Programmer blog

Git tag and versioning your project

2013-04-24T15:46:00.001+03:00

We use tags to version our project. I will set up here basic commands we use on everyday basis and may use to handle some issues.

Showing your tags

Listing the available tags in Git is straightforward. Just type git tag.

$ git tag
1.0.7
1.0.8
1.0.9
1.1.0

Adding tags

To add tags you may do git tag tagname. But better to specify args to be able to use in scripts:

$ git tag -a 1.1.1 -m "major improvements"
$ git tag
1.0.7
1.0.8
1.0.9
1.1.0
1.1.1

Here we have added a tag version 1.1.1 with commit message "major improvements".

Uploading to repository

I assume you have a GitHub repo. So to push your tags there run git push (for your tagged commits commits) and then push your tags:

$ git push
Counting objects: 120, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (59/59), done.
Writing objects: 100% (69/69), 390.70 KiB, done.
Total 69 (delta 41), reused 0 (delta 0)
To ****************.git
   ******..******  master -> master
$ git push --tags
Counting objects: 1, done.
Writing objects: 100% (1/1), 157 bytes, done.
Total 1 (delta 0), reused 0 (delta 0)
To ****************.git
 * [new tag]         1.1.1 -> major improvements

You may now check this is pushed by clicking your GitHub repo's Tags button.

Removing a Tag

In case you have pushed a wrong tag... For example, if you have created a tag called push in a Git repository you would remove it from your repository by doing the following:



git tag -d push
git push origin :refs/tags/push

Now it's up to you to handle it in a release/deployment/staging/whatever script. Hope it helps. Besides, this should handle 99% of your tasks with tags. If not feel free to use official docs ;)

Comments and suggestions are welcome.

Django 1.5 migrating to new url tag syntax

2013-04-04T17:58:00.001+03:00

Django 1.5 changed. The url tag syntax from {% url my_page %} into {% url "my_page" %}. Now all my templates should be changed accordingly to migrate to 1.5. You could do it with executing this command in your bash shell.
Assuming you are in the project folder. Assuming you have version control, or do know how to backup your work ;). Some people claim it is unsafe. But you may want to make it safer with adding '{%' to the beginning of the string. To make sure only templates arre changed. Or either execute this only in your templates folder to be sure.

$ find . -type f -print0 | xargs -0 sed -i 's/ url \([^" >][^ >]*\)/ url "\1"/g'

And one person did a django snippet for this operation. You may also find it useful. http://djangosnippets.org/snippets/2905/

Automating everyday system routine with Fabric (Python)

2013-02-17T14:05:00.003+02:00

Lets talk about your console work. I use it on everyday basis. I need to log on to my deployment/stage/whatever, server and do some redundancy. E.g. download some logs, clean up some caches and/or redeploy something. Here is the occasion, when Fabric comes handy. You can eliminate all the redundancy and cover tons of operations you need to do every day, using console, shortcutting them to one of your simple commands, like: fab deploy_production -H root@mydeployment.com interested? Let's move on then. Notie this command is not a masterpiece, but must give you the understanding of usual workflow.

1. Installation and purposes

Let's go ahead, installing Fabric and automating some simple operations. You can read official docs about alternative methods. But installation is fairly simple. And is the matter of typing:

$pip install fabric

Nothing more special is required. You will need fabric at system wide scale. I install it locally with superuser privilegies. Note I'm assuming installation of this on my MacBook and not on a production server or any remote destination. Fabric really is a shell wrapper with commands queue that executes your console tasks. Something similar to Django unit tests, if you know what I mean. E.g. Developing a fabric script is a process of copy-paste your console interactions and here you go... Similar to BASH/SH scripts. But why use fabric then?

How would you handle commands exit status in shell or bash? Write some error handling right? And this one comes out of the box with it ;)
Let's not just talk, but write some

2. Example usage

To write your commands you need to make a file called fabfile.py with syntax that is described in fabric documentation. But it's almost "pure" python. So everything must be understood easily.

Fabric supports all the python stuff. And good IMHO practice it to split your task to steps and execute them one by one. You can use def method to define them. E.g.:

"""Test script for article"""

from fabric.api import run

def set_up():
    """does some setup automation"""
    run('setup something')

def run():
    """main run method, that now calls set_up() method"""
    set_up()

Now you can run a command fab --list in this directory. Output will be like so. Note how it converted all your comments into handy help text explanation for your script.

leopard-2:article garmon$ fab --list
Test script for article

Available commands:

    run     main run method, that now calls set_up() method
    set_up  does some setup automation
leopard-2:article garmon$

Now we have got a handy script to setup something. But let's not get away with it and show you some real life example. It will also be converted to hide some (possible) sensitive information...

3. Real life example

"""Example script for article with deployment parts"""

from fabric.api import run, sudo, cd, get, local, lcd, prefix

def cleanup_environment(folder_path):
    """ deletes old deployment instance"""
    sudo('rm -rf %s' % folder_path)

def install(folder_path):
    """ Installs an instance of my website into production deployment server with recreating a virtual environment"""
    with cd(folder_path):
        sudo('virtualenv ve --no-site-packages')
    with prefix('source %sve/bin/activate' % folder_path):
        sudo('pip install mysite')

def restart():
    """Restarts a service of my website"""
    sudo('service mysite restart')

def stop():
    """Stops a service of my website"""
    sudo('service mysite stop')

def run():
    """Runs a service of my website"""
    sudo('service mysite run')

def download_logs(folder_path):
    """ Download logs files and db from server to local machine current directory
        This will download all the logs from deployment server
        into your current directory (locally) with current timestamp"""
    import datetime
    time_stamp = datetime.datetime.strftime(datetime.datetime.now(), "%Y-%m-%d-%HH-%MM-%SS")
    with cd(folder_path):
        sudo('tar -czvf /tmp/dms_logs.tgz ./log/dms.log*', user='user')
    local('mkdir ./%s/' % time_stamp)
    with lcd(time_stamp):
        get("/tmp/dms_logs.tgz", "./dms_logs.tgz")
        local("tar -xzvf dms_logs.tgz")
        local("rm -f dms_logs.tgz")
    sudo("rm -f /tmp/dms_logs.tgz")

def main():
    """Redeploy your Django powered website"""
    deployment_path = '/srv/www/django'
    download_logs(deployment_path)
    stop()
    cleanup_environment(deployment_path)
    install()
    run()

This script has main commands for me to run make redeployment of my website into a production web server, with archiving all the data (logs in this example) into a directory with current timestamp.

You must run it with a parameter -H specified. this parameter represents a host that our little script will ssh to. and is something like username@192.168.1.1, or something you can ssh to, using command ssh. E.g. -H stage-server in case you can locally type ssh stage-server. So the full command to run this will be something like: fab main -H my_deployment_server

Most fo the code in this script is given as an example. And may contain some logical errors, because it is taken from different parts of a real life deployment commands. But I really hope it will help you to master your own deployment with this handy tool.

Comments and suggestions are welcome...

Raspberry Pi boot applications Autorun

2013-01-04T17:28:00.001+02:00

I had a problem with running required programs upon system startup. I had to set up hdparm utility each time system boots up. I have 2 external HDD's connected and require setting their sleep time for 10 minutes each boot. So setting this up. The answer is found at Debian administration guides. I'll provide it in the end of the article. Here is my decision based on that:

So to set up a program to run on system boot. (In fact one of the system run-levels). You can add it's name. But a good practice will be to add an sh script with execution of this utility an all the parameters.
Sample script is:

#! /bin/sh
# /etc/init.d/blah
#

# Some things that run always
touch /var/lock/blah

# Carry out specific functions when asked to by the system
case "$1" in
  start)
    echo "Starting script blah "
    echo "Could do more here"
    ;;
  stop)
    echo "Stopping script blah"
    echo "Could do more here"
    ;;
  *)
    echo "Usage: /etc/init.d/blah {start|stop}"
    exit 1
    ;;
esac

exit 0

You can insert your utility name instead of "blah" and give it a spin. My script looked like:
And make sure it is owned by root user:

sudo chmod 755 /etc/init.d/blah
sudo chown root:root /etc/init.d/blah

So my script began to look like:

#! /bin/sh
# /etc/init.d/hdparams
#

# Carry out specific functions when asked to by the system
case "$1" in
  start)
    echo "Setting Connected HDD sleep timeout to 10 Minutes."
    hdparm -S 120 /dev/sda1
    hdparm -S 120 /dev/sdb1
    ;;
  stop)
    ;;
  *)
    echo "Usage: /etc/init.d/hdparams {start|stop}"
    exit 1
    ;;
esac

exit 0

Because I really don't care about hdd spin down time at shut down and/or system reboot.
Original article is: http://www.debian-administration.org/articles/28
Hope it helped somebody. Please comment if you used/found some inconsistency...

Raspbery Pi as a home file server

2012-12-30T21:10:00.001+02:00

I've wanted more from my PI, besides Time Machine functions, established in my earlier articles. So I've decided to make it a SAMBA server. I have a Mac and a Windows PC. And several iPads/iPhones. So the decision was obvious. To access my storage content I'd need a samba share. Because it is recognized by all this tech. iPad has Oplayer, to watch movies without conversion... GoodReader for docs. One word - SAMBA share. SO this article will be about adding a samba share to your Raspberry Pi.
Note I have a Time Machine already set up. But it won't matter much. I only have another drive index letters and so on. I'll try to cover this as much as possible. And macs somewhat easily read NTFS filesystems. So we will be mounting NTFS filesystem volume to a Raspberry Pi with SAMBA network sharing. Let's get on to it.

To set up a Raspbery Pi home network Samba server you will need to:
- Connect an external drive (USB HDD in my case).
- Update fstab for auto-mounting of it on system boot
- Install and configure samba
- Enjoy

1. Connecting an external USB drive.

Assuming you hava an external hard drive connected type the command:

pi@raspberrypi /mnt $ sudo blkid
/dev/mmcblk0p1: SEC_TYPE="msdos" UUID="XXXX_XXXX" TYPE="vfat" 
/dev/mmcblk0p2: UUID="XXXXXX_XXXXXX_XXXXX_XXXX" TYPE="ext4" 
/dev/sda1: UUID="XXXXXXX-XXXXXXX-XXXXXXX-XXXXX" LABEL="Time Machine" TYPE="hfsplus" 
/dev/sdb1: LABEL="Data1" UUID="5XXXXXXXXXX1" TYPE="ntfs"

Note I have sda1 drive with hfsplus filesystem for AFP shares so my NTFS drive is named sdb1. Yours will probably be sda1 out of the box. Note that. We now see the NTFS hdd. Lets make it mounted. Typing mount command will show it's not yet there.

pi@raspberrypi /mnt $ mount
/dev/root on / type ext4 (rw,noatime,user_xattr,barrier=1,data=ordered)
# ...
/dev/sda1 on /mnt/TimeMachine type hfsplus (rw,nosuid,nodev,noexec,relatime,umask=0,uid=0,gid=0,nls=utf8)

2. Update fstab for auto-mounting of it on system boot

Lets make a directory for mounting a drive:

pi@raspberrypi /mnt $ sudo mkdir /mnt/data

Now lets update fstab to add a setting for proper auto-mount of our drive on boot.

pi@raspberrypi /mnt $ sudo nano /etc/fstab

Lets add a line. Something like.

UUID="XXXXXXXXXXXXXX" /mnt/data ntfs rw,auto 0 0

Where UUID is your device UUID, that you can find out using sudo blkid command.

Now we can either reboot or mount a hard drive. I choose to mount. Lets type:

pi@raspberrypi /mnt $ sudo mount /dev/sdb1
pi@raspberrypi /mnt $ mount
/dev/root on / type ext4 (rw,noatime,user_xattr,barrier=1,data=ordered)
# .....
/dev/sdb1 on /mnt/data type fuseblk (rw,nosuid,nodev,relatime,user_id=0,group_id=0,allow_other,blksize=4096)

And our drive is mounted and will be auto-mounted on each reboot.

3. Install samba

Now lets move on to installing samba server for our network sharing. Lets do it with apt-get:

pi@raspberrypi /mnt $ sudo apt-get install samba

Confirm additional package downloads by typing y to the prompt of this command. And wait while it installs some demons and defaults.

Now lets edit the default samba config:

pi@raspberrypi /mnt $ sudo nano /etc/samba/smb.conf

And add the config of our drive to the end of this config file. (After numerous other parameters):

[Data]
comment = DataDrive
read only = no
locking = no
path = /mnt/data     
guest ok = yes
force user = pi

After this is done simple restart of samba server will be sufficient:

pi@raspberrypi~$ sudo service samba restart

4. Enjoy

Now I have a file dump server I ever dreamed of. I can connect as guest with all my devices and have a cross platform decision built on a PI platform.

And this all stuff looks like this now. With a second ntfs external HDD and my Pi in Punnet MK1 paper case.

Raspbery Pi Time Capsule HDD managing power saving mode

2012-11-27T19:45:00.003+02:00

I want to sugest power saving changes after installing and handling creation of Time Capsule for Time Machine for my home network (form my previous article). I like when it's quiet. And like when disks does not spin, when it's not necessary to do so. So if you usually backup once in some hours and main time (Night and a part of the day) you do not touch a Time Machine... You would probably want to slow down your disks for that time. Fortunately Linux does it very well. And support of it is encoded into Linux core (kernel). You just would need a UI for this possibilities and a bit of knowledge and luck ;)

Warning: Operations described farther may be potentially hazardous to your hardware.

So now after you are warned ;) let's continue. I guess I will not touch those hazardous functions, but you may, typing wrong letters to this command in Linux shell... So be careful. Better read manuals for farther commands and so on...

Anyway what we need is a UI. And we have it in form of nice utility called hdparm. It is not included in current Raspbian distribution (and it should not be perhaps) so you would probably need to install it by yourself. In fact you can find a package with it. But I'd recommend installing from latest version. It is http://hdparm.sourceforge.net/ situated. You can download it there. Using wget, for e.g. So move on to downloading and unpacking it. If you have a default Raspbian kernel, or did not screw up something useful while configuring and compiling your on one... You would probably simply enter a directory and type something like this:

pi@raspberrypi ~ $ wget http://switch.dl.sourceforge.net/project/hdparm/hdparm/hdparm-9.43.tar.gz
pi@raspberrypi ~ $ tar -zxvf hdparm-9.43.tar.gz
pi@raspberrypi ~ $ cd hdparm-9.43
pi@raspberrypi ~/hdparm-9.43 $ ./configure
pi@raspberrypi ~/hdparm-9.43 $ make
pi@raspberrypi ~/hdparm-9.43 $ sudo make install
pi@raspberrypi ~/hdparm-9.43 $ hdparm

hdparm - get/set hard disk parameters - version v9.43, by Mark Lord.

Usage:  hdparm  [options] [device ...]

Options:
...

Hurray! we have our desired utility. Reading through it's help you would probably find a useful for us paramether: -S Set standby (spindown) timeout
Here is our goal. I've set mine to 120 (10 minutes) by executing:

pi@raspberrypi ~ $ sudo hdparm -S 120 /dev/sda1

/dev/sda1:
 setting standby to 120 (10 minutes)
pi@raspberrypi ~ $

I have a WD "green" drive and will benefit from this. But you may harm your drive with often standbys and spinups. So choose wisely and read man hdparm for more ;).
Hope you would benefit from my finding. Please comment in either way.

UPD: to persist this upon boots see article:
http://garmoncheg.blogspot.com/2013/01/raspberry-pi-boot-applications-autorun.html

Time Capsule for $25

2012-11-27T19:06:00.002+02:00

The real article name might be something like: Configuring Raspbery Pi to serve like a Time Capsule with Netatalk 3.0 for Mountain Lion. But it's too long ;)

Here I will describe the process of using Raspberry Pi like a Time Machine in my network. To be able to backup your MAC's remotely (Like it would be NAS of some kind). It assumes you have a Raspberry Pi and have installed a Raspbian there and have a ssh connection, or somehow having access to it's console. Refer to my previous article for details.

Now that we have a Pi that is ready for action let's animate it. So to make it suit you as a Time Capsule (NAS) for your MAC's you need to do those basic steps:
- connect and configure USB hard drive(s)
- install support of HFS+ filesystem to be able to use MAC's native filesystem
- make mount (auto-mount on boot) of your hard drive
- install Avahi and Netatalk demons
- configure Netatalk daemon to make it all serve as a Time Machine
- configure avahi demon
- put avahi and netatalk demons into autolaunch on system boot
- ENJOY.

1. Setup a Hard drive.

I assume you know what you are doing and understand you have to have a hard drive formatted under your mac's HFS+ filesystem. That you, like me, used as an external hard drive for backups before this. Just for example. Or bought it... Whatever. One word you will need (or assume you have) an external USB drive to follow me farther.

Start your drive (in case it has external power). I think Pi would handle only drives with external power. But who knows ;). And connect it to your Pi. Let it spin for some seconds to be recognized.
Now connect to console of your Pi and let's roll.

leopard$ ssh pi@192.168.1.140
pi@192.168.1.140's password: 
Linux raspberrypi 3.2.27+ #250 PREEMPT Thu Oct 18 19:03:02 BST 2012 armv6l

The programs included with the Debian GNU/Linux system are free software;
the exact distribution terms for each program are described in the
individual files in /usr/share/doc/*/copyright.

Debian GNU/Linux comes with ABSOLUTELY NO WARRANTY, to the extent
permitted by applicable law.
Last login: Mon Oct 29 01:04:39 2012
pi@raspberrypi ~ $

As you can see I have a clean system installed from latest (as for 24.11.2012) downloaded image.

Let's do some magic. Because Raspbian does not support Apple's filesystem out of the box. At least we need to mount it write enabled. Which is not set up by default too. SO let's start from typing command blkid under root and looking at output.

pi@raspberrypi ~ $ sudo blkid
/dev/mmcblk0p1: SEC_TYPE="msdos" UUID="8B12-9112" TYPE="vfat" 
/dev/mmcblk0p2: UUID="AAAAAAA-1111-1111-......." TYPE="ext4" 
/dev/sda1: UUID="AAAAAAA-1111-1111-......." LABEL="Time Machine" TYPE="hfsplus" 
/dev/sda2: LABEL="Data" UUID="1234567890......." TYPE="ntfs" 
pi@raspberrypi ~ $

Note UUID of your device. Mine is called sda1 and has a partition type hfsplus. Hurray! Drive is connected and is recognized by Pi.

2. Install support of AFP filesystem (MAC's native).

Now let's install tools we need for HFS+ filesystem mounting. We need to mount it RW to be able to backup.

pi@raspberrypi ~ $ sudo apt-get install hfsplus hfsutils hfsprogs

Press "y" in the process of install. You will get HFS+ support libraries installed. So let's move on to mounting.
After this you should be able to mount your HDD. You may try it or skip to making permanent mounting config via /etc/fstab (Step 3).

pi@raspberrypi ~ $ sudo mount -o force /dev/sda1 /mnt/TimeMachine
pi@raspberrypi ~ $ cd /mnt/TimeMachine
pi@raspberrypi /mnt/TimeMachine $ ls -l
total 116
drwxr-xr-x 1 root      99      6 Jan 11  2012 Backups.backupdb
drwxrwxrwx 1  501 dialout     31 Feb  1  2012 !DVD-s
-rwxr-xr-x 1 root root    115716 Sep  1 07:23 tmbootpicker.efi
pi@raspberrypi /mnt/TimeMachine $

You may see your files from a HDD. Hurray! Again ;). Btw you can look at how is your device was mounted by typing mount command into prompt. It will show your mounted volume and mode it was mounted. So you will see something like this:

pi@raspberrypi /mnt/TimeMachine $ mount
/dev/root on / type ext4 (rw,noatime,user_xattr,barrier=1,data=ordered)
devtmpfs on /dev type devtmpfs (rw,relatime,size=118872k,nr_inodes=29718,mode=755)
tmpfs on /run type tmpfs (rw,nosuid,noexec,relatime,size=23788k,mode=755)
tmpfs on /run/lock type tmpfs (rw,nosuid,nodev,noexec,relatime,size=5120k)
proc on /proc type proc (rw,nosuid,nodev,noexec,relatime)
sysfs on /sys type sysfs (rw,nosuid,nodev,noexec,relatime)
tmpfs on /run/shm type tmpfs (rw,nosuid,nodev,noexec,relatime,size=47560k)
devpts on /dev/pts type devpts (rw,nosuid,noexec,relatime,gid=5,mode=620)
/dev/mmcblk0p1 on /boot type vfat (rw,relatime,fmask=0022,dmask=0022,codepage=cp437,iocharset=ascii,shortname=mixed,errors=remount-ro)
/dev/sda1 on /mnt/TimeMachine type hfsplus (rw,relatime,umask=22,uid=0,gid=0,nls=utf8)

Note last line with /dev/sda1 mounted in mode "rw". Yes everything works well. Let's move on.

3. Make mount (auto-mount on boot) of your hard drive

Let's create a directory for our new tome. We will need this to mount filesystem.

pi@raspberrypi ~ $ sudo mkdir /mnt/TimeMachine

Lets unmount our tome (in case you did tested it exists with me).

pi@raspberrypi ~ $ sudo umount /dev/sda1

And add a proper fstab entry to auto-mount this hard drive on boot. Open fstab:

pi@raspberrypi ~ $ sudo nano /etc/fstab

And add a line to the end of the file, indicating our Time Machine tome mounting preferences. Mine was:

UUID="AAAA-BBBB..." /mnt/TimeMachine hfsplus rw,force,exec,auto,users 0 3

Where AAAA-BBBB... is your devise UUID from that blkid command remember?
I used UUID to be sure it won't matter if I'll switch hard drives places in USB ports. But you may use /dev/sdXX, where XX is your device counting numbers in /dev. Mine in this case was /dev/sda1.

Note: You can skip reboot.

Let's now do a reboot to make sure it all spins in automatic mode ;).

pi@raspberrypi ~ $ sudo reboot

Broadcast message from root@raspberrypi (pts/0) (Wed Nov 21 19:13:49 2012):
The system is going down for reboot NOW!

After system boots up run mount command to make sure we are ok in rw mode. Btw you may play with commands:

# to unmount volume:
pi@raspberrypi ~ $ sudo umount /dev/sda1
# to mount again:
pi@raspberrypi ~ $ sudo mount UUID="AAAA-BBBB..."

While you will find your string. BTW you can read about how and why in /etc/fstab manual here... Or google about it.

Note: if your HFS+ tome does not mount and/or mounted read only. You should try to run fsck for hfsplus partition type manually.

pi@raspberrypi ~ $ sudo fsck.hfsplus -f /dev/sda1

Anyway we must now have an rw mode HFS+ volume mounted and ready for our Time Machine. So Lets move on...

Checking your user permissions. As we will use user "pi" farther along the article we must add him ability to manage Time Machine data. (In case you are not mounting empty drive). You should consider changing all the data owner to user "pi" to make sure we will be abler to backup in future:

pi@raspberrypi ~ $ sudo chown -R pi /mnt/TimeMachine

4. Install and configure Avahi and Netatalk demons

First make sure you have the latest packages lists, running:

pi@raspberrypi ~ $ sudo apt-get update

And updating what necessary.

Now install all the required packages with db and encryption support in order Avahi and Netatalk demons to support for HFS+ filesystem:

pi@raspberrypi ~ $ sudo apt-get install avahi-daemon libavahi-client-dev libdb5.3-dev db-util db5.3-util libgcrypt11 libgcrypt11-dev

After this is complete we will have almost everything required for our little demon.

Download and unpack 3.0.0 (latest, as for now is 3.0.1 but I had problems with it spinnig) Netatalk demon sources. For e.g. using command: $ wget http://somesite.com/photos.zip You may get Netatalk at SourceForge. And unpack using e.g. tar: $ tar -xvf netatalk-3.0.0.tar.bz2

After you are ready with unpacked netatalk distribution go ahead and make custom configuration of it. Enter directory of unarchived netatalk and execute command:

pi@raspberrypi ~/netatalk-3.0 $ ./configure --with-init-style=debian --with-zeroconf

It will do the proper configuration for our needs. After this is done run:

pi@raspberrypi ~/netatalk-3.0 $ make
pi@raspberrypi ~/netatalk-3.0 $ sudo make install

5. Configure Netatalk daemon to make it all serve as a Time Machine

Now when you will have all this done go edit a configuration file of Netatalk. From version 3.0 it is located in /usr/local/etc and has Samba-like look. You can symlink this file to /etc/afp.conf if you wish...

pi@raspberrypi ~ $ sudo nano /usr/local/etc/afp.conf

And insert this sample configuration. This config file works with latest edition ( now it is 10.8.2 ) of OS X Mountain Lion. So the config:

;
; Netatalk 3.x configuration file
;

[Global]
; Global server settings
; Name of your computer in apple devices network
hostname = TimeMachine Pi
; IP of your Pi
afp listen = 192.168.1.140
; logging config
log file = /var/log/netatalk.log
log level = default:info

[Homes]
basedir regex = /home
cnid scheme = dbd
; Display each user home directory in this format
home name = Home: $u
[Time Machine]
; Our Time Machine volume
path = /mnt/TimeMachine
cnid scheme = dbd
file perm = 0660
directory perm = 0770
time machine = yes

; Example to add a new static share:
; [My AFP Volume]
; path = /path/to/volume

It is quite well commented so you will be able to copy-paste and edit ;).

6. Configuring AVAHI services

Now you would probably want to have a Time Capsule as a separate share in your network shares. Note this step is optional. But if you plan to use your pi for more than 1 thing, like I do... I'd recommend to do it. Create a special afpd services configuration file by executing:

pi@raspberrypi ~ $ sudo nano /etc/avahi/services/timecapsule_afpd.service

Then add this contents to file:

<?xml version="1.0" standalone='no'?>
<!DOCTYPE service-group SYSTEM "avahi-service.dtd">
<service-group>
      <name replace-wildcards="yes">TimeCapsule %h</name>
      <service>
          <type>_afpovertcp._tcp</type>
          <port>548</port>
      </service>
      <service>
          <type>_device-info._tcp</type>
          <port>0</port>
          <txt-record>model=TimeCapsule</txt-record>
      </service>
</service-group>

After this is configured for your desired settings. Go run your demons:

pi@raspberrypi ~ $ sudo /etc/init.d/netatalk start
pi@raspberrypi ~ $ sudo /etc/init.d/avahi-daemon start

Avahi demon was installed from package. So you might need to actually run restart instead of start command. Now go check your Mac's Network Shares.

7. Put avahi and netatalk demons into autolaunch on system boot

We have installed here an avahi demon from package. So it must be written to all system run-levels during installation. Only thing here is compiled Netatalk 3.0 so you can make a proper trick here by putting it to run with default init.d command:

pi@raspberrypi ~ $ cd /etc/init.d/
pi@raspberrypi ~ $ sudo update-rc.d netatalk defaults
update-rc.d: using dependency based boot sequencing
update-rc.d: warning: default stop runlevel arguments (0 1 6) do not match netatalk Default-Stop values (1)
pi@raspberrypi /etc/init.d $

You can read about system runlevels and why this command and not another in this article:
http://www.debuntu.org/how-to-manage-services-with-update-rc.d I find it nice simple and easy.
You can also read about Time Capsule theory and find answers to many "why?" questions here: http://buffalo.nas-central.org/wiki/Time_Machine_&_Time_Capsule_support_on_your_LinkStation

Result:

TADAM. The moment of truth. You can look into your Network Shares on a MAC. Select a virtual Time Capsule and hit "Connect as..." button. Login to your raspberry with user pi and saving password. Enter Time Machine directory... And Time Machine share will appear on a desktop (If it is configured to appear for newly mounted volumes...). You will experience something like this:

After all of this is done. Go enter your Time Machine settings and add a new drive. Bckuping on a time capsule is some kind of different to backing up into external (e.g. USB) hard disk drives. Using external HDD as a Time Machine backup disk will store folder Backups.backupdb into the root of your drive. But backup into time capsule (network backup) creates a volume like leopard.sparsebundle, where this leopard is your MAC's name. And connects it as a file system image, storing this folder inside. So to make initial backup connecting through USB and then continue making backups is possible but irrelevant. Because you would have to mount this image and change permission of all files. Then move them into mounted virtual image under Pi or over network. So the advantages of this idea are controversial.

It was looking like so for me:

And my pi, serving like a Time Capsule and external HDD:

Please comment your thoughts about all of this. Every comment is appreciated. Thanks. And hope someone would benefit from those my findings.

Raspberry Pi first steps and basic network configuration on a Mac

2012-11-19T10:44:00.002+02:00

Here are my first steps. And I hope you will find something useful here, while configuring your Pi...
First of all. Mine have been bought on ebay, from resellers. And were delivered a while ago. Main purpose of this purchase war to attach a headless server to my router. I have 2 external HDD's and would like to have torrents, Time Mashine fro my macs and so on. So buying a handheld computer like this would be a bargain for me.

As for built in decisions like some kinds of NAT devices and different routers with external HDD features... They are either cost a lot or lack some kinds of desired functionality. So the goal is to make some kind http/api manageable server in my local network with Time Mashine and file storage/backup. Just for fun. And to have only laptop on my work table.

SO back to the Pi. If you are buying "device only" configuration, like I did. First of all you'll need some different kind of things many computer fans usually have for backup purposes, or just in case... Those are device with HDMI port or either a TV set with AV input, e.g. SCART or Video in port. A keyboard (USB, of course) and SD card usually bigger then 2 Gigabytes.

I had this all. So I've connected my PI and tried to flash a card, having problems doing it. I'll try to cover most of them here.

1. Install a system. Usually it is a simple "burning an SD card" with Raspbian Linux distribution. To do this you need to download it from: HERE and do actions from official manual here with looking further note first.

Problem: "dd" does not work like intended, because volume is bisy.
Solution: We need to unmount volume, using diskutil command with "force" flag for e.g.:
sudo diskutil umount force /Volumes/my_flash_drive

Problem: Doing "dd" command under mac environment did write the card, but took ages on my mac's config. So the...
Solution: Is modified dd command call with manually specifying block size. This will speed up sd card write tremendously.
dd bs=1m if=~/Downloads/debian6-19-04-2012/debian6-19-04-2012.img of=/dev/rdisk1
Note "bs=1m" here. It helped my SD card burning not to hang out to infinite time.

After SD card was written and you are good to go...

2. Pi initial config. Appears after first boot and looks like so:

You can run it afterwards using command "sudo raspi-config". Just in case you would missed something hitting Esc or similar. It will be available via ssh in future also.

Tools you will use here are similar to bios config on a PC. So to get started:

Here you need to expand rootfs (to use entire SD card space) I doubt you have original Pi Raspbian card. But you may skip this step in case you do. This will be executed on next start. So feel free to do it first thing.

Next important option is change pass and you need it no-mater what. It will change your default user's with username pi password. It is similar to user with root permissions... So be sure to do it. Note password symbols are not displayed during entry.

change timezone goes next. If you do not leave in other timezone, like I do, change it. ;) Note pi does not have an internal "bios" clock so you would need to do it. Because time is used from internet atomic clocks...

memory split is also important option that help you to manage precious pi's RAM. I've used 16 because I will not need GUI and will run "headless", using it via ssh/network services. To use GUI consider 32+ Mb or RAM for video memory. But note RAM is really in lack for most modern services. and you may need every Kb of it in future.

ssh is our main goal. To access it/configure it in future. Hit it to enable ssh connections.

and you can run update if you have already connected network somehow :).

That's it with basic config for our needs. Here goes...

3. Static network configuration.
I think this is the most important part. Because I have googled a lot to find a proper solution for my needs.

Pi is connected via dhcp IP address, usually given randomly by your router. And you need to make it static to know you will always have pi under certain network address.

You may know your Pi's ip address by entring route command into prompt. Yo will see something like this.

pi@raspberrypi ~ $ route
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
default         192.168.1.1     0.0.0.0         UG    0      0        0 eth0
192.168.1.0     *               255.255.255.0   U     0      0        0 eth0

Here you can see your network is in range 192.168.1.x. Biut if you are configuring your Pi without actual network config, like I did, you will need to do something like this:

You need to know it from your router config. And you can simply connect it via network, if you did it properly in past. Or either connect, like you did for initial router config. You need something like this, saying about your LAN network address. I have a wired mix of different network decisions in my router. So note your typical network router address here:

Here you can know your router's DHCP address. And usually you can do it by examining your router's user manual.
Anyway this tells me that I have a wire (LAN) IP network with adress range 192.168.1.x where x is between 1 and 254. Your DHCP in router may work in another way and have different interface. So you may have other IP range set. E.g. Mine had a 100-199 range by default:

I've chosen a 192.168.1.140 IP for my PI, because I would not have 40 devices in my network and would not have any possible IP conflicts this way.

This are data required for our Pi network config. We need to manually change this in our Pi's config files.

Try to do this:

cd /etc/network
sudo nano interfaces

This will start default Pi's text editor, called "nano", pointing to networks configuration file. Here is how it might look like:

You can notice nice hotkeys here in the bottom. So I think you will be pleased with this user interface. Let's replace the line iface eth0 inet dhcp with lines.

iface eth0 inet static
address 192.168.1.140
netmask 255.255.255.0
gateway 192.168.1.1

This will create a static address for our Pi. Now to make sure we have a proper DNS. Go to editing /etc/resolv.conf. E.g. type:

sudo nano /etc/resolv.conf

Mine had a a line:

nameserver 8.8.8.8

Pointing to google's DNS. And I've left it, adding my router as a first DNS. So my resolv.conf became looking like:

nameserver 192.168.1.1
nameserver 8.8.8.8

And that's it. After doing this you would have to restart network interfaces. Or better do "sudo reboot" to make sure everything is configured. At this point you are OK to go running headless already. You can now do ssh connections to static ip at your pi.

So to use it you would need only to type in your terminal:

ssh pi@192.168.1.140

This will connect you to your pi. And you are ready to go ahead.
you may see something like this:

And to finish here is how my Pi looks like in a "printable case" The Punnet from user "E". Thanks to him.

Migration from HDD to SSD with Mac OS X Lion

2012-11-03T17:21:00.000+02:00

Today I'll try to cover my migration experience. I have bought myself an OCZ Technology 128GB Vertex 4 SSD from Amazon. And I also have an external 1 Tb HDD. I'm using it for migration purposes as a backup Time Mashine. So to migrate yourself without reinstalling operating system and so on you need:

1. Compact your data. Usually HDD's, installed in Mac's, are much bigger then SSD's you buying. And to tell you the truth, you probably do not need a 512 Gigabytes SSD at all. Most of data are accessed rarely and do not require speedy storage. So COMPACT them. Move to an external HDD and/or delete. Time to cleanup one word. There are also several places that can be easily deleted without much troubles. They are iTunes backup caches for your iOS devices, like iPhones and iPads.
they are usually sotred on your user library path.
You can delete them without sorry from either:
~/Library/Application Support/MobileSync/Backup
folder or in your iTunes => Preferences => Devices section, hitting delete on each backup.

They will be created from scratch on next sync. So no worries about them. Your data are safe on your mobile devices. If you do not screw things up on next sync (:

Besides this you have ability to link files. For e.g. I have an iTunes library of around 100 Gigs of songs, books and etc. So storing it by default path is bad idea. My SSD only has 128 Gigabytes. Anyway I'll tell what to do with this later on. For now I'll just say you can move iTunes Libraries and relink them. And you can move iPhoto library. I have around 100 Gigs of photos too... So those 2 are moved to external HDD. Leaving me 200 gigabytes of data. After deleting iTunes backups I've fried around 60+ gigs. I have had several iPhones and their backups also were stored there. So the list was big.

Anyway. The file system started to look like this:

But I've fried this space, and made a backup. Why the hell I have no free space. And my Mac says I have only 70+ gigs of space used. The catch is in .MobileBackups folder. Mac stores your Time Mashine backups to local hard drive. Backing up data locally. What a mess :). You can disable it by hitting terminal command:

$ sudo tmutil disablelocal

this will disable local Time Mashine backups. TADAAM. We really have a hard drive with around 78 gigs of space used for system and files.

Now make a Time Mashine backup to external HDD, or second hdd, that will not be formatted and can be accessed during setup. Math is simple. Last backup must be smaller then your SSD capacity. 78 is quite ok for me.

In case you need to fit something on your HDD and have all the space used by Time Mashine backups, like I did, you can delete old Time Mashine backups in manual mode. Simply go to folder on your Time Mashine drive, called Backups.backupd => YourUserNameHere. You will see a set of date stamped folders. Those are Time Mashine backup dates.

You can free them up by deleting each folder and cleaning the trash. Note it will come to cleaning Trash with - amount of objects, e.g.: "deleting file -123". Don't bother with that and simply wait for cleanup.

Now that we have a most recent backup in Time Mashine of all our system HDD. We can proceed. Power of your Mac/Macbook and insert a SSD drive, if you have not done so before.

I have used an Second SSD/HDD drive Caddy that is around 10$ at eBay.

Looks something like this:

I've plugged my HDD there. And SSD came in place of HDD. But it's a metter of taste and you can do as you wish here.

Now time to boot. Boot, holding "Option" key. You will see the menu with OS boot choice, Looking something like that:

Choose Recovery HD. This will be mush simpler/quicker then making any external boots. But you may have to do it in case you do not have this option here. Anyway booting this option will lead you to a screen with recover options.

Choose Restore From Time Mashine Backup. Will lead you to the drive with your Time Mashine selection, in case you have several. Afterwards select your LATEST! backup. And choose where to restore to. Be careful to choose your SSD. And hit Fire. :)

You will wait several minutes, depending on your Time Mashine drive speed with this screen:

After load you will have fully restored system with all your data on a SSD.

Now be sure you have boot from an SSD ( It is usually hard to make a mush here because of boot speed :D ). But make sure you are on SSD now and go ahead to diskutil and format your HDD. You will not need it anymore. TADAM. You're done.

Thanks for reading this far and comments with your success are welcome, as long as critics apriciated ;).

Django Log Files Viewer documents

2012-09-25T14:53:00.002+03:00

Django Log File Viewer.

This is a PYPI package django-log-file-viewer documents.
Github repo: django-log-file-viewer@garmoncheg.github.com

Usage:

Useful to add log files view functionality to your Django admin web site.
Instead of using database log files storage, it gives you ability to store/view log files through GUI.
It requires a directory with Django log files to function. E.g. directory structure:

$ project_dir/logs/:
       applog.log
       applog.log.2012-09-22
       ...
       errors.log
       applog.log.2012-09-22
       ...

Screenshots:

To parse/display these log files you need:

1. Install an app and add it to your settings.py INSTALLED_APPS section:

# settings.py
INSTALLED_APPS = (
    # ...
    'django-log-file-viewer',
    # ...
)

2. Set UP 2 django variables in settings.py:

# settings.py:
LOG_FILES_DIR = '/path/to/your/log/directory'
    # Relative or static path string of your log files directory.
    # I recommend using more pythonic way of defining tis with os module. E.g. :
    # settings.py:
    #
    # LOG_FILES_DIR = os.path.join(APP_PATH, 'testdata', 'log')
    # 
    # where APP_PATH is your app's/project's path.

LOG_FILES_RE = '(?P<date>\d{4}-\d{2}-\d{2} \d{2}:\d{2}:\d{2},\d{3})\s\[(?P<type>[A-Z]+)\]\s(?P<message>.+)'
    # Is a regex to parse your log file against. It completely depends of your Django logging settings.
    # And table column names (in a parsed logfile) depend from group names you provide in the regexp.
    # E.g. for Django logging server to parse with this regexp you need to have log, as in example
    # django_log_file_viewer/testdata/testing.log file.

    # to produce this log I've added this formatter to my website.
    
    #'formatters': {
    #    'verbose': {
    #        'format': '%(asctime)s [%(levelname)s] %(message)s'
    #    },
    #},

3. And add urls to your main urls section:

# urls.py
urlpatterns = patterns('',

    # Include this before admin to enable app admin url overrides
    # Note url must be the same as admin
    # This is required step
    url(r'^admin/', include('django-log-file-viewer.admin_urls')),
    url(r'^admin/', include(admin.site.urls)),

    # To view with custom views:
    # Optional step
    # Will ad urls like www.example.com/logfiles/
    url(r'', include('django-log-file-viewer.urls')),
)

TODO's:

Add pagination to both log files list and log file content

Reference:

Info on how to add/config logging for your Django based website here: https://docs.djangoproject.com/en/dev/topics/logging/ (Official Django documentation)
Parsing/testing Regexp is useful with: http://gskinner.com/RegExr/ (real time regexp testing)

SSH Unix/Lunux Recipes for daily usage.

2012-09-19T21:22:00.002+03:00

I've decided to create a post with several recipes for commands I need to use daily. I use them from a Mac OS X Lion default console, except for some mac ports installed... Remote is special linux distribution.

How do I Compress a Whole Linux or UNIX Directory?

You need to use tar command as follows (syntax of tar command):

tar -zcvf archive-name.tar.gz directory-name
Where,

-z: Compress archive using gzip program
-c: Create archive
-v: Verbose i.e display progress while creating archive
-f: Archive File name

For example, you have directory called /home/garmoncheg/data and you would like to compress this directory then you can type tar command as follows:

$ tar -zcvf data.tar.gz /home/garmoncheg/data/

Above command will create an archive file called data.tar.gz in current directory. If you wish to restore your archive then you need to use following command (it will extract all files in current directory):

$ tar -zxvf data.tar.gz

Where,

-x: Extract files

Put or Retrieve a file from/to remote directory via SSH

For all those operations you need to use scp command. It is like a cp command under local unix/linux console. Except it does remote transfers.

So to put one file:

scp file.dat garmoncheg.google.com:NewData.txt

copy file "file.dat'' from your current computer/directory to your home directory on garmoncheg.google.com, naming the copy on remote server "NewData.txt''.

scp file.dat garmoncheg.google.com:

is the same thing, except file will be named like original at local computer HD.

scp file.dat garmoncheg.google.com:/home/garmoncheg/data.txt

with specifying original path to store files at.

To get files you need to reverse paths given via arguments to scp. E.g.:

scp garmoncheg.google.com:data.txt file.txt

This will copy file from your default ssh directory named "data.txt" to local (current directory) file called "file.txt"

To work with directory (get/put files) recursively there is an -r switch:

scp -r project-stage:logfiles /home/garmoncheg/logfiles

This will download directory "logfiles" from server with ssh shortcut "project-stage" into current directory naming it "logfiles"

That's it. There are many more decisions and recipes for those typical enough operations. Dont hesitate to comment with yours ;)

Python: simple recipe to measure your function's execution time

2012-08-06T10:38:00.000+03:00

We always write something unusual while doing basic things. Our own bicycles and crutches to work out some unusual situation. Here is another recipe to do a thing like so. I have tried several libraries and readymade decisions. But it assumes you have them installed. And you often try things in console, don't you?

Anyway the recipe is simple and quite straightforward.

import datetime

# Getting first timestamp
t1 = datetime.datetime.now()
# Your function e.g.:
data = [g.name for g in request.user.groups.all()]
# Second timestamp
t2 = datetime.datetime.now()

print "Execution time: %s" % (t1-t2)

It is rude and quite simple but may often suit you well to measure execution time in a simple and straightforward manner.
Also nice idea to write down this function into logs. It may be handy on refactoring of your core app for e.g.;

import datetime
import logging
log = logging.getLogger('mylogger')

# Getting first timestamp
t1 = datetime.datetime.now()
# Your function e.g.:
data = [g.name for g in request.user.groups.all()]
# Second timestamp
t2 = datetime.datetime.now()
# Using Logging to write down calls
log.debug("my_function Execution time: %s" % (t1-t2))

This also can grow up to Django decorator that can log time of function execution into console/log file/whatever. But it's quite another story.
Those snippets are quite Django based. But, I think, the core is pure python, if you've got the core concept of this.

Used, have thoughts, know how to do it in a better/simpler manner? I'm wrong? Please comment!

Django: Resetting Passwords (with internal tools)

2012-07-21T16:55:00.001+03:00

I have had a task recently. It was about adding a forms/mechanism for resetting a password in our Django based project. We have had our own registration system ongoing... It's a corporate sector project. So you can not go and register yourself. Admins (probably via LDAP sync) will register your email/login in system. So you have to go there and only set yourself a password. For security reasons you can not register. One word.

First I've tried to find standart decision. From reviewed by me were: django-registration and django password-reset. These are nice tools to install and give it a go. But I've needed a more complex decision. And the idea was that own bicycle is always better. So I've thought of django admin and that it has all the things you need to do this yourself in no time. (Actually it's django.contrib.auth part of django, but used out of the box in Admin UI) You can find views you need for this in there. they are:

password_reset
password_reset_done
password_reset_confirm
password_reset_complete

To use this method you need:

1. Emails config:

Your typical Django project has settings.py variables to be set for enabling SMTP server. Usually you need this on production to send admins emails about 500 errors and/or system warnings. For our needs python has an internal mail server. I prefer to use it for debug purposes. So good method would be to use it only in Debug case. You can do it by adding those lines to your settings.py:

if DEBUG:
    EMAIL_HOST = 'localhost'
    EMAIL_PORT = 1025
    EMAIL_HOST_USER = ''
    EMAIL_HOST_PASSWORD = ''
    EMAIL_USE_TLS = False
    DEFAULT_FROM_EMAIL = 'testing@example.com'

Here you will have this custom email server set to alternative port. For production you should enter your Sendmail (or whatever you use there) parameters...

Now you will be able to start it running the terminal and executing:

python -m smtpd -n -c DebuggingServer localhost:1025

You can also read abut it here. You will be able to copy/paste email links that should be sent via Django email. Now that you have everything setup for experiments...

2. Config Url's:

urls.py

urlpatterns = patterns('',
    url(r'^user/password/reset/$', 
        'django.contrib.auth.views.password_reset', 
        {'post_reset_redirect' : '/user/password/reset/done/'},
        name="password_reset"),
    (r'^user/password/reset/done/$',
        'django.contrib.auth.views.password_reset_done'),
    (r'^user/password/reset/(?P<uidb36>[0-9A-Za-z]+)-(?P<token>.+)/$', 
        'django.contrib.auth.views.password_reset_confirm', 
        {'post_reset_redirect' : '/user/password/done/'}),
    (r'^user/password/done/$', 
        'django.contrib.auth.views.password_reset_complete'),
    # ...
)

3. Add Templates:

You have templates for your reset password forms already. Django admin templates/css/js are used there. But I think you would want custom one's for your needs. SO basically all you may need is overriding those templates, specified down here and place them on your templates Django path.
You need to override:
Django templates:

registration/password_reset_form.html
registration/password_reset_done.html
registration/password_reset_confirm.html
registration/password_reset_complete.html

Email template:

registration/password_reset_email.html

I'll write down here one's I've used, according to Django's "Batteries included" philosophy.

registration/password_reset_form.html

{% extends "base.html" %}

{% block title %}Reset Password{% endblock %}

{% block content %}
<p>Please specify your email address to receive instructions for resetting it.</p>

<form action="" method="post">
    <div style="display:none">
        <input type="hidden" value="{{ csrf_token }}" name="csrfmiddlewaretoken">
    </div>
     {{ form.email.errors }}
    <p><label for="id_email">E-mail address:</label> {{ form.email }} <input type="submit" value="Reset password" /></p>
</form>
{% endblock %}

registration/password_reset_done.html

{% extends "base.html" %}

{% block title %}Password reset successful{% endblock %}

{% block content %}
<p>We've e-mailed you instructions for setting your password to the e-mail address you submitted.</p>
<p>You should be receiving it shortly.</p>
{% endblock %}

registration/password_reset_confirm.html

{% extends "base.html" %}
{% block title %}Setting New password{% endblock %}

{% block content %}
    {% if validlink %}
        <p>Please enter your new password twice.<br />
           So we can verify you typed it in correctly.</p>
        <form action="" method="post">
            <div style="display:none">
                <input type="hidden" value="{{ csrf_token }}" name="csrfmiddlewaretoken">
            </div>
            <table>
                <tr>
                    <td>{{ form.new_password1.errors }}
                        <label for="id_new_password1">New password:</label></td>
                    <td>{{ form.new_password1 }}</td>
                </tr>
                <tr>
                    <td>{{ form.new_password2.errors }}
                        <label for="id_new_password2">Confirm password:</label></td>
                    <td>{{ form.new_password2 }}</td>
                </tr>
                <tr>
                    <td></td>
                    <td><input type="submit" value="Change my password" /></td>
                </tr>
            </table>
        </form>
    {% else %}
        <h1>Password reset unsuccessful</h1>
        <p>The password reset link was invalid, <br />
        possibly because it has already been used. <br />
        Please request a new password reset.</p>
    {% endif %}
{% endblock %}

registration/password_reset_complete.html

{% extends "base.html" %}

{% block title %}Password reset complete{% endblock %}

{% block content %}
<p>Your password has been set.  You may go ahead and log in now.</p>
<p><a href="{{ login_url }}">Log in</a></p>
{% endblock %}

registration/password_reset_email.html

{% autoescape off %}
You're receiving this e-mail because you requested a password reset for your user account at {{ site_name }}.

Please go to the following page and choose a new password:
{% block reset_link %}
{{ protocol }}://{{ domain }}{% url django.contrib.auth.views.password_reset_confirm uidb36=uid, token=token %}
{% endblock %}

Your username, in case you've forgotten: {{ user.username }}

Thanks for using our site!

The {{ site_name }} team.

{% endautoescape %}

SO you're getting password reset system out of the box using only Django included tools. In no time.
Comment me...!

Django: Unittest for HttpResponseRedirect method recipe

2012-07-15T12:22:00.000+03:00

I like to invent bicycles in my code. They often come in handy and you basically do not rely on Django version... So upon updating your project's Django version in future times you will not have to refactor half of all the code. So...

How do you test http redirects? I usually restrict my time in thinking and write code like if redirect target is hardcoded. And it really is in most of the times. But imagine if you decide to redirect your view response in several places. Or imagine if you're not the only one who owns you'r project's code and there are many collaborators that can change things without warning... And you need to have a test for that view that will test redirects and complex code behavior.

I used to check redirects like so:

response = self.cliet.get('/myview')
self.assertEqual(response.status_code, 302)
# View redirected... All ok...

But if you have some view that can redirect to things that you need to check. How would you test that? E.g. your view:

import random

def myview(request):
    url_end = random.choice([1,2,3,4])
    url = '/someplace/' + str(url_end)
    return HttpResponseRedirect(url)

It can produce many urls and you basically can not predict which one it will produce this time. Maybe it's not the best example. However it's quite simple and understandable IMHO. You're view can redirect you to:

/someplace/1

/someplace/2
...

Imagine if your program results may depend on that and you must check if the page rendering is properly relayed on that URL that you're redirecting.

How would you check that without relying much into Django variables and producing hacks? Exactly! You name it... Regexp's:

class MyTestCase(TestCase)

    def setUp(self):
        # We are using only logged in client in this test
        self.client.login(username=username, password=password)

    def test_mytest(self):
        response = self.client.get(reverse('myview'))
        # Following redirect in tests
        new_url = self._retrieve_redirect_response_url(response)
        response = self.client.get(new_url)
        self.assertEqual(response.status_code, 200)
        # No errors appeared
 
    def _retrieve_redirect_response_url(self, response):
        """
        My handy helper:
        
        parses 302 response object.
        Returns redirect url, parsed by regex.
        """
        self.assertEqual(response.status_code, 302)
        new_url = re.search(
                            # Our REGEXP:
                            "(?P<url>https?://[^\s]+)", 
                            str(response)).group("url"
                            )
        return new_url

A lot of third party logic exists in this test. But it's partially a live tests snippet with names changed. So you may assume it should be like so and concentrate your thoughts on this approach in general, rather than possible inconsistencies.

Main deed does regexp: "(?P<url>https?://[^\s]+)". Other code is only possible solution to use it in a handy way...

Anyway, now you can use this snippet in your code to check url that view redirected you to. Only concern that I have here that there are much simpler approaches. But this one is almost Django updates independent. It stringifies your response and parses for url there.

If you want to know about re.search more: http://docs.python.org/howto/regex.html
Writing tests in general: https://docs.djangoproject.com/en/dev/topics/testing/?from=olddocs
Hope this will help someone. Comment your ideas on this! Appreciate your every thought...

Blocking iOS to acces Contacts freely

2012-07-09T13:36:00.000+03:00

There are lots of occasions when you want to be private in the net. Now it's common to say "Anonymous". Many people do not even think it's important. I, personaly, dislike facebook, skype, twitter, any else third party... having my contacts list synched to their server without even telling me... I do not want to be truly anonymous, you know... Have nothing to hide really. But I want that really smal confirm that Skype, for e.g. would take my contacts and for what reason.

Facebook was first, as usual... They've done it. I really deleted their app from my iOS devices first. But instead of taking lessons from their mistakes... All huge companies gone wild about this possibility to copy some new ammount of user data... Skype was second (for me)... Then LinkedIn... However they warned and I've agreed... So nothing really bad to say about them. But MS is such a Microsoft... They want me to "skype" instead of "calling" my friends through my Apple/Android device :)...

Anyway I've decided I really need something like a hardware switch to grant access to my contact list on iDevices... And it is really done for us already...

You need a Jailbreak for this! (As I had one already wont mention it here. I think there are lots of info in the net about it)

ContactPrivacy tweak. For now, that I'm writing this article... It exists in BigBoss repo, that is default set in Cydia AFAIK. And you suppose to have access to it without adding any repositories. So typing in Cydia search: "ContactPrivacy" (without scopes) should reveal it without any problems.

After installing it you will probably see something like this:

So ...... you Skype after installing it. :)

Note: That I think it's good idea to install this tweak even before any other APP is installed. I'd do so if I only knew before.

Heard somewhere about Twitter is going to have my contacts too... Wtf? Why Twitter wants them too... Can't even imagine... Who's next. "Corporation of all good". Nahh... I think they already have them but not so dumb to tell >:) But somehow it does disturb me less...

BTW it's a nice reason to make a jailbreak for this tweak. Why not make it standard in iOS?
How do you think?

Twitter official client for Mac problem "Unauthorized" solving.

2012-06-26T12:04:00.000+03:00

I had a problem "Unauthorized" in official twitter client for Mac.

Tried several instal/reinstall. Deleted/Downloaded again from Apple App Store. It all didn't help. They seem to change something in authentication. So I've tried deleting all the config files in ~/Library/
Main solution that helped me:
1. Switch off the client. (be sure to Close an app)
2. Delete:

~/Library/Preferences/com.twitter.twitter-mac.plist.lockfile

~/Library/Preferences/com.twitter.twitter-mac.plist

3. Launch Twitter app and login from scratch.

I have revoked access this twitter client too in my https://twitter.com/settings/applications but it seem to make no effect.

Making similar actions to TweetDeck for Mac does not help. So it seems like problem is somewhere deeper. Hope it will help you solve similar issue.

Pretty git Log

2012-06-19T14:01:00.002+03:00

SO you dislike git log output in console like me and do not use it... Because it looks like so:

How about this one?

It's quite easy...
Just type:

git log --graph --pretty=format:'%Cred%h%Creset -%C(yellow)%d%Creset %s %Cgreen(%cr) %C(bold blue)<%an>%Creset' --abbrev-commit --

It may be hard to enter such an easy command every time. Let's make an alias instead... Copypaste this to your terminal:

git config --global alias.lg "log --color --graph --pretty=format:'%Cred%h%Creset -%C(yellow)%d%Creset %s %Cgreen(%cr) %C(bold blue)<%an>%Creset' --abbrev-commit --"

And use simple command to see this pretty log instead:

git lg

Now in case you want to see lines that changed use:

git lg -p

In order for this command to work remove the -- from the end of the alias.

May the code be with you!

NOTE: this article is a rewritten copy of http://coderwall.com/p/euwpig?i=3&p=1&t=git and have been put here only in purpose of note to myself. However you may use for your own needs...

UPD2: And unless you're a contributor of the git project and have written at least part of original code and/or documentation that does this trick... Then you probably stole the idea. ;) (IMHO)

Python sort() patterns

2012-06-07T22:35:00.000+03:00

Python has a lot of sorting patterns. Let's make a short a list.

1. Sorting list by it's element. Simple case. Should simply do:

>>> exmpl_list =['a', 'c', 'B', 'd']
>>> exmpl_list.sort()
['a', 'B', 'c', 'd']

However this example does not take locale into account and works only for ASCII characters.

2. Sorting list of sub elements.

exmpl_list = [{'name':'Homer', 'age':39}, {'name':'Bart', 'age':10}]  
# Sorting by 'name'  
newlist = sorted(exmpl_list, key=lambda k: k['name'])
# Better way to use itemgetter():
from operator import itemgetter
newlist = sorted(exmpl_list, key=itemgetter('name'))

Note that it is equivalent to:

exmpl_list.sort(key=lambda k: k['name'])
# OR:
exmpl_list.sort(key=itemgetter('name'))

if you do not need new list in result.

3. Local aware decision

import locale
locale.setlocale(locale.LC_ALL, 'en_US.UTF-8') # vary depending on your lang/locale
assert sorted((u'Ab', u'ad', u'aa'), cmp=locale.strcoll) == [u'aa', u'Ab', u'ad']

# Without using locale.strcoll you get:
assert sorted((u'Ab', u'ad', u'aa')) == [u'Ab', u'aa', u'ad']

4. Sorting list of tuples/lists by known element position:

data = [[1,2,3], [4,5,6], [7,8,9]] 
data = [(1,2,3), (4,5,6), (7,8,9)]
 
sorted_by_second = sorted(data, key=lambda tup: tup[1])
data.sort(key=lambda tup: tup[1])  # sorts in place

BTW. Python How To: Sorting

Python super() method and class inheritance

2012-06-06T08:12:00.000+03:00

There is one more useful method in python. super()

It helps to handle inheritance of ancestor class. This sounds a bit of messy. So I'll try to explain.

You have class A that does some useful things. And you have to use this class A everywhere in your application with adding some piece of functionality. You may inherit this class B(A) and expand it's functionality. Now you can use class B instead of class A everywhere and get rid of redundant operations.

# Main class (parent)
class A(object):
    def __init__(self):
        print(u'class A constructor')

# Main class ancestor (inheriting main class)
class B(A):
    def __init__(self):
        print(u'class B constructor')
        super(B,self).__init__()

For live example:
NOTE! Code is theoretical for you because it is taken from live project. I'm using many own written methods here. Neither document not it's methods are not django standard and given here to get the point of it all in real world example.
You have Django HttpResponse() object. Your Django application creates (instantiates) this special HttpResponse() object and afterwards populates some file into it. you may often do redundancy in your code. One way: you can create a manager to instantiate HttpResponse() object from scratch and then populate it with your data. But it's making your code use some additional function that instantiates and then operates object. Better (proper from python perspective) to inherit HttpResponse() object and create your own, say DocumentResponse() object that will be in fact standard HttpResponse() that will populate itself with required data from your provided object. This will lead to code simplification.
Let's provide some code for this object's declaration:

# Using DJango internal object
from django.http import HttpResponse

# Inheriting it and populating with our data.
class DocumentResponse(HttpResponse):
    """
    HttpResponse() object containing file.
    """
    def __init__(self, document):
        # our redundant actions we want to get rid...
        content = document.get_file_object()
        mimetype = document.get_mimetype()
        filename = document.get_full_filename()
        # Instantiating ancestor class HttpResponse()
        # with our special parameters.
        super(DocumentResponse, self).__init__(content=content, mimetype=mimetype)
        # Making some actions with "self" like a new ancestor class instance
        # now: self == HttpResponse() 
        self["Content-Length"] = len(content)

TADA! We have ability now to create our custom HttpResponse() from internal file object (document) everywhere in our imaginary project.
we can now write something like:

response = DocumentResponse(document)

and it will provide a proper HttpResponse() object specially transformed for our needs. Django likes it and returns properly instead of HttpResponse(). It has all of it's methods after all :).
One more thing to note that python will not run constructor of ancestor here until we will tell him to do it with our super() method. And afterwards we can modify it too.
Naming this method may sound syntactically "to heavy". This problem is spolved in python 3.0. However there is old implementation, sometimes called 'old school':

# Parrent
class A(object):
    def __init__(self):
        print(u'class A constructor')

# Ancestor
class B(A):
    def __init__(self):
        print(u'class B constructor')
        A.__init__(self)

Helped? Improper? Please leave a comment below...

Templates. New in Django 1.4

2012-04-16T11:25:00.000+03:00

Continuing those series of tasty features in Django 1.4. It's templates and newly added templatetags this time. Many things improved in template logic now. Lets move on point by point...

Tag {% elif %}

For most of us it is almost killer feature of new Django 1.4. You will have simpler logic in your templates, if you have advanced logic there, of course. ;)

Key-value arguments for user tags

You can use key-value for tags along with old, positional, arguments, e.g.:

{% custom_tag 123 "test" thing.var text=msg|upper username=user.get_full_name %}

All of this implemented using common "args" and "kwargs":

@register.simple_tag
def custom_tag(a, b, *args, **kwargs):
    warning = kwargs['text']
    profile = kwargs['username']
    ...
    return ...

assignment_tag

Some tags are not built to output something in template, like simple_tag. They only have to change something in context during rendering of templates. (Built-in {% with %} tag, for e.g.).
You can make it output time like so:

{% get_common_time "%Y-%m-%d %I:%M %p" as object_time %}
The time is {{ object_time }}.

SO now you can use "assignment_tag", to make creating tags like this simpler:

@register.assignment_tag
def get_common_time(time_format):
    return datetime.datetime.now().strftime(time_format)

Translation marks in context

Django 1.3 appeared to have functions pgettext() and ngettext() to indicate context for translatable strings. It solved problem of short text phrases partially. It suited phrases that can have different translation depending on context. Now you can transfer it in template directly:

{% trans "Iurii" context "username" %}
{% trans "Do" context "verb" %}
{% blocktrans with name=user.username context "hello message" %}Hi {{ name }}{% endblocktrans %}

New {% static %} tag

You can now share STATICFILES_STORAGE with special tag:

{% load static from staticfiles %}
<img src="{% static "app/css/base.css" %}" />

Most profit will get those one's that for e.g. use CDN on S3, or another not "standart" way. {{STATIC_URL}} is equivalent to {% static %} in other cases.

You also get new storage.CachedStaticFilesStorage, that can add md5-hash to filenames from static, to make cash for css, js, images and so on definitely invalid.

|truncatechars filter

Incredible but true. After 4,5 years of Django dev you received |truncatechars in default filers list. It can trim string to provided amount of characters and change the end of it into three dots... Anyway |truncate_words is also available.

Almost all "IMHO tasty new features in 1.4 templates" covered. The farther the more will come. Stay tuned.

Timezones. New in Django 1.4.

2012-03-29T11:33:00.000+03:00

What and Why?

Under Django version 1.4 all dates stored and displayed for a single timezone. The one that is specified in project.settings.TIME_ZONE. Now you gain ability to store dates in UTC and render it with timezone correction.

Problems with localtime bypass is an additional plus. They can happen once a year. For e.g. 31 of November 2012 in Russia. Time from 2:00am to 3:00am in fact goes by twice. It may not be a problem for 99% of users. But it can become a nightmare for billing systems. So it's better to store time in UTC and display with user Time Zone correction. So "02:15am 31 November 2012" will become "2012-10-30T22:15:00+04:00" and "2010-10-30T23:15:00+03:00" that is so handy for programmers ;).

Concepts

datetime objects in Python support timezones with attribute tzinfo. If this attribute is filled out it is called "timezone-aware", otherwise it's "naive" date.
Django uses timezone-aware dates, if you've used new feature in Django 1.4 while update. (settings.USE_TZ=True)
"Default timezone" in Django is the one that is specified in settings.TIME_ZONE. (E.g. 'Europe/Kiev'). While "current timezone" is the one that activated this moment.

Current timezone

You need to know where is user from to show him his date and time. You can ask him directly and store his timezone in session BTW. You can use DB's like "maxmind", guessing his timezone by IP or even use JavaScript: Date.getTimezoneOffset(). One or another way we need to activate user's timezone to get access to rendering of timezone aware date and time in templates and forms.

Example of setting current timezone (from Django docs):

from django.utils import timezone

class TimezoneMiddleware(object):
    def process_request(self, request):
        tz = request.session.get('django_timezone')
        if tz:
            timezone.activate(tz)

Main functionality we need for this issue is concentrated in 'django.utils.timezone'.

Templates

Those dates workout is widely supported in templates along with Python code.

# Converting to Paris timezone
{% timezone "Europe/Paris" %}
    Paris time: {{ value }}
{% endtimezone %}

# default date display, like they're actually stored (UTC)
{% localtime off %}
    {{ value }}
{% endlocaltime %}

Migration

Setting settings.USE_TZ = True
pip install pytz (read farther)
Convert DB datetime values into UTC time. (If it is different)

Now your code supports timezones. You have to repair some old places where Django does not handle it for us. We have to convert naive dates into timezone-aware. Otherwise you can get wired bugs while changing Winter/Somer time. To complete the migration you have to use 'django.utils.timezone' and debug places where dates in new format are compared to "new" one's.

pytz

Library that handles handy timezone workout in python. Olson database and handy API are included for different timezone calculations. Despite it is optional in Django 1.4 developers recommend to include it. It helps Framework not to guess default timezones while calculating time and provides timezone list to render for users.

Conclusion

Those new features make me proud user of this framework and simplify many holes that I had to handle myself. E.g. with hosting app using Australia/Sydney Timezone and working with it in Europe/Kiev. ;)

Django template tags to find out field type

2012-03-28T16:03:00.000+03:00

Sometimes you have the task to decorate django form fields. In our case it was need to add different widget to form field depending on field type.

You can know your field type by refering to it's python class name, e.g.:

field.__class__.__name__

You will get wrong symbols error if you try to refer this python variable in templates directly. So you need a custom template tag to serve here. Decision was not to return variable for tag but to add context variable, cause I had need to perform exact actions on certain field types.

So tag example:

@register.simple_tag(takes_context=True)
def set_this_field_type(context, field):
    """
    Adds to context given field type variable
    variable named "this_field_type"
    """
    context["this_field_type"] = field.field.__class__.__name__
    return ''

Now you can add this variable to your if statements. E.g. usage:

{% load my_template_lib %}
{% for field in form %}
    {# Setting this field type in context #}
    {% set_this_field_type field %}
{% if not this_field_type = "DateField" %}
    {# Rendering specific for Date fields (With DatePicker widget) #}
    <div class="field-wrapper {{ this_field_type }}"
        <label class="control-label">{{ field.label }}</label>
        <input id="{{ field.id_for_label }}" class="datepicker-widget" type="text" name="{{ field.name }}">
    </div>
{% else %}
    {# Default field rendering #}
    <div class="field-wrapper {{ this_field_type }}"
         <label class="control-label">{{ field.label }}</label>
        {{ field }}
    </div>
{% endif %}
{% endfor %}

You will get your form fields rendered in custom manner. You will have the CSS class with field type for styling or you can use this variable in context to render various widgets for e.g.

Hope you've got the main concept here. Now this idea must be limited only by your imagination.

Including those "simple" approaches kindly suggested by guys at 'reddit.com' They are more simple and can play in your case maybe ;)
Using the filter: (Thanks to POTUS @reddit.com)

@register.filter()
def field_type(field):
    return field.field.__class__.__name__

{% if not field|field_type ="DateField" %}
    <span>Whatever</span>
{%endif%}

Using the widget: (Thanks to tgerdes @reddit.com)

from django import forms
class DateForm(forms.Form):
    date = forms.DateField(
      widget=forms.DateInput(attrs= {'class': 'datepicker-widget' } ))

form = DateForm()
print form.as_p()

outputs: <p><label for="id_date">Date:</label> <input id="id_date" type="text" class="datepicker-widget" name="date" /></p>

However this approach in not this topics main cause it concludes we only have to style input. But in some cases it may be quite enough...

Thanks guys for correction!

Please drop me a comment here if you've liked/used this decision(s).

Delete all '.pyc' files recursively OS X

2012-02-24T08:20:00.000+02:00

We often get strange errors with python. When you changing the source and cannot figure out why files are used from past. So if you ever get this dejavu feeling, try using this command:

find . -type f -name '*.pyc' -exec rm {} \;

It will delete all precompiled python sources. (Because with python you have ability to distribute your files in binaries in fact.) Those errors usually appear if you move/delete *.py files with version control. You usually exclude *.pyc files from list. Old links used this way. Debugger does not suggest wrong imports because your files are in fact there, just precompiled.
Anyway I've just put this command to my blog for memorizing. It often helps me. Maybe it can for you.

Python / Django UnicodeEncodeError hacks.

2012-01-26T12:28:00.000+02:00

Python plays with unicode nicely nowadays. Bt what if you must deal with old time formats conversion, or ASCII files exporting for e.g. You may also use software that is out of date but is too long to rewrite... Here often errors occur. I have received mine at copy-pasting from MS Word into Django admin UI by stupid users. Most of the website played nicely with this fancy characters, but exporting to CSV failed due to non ASCII characters support. Google said nothing special. Python docs about unicode usage briefly cover this type of events. So here is the result of some hours of experiments. I've decided to rewrite some of the python functionality to create decode function with behavior for my needs. Hopefully they will shorten you some time with those collisions you may get in your Django apps...

Anyway I've started to receive errors like:

Exception Type:  UnicodeEncodeError
Exception Value: 'ascii' codec can't encode character u'\u2013' in position 17: ordinal not in range(128)

So I had a list of u'' values that contained special characters " ordinal not in range(128) ". Requires no imports... Pure python:

        values == [
          u'Some fancy text \u2013 something', 
          u'some normal, easy convertible text', 
          u'some more normal text'
          ]      
        HACK: entry cleanup for special characters (Fixing Bug #...)
        # entry cleanup for special characters
        i = 0
        for value in values:
            try:
                # if string can be encoded to 'ascii' pass
                unicode(value).encode('ascii')
            except UnicideEncodeError:
                val_temp = unicode(value)
                # cleaning up string with escaping non convertible characters
                result = []
                for symbol in val_temp:
                    try:
                        symbol.encode('ascii')
                        result.append(symbol)
                    except UnicodeEncodeError:
                        pass
                # rewriting wrong value in values array
                val_temp = ''.join(result)
                values[i] = val_temp
                pass
            i = i+1
        # normally work with our list... it's safe now...
        values == [
          u'Some fancy text  something', 
          u'some normal, easy convertible text', 
          u'some more normal text'
          ]

This code is a bit complicated due to mine specific task and has iterations in iterations etc... But it's from a working app and checked working. However here is the theoretical example that must clean up a single string:

value = u'Some fancy text \u2013 something'
try:
    # if string can be encoded to 'ascii' pass
    value.encode('ascii')
except:
    # cleaning up string with escaping non convertible characters
    result = []
    for symbol in val_temp:
        try:
            symbol.encode('ascii')
            result.append(symbol)
        except UnicodeEncodeError:
            pass
    # rewriting our variable with safe one
    value = ''.join(result)
    pass
# normally work with our unicode string... it's safe now...
value = u'Some fancy text  something

So the technique here is simple. We are checking if this unicode string can be converted to 'ascii' python encoding without errors we simply passing through. And if it's not... Converting it to 'ascii' string symbol by symbol. Symbols that will fail will be gracefully omitted. You can create a function from all of this, like 'my_decode_cleanup' or something and use whenever needed...

Hope this will help you to save some precious time during your python development.

Helped? I'm wrong somewhere? Please comment!

Installing CouchDB for Mac OS X. Nuances.

2012-01-16T11:38:00.001+02:00

Our project will have a new architecture changes in future. We've chosen from different tech. But now we 're experimenting with CouchDB a bit. It may become a part of our new architectural Changes. I'll try to cover most of newbies observations with CouchDB for Python(Django) programmer, planning to use it in your project.

First article will be, unsurprisingly, installing CouchDB under Mac OS X. It, sure, has some inches that you need to scratch. So let's roll...

CouchDB is not a simple SQL database like you used to install before. It has own server to accept connections. And, in fact, you can work with database from your browser GUI (without other server proxy, like Django or ...). Anyway it will not be our target. Building JS UI is quite easier with CouchDB and changes your logic dramatically. To build UI with Python we need some proxy parts. Anyway, again, CouchDB is a standalone server, that runs on your server's port and talks to your APP/Browser/whatever... through JSON API. It's quite a bad thing, because it's a bit slow, but a good thing, because it's extremely easy to implement. Most programming languages have adaptors to JSON. CouchDB is quite a new tech. And it is highly developed and "pushed through" tech for now...

Ok enough preface... Actual install is described in official DOCS HERE. I'm using "mac ports" for my project already. So it's my choice of install method by all means. I'll cover docs here in brief to be consistent:

1. Install it through mac ports.

$ sudo port install couchdb

Will start huge and massive install of required ports for this DB.

2. Good idea would be to update Mac ports. If you have some ports installed already, like I do:

$ sudo port upgrade couchdb

Will upgrade only ports required by "couchdb" port.
It's also a good idea to selfupdate ports if you already have them installed:

$ sudo port selfupdate

This must be sufficient for all of our needs.

At this point you must be able to run CouchDB from command line running:

$ couchdb

command in Terminal. If it runs with error, no worries. Read on. There is a known issue!

3. Bug occurs in MacPorts itself with "couchdb" package. It does not create some paths in required dir. You need to manually create those paths and give CouchDB permissions to write there. Mac Ports Track has the solution. Or you can manually (better IMHO) make sure all your permissions and directories exist.
You need to make sure those directories exist, and create them if they don't:

/opt/local/var/lib/couchdb/
/opt/local/var/log/couchdb/
/opt/local/etc/couchdb/

and have permission to work with CouchDB:

$ sudo chown -R couchdb:couchdb /opt/local/var/lib/couchdb/
$ sudo chown -R couchdb:couchdb /opt/local/var/log/couchdb/
$ sudo chown -R couchdb:couchdb /opt/local/etc/couchdb/

(Those paths are ok for a standart "mac ports" install.. If your directories differ... than change those paths accordingly)
After those actions, running couchdb from console should led to normal startup sequence output:

$ couchdb
Apache CouchDB 1.1.1-incubating (LogLevel=info)
Apache CouchDB is starting.
Apache CouchDB has started. Time to relax.

This will signal that your CouchDB install is almost OK and you may check that it is running:

http://127.0.0.1:5984/

In your browser. It must Show output JSON. Something like:

{"couchdb":"Welcome","version":"1.1.1"}

You may also examine it's web admin UI, typing in browser:

http://127.0.0.1:5984/_utils/index.html

Should show handy Admin UI. You may examine/edit DB by hand there... (Something similar to PHP MyAdmin or whatever you used to use on SQL DB's)

4. Add autorun command (optionally). It's good to run every time you need it for experiments, etc... But if you plan to Debug test apps on that platform it's good idea to make it start with system every toime it runs. To make that you need to type this command into console:

$ sudo launchctl load -w /opt/local/Library/LaunchDaemons/org.apache.couchdb.plist

THE END.

Comments and suggestions are welcome!
In future I plan to make some articles on actual befriending CouchDB with Python. Stay tuned and thanks for reading that far.