My Meaningless Scribbles ...

GTK Theme is not being followed by all GTK Applications (on LXDE)

2012-01-20T13:05:00.002+11:00

On my LXDE (Ubuntu Oneiric 11.10) setup, I have recently discovered that not all GTK applications are getting themed with the GTK theme (Adwaita) of my preference.

My first obvious attempt was to reset theme with LXDE Menu > Preferences > Customize Look and Feel. I set Adwaita in the "Gadget" tab. That Didn't help. Many "sudo" applications along with some other normal GTK-3.0 applications (e.g. evince) stills stayed un-themed.

Ubuntu repository apparently has two tools to set GTK-2.0 themes:

gtk-chtheme
gtk-theme-switch

Gtk-chtheme is much better in my opinion. So, I gave it a shot and selected "Adwaita" from the list of installed GTK theme. Still no change.

Then as a last resort, I wanted check which GTK theme files are actually being accessed during the initialization of the applications. For that, I started Evince from the console with strace as below:

strace -e trace=access evince 2>&1 | less

After a quick checking through the output it was obvious that my GTK-3.0 themes are not being found by the applications and the theme files are being searched for both in:

/home/USER-NAME/gtk-3.0
/home/USER-NAME/.config/gtk-3.0

On Ubuntu, GTK themes are installed in /usr/share/themes. So, the quick, easy and clean solution to the problem was to create a sym-link to the GTK-3.0 theme of my preference as shown below:

ln -s /usr/share/themes/THEME-NAME/gtk-3.0 /home/USER-NAME/gtk-3.0

ln -s /usr/share/themes/THEME-NAME/gtk-3.0 /home/USER-NAME/.config/gtk-3.0

After creating the sym-links, I restarted Evince and voila! it's perfectly themed with Adwaita now.

P.S.
If you ware selecting another themes other than Adwaita, make sure /usr/share/themes/THEME-NAME/gtk-3.0 actually exists. Not all themes in /usr/share/themes/ are GTK-3.0 compatible and they will miss the gtk-3.0 directory inside the theme folder.

Compile NS-2.34 and 2.35 with GCC-4.6

2011-11-11T15:20:00.001+11:00

I stumbled into the compilation issue after upgrading to Ubuntu Oneiric 11.10. The error occurs in "mac-802_11Ext.h" saying something like the following:

mac/mac-802_11Ext.h: In member function 'u_int32_t PHY_MIBExt::getHdrLen11()':mac/mac-802_11Ext.h:175:19: error: expected primary-expression before 'struct'mac/mac-802_11Ext.h:175:41: error: 'dh_body' was not declared in this scopemac/mac-802_11Ext.h:175:51: error: 'offsetof' was not declared in this scope

I found the solution in Ubuntu's patched version of NS-2.35-RC6. The direct link to the patch is here: http://launchpadlibrarian.net/74445224/ns2_2.35~RC6-4_2.35~RC6-4ubuntu1.diff.gz

It's really simple. All you have to do is to add #include cstddef as shown below:

#ifndef ns_mac_80211Ext_h
#define ns_mac_80211Ext_h
#include cstddef
#include "marshall.h"
#include "timer-handler.h"

Now, NS-2.34 should compile fine with GCC-4.6. Enjoy your simulation!

Google Chrome (Chromium) font anti-aliasing is wrong on LXDE (L)Ubuntu Oneiric (11.10)

2011-10-21T17:09:00.000+11:00

The solution was to add the following in the ~/.fonts.conf

https://gist.github.com/1318856

xml version="1.0"?>DOCTYPE fontconfig SYSTEM "fonts.dtd">
<fontconfig>
 <match target="font">
  <edit mode="assign" name="rgba">
   <const>noneconst>
  edit>
 match>
 <match target="font">
  <edit mode="assign" name="hinting">
   <bool>truebool>
  edit>
 match>
 <match target="font">
  <edit mode="assign" name="hintstyle">
   <const>hintslightconst>
  edit>
 match>
 <match target="font">
  <edit mode="assign" name="antialias">
   <bool>truebool>
  edit>
 match>
 <match target="font">
   <edit mode="assign" name="lcdfilter">
    <const>lcddefaultconst>
   <edit>
 <match>
</fontconfig>

One may have to play around with the "hintstyle" to get it right. The available options are:

hintslight
hintmedium
hintfull
none

Ubuntu Oneiric (11.10) DPKG error removing old Linux kernel

2011-10-19T14:10:00.001+11:00

I stumbled into another trouble removing my obsolete packages after upgrading into Ubuntu Oneiric 11.10 with Linux kernel 3.00-12. The problem is with removing the old 64bit kernel 2.6.35-24. The Debian package name is 2.6.35-24. The error I got is following:

[13:41:34] user@machin:[/var/lib/dpkg/info]$ sudo apt-get -f install
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following packages will be REMOVED:
linux-image-2.6.35-24-generic
0 upgraded, 0 newly installed, 1 to remove and 31 not upgraded.
1 not fully installed or removed.
After this operation, 139 MB disk space will be freed.
Do you want to continue [Y/n]?
(Reading database ... 248250 files and directories currently installed.)
Removing linux-image-2.6.35-24-generic ...
Examining /etc/kernel/postrm.d .
run-parts: executing /etc/kernel/postrm.d/initramfs-tools 2.6.35-24-generic /boot/vmlinuz-2.6.35-24-generic
update-initramfs: Deleting /boot/initrd.img-2.6.35-24-generic
run-parts: executing /etc/kernel/postrm.d/zz-extlinux 2.6.35-24-generic /boot/vmlinuz-2.6.35-24-generic
P: Checking for EXTLINUX directory... found.
P: Writing config for /boot/vmlinuz-3.0.0-12-generic...
E: /usr/share/syslinux/debian/extlinux: No such file or directory
run-parts: /etc/kernel/postrm.d/zz-extlinux exited with return code 1
Failed to process /etc/kernel/postrm.d at /var/lib/dpkg/info/linux-image-2.6.35-24-generic.postrm line 328.
dpkg: error processing linux-image-2.6.35-24-generic (--remove):
subprocess installed post-removal script returned error exit status 1
Errors were encountered while processing:
linux-image-2.6.35-24-generic
E: Sub-process /usr/bin/dpkg returned an error code (1)

The highlighted lines told me the error is caused by the postrm script called zz-extlinux located in /etc/kernel/postrm.d/. I don't know how I managed to have extlinux installed on my system. If I am not mistaken booting is still managed by grub-2. So, I decided to remove both extlinux and linux-image-2.6.35-24-generic. But, since my DPKG is already going crazy removing linux-image-2.6.35-24-generic and failing on zz-extlinux, I have to remove the old Linux image first. I quickly patched the zz-extlinux script with an "exit 0" at top followed by an "sudo apt-get -f install". One could also make it non-executable as following as a solution, followe:

sudo chmod -x /etc/kernel/postrm.d/zz-extlinux; sudo apt-get -f install

This solved my problem removing the Linux image. Now, I guess one can also remove extlinux considering Grub-2 is the bootloader.

sudo apt-get remove --purge extlinux

Changing Gnome-Screensaver Background on Lubuntu 11.10 (Oneiric)

2011-10-18T17:18:00.002+11:00

I had to find a way on Lubuntu flavoured LXDE to lock/unlock the current running desktop session. The obvious easy way is to use xscreensaver and/or gnome-screensaver and do one of the following:

xscreensaver-command -lock
gnome-screensaver-command --lock

I chose to keep with gnome-screensaver since it is prettier. But immediate annoyance was after locking screen, gnome-screensaver would not retain the same background as current Lubuntu session. After trying so many options (gconf, editing files in /etc), finally I figured - the solution is rather easy.

It appears Lubutu session keeps a separate sets of options from LXDE. When gnome-screensaver is launched, it takes the background option of LXDE (not Lubuntu). To match same background both on LXDE and Lubuntu, which also matches gnome-screensaver background inside Lubuntu session, the steps are below:

Set preferred desktop background on running Lubuntu session by right-click > Desktop Preferences > Background.
Log-out from Lubuntu session.
On the log-in prompt, select LXDE as desktop session to log into.
Once inside LXDE, you'll see the desktop background that gnome-screensaver uses as background.
From the right-click menu, change the background to match your preference.

That's all !!! Now, you can log in back into Lubuntu and lock screen by initiating gnome-screensaver (I use a Compiz command shortcut with Ctrl+Alt+L which triggers "gnome-screensaver-command --lock") and desktop background will match.

Ubuntu Oneiric 11.10 - problem with package dictionaries-common

2011-10-18T15:33:00.000+11:00

I started getting troubles with dictionaries-common package right from the beginning of upgrading into Ubuntu Oneiric (11.10). The error message looks kind of like the following:

Setting up dictionaries-common (1.11.5ubuntu1) ...
update-default-wordlist: Question empty but elements installed for class "wordlist"
  dictionaries-common/default-wordlist: return code: "0", value: ""
  Choices: , Manual symlink setting
  shared/packages-wordlist: return code: "10" owners/error: "shared/packages-wordlist doesn't exist"
  Installed elements: american (American English)
  Please see "/usr/share/doc/dictionaries-common/README.problems", section
  "Debconf database corruption" for recovery info.
update-default-wordlist: Selected wordlist "" 
does not correspond to any installed package in the system
and no alternative wordlist could be selected.
dpkg: error processing dictionaries-common (--configure):
 subprocess installed post-installation script returned error exit status 255
dpkg: dependency problems prevent configuration of xscreensaver-data-extra:
 xscreensaver-data-extra depends on dictionaries-common; however:
  Package dictionaries-common is not configured yet.
dpkg: error processing xscreensaver-data-extra (--configure):
 dependency problems - leaving unconfigured
No apport report written because the error message indicates its a followup error from a previous failure.
Errors were encountered while processing:
 dictionaries-common
 xscreensaver-data-extra
E: Sub-process /usr/bin/dpkg returned an error code (1)

May be I was too quick to have skimmed over the error message and did not realize what is missing. It looks like the problem actually is that I do not have any word-list installed (or, somehow uninstalled during the distribution upgrade process) on my machine for the dictionary to point to. The quick solution is to install the preferred word-list as following:

sudo apt-get install wamerican-insane

This should remove fix the package installation error with dictionaries-common package.

Compiz Config Settings Manager (CCSM) crashes with SIGSEGV on (L)ubuntu Oneiric (11.10)

2011-10-18T14:16:00.001+11:00

On my latest Lubuntu 11.10 installation, CCSM stumbles into a segmentation violation (SIGSEGV) if "Desktop Cube" is enabled in Compiz plugins and I try to disable it through CCSM. Since, I managed to convince myself that "Expo" is capable of providing what I really need, I needed a way to disable it without using CCSM. So, for my own future reference and someone who's going through the same kind of nagging crash, the Compiz config file is located here:

/home/$USER/.config/compiz-1/compizconfig/Default.ini

My list of enabled plugins are below:

[core]
s0_active_plugins = core;detection;gtkloader;composite;opengl;copytex;decor;mousepoll;commands;imgjpeg;regex;addhelper;annotate;resize;grid;compiztoolbox;place;imgsvg;imgpng;notification;dbus;gnomecompat;resizeinfo;opacify;wobbly;move;text;loginout;scale;animation;kdecompat;thumbnail;workarounds;scaleaddon;scalefilter;animationaddon;expo;ezoom;switcher;

All one has to do is remove "cube" (or whichever is causing trouble) from "s0_active_plugins".

LightDM Fails to Load with (L)ubuntu Oneiric (11.10)

2011-10-18T11:37:00.001+11:00

After my migration from default Ubuntu with Unity and Gnome3 to Lubuntu with LXDE and OpenBox, I realized I have managed to sacrifice beautiful LightDM, the new default login manager on 11.10 and instead, running not so pretty LXDM.

After realizing that LightDM is missing the first thing to do was obviously:

sudo apt-get install lightdm

Followed by:

sudo dpkg-reconfigure lightdm
sudo service lxde stop
sudo service lightdm start

Not much to my surprise on 11.10, LightDM didn't start. After a little digging through the log files located at /var/log/lightdm, I came across the following:

"lightdm can not write to /var/lib/lightdm/.Xauthority"

Time to check /var/lib/lightdm directory:

[11:13:26] user@machine:[/var/lib/lightdm]$ sudo ls -alh
total 44K
drwxr-x--- 9 109 130 4.0K 2011-10-14 20:08 .
drwxr-xr-x 74 root root 4.0K 2011-10-17 16:56 ..
drwx------ 4 119 136 4.0K 2011-10-14 15:28 .cache
drwx------ 4 119 136 4.0K 2011-10-14 15:28 .config
drwx------ 3 119 136 4.0K 2011-10-14 15:28 .dbus
-rw------- 1 119 136   16 2011-10-14 15:32 .esd_auth
drwx------ 2 119 136 4.0K 2011-10-14 18:29 .gconf
-rw------- 1 119 136    0 2011-10-14 19:55 .goutputstream-54JJ3V
drwx------ 2 119 136 4.0K 2011-10-14 15:28 .gvfs
drwxrwxr-x 3 119 136 4.0K 2011-10-14 15:32 .local
drwx------ 2 119 136 4.0K 2011-10-14 16:22 .pulse
-rw------- 1 119 136 256 2011-10-14 15:32 .pulse-cookie
-rw------- 1 119 136    0 2011-10-14 20:08 .Xauthority

After checking /etc/passwd and /etc/group, it confirmed that somehow user-id 119 and groupd-id 136 has gone missing from my system. So, as a quick roll-back, I did the following:

sudo apt-get remove --purge lightdm

sudo mv /var/lib/lightdm /var/lib/lightdm.bak

sudo apt-get install lightdm

sudo service lxdm stop

sudo apt-get remove --purge lxdm

sudo service lightdm start

Voila, LightDM showed up with a nice log-in prompt. Re-installing LightDM, added a new user and group with ids 109 and 130 respectively. Also, the ls output in /var/lib/lightdm looks correct.

[11:33:08] user@machine:[/var/log/lightdm]$ sudo ls -al /var/lib/lightdm/
total 44
drwxr-x--- 9 lightdm lightdm 4096 2011-10-18 11:14 .
drwxr-xr-x 74 root    root    4096 2011-10-18 11:15 ..
drwx------ 4 lightdm lightdm 4096 2011-10-18 11:14 .cache
drwx------ 3 lightdm lightdm 4096 2011-10-18 11:14 .config
drwx------ 3 lightdm lightdm 4096 2011-10-18 11:14 .dbus
-rw------- 1 lightdm lightdm   16 2011-10-18 11:14 .esd_auth
drwx------ 5 lightdm lightdm 4096 2011-10-18 11:15 .gconf
d????????? ? ?       ?          ?                ? .gvfs
drwxrwxr-x 3 lightdm lightdm 4096 2011-10-18 11:14 .local
drwx------ 2 lightdm lightdm 4096 2011-10-18 11:14 .pulse
-rw------- 1 lightdm lightdm 256 2011-10-18 11:14 .pulse-cookie
-rw------- 1 lightdm lightdm   49 2011-10-18 11:14 .Xauthority

Edit (20 Jan 2012):
I stumbled into another problem with LightDM after uninstalling gnome-settings-daemon. It looks like the background file to set for LightDM is requested through gnome-settings-daemon for some reason on my system. So, if you have the same problem, a quick apt-get install gnome-settings-daemon will solve the problem.

Removing "RC" packages on Debian/Ubuntu

2011-09-27T13:24:00.001+10:00

If you do "dpkg -l", sometimes you will get to see a long list of packages with the following details:

rc package-name 0.4-3 package-details

Note the "rc" status at the beginning of the package description. Some packages will show up in the "dpkg -l" output even after removing them. The reason behind this is - packages with "rc" status indicate that the packages have been removed but some configuration files for those packages are left on the system.

To remove these packages, do the following:

sudo dpkg --purge package-name

For a list of packages, run the following:

dpkg -l | grep ^rc | cut -d ' ' -f3 | xargs sudo dpkg --purge

Forcing Vim's syntax highlighting (for odd file extensions)

2011-09-02T16:58:00.001+10:00

I often tend to keep backup copies of my C/C++ source files with odd extensions (i.e. something that does not end with CC or C). If these files are opened with Vim, they wouldn't be recognized as they should, which means my CTAGS extension and colored syntax highlighting would both refrain to work. The quick solution to this problem is to tell Vim explicitly about the file type. Here is how one can do it:

:setfiletype c

More information on it, as always:

:help setfiletype

Remove Firefox add-ons installed by 3rd Party Applications

2011-08-24T12:27:00.003+10:00

The list of add-ons with pointers to the XPI code is here:

HKEY_CURRENT_USER\Software\Mozilla\Firefox\Extensions.

Just remove the unwanted entry from the list. Cheers :).

Multiple Java Console add-on in Firefox

2011-08-23T17:57:00.002+10:00

I recently discovered more than one Java console add-on inside my add-ons manager in my Windows 7 machine. Since I only had only one version of JRE (Java Runtime Environment) updated to the latest version, this did not make sense. To make it more of an eye-sore, these add-ons did not have an associated "Remove" button in the add-on manager.

Some Google-ing revealed these are Java update left-overs. They do not have the "Remove" button because they are installed inside the %ProgramFiles(x86)%\Mozilla Firefox and not inside the %AppData%\Mozilla\Firefox profile directory.

In my installation, I had the following directories inside %ProgramFiles(x86)%\Mozilla Firefox\extensions directory:

{CAFEEFAC-0016-0000-0026-ABCDEFFEDCBA}
{CAFEEFAC-0016-0000-0023-ABCDEFFEDCBA}
{CAFEEFAC-0016-0000-0017-ABCDEFFEDCBA}

The blue numbers are indicative to the Java versions. To remove older left-overs, just keep the folder with associated with the latest Java version, which usually is the highest number in the list and delete the others.

Evince would not start over SSH X-forwarding

2011-07-14T15:19:00.003+10:00

Evince, which is the primary PDF and PostScript viewer on Ubuntu, recently stopped working on my Ubuntu/Natty (11.04) over SSH X-forwarding. It works flawlessly inside 'gnome-session'. But, would not run as a regular application (e.g. xpdf or gv) over X-forwarded SSH.

After digging up a little bit, I found a lots of these message in my dmesg output:

[1305902.871825] type=1400 audit(1310620543.587:75): apparmor="STATUS" operation="profile_remove" name="/usr/bin/evince" pid=20166 comm="apparmor

It was obvious that apparmor [1] was blocking Evince to run outside Gnome session over SSH. A solution to disable the AppArmor block on Evince over SSH connection was suggested here [2]:


#inside the gnome-terminal
sudo ln -s /etc/apparmor.d/usr.bin.evince /etc/apparmor.d/disable
sudo /etc/init.d/apparmor restart

And voila! Now Evince can again be launched over SSH.

[1] http://en.wikipedia.org/wiki/AppArmour
[2] http://ubuntuforums.org/showthread.php?t=1472642

No need to multi-thread if one dones't have to ...

2011-01-27T13:53:00.011+11:00

I like spend time writing, re-writing, reversing and hacking into codes. I consider myself a learner. Regarding multi-threaded codes, I keep a simple belief that almost all of them could be written using a single thread doing effectively the same thing. I realize the world isn't going to change because of what I believe. I wrote the following little piece to demonstrate a friend of mine that a multi-threaded version of one of his little projects which has a functions that blocks and responsible for a few other things (eg. write_db(), on_demand() ) could also have been written using a single-thread. For whatever it is worth, here is my little demo code:


#include < stdio.h >
#include < sys/time.h >
#include < signal.h >
#include < setjmp.h >

typedef void (*sighandler_t) (int);
sigjmp_buf saved_state;

struct data 
{
    unsigned int count;
    unsigned int epoch;
    double watt;
} d;

/* 
 * This value (in seconds) defines how often an 
 * attempt is made to read the serial port.
 */
unsigned int sleep_time = 10;
unsigned int sleep_start = 0;   // in epoch seconds

/* 
 * Function to write into SQLite/RRD database.
 */
int write_db()
{
    printf("[*] write_db()\n");
    return 0;
}

/* 
 * Function to serve on-demand query.
 */
void on_demand()
{
    printf("[*] on_demand()     :: ");
    printf("count %u now %u watt %.2f\n", d.count, d.epoch, d.watt);
    fflush(stdout);
    siglongjmp(saved_state, 1);
}

/*
 * Function to read from serial port.
 */
int blocking_read()
{
    struct timeval timeout;

    printf("[*] blocking_read() :: ");

    /* simulate the read() blocking-time */
    timeout.tv_sec = 0;
    timeout.tv_usec = 500;

    select(0, NULL, NULL, NULL, &timeout);
    d.count = d.count + 1;
    d.epoch = time(NULL);
    d.watt = random();

    printf("count %u now %u watt %.2f\n", d.count, d.epoch, d.watt);
    fflush(stdout);
    return 0;
}

int main()
{
    printf("[*] main() -- pid: %u\n\n", getpid());
    fflush(stdout);

    signal(SIGUSR1, (sighandler_t) on_demand);

    while (1) 
    {
        /* 
         * If we have just came out from a 
         * signal, perform the pending delay.
         */
        sigsetjmp(saved_state, 1);
        if ((int)(sleep_start + sleep_time - time(NULL)) > 0) {
            printf("[-] pending delay %u secs\n",
                   (int)(sleep_start + sleep_time - time(NULL)));
            fflush(stdout);
            sleep((sleep_start + sleep_time) - time(NULL));
        }

        blocking_read();
        write_db();
        printf("\n");
        fflush(stdout);

        /* 
         * collect serial port data at 
         * each 'sleep_time' seconds.
         */
        sleep_start = time(NULL);
        sleep(sleep_time);
    }
    return 0;
}

Keyboard num-keys not working in VirtualBox Ubuntu guest

2011-01-14T16:16:00.004+11:00

The solution that works for me is by disabling 'Mouse Keys'.

Go to System > Preferences > Keyboard.
In Mouse Keys tab, un-tick 'Pointer can be controlled using keypad'

Captcha Prompter

2010-11-25T16:24:00.009+11:00

latest version is available here: http://gsbabil.googlecode.com/svn/trunk/prompt_captcha.py
This tool is written using PyGTK2. It takes an image file name as input, shows the image to the user and asks him to type the text shown in the image. The goal is to use it inside the Unix scripts. For example, one could do 'CAPTCHA_TEXT=$(python prompt_captcha.py captcha.jpg)' in Bash and perform further actions with $CAPTCHA_TEXT. This little tool will accept png, jpg, gif and most other commonimage formats.
usage: python prompt_captcha.py captcha_image

#!/usr/bin/env python
# -*- coding: utf-8 -*-

"""
prompts user to solve captcha image.

This tool is written using PyGTK2. It takes an image file name as input, shows the image to the user and asks him to type the text shown in the image. The goal is to use it inside the Unix scripts. For example, one could do 'CAPTCHA_TEXT=$(python prompt_captcha.py captcha.jpg)' in Bash and perform further actions with $CAPTCHA_TEXT. This little tool will accept png, jpg, gif and most other common image formats.

[*] usage: python prompt_captcha.py captcha_image
"""

__author__ =  'Babil (Golam Sarwar)'
__version__=  '0.1'

import sys
import time
import Image
import gtk
import pygtk


pygtk.require('2.0')

class getCaptcha(object):
    """performs all the GTK2 gui actions"""

    def enter_callback(self, widget, textbox):
        """callback funcion for when user presses the enter key"""
 textbox_text = textbox.get_text()
        print textbox_text
 gtk.main_quit()

    def __init__(self, captcha=None):
        #
 if captcha == None:
  print >> sys.stderr, "[*] usage: python prompt_captcha.py captcha_image"
  sys.exit(-1)

 width, height = Image.open(captcha).size
 
 window = gtk.Window(gtk.WINDOW_TOPLEVEL)
 window.set_geometry_hints(window,width+15,height+15,width+30,height+30)
        window.set_size_request(width, height)
 window.resize(width+15, height+15)
        window.set_title("GTK textbox")
        window.connect("delete_event", lambda w,e: gtk.main_quit())
 window.set_title("Enter Captcha")
 window.set_keep_above(True)
 window.set_urgency_hint(True)
 window.move(gtk.gdk.screen_width()/2 - (width+15)/2, gtk.gdk.screen_height()/2 - (height+15)/2)


        vbox = gtk.VBox(False, 0)
        window.add(vbox)

 image = gtk.Image()
 image.set_from_file(captcha)

 button = gtk.Button()
 button.set_image(image)
 button.set_relief(gtk.RELIEF_HALF)
 #button.set_tooltip_text("Type captcha then press enter \n or click the image.")
        vbox.pack_start(button, True, True, 0)

 textbox = gtk.Entry()
        textbox.set_max_length(50)
        textbox.connect("activate", self.enter_callback, textbox)
 textbox.set_text("Enter captcha & then click on the image :-) ")
        textbox.select_region(0, len(textbox.get_text()))
 vbox.pack_start(textbox, True, True, 10)


 button.connect("clicked", self.enter_callback, textbox)
 window.show()
        vbox.show()
 button.show()
 image.show()
 textbox.show()
 textbox.grab_focus()



def main():
    gtk.main()
    return 0

if __name__ == "__main__":
 if len(sys.argv) >=2: 
  getCaptcha(sys.argv[1])
 else: 
  getCaptcha()
 main()

Download Fileserve links from shell

2010-11-22T12:03:00.008+11:00

The script assumes, you have

ImageMagick
Curl
Common Unix utilities (eg. sed, awk, cut etc.)

... are installed in the path. The captcha is shown using 'display' component from ImageMagick. You need X forwarding if you are running the script over ssh.

Download source here: http://gsbabil.googlecode.com/svn-history/r8/trunk/fileserve.sh


#!/bin/bash
#source /home/babil/Desktop/recaptcha/plowshare-0.9.4/src/lib.sh

PUB_URL='http://www.fileserve.com/file/zXvrJS2'
FILESERVE_RECAPTCHA_PUBKEY='6LdSvrkSAAAAAOIwNj-IY-Q-p90hQrLinRIpZBPi'
IMAGE_VIEWER="display"

CAPTCHA_IMAGE=$(mktemp) #'captcha.jpg'
COOKIE=$(mktemp) #'tmp_cookie.txt'
HTML=$(mktemp) #'tmp_fileserve.htm'

AGENT='Mozilla/5.0 (X11; U; Linux x86_64; en-AU; rv:1.9.2.12) Gecko/20101027 Ubuntu/10.10 (maverick) Firefox/3.6.12'
CUSTOM_HEADERS='-H "Accept: text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8" -H "Accept-Language: en-us,en;q=0.5" -H "Accept-Encoding: gzip,deflate" -H "Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7" -H "Keep-Alive: 115" -H "Connection: keep-alive"'
XML_HTTP_REQUEST='-H "X-Requested-With: XMLHTTPRequested"'

RED='\e[0;31m'
CYAN='\e[0;36m'
NC='\e[0m' # No Color


function regex_replace() {
    RE='s/\(\/\|\\\|&\)/\\&/g'
    FROM=$(sed -e "$RE" <<< "$1")
    TO=$(sed -e "$RE" <<< "$2")
    sed -e "s/$FROM/$TO/g"
}

# Delete leading and trailing spaces, tabs, \r, ...
function strip_whiteSpace() {
    echo "$1" | sed "s/^[[:space:]]*//; s/[[:space:]]*$//"
}

# Return uppercase string
function uppercase() {
    tr '[a-z]' '[A-Z]'
}

function clean_up()
{
 rm -f $COOKIE
 rm -f $CAPTCHA_IMAGE
 rm -f $HTML
}

function find_fileName()
{
 
 if [ -z $FILE_NAME ]
 then
  FILE_NAME=$(cat $HTML | grep 'Forum (BBCode)' | sed -n 's/.*File name: \(.*\) File size.*/\1/p')
 fi
 
 if [ -z $FILE_NAME ]
 then
  FILE_NAME=$(cat $HTML | grep 'Forum (BBCode)' | sed -n 's/.*url=\(.*\)\]\[b\].*/\1/p')
 fi
 
 if [ -z $FILE_NAME ]
 then
  FILE_NAME=$(cat $HTML | grep 'Website (HTML)' | sed -n 's/.*href=.\(.*\).>.b.File name.*/\1/p')
 fi

 if [ -z $FILE_NAME ]
 then
  FILE_NAME=$(cat $HTML | grep 'Text link' | sed -n 's/.*value="\(.*\)".*/\1/p')
 fi

 FILE_NAME=$(strip_whiteSpace $FILE_NAME)

}

function delay()
{
 for((i=$1;i>=1;i--))
 do
  echo -n "$i "
  sleep 1s
 done
 echo
}

function check_waitTime()
{
 WAIT_TIME=$(cat $HTML | sed -n 's/.*You need to wait \(.*\) seconds to start another download.*/\1/p')
 if [ -z $WAIT_TIME ]
 then
  WAIT_TIME=3
 fi
 WAIT_TIME=$(strip_whiteSpace $WAIT_TIME)
 echo $WAIT_TIME
}

function solve_recaptcha()
{
 CAPTCHA_URL='http://www.google.com/recaptcha/api/challenge?k='$FILESERVE_RECAPTCHA_PUBKEY
 echo -e "$RED [*] captcha url: $NC $CAPTCHA_URL "

 CHALLENGE=$(wget -q --referer $PUB_URL "$CAPTCHA_URL" -O - |  grep challenge | sed -n "s/.*'\(.*\)',/\1/p")
 echo -e "$RED [*] captcha challenge: $NC $CHALLENGE"
 wget -q --referer $PUB_URL http://www.google.com/recaptcha/api/image?c=$CHALLENGE -O $CAPTCHA_IMAGE 
 
 #convert -trim $CAPTCHA_IMAGE $CAPTCHA_IMAGE
 #$IMAGE_VIEWER "$CAPTCHA_IMAGE" &
 #pid=$!
 #read -p "[*] enter captcha response: " RESPONSE
 #RESPONSE=$(echo $RESPONSE | sed 's/ /+/g')
 #disown $(kill -9 $pid) 2>&1 1>/dev/null

 echo -e -n "$RED [*] captcha input : $NC"
 RESPONSE=$(prompt_captcha.py $CAPTCHA_IMAGE | sed 's/ /+/g')
 echo "$RESPONSE"

 RESULT=$(curl -s --user-agent "$AGENT" --cookie-jar $COOKIE --referer $PUB_URL http://www.fileserve.com/checkReCaptcha.php --data "recaptcha_challenge_field=$CHALLENGE&recaptcha_response_field=$RESPONSE&recaptcha_shortencode_field=$SHORTENCODE")
}


### - 0
if [ -z "$1" ]
then
 echo
 echo "[*] usage: $(basename $0) $PUB_URL"
 echo
 exit -1
else
 PUB_URL="$(strip_whiteSpace $1 | sed 's#//fileserve.com#//www.fileserve.com#g')"
fi
echo -e "$RED [*] attempting to download: $CYAN $PUB_URL $NC"


### - 1
clean_up

### - 2
SHORTENCODE=$(echo $PUB_URL | sed -n 's/.*file\/\(.*\)/\1/p' | cut -d '/' -f 1)
SHORTENCODE=$(strip_whiteSpace $SHORTENCODE)

### - 3 save cookie
curl -s --user-agent "$AGENT" --cookie-jar $COOKIE --referer $PUB_URL $PUB_URL --data 'checkTimeLimit=check' --output $HTML

### - 4 immitating "slow download".click()
curl -s --user-agent "$AGENT" --cookie $COOKIE --referer $PUB_URL $PUB_URL --data "recaptcha_shortencode_field=$SHORTENCODE" --output $HTML

### - 5 
if [ -z "$2" ]
then
 FILE_NAME="$(strip_whiteSpace $2)"
else
 FILE_NAME=""
fi
find_fileName
if [ -z $FILE_NAME ]
then
 echo "[*] can not find file-name, exiting ..."
 exit -1
else
 echo -e "$RED [*] file-name found: $NC $FILE_NAME"
fi

### - 6 solve captcha
RESULT="(null)"
while [ $RESULT != "success" ]
do
 solve_recaptcha
 RESULT=$(echo $RESULT | sed -n 's/{.*"\(.*\)".*/\1/p')
done
echo -e "$RED [*] captcha solved :)$NC"

### - 7 free-user delay
WAIT_TIME=$(curl -s --user-agent "$AGENT" --cookie $COOKIE --referer $PUB_URL $PUB_URL --data "downloadLink=wait" --output -)
echo -e "$RED [*] free-user delay: $NC $WAIT_TIME"
WAIT_TIME=$(echo $WAIT_TIME | sed 's/[^0-9]//g' | bc)
delay $WAIT_TIME
$(curl -s --user-agent "$AGENT" --cookie $COOKIE --referer $PUB_URL $PUB_URL --data "downloadLink=show" --output /dev/null)


### - 8 final download ... should give a 302 redirection
echo -e "$RED [*] Starting download ...$NC"
REDIR=$(curl --user-agent "$AGENT" --cookie $COOKIE --referer $PUB_URL $PUB_URL --data "download=normal" --location --location-trusted --output $FILE_NAME --write-out 'REDIRECT-URL:[%{url_effective}]' | sed -n 's/.*:\[\(.*\)\]/\1/p')

if [[ ! "$REDIR" =~ "/dl/" ]]
then
 echo -e "$RED [*] incorrect redirection; delay then restart ...$NC"
 mv -f $FILE_NAME $HTML
 WAIT_TIME=$(check_waitTime)
 delay $WAIT_TIME
 $0 $PUB_URL
else
 echo -e "$RED [*] done :D. $NC"
 clean_up
 exit
fi

IDLE not working in ActivePython

2010-10-19T15:14:00.000+11:00

The problem is that "idle.bat" is missing some double-quotes. If you have installed Python in "C:\Programs files\Python27" instead of "C:\Python27\", "idle.bat" will fail to execute "idle.pyw".

Copy and paste the follow 2 lines in "C:\Program Files\Python27\Lib\idlelib\idle.bat" (or wherever "idle.bat" is located) to fix it:

@echo off
rem Start IDLE using the appropriate Python interpreter
set CURRDIR=%~dp0
start "%CURRDIR%..\..\pythonw.exe" "%CURRDIR%idle.pyw" %1 %2 %3 %4 %5 %6 %7 %8 %9

"Unlocker" not working on Windows 7 ?? Try "LockHunter"

2010-10-18T16:35:00.001+11:00

Unlocker by http://ccollomb.free.fr/unlocker/ was a nice tool to find, unlock, delete files when Windows explorer can't delete them (as they are locked by other processes).

After my upgrade to Windows 7 - 64bit, Unlocker tool stopped working. Even the 64bit from author's official site doesn't work at all (no window shows up after running the application).

So, for whatever it is worth and whoever is looking for an alternative, LockHunder by http://lockhunter.com is a great alternative, works much like Unlocker and is working without any hiccup (so far) on my Windows 7.

MikTex 2.8 Downloading/Missing "ifxetex.sty" On Every Run

2010-10-14T20:49:00.001+11:00

I debugged this problem after seeing my LyX on my Windows 7 machine taking so long to finish. I saw "latex" process was taking too long to finish. The I ran "latex" manually and the problem was obvious - "ifxetex" is being downloaded in every run.

This seems to be a bug in MikTex 2.8. The bug report can be found here: http://sourceforge.net/tracker/?func=detail&aid=3067362&group_id=10783&atid=110783

The solution is:

Copy ...
from: C:\Program Files (x86)\MikTeX 2.8\tex\generic\ifxetex\ifxetex.sy
to: C:\Program Files (x86)\MikTeX 2.8\tex\latex\ifxetex\ifxetex.sty
From Start-Menu > MikTex 2.8 > Maintenance (Admin), run "Update (Admin)" and update all pending update-able packages.
Start-Menu > MikTex 2.8 > Maintenance (Admin), run "Settings (Admin)". and Then from "General" tab, click "Refresh FNDB" and "Update Formats"

MikTex should not download "ifxetex.sty" in every run now ;-).

Post to PasteBin from Command-Line

2010-10-08T20:04:00.005+11:00

Full credit for "urlencode" script goes to the author.

I modified original "urlencode" script slightly to meet my

needs. Original script can be downloaded at:

http://www.shelldorado.com/scripts/cmds/urlencode

============ 

pastebin.sh
============

#!/bin/bash
#

name="your_name"
format="text"
expire_date="N"
private="0"


if [ -z $1 ]; then
 echo
 echo "[*] usage:  `basename $0`  filename"
 echo
 exit
fi


#DATA=`cat "$@" | uuencode -m -`
DATA=`cat "$@" | urlencode.sh -l`

curl -s -d paste_code="$DATA" -d paste_name="$name" -d paste_format="$format" -d
paste_expire_date="$expire_date" -d paste_private=$private --include http:
//pastebin.com/api_public.php | grep 'http://pastebin.com/'

============ 

urlencode.sh
============

##########################################################################
# Title      : urlencode - encode URL data
# Author     : Heiner Steven (heiner.steven@odn.de)
# Date       : 2000-03-15
# Requires   : awk
# Categories : File Conversion, WWW, CGI
# SCCS-Id.   : @(#) urlencode 1.4 06/10/29
##########################################################################

PN=`basename "$0"`   # Program name
VER='1.4'

: ${AWK=awk}

Usage () {
    echo >&2 "$PN - encode URL data, $VER
usage: $PN [-l] [file ...]
    -l:  encode line endings (result will be one line of output)

The default is to encode each input line on its own."
    exit 1
}

Msg () {
    for MsgLine
    do echo "$PN: $MsgLine" >&2
    done
}

Fatal () { Msg "$@"; exit 1; }

set -- `getopt hl "$@" 2>/dev/null` || Usage
[ $# -lt 1 ] && Usage   # "getopt" detected an error

EncodeEOL=no
while [ $# -gt 0 ]
do
    case "$1" in
     -l) EncodeEOL=yes;;
 --) shift; break;;
 -h) Usage;;
 -*) Usage;;
 *) break;;   # First file name
    esac
    shift
done

LANG=C export LANG
$AWK '
    BEGIN {
 # We assume an awk implementation that is just plain dumb.
 # We will convert an character to its ASCII value with the
 # table ord[], and produce two-digit hexadecimal output
 # without the printf("%02X") feature.

 EOL = "%0D%0A"  # "end of line" string (encoded)
 split ("1 2 3 4 5 6 7 8 9 A B C D E F", hextab, " ")
 hextab [0] = 0
 for ( i=1; i<=255; ++i ) ord [ sprintf ("%c", i) "" ] = i + 0
 if ("'"$EncodeEOL"'" == "yes") EncodeEOL = 1; else EncodeEOL = 0
    }
    {
 encoded = ""
 for ( i=1; i<=length ($0); ++i ) {
     c = substr ($0, i, 1)
     if ( c ~ /[a-zA-Z0-9.-]/ ) {
  encoded = encoded c  # safe character
     } else if ( c == " " ) {
  encoded = encoded "+" # special handling
     } else {
  # unsafe character, encode it as a two-digit hex-number
  lo = ord [c] % 16
  hi = int (ord [c] / 16);
  encoded = encoded "%" hextab [hi] hextab [lo]
     }
 }
 if ( EncodeEOL ) {
     printf ("%s", encoded EOL)
 } else {
     print encoded
 }
    }
    END {
     #if ( EncodeEOL ) print ""
    }
' $@

Executing commands in shell and accepting the output in Vim

2010-09-27T12:10:00.000+10:00

Source: http://www.linux.com/archive/feed/57727

:r ! ls

It is also possible to select certain lines in Vim with v, V or ctrl+v and send it to a Bash command. The selected lines (input for Bash command) will be replaced by the output from Bash. Typical line number specification is also allowed.

:5,7 ! xargs ls -alh

Forcing Floating Point Calculation in Python

2010-09-24T16:10:00.000+10:00

Sources:

http://docs.python.org/tutorial/floatingpoint.html
http://stackoverflow.com/questions/1267869/how-can-i-force-division-to-be-floating-point-in-python

Method-1:Just write any of the parameters in the division operation in floating-point format (eg. 4.0, 3.0)

>>> 4.0 / 3
or, >>> 4 / 3.0

Method-2:

from __future__ import devision

Method-3:

>>> from decimal import Decimal
>>> Decimal(2.675)
Decimal('2.67499999999999982236431605997495353221893310546875')

Latex: Incorrect references for Tables, Figures etc.

2010-09-21T12:32:00.000+10:00

Just put the \label{} inside \caption{}.
See the following example.

\begin{table}[!ht]
\begin{center}
\caption{ \label{tab:blah} Blah Caption}
\begin{tabular}{c|c|c|c||c} A1 & B1 & C1 & D1 & E1 & F1 & G1 & & H1
\hline

aa &bb &cc &dd &ff \\

aa &bb &cc &dd &ff \\

\end{tabular}
\end{center}
\end{table}

Installing DMG or PKG from command-line on Mac, Apple TV

2010-09-20T23:05:00.002+10:00

mounting dmg

hdid package.dmg
or, hdiutil attach package.dmg
or, hdiutil mount package.dmg

change directory

cd /Volumes/package/

install pkg

sudo installer -verbose -pkg package.pkg -target /
or, sudo /usr/sbin/installer -verbose -pkg package.pkg -target /

unmounting dmg

hdiutil detach /Volumes/package
or, hdiutil detach /Volumes/package -force

Install Windows from portable USB drive

2010-09-08T16:16:00.001+10:00

This Windows tool [1] can apparently convert a Windows CD/ISO image (and also other Live CD/DVD images) into a boot-able USB stick. For whatever it is worth, it's Good to know there exists such a thing for Windows installation.

[1] http://www.wintoflash.com/home/en/

Installing Ubuntu with AARNET mirror

2010-09-06T13:31:00.002+10:00

This post applies to you, only if ...

you are in Australia
and you want to use AARNET's mirror (which for many people would work a lot faster than the standard Australian Ubuntu mirror)

Default Ubuntu installation provides the standard AU mirror for Australians. To use AARNET, you have to enter the following details manually during installation:

Protocol: http (ftp should also be supported but I haven't tested it)
Mirror address: mirror.aarnet.edu.au
Mirror path: /pub/ubuntu/archive/ (don't drop the slashes, type them as shown)

Sydney Cityrail (131500) Timetable on Command-line (Python version) v0.2

2010-09-02T15:02:00.014+10:00

This python script does the same thing as my previous Bash script: http://babilonline.blogspot.com/2010/08/sydney-cityrail-131500-timetable-on.html
The script depends on "BeautifulSoup" Python module.
Download it from here: http://www.crummy.com/software/BeautifulSoup/#Download
v0.2 changes:-- fixed the problem with "Sydney Terminal" and "Central Station". Specifying either of them works now.

    1 #!/usr/bin/python
    2 
    3 import sys
    4 import getopt
    5 import time
    6 import re
    7 
    8 from BeautifulSoup import BeautifulSoup, Comment
    9 import urllib
   10 import urllib2
   11 
   12 
   13 def strip_comments(soup):
   14  comments = soup.findAll(text=lambda text:isinstance(text, Comment))
   15  [comment.extract() for comment in comments]
   16  return soup
   17 
   18 def show_timetable(html, from_sta):
   19 
   20  ## PARSE
   21  soup = BeautifulSoup(BeautifulSoup(html).prettify())
   22  regex = re.compile("\&nbsp",re.I)
   23 
   24  ## FINDING OPTIONS
   25  options = []
   26  div1s = soup.findAll("div",{ "class" : "boxhead tal iewfix2" })
   27 
   28  for d1 in div1s:
   29   option = d1.find("h2").renderContents().encode("ASCII").strip()
   30   options.append(option)
   31 
   32 
   33  ## FINDING TIMETABLE
   34  timetable = []
   35  div2s = soup.findAll("div",{ "class" : "boxbody iewfix" })
   36 
   37  for d2 in div2s:
   38   bs = d2.findAll("b")
   39   for b in bs:
   40    tmp = b.renderContents().encode("ASCII").strip()
   41    tmp = regex.sub("",tmp)
   42 
   43    # fixing "Sydney Terminal" / "Central Station" dilemma
   44    if tmp.find("Sydney Terminal") > 0:
   45     tmp = tmp.replace("Sydney Terminal", "Central Station (Sydney Terminal)
   46     ")
   47    elif tmp.find("Central Station") > 0:
   48     tmp = tmp.replace("Central Station", "Central Station (Sydney Terminal)
   49     ")
   50 
   51    timetable.append(tmp)
   52 
   53 
   54  ## DIPLAY
   55  a_regex = re.compile(from_sta,re.I)
   56 
   57  j=0
   58  for i in xrange(0,len(timetable)):
   59   if a_regex.search(timetable[i]):
   60    print "\n  [[ %s ]]" % (options[j])
   61    j = j + 1
   62 
   63   print timetable[i]
   64 
   65 def usage():
   66  print """
   67  [*] usage: %s  options
   68 
   69  OPTIONS:
   70  -u      print url
   71  -h      show this message
   72 
   73  -a      set from station
   74  -b      set dest. station
   75 
   76  -t      set search time (08:30:AM)
   77    hour, min, sec must be seperated by colon (:)
   78 
   79  -d      set search date (yyyy/mm/dd)
   80    yyyy = 4 digits
   81    mm = 2 digits
   82    dd = 2 ditigs
   83 
   84  EXAMPLE:
   85  %s -a central -b 'town hall' -d 2010/09/02 -t 05:30:PM
   86   """ % (sys.argv[0], sys.argv[0])
   87 
   88 def prepare_param(args):
   89  ## PRESET VARIABLES
   90  print_url = False
   91  from_sta = "redfern"
   92  to_sta = "blacktown"
   93  now_time = time.strftime("%H:%M:%p")
   94  now_date = time.strftime("%Y/%m/%d")
   95 
   96 
   97  ## PARSE COMMAND LINE ARGS
   98  try:
   99   opts, args = getopt.getopt(sys.argv[1:], "a:b:d:t:uh")
  100 
  101  except getopt.GetoptError, err:
  102   print str(err)
  103   usage()
  104   sys.exit(2)
  105 
  106  for opt, arg in opts:
  107   if opt in ("-a"):
  108    from_sta = arg
  109   elif opt in ("-b"):
  110    to_sta = arg
  111   elif opt in ("-h"):
  112    usage()
  113    sys.exit(-1)
  114   elif opt in ("-u"):
  115    print_url = True
  116   elif opt in ("-t"):
  117    now_time = arg
  118   elif opt in ("-d"):
  119    now_date = arg
  120   else:
  121    assert False, "unhandled option"
  122 
  123  ## BREAK TIME INTO HOUR, MIN, AM/PM
  124  try:
  125   now_hour = re.split(":", now_time)[0]
  126   now_min = re.split(":", now_time)[1]
  127   now_ampm = re.split(":", now_time)[2]
  128  except:
  129   print "Incorrect time format"
  130   sys.exit(-1)
  131 
  132  now_date = re.sub("[^0-9]","",now_date)
  133 
  134  return from_sta, to_sta, now_date, now_hour, now_min, now_ampm, print_url
  135 
  136 def fetch_html( from_sta,
  137     to_sta,
  138     now_date,
  139     now_hour,
  140     now_min,
  141     now_ampm):
  142 
  143  url = "http://www.131500.com.au/plan-your-trip/trip-planner?"
  144  values = { "session" : "invalidate",
  145     "itd_cmd" : "invalid",
  146     "itd_includedMeans" : "checkbox",
  147     "itd_inclMOT_5" : "1",
  148     "itd_inclMOT_7" : "1",
  149     "itd_inclMOT_1" : "1",
  150     "itd_inclMOT_9" : "1",
  151     "itd_anyObjFilter_origin" : "2",
  152     "itd_name_origin" : from_sta ,
  153     "itd_anyObjFilter_destination" : "2",
  154     "itd_name_destination" : to_sta,
  155     "itd_itdDate" : now_date,
  156     "itd_itdTripDateTimeDepArr" : "dep",
  157     "itd_itdTimeHour" : now_hour,
  158     "itd_itdTimeMinute" : now_min,
  159     "itd_itdTimeAMPM" : now_ampm,
  160     "x" : "56",
  161     "y" : "11"
  162     }
  163  data = urllib.urlencode(values)
  164  try:
  165   req = urllib2.Request(url, data)
  166  except:
  167   print "Server Error", error
  168   sys.exit(-1)
  169 
  170  response = urllib2.urlopen(req)
  171  html = response.read()
  172 
  173  return html, url, data
  174 
  175 def main(args):
  176  from_sta, to_sta, now_date, now_hour, now_min, now_ampm, print_url =
  177  prepare_param(args)
  178 
  179  print
  180  print "From Station:", from_sta.title()
  181  print "To Station:", to_sta.title()
  182  print "Time:", "%s:%s:%s" % (now_hour, now_min, now_ampm)
  183  print "Date: %s/%s/%s" % (now_date[:4], now_date[4:6],now_date[-2:])
  184 
  185  html, url, data = fetch_html(from_sta, to_sta, now_date, now_hour, now_min,
  186  now_ampm)
  187  if print_url:
  188   print "\n%s%s" % (url, data)
  189 
  190  show_timetable(html, from_sta)
  191 
  192 
  193 if __name__ == "__main__":
  194   main(sys.argv[1:])

Sydney Cityrail (131500) Timetable on Command-line (ver: 0.2)

2010-08-24T16:37:00.002+10:00

Sydney CiryRail (www.131500.info) changed how parameters are fed in the URL from user's browser which broke my previous command line attempt.

Changes in this version:

Time format to use with '-t' is: "16:00" or "04:08:pm"
Date format to use with '-d' is: "2010:08:24"

Here is a fixed version (0.2) of the script:

    1 #!/bin/bash -e
    2 
    3 print_url=0
    4 timeout=10
    5 
    6 from_sta='Central'
    7 dest_sta='Town+Hall'
    8 
    9 old_now_date=$(date +%d%%2F%m%%2F%y)
   10 old_now_time=$(date +%I%%3A%M%p)
   11 
   12 now_date=$(date +%Y%m%d)
   13 now_hour=$(date +%I)
   14 now_min=$(date +%M)
   15 now_ampm=$(date +%P)
   16 
   17 function usage()
   18 {
   19 echo
   20 cat <<EOF
   21 [*] usage: `basename $0`  options
   22 
   23 OPTIONS:
   24  -u print url
   25  -h show this message
   26  -a set from station
   27  -b set dest. station
   28  -t set search time (08:30AM)
   29  -d set search date (21/5/10)
   30 EOF
   31 echo
   32 exit
   33 }
   34 
   35 function get_cookie()
   36 {
   37  wget -q -O /dev/null --keep-session-cookies --save-cookies="cookie.txt" "${
   38  1}"
   39 }
   40 
   41 while getopts "uha:b:t:d:" flag
   42 do
   43  if [ "$flag" == 'u' ]
   44  then
   45   print_url=1
   46  fi
   47 
   48  if [ "$flag" == 'h' ]
   49  then
   50   usage
   51  fi
   52 
   53  if [ "$flag" == 'a' ]
   54  then
   55   A=$(echo -n "${OPTARG:0:1}" | tr "[:lower:]" "[:upper:]")
   56   from_sta="$(echo -n "${A}${OPTARG:1}" | sed 's/ /+/g')"
   57  fi
   58 
   59  if [ "$flag" == 'b' ]
   60  then
   61   A=$(echo -n "${OPTARG:0:1}" | tr "[:lower:]" "[:upper:]")
   62   dest_sta="$(echo -n "${A}${OPTARG:1}" | sed 's/ /+/g')"
   63  fi
   64 
   65  if [ "$flag" == 't' ]
   66  then
   67   now_time=$(echo -n "$OPTARG" | tr '[:lower:]' '[:upper:]')
   68   now_time=$(echo "$now_time" | sed 's/://g;s/\-//g;s/\///g;s/\\//g')
   69 
   70   now_hour=$(echo -n "${now_time:0:2}")
   71   now_min=$(echo -n "${now_time:2:2}")
   72   now_ampm=$(echo -n "${now_time:4:2}")
   73  fi
   74 
   75  if [ "$flag" == 'd' ]
   76  then
   77   now_date="$OPTARG"
   78   now_date=$(echo "$now_date" | sed 's/://g;s/\-//g;s/\///g;s/\\//g')
   79  fi
   80 done
   81 
   82 old_url="http://www.131500.com.au/fullEnquiry.asp?&Vehicle=Bus,Train,
   83 Ferry&WalkSpeed=NORMAL&&IsAfter=A&MaxChanges=-
   84 1&FromLocType=s&ToLocType=s&x=64&y=11&FromLoc=$from_sta+Station~~;$from_sta+
   85 Station;$from_sta+Station~~LOCATION&ToLoc=$dest_sta+Station~~;$dest_sta+
   86 Station;$dest_sta+Station~~LOCATION&Time=$now_time&Date=$now_date"
   87 
   88 ori_url='http://www.131500.com.au/plan-your-trip/trip-
   89 planner?session=invalidate&itd_cmd=invalid&itd_includedMeans=checkbox&itd_in
   90 clMOT_5=1&itd_inclMOT_7=1&itd_inclMOT_1=1&itd_inclMOT_9=1&itd_anyObjFilter_o
   91 rigin=2&itd_name_origin=central&itd_anyObjFilter_destination=2&itd_name_dest
   92 ination=redfern&itd_itdDate=20100824&itd_itdTripDateTimeDepArr=dep&itd_itd
   93 TimeHour=2&itd_itdTimeMinute=0&itd_itdTimeAMPM=pm&x=56&y=11'
   94 
   95 url="http://www.131500.com.au/plan-your-trip/trip-
   96 planner?session=invalidate&itd_cmd=invalid&itd_includedMeans=checkbox&itd_in
   97 clMOT_5=1&itd_inclMOT_7=1&itd_inclMOT_1=1&itd_inclMOT_9=1&itd_anyObjFilter_o
   98 rigin=2&itd_name_origin=${from_sta}
   99 &itd_anyObjFilter_destination=2&itd_name_destination=${dest_sta}
  100 &itd_itdDate=${now_date}&itd_itdTripDateTimeDepArr=dep&itd_itdTimeHour=${
  101 now_hour}&itd_itdTimeMinute=${now_min}&itd_itdTimeAMPM=${now_ampm}
  102 &x=56&y=11"
  103 
  104 if [ $print_url == '1' ]
  105 then
  106  echo $url
  107 fi
  108 
  109 echo
  110 echo "From: $from_sta Station"
  111 echo "Dest: $dest_sta Station"
  112 echo "Time: $(echo $now_hour:$now_min $now_ampm | sed 's/\%3A/:/g')"
  113 echo "Date: $(echo ${now_date:0:4}:${now_date:4:2}:${now_date:6:2} | sed
  114 's/\%2F/\//g')"
  115 echo
  116 
  117 referer='http://www.131500.com.au/'
  118 agent='Mozilla/5.0 (Windows; U; Windows NT 6.1; en-GB; rv:1.9.2.8)
  119 Gecko/20100722 Firefox/3.6.8 (.NET CLR 3.5.30729)'
  120 
  121 #get_cookie $url
  122 #wget $url -q -O - --referer "$referer" --load-cookies=cookie.txt --user-
  123 agent="$agent" | html2text | grep 'Option \([0-9]\+\)\|Arr:\|Dep:'
  124 lynx -noredir -dump $url | grep 'Option \([0-9]\+\)\|Arr:\|Dep:'
  125 
  126 echo

DokuWiki Beautifier: Floating Table of Contents (TOC) and others

2010-08-24T12:37:00.007+10:00

Just put the following code inside conf/userstyle.css inside DokuWiki directory.

P.S. What you get from the code:

The table of content moves up/down as you scroll.
"Edit" buttons are a bit more stylish. (comment out if you don't like it)

    1 /* Place customisations to screen mode style sheet here */
    2 
    3 /* table-of-content sidebar*/
    4 div.toc {
    5   position: fixed !important;
    6   top: 1em !important;
    7   bottom: 3em !important;
    8   right: 0px !important;
    9   overflow-y: auto !important;
   10   display: -moz-box !important;
   11   height: 90% !important;
   12   width: 16% !important;
   13 }
   14 
   15 /* reducing page width to save it from going under toc */
   16 div.header, div.bar, div.doc, div.breadcrumbs {width:82% !important}
   17 div.page {
   18   width: 80% !important;
   19   border: 0px #6C9AD0 solid !important;
   20   padding: 1em !important;
   21 }
   22 
   23 /* styling the 'Edit' button */
   24 input[value="Edit"] {
   25 font-weight: bold !important;
   26 font-size:1em !important;
   27 border-width: 0.2em !important;
   28 }
   29

Mounting HTC Android (G1) SD card on Windows 7

2010-08-18T17:46:00.004+10:00

Connect phone to the Windows 7 computer with USB cable
On phone, select "Turn on USB storate" from status bar (slides down from top)
On Windows 7, go to Control Panel > Device Manager
Expand/Click on "Disk Drives" and select "HTC Android Phone USB Device"
Right click on "HTC Android Phone USB Device" and Click "Disable"
Now right click again on "HTC Android Phone USB Device" and Click "Enable"
Windows 7 should show you the SD card files now ;)

Debian/Ubuntu: Delay in getting SSH login prompt

2010-08-17T13:31:00.006+10:00

On some of my Debian/Ubuntu machines, I get an annoying 10-20 seconds delay in seeing the prompt while trying to login. My first guess was - something wrong in key exchange and it's must be falling back with standard password authentication, as can be seen with ssh -vv babil@servername

But the solution seems to be in disabling DNS resolution for the SSH server. I found the solution here: http://www.unix.com/linux/96010-delay-getting-ssh-login-prompt.html. /etc/ssh/sshd_conf needs a "UseDNS no" line, followed by a "sudo /etc/init.d/ssh restart".
Cheers.

Stopping Gnuplot from flashing a plot window and disappear

2010-08-12T13:49:00.002+10:00

In Ubuntu/Debian, Gnuplot tends to use terminal 'X11' if no terminal has been defined before the 'plot' command. The reason for not specifying a terminal could be various (eg. being lazy and ignorant, subsequent use of 'replot' command using other terminal types in several other scripts)

In my case I had a loop of 5x5x5x5 times in my NS-2 script invoking Gnuplot 6 times with every run. The flashing window literally leaves the screen unusable as the Gnuplot X11 terminal window keeps popping up and out in a blink.

A temporary fix (which doesn't require rewriting my whole Gnuplot script set) is to use:

set terminal unknown

Doing this stops the flash with the first 'plot' command. The subsequent 'replot' commands can define their own terminals as needed.

Suppressing page number in LaTeX

2010-08-01T22:11:00.003+10:00

\thispagestyle{empty}

\pagestyle{empty}

Bash: simple wrapper script for java jar files

2010-07-27T14:12:00.003+10:00

There could be thousands of other (probably better) ways of writing a wrapper bash script for executing jar files on command line (instead of having to do "java -jar tool.jar"). For whatever it is worth, here's what I came up in quick 2 minutes. One can optimize according to his/her needs inside the for loop:

    1 #!/bin/bash
    2 
    3 path='/home/USERNAME/android-sdk-linux_86/tools'
    4 
    5 args=''
    6 for ((i=$#;i>0;i--))
    7 do
    8  args="${args} ${1}"
    9  shift
   10 done
   11 
   12 java -jar $path/tool.jar "$args"
   13 
   14

Windows 7: Windows Media Player 12 "Server Execution Failed" problem

2010-07-23T22:25:00.003+10:00

I found the solution here: http://social.answers.microsoft.com/Forums/en-US/w7music/thread/ecc08c3b-d445-48c2-b07d-9df48500434f

The solution that worked for me:

Unpin media player from taskbar.
Pin it back.

lol, yes, that literally solved the problem for me.

Android: Sliding/Rotating keyboard back 'n forth in android emulator

2010-07-22T15:40:00.004+10:00

It may be obvious to many, but just as a record - to rotate the keyboard in android emulator, the keyboard short-cut is:

Ctrl + F11

Sydney Cityrail (131500) Timetable on Command-line

2010-06-30T16:55:00.008+10:00

You need "Lynx" installed. On Debian/Ubuntu, you can do "apt-get install Lynx".

Example usage:

Put a '+' between the words, if the station name is not a single word.
'-h' shows you the help.
By default, it uses current date,time and Redfern-Blacktown stations as from-to combination. The station-name variables are easily editable on top of the script.

babil@quad:~$ train.sh -h

[*] usage: train.sh options

OPTIONS:
-u print url
-h show this message
-a set from station
-b set dest. station
-t set search time (08:30AM)
-d set search date (21/5/10)

babil@quad:~$ train.sh -a redfern -b north+sydney -d 2/7/2010 -t 14:00

From: Redfern Station
Dest: North+sydney Station
Time: 14:00
Date: 2/7/2010

Option 1 : 15 minutes
Dep: 2:00pm Redfern Station Platform 3
Arr: 2:15pm North Sydney Station Platform 4
Option 2 : 15 minutes
Dep: 2:04pm Redfern Station Platform 3
Arr: 2:19pm North Sydney Station Platform 4
Option 3 : 15 minutes
Dep: 2:15pm Redfern Station Platform 3
Arr: 2:30pm North Sydney Station Platform 4
Option 4 : 15 minutes
Dep: 2:19pm Redfern Station Platform 3
Arr: 2:34pm North Sydney Station Platform 4
Option 5 : 21 minutes
Dep: 2:20pm Redfern Station Platform 5
Arr: 2:28pm Wynyard Station Platform 6
Dep: 2:35pm Wynyard Station Platform 4
Arr: 2:41pm North Sydney Station Platform 4

Here's is the script:

    1 #!/bin/bash
    2 
    3 print_url=0
    4 
    5 from_sta='Central'
    6 dest_sta='Town+hall'
    7 now_date=$(date +%d%%2F%m%%2F%y)
    8 now_time=$(date +%I%%3A%M%p)
    9 
   10 function usage()
   11 {
   12 echo
   13 cat <<EOF
   14 [*] usage: `basename $0`  options
   15 
   16 OPTIONS:
   17  -u print url
   18  -h show this message
   19  -a set from station
   20  -b set dest. station
   21  -t set search time (08:30AM)
   22  -d set search date (21/5/10)
   23 EOF
   24 echo
   25 exit
   26 }
   27 
   28 while getopts "uha:b:t:d:" flag
   29 do
   30  if [ "$flag" == 'u' ]
   31  then
   32   print_url=1
   33  fi
   34 
   35  if [ "$flag" == 'h' ]
   36  then
   37   usage
   38  fi
   39 
   40  if [ "$flag" == 'a' ]
   41  then
   42   A=$(echo -n "${OPTARG:0:1}" | tr "[:lower:]" "[:upper:]")
   43   from_sta="$(echo -n "${A}${OPTARG:1}" | sed 's/ /+/g')"
   44  fi
   45 
   46  if [ "$flag" == 'b' ]
   47  then
   48   A=$(echo -n "${OPTARG:0:1}" | tr "[:lower:]" "[:upper:]")
   49   dest_sta="$(echo -n "${A}${OPTARG:1}" | sed 's/ /+/g')"
   50  fi
   51 
   52  if [ "$flag" == 't' ]
   53  then
   54   now_time="$(echo -n "$OPTARG" | tr "[:lower:]" "[:upper:]")"
   55  fi
   56 
   57  if [ "$flag" == 'd' ]
   58  then
   59   now_date="$OPTARG"
   60  fi
   61 done
   62 
   63 url="http://www.131500.com.au/fullEnquiry.asp?&Vehicle=Bus,Train,
   64 Ferry&WalkSpeed=NORMAL&&IsAfter=A&MaxChanges=-
   65 1&FromLocType=s&ToLocType=s&x=64&y=11&FromLoc=$from_sta+Station~~;$from_sta+
   66 Station;$from_sta+Station~~LOCATION&ToLoc=$dest_sta+Station~~;$dest_sta+
   67 Station;$dest_sta+Station~~LOCATION&Time=$now_time&Date=$now_date"
   68 
   69 if [ $print_url == '1' ]
   70 then
   71  echo $url
   72 fi
   73 
   74 echo
   75 echo "From: $from_sta Station"
   76 echo "Dest: $dest_sta Station"
   77 echo "Time: $(echo $now_time | sed 's/\%3A/:/g')"
   78 echo "Date: $(echo $now_date | sed 's/\%2F/\//g')"
   79 echo
   80 
   81 lynx -dump $url | grep 'Option \([0-9]\+\)\|Arr:\|Dep:'
   82 
   83 echo

Gnuplot: skipping header line from input data file

2010-06-23T15:50:00.004+10:00

The trick is in gnuplot command - "every"

The syntax of "every" is:

plot 'file' every {point_incr}
{:{block_incr}
{:{start_point}
{:{start_block}
{:{end_point}
{:end_block}}}}}

So, in order to skip the first two lines the gnuplot command would be:

plot "data.txt" using 1:5 every ::2 with linespoints linestyle 7

This will skip the first line of the data file

Tab, Newline and other tricks in Gnuplot graph title

2010-06-09T15:30:00.004+10:00

Example:

plot 'data.txt' using 1:2 with lines points title sprintf("AAA vs. BBB \n (2nd line)")

Producing greppable output with tshark

2010-05-11T18:13:00.004+10:00

Tshark is the command-line cousin of open-source packet capturing software - Wireshark. The trick is to use the '-e' and '-Tfields' parameters to get the desired fields. The names of the fields can be found by opening the 'input.pcap' file with wireshark and clicking "+expression" button next to filter text box.

Example:

tshark -e frame.time_relative -e frame.number -e frame.len -e ip.src -e ip.dst -e dccp.ccval -e dccp.type -e dccp.seq -e ip.len -Tfields -r input.pcap

Splitting 4 GB or bigger files for Fat32

2010-05-03T23:21:00.004+10:00

Fat32 can not handle files bigger than 4 gigabytes. Surprisingly Fat32 still prevails on almost every USB stick. 'Split' command in Linux can be very handy to transfer files bigger than 4 gig using USB sticks.

Copy & paste the following to split file on Linux:

split --verbose --numeric-suffixes --bytes=3990000000 'large_input_file.mkv' 'output_'

Joining can be done by 'cat' or 'copy /B' on Windows.

Elsevier style with Lyx on Ubuntu 9.10

2010-04-26T02:22:00.002+10:00

Elsevier style doesn't work out of the box on Ubuntu 9.10. Follow the steps below:

mkdir elsarticle
cd elsarticle
wget http://www.elsevier.com/framework_authors/misc/elsarticle.zip
unzip elsarticle.zip
sudo cp *.sty /usr/share/texmf-texlive/tex/latex/elsevier/
sudo cp *.cls /usr/share/texmf-texlive/tex/latex/elsevier/
sudo cp *.bst /usr/share/texmf-texlive/tex/latex/elsevier/
sudo texhash
Run LyX, click Tools > Reconfigure
Restart LyX
File > New Document from Template ... select elsarticle.lyx ... it should work now ;-)

IEEEtran with Lyx on Ubuntu 9.10

2010-04-26T00:48:00.004+10:00

If you are fighting to get IEEEtran.sty install and work with lyx, install the following packages to get (almost) everything working straight out of the box:

sudo apt-get install texlive texlive-base texlive-base-bin texlive-base-bin-doc texlive-bibtex-extra texlive-binaries texlive-common texlive-doc-base texlive-extra-utils texlive-font-utils texlive-fonts-extra texlive-fonts-extra-doc texlive-fonts-recommended texlive-fonts-recommended-doc texlive-formats-extra texlive-generic-extra texlive-generic-recommended texlive-humanities texlive-humanities-doc texlive-latex-base texlive-latex-base-doc texlive-latex-extra texlive-latex-extra-doc texlive-latex-recommended texlive-latex-recommended-doc texlive-latex3 texlive-luatex texlive-math-extra texlive-metapost texlive-metapost-doc texlive-omega texlive-pictures texlive-pictures-doc texlive-plain-extra texlive-pstricks texlive-pstricks-doc texlive-publishers texlive-publishers-doc texlive-science texlive-science-doc texlive-xetex latex-beamer latex-xcolor latex-xft-fonts dblatex preview-latex-style

Ubuntu/Debian: Flash not responding to any click

2010-04-13T21:15:00.002+10:00

I found the fix here: http://helpforlinux.blogspot.com/2009/11/i-cannot-click-on-flash-in-ubuntu.html

As instructed, perform the following and restart your firefox:

Hit ALT+F2 and enter
gksudo gedit /usr/lib/nspluginwrapper/i386/linux/npviewer
add the following line BEFORE the last line of text
export GDK_NATIVE_WINDOWS=1
Save.
Restart any applications using flash

Ubuntu (jaunty/9.04) - Delay in print dialog

2010-03-31T15:14:00.003+11:00

We use a 'Konica Minolta c451' printer at work and I have been having this issue since I upgraded to Ubuntu 9.04 (Jaunty). The bug has been reported here: https://bugs.launchpad.net/ubuntu/+source/gtk+2.0/+bug/475845

The temporary recommended solution is to comment out the following in /etc/cups/cupsd.conf:

##Listen /var/run/cups/cups.sock

Extracting subtitles with MKVmerge

2010-03-16T14:40:00.002+11:00

check the tracks with:

mkvinfo "filename.mkv"

extract the subtitle track with:

mkvextract tracks "filename.mkv" 3:subs.srt

Decode captcha in bash

2010-02-22T19:02:00.003+11:00

convert input image to grayscale using imagemagik
convert input.png -colorspace Gray output.tif
decode with tesseract
tesseract output.tif result; cat result.txt

Bash wordlist/dictionary generator for brute-forcing

2010-02-22T15:20:00.003+11:00

This example code will generate passwords of 4 characters only. To generate bigger passwords, there has to be more loops (pretty obvious in the code below). Happy Brute-forcing !!

    1 #!/bin/bash
    2 
    3 list=`echo {0..9} {a..z} {A..Z}`
    4 
    5 for c1 in $list
    6 do
    7  for c2 in $list
    8  do
    9   for c3 in $list
   10   do
   11    for c4 in $list
   12    do
   13     echo $c1$c2$c3$c4
   14    done
   15   done
   16  done
   17 done
   18

Cutting MKV files with 'mkvmerge'

2010-02-19T19:30:00.003+11:00

mkvmerge --split timecodes:00:28:00,00:34:42 'input.mkv' -o 'output.mkv'

Ubuntu/Debian :: watching trailers from Apple

2010-02-18T20:25:00.001+11:00

apt-get install mplayer mozilla-mplayer

Getting Broadcom CrystalHD into AppleTV [[ A 2 Z ;-) ]]

2010-02-13T02:42:00.023+11:00

If you don't happen to know much about "crystalHD" yet, you will find the gist of the story here: http://xbmc.org/davilla/2009/12/29/broadcom-crystal-hd-its-magic/. In short, it's a mini pci-e card which lets you decode "H264" in hardware.

Can't wait to see the proper driver and XBMC version which will leave everyone with the possibility to have a kick-ass media center for the living room:

Which card to buy: BCM970012 or BCM970015 ?

Answer: BCM970015.
Why: http://www.logicsupply.com/blog/2010/08/09/next-gen-broadcom-hardware-decoder-one-chip-many-advantages/

Installing hardware
http://rufn.it/aTV/
http://patchstick.wikispaces.com/Broadcom+Installation
Installing driver
http://todayguesswhat.blogspot.com/2010/01/thanks-davilla-and-xbmc-team-apple-tv.html
Installing XBMC

(manual process)
replace "latest.dmg" with the version you want. You can get a list of version on this website: http://mirrors.xbmc.org/

------------------------------------------------------------------------------SSH into ATV (username: frontrow, password: frontrow)

cd Downloads

wget http://mirrors.xbmc.org/nightlies/osx/latest.dmg
or, curl http://mirrors.xbmc.org/nightlies/osx/latest.dmg

sudo hdutil mount latest.dmg
cp -r /Volumes/XBMC/XBMC.app /Users/frontrow/Applications/
sudo reboot
------------------------------------------------------------------------------

(automatic process)

Installing XBMC and CrystalHD drivers by "Launcher":

DON'T RUN THE "DEFAULTS" COMMANDS WITH SUDO.

defaults write com.teamxbmc.xbmclauncher XBMCAdditionalDownloadPlistURLs -array-add http://mirrors.xbmc.org/nightlies/osx/update_urls.plist

defaults write com.teamxbmc.xbmclauncher XBMCAdditionalDownloadPlistURLs -array-add http://crystalhd.sartori.at/files/crystalhd.plist
defaults write com.teamxbmc.xbmclauncher XBMCAdditionalDownloadPlistURLs -array-add http://www.sshcs.com/xbmc/Info.asp
REBOOT. THERE SHOULD BE TWO NEW ENTRIES UNDER "LAUNCHER" MENU NAMELY - NIGHTLIES, CRYSTAL-HD.

Checking Crystal-HD:

grep -i crystal /Users/frontrow/Library/Logs/xbmc.log
dmesg | grep -i crystal

Getting Other log files:

xbmc.log: /Users/frontorw/Library/Logs/xbmc.log
Console log: /Library/Logs/Console/501/console.log
Crashreporter log: /Users/frontrow/Library/Logs/CrashReporter/CrashReporter.log

More Info:

http://wiki.awkwardtv.org/wiki/Broadcom_HD_Decoder

CrystalHD-UBUNTU:This is a live ubuntu with crystalHD driver and xbmc, bootable from USB stick.

Source: http://forum.xbmc.org/showthread.php?t=74992
Download:
- http://www.surfanddestroy.org/resources/xbmc-r28256-crystalHD-1.0.3-alsa-1.019-nvidia100.14.19-hardy-cust_machkernel-2.6.24.27.rar
- http://stmlabs.com/resources/xbmc-r28256-crystalHD-1.0.3-alsa-1.019-nvidia100.14.19-hardy-cust_machkernel-2.6.24.27.rar
  
  MD5 hash: d4cb24bca0042e84425ab23454a5bab4

Installing Ubuntu on AppleTV:

There is a very nicely written wiki located here:

http://wiki.xbmc.org/index.php?title=HOW-TO_install_a_Miminal_Ubuntu_and_XBMC_with_sound_over_HDMI_on_the_AppleTV

Related Threads on XBMC Forum:

http://forum.xbmc.org/showthread.php?t=81209

Quotes:

atv(osx) has a quirk that limits display to 720p, it's been discussed about a million times and here's a trac ticket for it.

atv(linux) has no such limit, xorg controls display resolutions that are available.
--
davilla
Team-XBMC Developer
2010-09-20, 10:21

Debian/Ubuntu: highlight and publish source codes

2010-02-09T17:24:00.002+11:00

The tool is called - "highlight"

apt-get install highlight
man highlight

Perfect for publishing, integrating source-codes in tex and on websites. I like the following for my codes:

highlight --pretty-symbols --inline-css -I -l _FILE_NAME_ --style $ide -V > tmp; firefox tmp &

In firefox, one can do "View Page Source" and copy&paste.

Automate downloading from hotfile.com

2010-02-09T17:10:00.004+11:00

    1 #!/bin/bash
    2 
    3 RED='\e[1;31m'
    4 CYAN='\e[1;36m'
    5 NC='\e[0m' # No Color
    6 
    7 if [ -z "$1" ]
    8 then
    9  echo
   10  echo -e "${RED}[+] usage: ${CYAN}`basename $0`   http://hotfile.com/...
   11  /filename.rar.html${NC}"
   12  echo
   13  exit
   14 fi
   15 
   16 url="$1"
   17 
   18 echo
   19 echo -e "[+] Input URL : ${RED}$url${NC}"
   20 
   21 tm_='1245072880'
   22 tmhash_='e5b845119f0055c5d8554ee5f2ffc7b2d5ef86d7'
   23 wait_='30'
   24 waithash_='3bf07c5d83f2e652ff22eeaee00a6f08d4d2409a'
   25 action_='capt'
   26 
   27 line_with_url='http://hotfile.com/get/'
   28 
   29 url2=`wget -q "$url" --post-data "tm=${tm_}&tmhash=${tmhash_}&wait=${wait_}
   30 &waithash=${waithash_}&action=${action_}" -O - | grep "$line_with_url" |
   31 sed -s 's/.*<a href="\([^"]*\)".*$/\1/g'`
   32 
   33 echo -e "[+] Real URL  : ${RED}$url2 ${NC}"
   34 echo
   35 
   36 if [ ! -z "$url2" ]
   37 then
   38  wget "$url2"
   39 fi
   40

Debian/Ubuntu: stopping a package from upgrade

2010-02-09T15:45:00.007+11:00

Find package name
dpkg -l | grep "partial-name"
Set package to hold
echo "full-package-name hold" | dpkg --set-selections
Set package to un-hold
echo "full-package-name install" | dpkg --set-selections

Java is not working in Windows Firefox ...

2010-01-24T20:34:00.002+11:00

1. Reinstalling Java fixed it for me. You can download it from here:

http://www.java.com/en/download/

2. Some people also recommended deleting "NPOJI610.dll" from Firefox/Plugins directory.

Pasting text in Vim from webpages

2010-01-20T19:08:00.000+11:00

Vim auto-indents texts/code-snippets while pasting from websites. The trick is to take vim in "paste" mode to stop auto-indenting:

while pasting ...

:set paste

when pasting is doen ...

:set nopaste

Installing/compiling ns-2.34 on Ubuntu/Debian

2010-01-12T20:05:00.002+11:00

I had to install the following dev's to get ns-2.34 "all-in-one" package to work:

apt-get install libx11-dev xorg-dev

Recently stumbled upon another problem compiling otcl. The error message looks like the following:

ld: libotcl.so: hidden symbol `__stack_chk_fail_local' isn't defined

Apparently this an incompatibility between ld and gcc-4.4. The quick fix will be to replace "SHLIB_LD" variable for "Linux*)" targets from ~~"ld -shared"~~ to "${CC} -shared".

Fix broken pixels on LCD/Plasma

2010-01-12T13:28:00.000+11:00

From http://www.jscreenfix.com/basic.php

"The JScreenFix Java applet helps fix stuck pixels and image persistence on LCD screens and reduces burn-in on plasma screens. The applet runs safely inside a web browser, does not require installation and works with Windows, Linux and Mac computers.

JScreenFix repairs most stuck pixels in under 20 minutes, and mild screen burn-in within 12 hours..."

Removing ads from Prothom-alo.com

2009-12-15T14:28:00.020+11:00

I wrote the following style to remove the *UGLY* ads from prothom-alo.com. You are going to need the followings to be able to use it:

Firefox ...best browser on the planet
Stylish or GreaseMonkey ...lets you apply user provided styles/scripts which overrides the server provided stuffs for a particular webpage.
My CSS script ...Removes ads from prothom-alo.com

A complete screen-shot is here: http://imgur.com/U6czu.png

OneManga.com leecher

2009-12-15T14:18:00.005+11:00

This one leeches manga from www.onemanga.com. I was too lazy to implement the chapter ripper part. It will only leech the whole series with the right URL as argument.

** code comes with no warranty or whatsoever. I'm not responsible if it breaks anybody's computer. Use it at your own risk. You have been warned !!! **

Example:

onemanga.sh http://www.onemanga.com/School_Rumble/

    1 #!/bin/bash
    2 
    3 RED='\e[0;31m'
    4 CYAN='\e[0;36m'
    5 NC='\e[0m' # No Color
    6 
    7 if [ -z $1 ]
    8 then
    9  echo
   10  echo "[*] usage: `basename $0` manga_url"
   11  echo
   12  exit
   13 else
   14  manga_name=`echo $1 | awk -F '/' '{for(i=NF;i>=0;i--){if(length($i)>0){
   15  print $i;break;}}}'`
   16 fi
   17 
   18 main_url="http://www.onemanga.com"
   19 
   20 rm -rf ${manga_name}
   21 
   22 ##finding list of chapters
   23 echo -n -e "${CYAN}[*]${RED} Finding total chapters in ${CYAN} $manga_name
   24 ${NC}= "
   25 wget -q -nv ${main_url}/${manga_name} -O tmp.txt
   26 chapters=`cat tmp.txt | grep '<td class="ch-subject"><a href="/' | awk -F
   27 '"' '{print $4}'`
   28 
   29 
   30 count=0
   31 for c in $chapters
   32 do
   33  mkdir -p ./$c
   34  count=$((count+1))
   35 done
   36 echo -e "${CYAN}${count}${NC}"
   37 ##
   38 
   39 ##parse chapter and download
   40 for chapter in $chapters
   41 do
   42  pwd=`pwd`
   43 
   44  cd ./$chapter
   45 
   46  ## initial wget
   47  echo -e "${CYAN}[*]${RED} Trying to find the image base url${NC}"
   48 
   49  ## find collect the first page in the chapter
   50  wget -q -nv $main_url/$chapter -O tmp.txt
   51  page=`cat tmp.txt | grep "Begin reading" | awk -F '"' '{print $2}'`
   52 
   53  ## now go to that page & find image base
   54  wget -q -nv ${main_url}${page} -O tmp.txt 2>/dev/null
   55  image=`cat tmp.txt | grep "img_url" | awk -F '"' '{for(i=1;i<NF;i++){if($i
   56  ~ "jpg"){print $i}}}' | awk -F '/' '{print $NF}'`
   57  image_base=`cat tmp.txt | grep "img_url" | awk -F '"' '{for(i=1;i<NF;i++){
   58  if($i ~ "jpg"){print $i}}}' | sed s/"$image"//g`
   59  echo -e "${RED}>>${NC} $image_base"
   60 
   61  ## download
   62  d=$((d+1))
   63  names=`cat tmp.txt | awk '{for(i=1;i<=NF;i++){if($i ~ "selected")go++}{if(
   64  go>1){print}}}' | grep -e "</option>" -e 'credits</option>' -e 'extra*<
   65  /option>' -e 'cover*</option>' | awk -F '"' '{print $2}'`
   66 
   67  n=0
   68  for k in $names
   69  do
   70   n=$((n+1))
   71  done
   72 
   73  echo -e "${CYAN}[*]${RED} Downloading ${CYAN}$n ${RED}images from chapter
   74  ${CYAN}$chapter ${RED} ... ##${CYAN}$((count-d+1))${RED}##${CYAN}$count${
   75  RED}##${NC}"
   76  for k in $names
   77  do
   78   #echo -e "${RED}>>${NC} ${image_base}${k}.jpg"
   79   wget -nv "${image_base}${k}.jpg"
   80  done
   81 
   82  cd $pwd
   83 done
   84 ##

MangaFox.com leecher

2009-12-15T13:59:00.006+11:00

This is a bash script to leech a whole series or a particular chapter from www.mangafox.com. I wrote this script for my niece, hope it helps you too. Using is simple, 1st argument is the url to a particular manga and 2nd (optional) argument is the the chapter in that manga.

** code comes with no warranty or whatsoever. I'm not responsible if it breaks anybody's computer. Use it at your own risk. You have been warned !!! **

Example:

mangafox.sh http://www.mangafox.com/manga/school_rumble/   282
or

mangafox.sh School_Rumble 282


or
mangafox.sh

http://www.mangafox.com/manga/school_rumble/

or
mangafox.sh

School_Rumble

    1 #!/bin/bash
    2 
    3 RED='\e[1;31m'
    4 CYAN='\e[1;36m'
    5 NC='\e[0m' # No Color
    6 YLW='\e[1;33m'
    7 WHITE='\e[0;37m'
    8 
    9 main_url="http://www.mangafox.com/manga"
   10 wget_param="--tries=10 --retry-connrefused"
   11 
   12 ## usage
   13 if [ -z $1 ]
   14 then
   15  echo
   16  echo -e "${CYAN}[*]${RED} usage: `basename $0` manga_url${NC}"
   17  echo
   18  exit
   19 else
   20  manga_name=`echo $1 | awk -F '/' '{for(i=NF;i>=0;i--){if(length($i)>0){
   21  print $i;break;}}}'`
   22  if [ ! -z "$2" ]
   23  then
   24   specific_chapter="$2"
   25  fi
   26 fi
   27 ##
   28 
   29 function find_chapters()
   30 {
   31  TMP="${manga_name}_find_chapters.tmp"
   32 
   33  echo -n -e "${CYAN}[*]${RED} Finding total chapters in ${CYAN} $manga_name
   34  ${NC}= "
   35  wget $wget_param -q -nv "${main_url}/${manga_name}/?no_warning=1" -O $TMP
   36  chapters=`cat $TMP | grep -e 'class="chico">' | grep -v -e '</td>' -e
   37  '#listing' | awk -F '"' '{print $2}' | sed 's/^\/manga\///g'`
   38 
   39  count=0
   40  for c in $chapters
   41  do
   42   mkdir -p ./$c
   43   #echo $c ##debug
   44   count=$((count+1))
   45  done
   46  echo -e "${CYAN}${count}${NC}"
   47 }
   48 
   49 
   50 function parse_chapter_n_download()
   51 {
   52  PAGES="pages.tmp"
   53  PAGE="page_html.tmp"
   54 
   55  for chapter in $chapters
   56  do
   57   pwd=`pwd`
   58 
   59   if [ "$specific_chapter" ]
   60   then
   61    mkdir -p "$specific_chapter" 2>/dev/null
   62    chapter=$specific_chapter
   63   fi
   64 
   65   cd ./$chapter
   66 
   67   ## find total number of pages in chapter
   68   echo -n -e "${CYAN}[*]${RED} Total pages in ${CYAN} $chapter ${NC}= "
   69   wget -q -nv $wget_param $main_url/$chapter -O $PAGES
   70   pages=`cat $PAGES | grep '^.*<option value=.*<\/select>.*$' -m1 | awk '{
   71   for(i=1;i<=NF;i++){if($(i-1)~"value"){print $i}}}' | sed -e
   72   's/selected//g;s/option//g;s/[<>\/"=]//g;'`
   73 
   74   n=0
   75   for k in $pages
   76   do
   77    #echo $k ##debug
   78    n=$((n+1))
   79   done
   80   echo -e "${CYAN}$n${NC}"
   81 
   82   ## now i have a list of (1,2,3...).html pages
   83   for p in $pages
   84   do
   85    wget $wget_param  -q -nv $main_url/$chapter/${p}.html -O $PAGE
   86    img_url=`cat $PAGE | grep 'onclick="return enlarge();' | awk '{for(i=1;i<
   87    =NF;i++){if($i~"http://"){print $i}}}' | sed 's/src=//g;s/["=]//g'`
   88    img=`echo $img_url | awk -F '/' '{print $NF}'`
   89    echo -e -n "${CYAN}>>${WHITE} $img_url ${RED} ... ${NC}"
   90    wget $wget_param -q -nv $img_url
   91    if [ -e $img ]
   92    then
   93     echo -e "${CYAN}[done]${NC}"
   94    else
   95     echo -e "${YLW}[failed]${NC}"
   96    fi
   97   done
   98 
   99   cd $pwd
  100 
  101   if [ "$specific_chapter" ]
  102   then
  103    exit;
  104   fi
  105 done
  106 }
  107 
  108 
  109 function main()
  110 {
  111  rm -rf ${manga_name}
  112  find_chapters
  113  parse_chapter_n_download
  114 }
  115 
  116 main

Listing directories only in bash

2009-12-10T17:17:00.003+11:00

Now this is embarrassing. All these years with an Unix operating system and I couldn't figure it out by myself. Anyways, here is how you do it:

Method 1:

ls -d */

Method 2:

echo */

Installing Flurry screensaver on Ubuntu/Debian

2009-07-22T17:22:00.008+10:00

[1] Download the binary from here (I don't know/care where in ubuntu/debian repository it can be found): http://ubuntu-ky.ubuntuforums.org/showthread.php?p=6930123

[2.1] Gnome: just copy the files from extracted "Flurry" directory in /usr/. I wrote a little script to make it easier.

##########################################################

#!/bin/bash

tar_file="Flurry.tar"

#extract
rm -rf Flurry
tar xvf "$tar_file"

#copy files
cp -u -v Flurry/usr/lib/xscreensaver/flurry /usr/lib/xscreensaver/
cp -u -v Flurry/usr/share/xscreensaver/config/flurry.xml
/usr/share/xscreensaver/config/
cp -u -v Flurry/usr/share/applications/screensavers/flurry.desktop
/usr/share/applications/screensavers/

##########################################################

[2.2] KDE: copy the files in /usr/ and create "/usr/share/applnk/System/ScreenSavers/flurry.desktop" with the following content. It will put flurry in the KDE screensaver configuration menu and let you set the options:

[Desktop Entry]
Encoding=UTF-8
Exec=flurry
Icon=kscreensaver
Type=Application
TryExec=xscreensaver
Actions=InWindow;Root;Setup;
X-KDE-Category=OpenGL Screen Savers
X-KDE-Type=OpenGL
Name=Flurry (GL)

[Desktop Action Setup]
Exec=kxsconfig flurry
Name=Setup...
Icon=kscreensaver

[Desktop Action InWindow]
Exec=kxsrun flurry -- -window-id %w
Name=Display in Specified Window
NoDisplay=true

[Desktop Action Root]
Exec=kxsrun flurry -- -root
Name=Display in Root Window
NoDisplay=true

Baba ...

2009-06-28T02:05:00.006+10:00

prothom-alo :: 26th June 2009

Amar Desh ...

2009-06-16T14:06:00.007+10:00

prothom-alo:: 16th June 2009

Debian/Etch fails to upgrade x11-common ??

2009-01-20T15:54:00.005+11:00

When I was trying to upgrade my etch to lenny, x11-common failed to upgrade properly. I was getting the following error message:

Error::
trying to overwrite `/usr/X11R6/bin', which is also in package aaa-base

For somebody who is really as annoyed (to hell) as me, this fixed the (!@#!$%$) problem for me:

Solution::
dpkg -i --force-overwrite /var/cache/apt/archives/x11-common_1%3a7.3+18_all.deb

Use pastebin.com as your online clipboard

2009-01-15T17:17:00.003+11:00

THE SCRIPTS ARE BROKEN AT THE MOMENT DUE TO CHANGE IN PASTEBIN.COM WEBSITE. I HAVE CREATED ANOTHER PASTEBIN SCRIPT USING THEIR API. FIND IT HERE: http://babilonline.blogspot.com/2010/10/post-to-pastebin-from-command-line.html

I created the following pair of scripts to paste text files on pastebin.com from console and fetching it may be from a different computer on a different network on console. Comes really handy when you just want to share some texts on two diffrent computers. And the best thing is it lets you paste things that are blocked by pastebin ;).

CHANGE THE VARIABLE "USER" TO CREATE YOUR OWN CLIPBOARD ON PASTEBIN.

=============
pastebin.sh
=============

1 #!/bin/bash
2 #
3
4 USER="YOURNAME_IN_ONE_WORD"
5
6 bold='\e[1;1m';
7 white='\e[1;37m'
8 yellow='\e[1;33m'
9 magenta='\e[1;35m'
10 blue='\e[1;34m'
11 red='\e[1;31m'
12 cyan='\e[1;36m'
13 reset='\e[0m'
14
15
16 if [ -z $1 ]; then
17 echo
18 echo "[*] usage: `basename $0` filename"
19 echo
20 exit
21 fi
22
23 case `file "$1"` in
24 *Bourne*)
25 TYPE="bash";;
26 *python*)
27 TYPE="python";;
28 *perl*)
29 TYPE="perl";;
30 *HTML*)
31 TYPE="html4strict";;
32 *)
33 TYPE="text";;
34 esac
35
36 DATA=`uuencode -m $1 $1`
37 #DATA=`cat "$@" | ./urlencode.sh`
38
39 url=`curl -d format=$TYPE -d code2="$DATA" -d poster=$USER -d expiry=expiry_day -d paste=Send -i -s http://babil.pastebin.com | grep Location | sed 's/Location:[[:blank:]]\+//g;'`
40
41 echo -e "${magenta} ${url} ${reset}"
42
43

============
getbin.sh
============

1 #!/bin/bash
2
3 bold='\e[1;1m';
4 white='\e[1;37m'
5 yellow='\e[1;33m'
6 magenta='\e[1;35m'
7 blue='\e[1;34m'
8 red='\e[1;31m'
9 cyan='\e[1;36m'
10 reset='\e[0m'
11
12 if [ -z $1 ];then
13 n=1
14 else
15 n=`echo $1/1 | bc`
16 fi
17
18 keys=`wget -nv -q -O - http://babil.pastebin.com | grep -E 'ago</li>$' | awk -F '"' '{print $2}'| awk -F '/' '{print $NF}'`
19
20 if [ -z "$keys" ]; then
21 echo
22 echo "[*] no saved post found in pastebin.com"
23 exit
24 fi
25
26 for i in $keys
27 do
28 echo -e "[[${cyan} $i ${reset}]]"
29 wget -nv -q http://babil.pastebin.com/pastebin.php?dl=$i -O - | uudecode -o -
30 n=$((n-1))
31 if [ $n -eq 0 ];then
32 exit
33 fi
34 done
35
36

100% cpu on Gmail.com with Ubuntu+Compiz

2009-01-13T17:22:00.001+11:00

I have both greasemonkey and stylish installed and yet I fell for "Better Gmail 2" extenstion. Apparently it turns my firefox into a total memory hog when I'm checking my mails at Gmail.com. Just incase anybody is having the same problem, disabling or uninstalling the extension fixes the problem for me.

Bash scripting :: Handling filenames with spaces

2008-12-29T23:24:00.000+11:00

The following script moves a specified amount of files from the currect directory to a specified directory. I had a lots of images saved in one folder. Accessing the folder used to take a long time. So I decided to put the files chronologically in seperate folders. The challenge was - lots of my files had spaces in their names. So, a normal "for i in `ls -t`" does not work. So, I came up with the following work around. So far it works pretty good ;)

1 #!/bin/bash
2
3 my_space="XXYYZZ"
4 me=`basename $0`
5
6 if [ -z $1 ] || [ -z $2 ];then
7         echo
8         echo "[*] usage:  $me  dirname how_many"
9         echo
10         exit
11 fi
12
13 if [ ! -d $1 ];then
14         echo
15         echo "[*] '$1' does not exist."
16         echo
17         exit
18 fi
19
20 ii=`ls -t1 | grep -v $me | sed "s/ /$my_space/g"`
21
22 c=0
23 for i in $ii
24 do
25         c=$((c+1))
26         n=`echo $i | sed "s/$my_space/ /g"`
27         echo -n "[$c]  "
28         mv -v -f "$n" $1
29         if [ $c -ge $2 ];then
30                 exit
31         fi
32 done
33

Ubuntu/Linux force monitor to turn on/off

2008-12-23T11:25:00.002+11:00

DPMS (Display Power Management Signalling) for vesa compliant devices can be controlled by two tools on linux:

xset
vbetool

man xset/vbetool for more details.

To force the monitor to stay turned on or off can be achieved by doing the following:

root@kubuntu:/home/babil# xset dpms force on/off
OR,
root@kubuntu:/home/babil# vbetool dpms on/off

Generating initrd on ubuntu

2008-12-16T15:21:00.001+11:00

compile the kernel

cd /usr/src/dccp_exp

make ; make modules ; make install ; make modules_install
generate initrd

update-initramfs -k 2.6.xx-xx -c
update grub entries to reflect the new kernel

update-grub

Reading BIOS

2008-12-12T14:11:00.001+11:00

dd if=/dev/mem bs=1k skip=768 count=256 2>/dev/null | strings -n 8

Awk pattern matching over a particular field

2008-11-27T12:09:00.001+11:00

1 #!/bin/bash
2
3 echo babbbbil 12223 | awk '{for(i=1;i<=NF;i++){if($i ~ "122"){print $i;i=NF+1}}}'
4

This piece of awk code will go through all the fields and look for pattern "122" and print them. The "i=NF+1" will work like "break;" in C which will make awk to stop printing after the first hit. Happy Awk'ing ;)