Timestamped, gzipped tarball of a directory

#!/bin/sh

# tarit
# tar a given directory with timestamp.

fname=$1-`date +%Y-%m-%dT%H-%M`.tgz
tar czvf  $fname $1
echo
echo $fname

Updates

CGI programming with THTTPD

thttpd is a small, fast and robust HTTP server that weighs in at less than 100KB of executable.

thttpd can make a fine web server for lean hardware (eg: single board computers).

Here, I am trying to serve a CGI application using tHTTPd.

Install thttpd

  $ cd
  $ wget http://acme.com/software/thttpd/thttpd-2.25b.tar.gz
  $ tar -zxvf thttpd-2.25b-tar.gz
  $ mv thttpd-2.25b thttpd
  $ rm thttpd-2.25b.tar.gz
  $ cd thttpd
  $ make  

Run thttpd

  $ sudo ./thttpd 

By default thttpd runs on port 80, for which you need to be root. Let’s kill the process and use a higher port, which is accessible to non-super users.

  $ sudo killall -9 thttpd
  $ ./thttpd -p 8000

Configuring thttpd

Let’s copy the supplied example config file to the current directory and modify it to suite our requirement.

 $ cp contrib/redhat-rpm/thttpd.conf . 
 $ mkdir htdocs
 $ mkdir htdocs/cgi-bin
 $ mkdir log
 $ mkdir run

Note: I’m taking a few liberties in setting up the server. My intention here is to get to cgi programming. Check thttpd docs for safer alternatives, including chrooting thttpd.

dir=/home/pradeep/thttpd/htdocs
user=pradeep
logfile=/home/pradeep/thttpd/log/thttpd.log
pidfile=/home/pradeep/thttpd/run/thttpd.pid
port=8000
host=0.0.0.0
charset=utf-8
cgipat=**.cgi

Let’s test the setup.

  $ cat - > htdocs/test.html
   <h1>Hello thttpd</h1>
   ^D
  $ ./thttpd -C thttpd.conf

Visiting http://localhost:8000/ should show you the newly created page.

Creating a CGI Application

  $ cat - > htdocs/cgi-bin/hello.cgi
#!/usr/bin/env python
print "Content-Type: text/html"     # HTML is following
print                               # blank line, end of headers
print "<TITLE>CGI script output</TITLE>"
print "<H1>This is my first CGI script</H1>"
print "Hello, world! <br/>"
for i in range(10):
    print i, '<br/>'
print 'Just to show that some stuff is "dynamically" generated server side<br/>'
^D

The above code is directly out of Python CGI library doc.

  $ chmod +x htdocs/cgi-bin/hello.cgi

Restart the server and visit http://localhost:8000/cgi-bin/hello.cgi to see this simple CGI script in action.

Notes on zsh

Building my .zshrc using some of the references below.

References

• prompt and colour
• prompt escape sequences

Beginning an Adobe Flex project on Linux

• Download and unzip Flex to $HOME/flex

• Add $HOME/flex/bin to your path

    $ cat ~/.bash_profile | grep flex
    export PATH=$HOME/flex/bin:$PATH;

• Create a new directory for the Flex Project

    $ mkdir hello_flex

• Edit HelloWorld.mxml

    <?xml version="1.0"?>
    <Application xmlns="http://www.adobe.com/2006/mxml">
      <Label text="Hello, world!" />
    </Application>

• Create a build file – build.mxml. Change paths to reflect your own. I haven’t investigated how to use environmental variables in ant scripts, yet.

    <project name="HelloWorld" default="compile">
    <property name="flex.mxmlc" location="/home/pgowda/flex/bin/mxmlc" />
    <property name="dest.dir" value="bin" />
    <property name="media.dir"
        value="/home/pgowda/web_project/htdocs/media" />
    <target name="init">
      <delete dir="${dest.dir}" />
      <mkdir dir="${dest.dir}" />
    </target>
    <target name="compile" depends="init">
      <exec executable="${flex.mxmlc}" failonerror="true">
        <arg line="-output ${dest.dir}/HelloWorld.swf"/>
        <arg line="HelloWorld.mxml"/>
      </exec>
      <copy file="${dest.dir}/HelloWorld.swf"
        tofile="${media.dir}/HelloWorld.swf" />
    </target>
    </project>

• Build the project

    $ ant 

    Buildfile: build.xml

    init:
       [delete] Deleting directory
       /home/pgowda/web_project/flexclient/bin
        [mkdir] Created dir:
       /home/pgowda/web_project/flexclient/bin
  
    compile:
         [exec] Loading configuration file
         /home/pgowda/flex/frameworks/flex-config.xml
         [exec]
         /home/pgowda/web_project/flexclient/bin/HelloWorld.swf
         (175845 bytes)
         [copy] Copying 1 file to
         /home/pgowda/web_project/htdocs/media

    BUILD SUCCESSFUL
    Total time: 6 seconds

On little known web formats.

I’d never paid much attention to the “web archive (.mht)” format used by Internet Explorer, because I do not use IE at all. Yesterday, I was forced to use the dinosaur of a web browser on a bank site, which shall remain nameless except that the programmers of that site would probably give you a “deer in headlights” look if I said anything about palindromes, web standards and cross browser compatibility. Sigh.

So, I saved a completed form in .mht format for later use, on the tiny EEE laptop on which I have windows XP, for it did not have CutePDF installed. Today, Opera on my MacBook opened this file to my relief, but it was to be short lived. Opera kept crashing every time I tried to print the file. I looked around the web for a converter from .mht to a more humane format. I found only paid software.

That’s when I decided to poke into the file itself and see whether I could salvage the document. Surprisingly, .mht file format is defined in RFC 2557. Jacob Palme, one of the proposers of the standard, has a web page explaining the MHTML format.

The format looks simple enough to be parsed using Python’s mimetypes library.

I’ve started a github project — pyMHTML to write a library to parse web archive files.

Hacking on this library is my fallback for the PyCon sprint sessions.