.

Spent a couple more hours on Monday modifying my project to mirror the configuration for asset directories that the build script uses as its defaults, such as "js" for the scripts directory, where before I was using "javascripts" based on the html5 boilerplate template that I installed with (I forget which) Scout or compass.app.  Thought it would build after changing to the structure it expects / expected.  Didn't happen.

Well, the woes seem to be caused by changes to the default structure that have happened at the same time as the splitting of the build script out into it's own project / repository.

Today I put a little bit more time into the build and got it working by removing my exclusions from the project.properties file and adding the slug.libs property in.

So that's the deal ... if you're going to use something other than "vendors" as your libs directory, you *must* specify not only the dir.js.libs property but the slug.libs property.

Here are the settings I have in my project properties, which seem to match the html5 boilerplate template:

dir.source = ./public
file.stylesheets  = 
dir.js = js
dir.js.main = ${dir.js}
dir.js.libs = ${dir.js}/libs
slug.libs = libs
dir.js.modules = ${dir.js}/modules
dir.css = css
dir.images = img
file.root.stylesheet = style.css
file.root.script = script.js

Good luck with your html5 boilerplate project!

UPDATE:

I also had issues with the error "js/modules/_all.js was specified as an input resource."
I could get past the error by doing an "ant clean" before the "ant build", but obviously the error means there's a problem.

To get past it, modify your build.xml around line 486 to fix the module concatenation which is used to create a checksum, as follows, adding an exclusion for _all.js.

<concat destfile="./${dir.intermediate}/${dir.js.modules}/_all.js" overwrite="no">
    <fileset dir="./${dir.intermediate}/${dir.js.modules}/">
        <include name="*.js">
        <exclude name="_all.js">
    </exclude></include></fileset>
</concat>


UPDATE the Second:

Subdirectories of /libs are not being copied.  :/  This might help ... testing it now...

<copy todir="${dir.publish}/${dir.js}">
  <fileset
        dir="${dir.intermediate}/${dir.js}"
        includes="${file.js.bypass}, ${slug.libs}/*/**, ${slug.modules}/*/**">
        <exclude name="scripts-concat.js"/>
        <exclude name="scripts-concat.min.js"/>
        <exclude name="otherscripts-concat.js"/>
        <exclude name="plugins.js"/>
        <exclude name="${file.root.script}"/>
    </fileset>
    <regexpmapper from="^([^/])*/(.*)$$" to="\1/\2" handledirsep="true"/>
</copy>

Looks good. That's a change for the task around line 532 of build.xml.

Wednesday, August 15, 2012

Making a Compass Site Relative

I'm creating what at least for now is a single-page webapp and using, among other things, Compass.

One of the extensions I'm using (via Compass.app) is thomas-mcdonald-bootstrap-sass.  I've modified it to use the Font-Awesome icon font which replaces and extends the standard twitter icon library, as well as to add an IcoMoon icon font for even more icons.

To add the icon fonts to the compass bootstrap extension, you need to specify the path to the font.  In the instructions, you're told to put the absolute path to the font (such as "/fonts/fontawesome".  But we don't want to do that, we want to have all relative paths in our files.  If we don't include a beginning "/" in the $fontAwesomePath variable, then Compass will assume that the path is relative to the compass project's font directory (because of how the font-file function is coded).

The default font directory is the css directory plus "/fonts".

That made things awkward for using relative paths, so I did some digging and found a post where someone mentions using "font_dir" to configure the font directory path ... well, it's "fonts_dir", folks, not "font_dir".

To make a long story short, using the image-url function for my image paths, and trying to remove the absolute pathing by changing "http_path" to "" didn't work.

These are the settings in my config.rb that *did* work:

http_path = "./"
css_dir = "css"
sass_dir = "sass"
images_dir = "../img"
javascripts_dir = "js"
fonts_dir = "../fonts"

Of course you should modify that to use the same paths that you're using, such as "javascripts" and "stylesheets" instead of "js" and "css".  Just make sure that you have "./" as the http_path and "../" in front of your images and fonts directory paths.  When those are used, it's in the CSS context and your generated CSS file is going to end up in a directory that's parallel to them.

If somehow your final CSS file is in a different place, make sure that the images and fonts variables are changed accordingly.

Monday, July 23, 2012

Pandora, Stardock Objectdock, Large Number of Chrome Processes

I've been using Stardock's Objectdock for a while, because I'm now coding on a Windows 7 machine.  I also have a subscription to Pandora One ... at some point I saw about a bajillion chrome processes coming up in the dock and making it useless.

Turns out that it's because of Pandora One.  Chrome is my default browser and it seems like Pandora is using it and not closing the handle or something.  Who knows.  Problem solved by not using their standalone Adobe AIR application and going back to having one window open for listening.

UPDATE:  Apparently you have to start Pandora One BEFORE you start Chrome and the problem will not occur.  (Thanks to Pandora tech support for providing the workaround)

Thursday, November 24, 2011

Simultaneous Background Process Monitoring in Bash using Multitail

I'm working on an installer for the development server environment we use at www.salsalabs.com

There are time-consuming tasks which are part of the installation process which can be done simultaneously: installation of program dependencies (with brew) and download of code repositories (using git).

I backgrounded the first task in my installation script using the usual & and let the script continue on and complete the second task "normally". After the second task completed, a wait command makes sure that the script does not proceed until the first task is also complete.

That's all well and good. It works just dandy... except that the output of these two processes is intermixed and due to each process providing feedback about the process of individual steps ... that output is very very very very very messy.

So I looked for a solution and I found MultiTail. Awesome! And available via brew! Awesome!

Except ... I don't have log files, and I don't want to generate log files. I want to have two different output streams and use multitail to view them simultaneously in one terminal window.

Two different output streams is easy enough, we've got file descriptors for that, right? (check out the exercise here, it makes my brain hurt). No, we need two different files that are going to act like streams. Oh! We want buffers!

(Cue the superhero music)

Here come FIFOs to the rescue! Use the handy-dandy mkfifo command and you can turn any output stream like stdout, stderr into a buffer that looks like a file to the operating system. Yay!!!!

So, our code, something like this:


install_dependencies.sh &
install_packages.sh
wait

install_something_else.sh


becomes:


mkdir /tmp/fifos
mkfifo "/tmp/fifos/Dependencies"
mkfifo "/tmp/fifos/Code Installation"

install_dependencies.sh > "/tmp/fifos/Dependencies" &
install_packages.sh > "/tmp/fifos/Code Installation"
wait

install_something_else.sh


And we add on multitail:


mkdir /tmp/fifos
mkfifo "/tmp/fifos/Dependencies"
mkfifo "/tmp/fifos/Code Installation"

install_dependencies.sh > "/tmp/fifos/Dependencies" &
install_packages.sh > "/tmp/fifos/Code Installation" &
multitail -ts --basename "/tmp/fifos/Dependencies" "/tmp/fifos/Code Installation"
wait

install_something_else.sh


The problem, now? multitail is never going to exit on it's own, we'll have to press "q" to end the process when the other stuff is done ... assuming we can tell for sure. So we never get to the wait command and we never get on to the next steps of our installation. Let's send an email to the developer and ask him for that feature ... but we'll make our own solution in the meantime.

So, here presented for you is my solution for monitoring a group of parallel background processes using multitail and ending the monitoring when all processes are complete.


PROCESS_COMPLETED_COUNT_TMP_FILE=$(mktemp -t "proc-count")
echo 0 > $PROCESS_COMPLETED_COUNT_TMP_FILE

function killtail {
    MUST_COMPLETE=$1
    PROCESS_COMPLETED=$(cat $PROCESS_COMPLETED_COUNT_TMPO_FILE)
    ((PROCESS_COMPLETED++))

    if [ PROCESS_COMPLETED -ge $MUST_COMPLETE ]; then
        # dangerous if there are multiple multitails running at the same time
        TAIL_PID=$(pidof multitail)
        kill $TAIL_PID
        rm -rf $PROCESS_COMPLETED_COUNT_TMP_FILE
    fi
}

mkdir /tmp/fifos
mkfifo "/tmp/fifos/Dependencies"
mkfifo "/tmp/fifos/Code Installation"

# subshell for everything involved with installing dependencies
(
    install_dependencies.sh > "/tmp/fifos/Dependencies"
    killtail 2
)> "/tmp/fifos/Dependencies" &

# subshell for everything involved with installing code
(
    install_packages.sh > "/tmp/fifos/Code Installation" &
    killtail 2
)> "/tmp/fifos/Code Installation" &

multitail -ts --basename "/tmp/fifos/Dependencies" "/tmp/fifos/Code Installation" 2> /dev/null
wait

rm "/tmp/fifos/Dependencies"
rm "/tmp/fifos/Code Installation"

install_something_else.sh


Now the subshells are backgrounded instead of the scripts themselves. Both "steps" are backgrounded, not just the first one. The command to increment the kill counter and check for the appropriate number of processes to have completed is added to the block of code. As well as backgrounding the subshells, we're directing stdout output from the subshells into our FIFOs, instead of directing the output of individual commands.

We pass the --basename option to multitail so that we see only the name of our FIFO in the multitail windows... and we've used nice human readable names for these temporary buffer files, so isn't that special?

When the kill counter hits the magic number, a SIGTERM will be sent to the multitail that we get back from pidof multitail. This makes multitail error, so we're throwing away the error output from multitail.

Finally, we still have a "wait" statement, just in case our multitail monitoring of the processes doesn't work for some reason. We still want to make sure they complete before moving forward with more steps. We remove the FIFOs and move on to the next steps in our installation.

Happy process monitoring!

Monday, December 21, 2009

Question: How do you convert a Java String object to a Javascript String in Rhino?

Convert any Java object with a toString() method into a JS string by calling String(object) in JS


var javaString = new java.lang.String("Java String");
var javascriptString = String(javaString);