- Simple file browser to flip through images
- Powerful searching via Mac OS X spotlight
- Displays basic EXIF information
- Displays embedded thumbnail (or generates one on the fly)
- Allows you to modify captions and keywords only
- Add photos to a pile and then batch resize and copy, ready for emailing
- Pseudo commandline interface: access all functions with short keystrokes
(With a cameo appearance from ffmpeg)
The nature of this program makes it useful/appealing only to people with some knowledge of programming and using the commandline on macs. Therefore I have not spent effort in making a nice package with all the dependencies etc. I'm listing the non-standard Python packages that Chhobi uses along with download links. Anybody with a moderate interest in computers should be able to get Chhobi up and running with these instructions. It should also be possible, with a little tweaking, to get this program running on Linux and Windows machines, but I have not tried it.
- PIL - needed for thumbnail display
- xattr - needed to write video metadata as Mac OS X extended attributes
- biplist - needed to write video metadata as Mac OS X extended attributes
- exiftool - Needed to read and write exiv photo meatadata
- ffmpeg - needed to generate display thumbnail for videos.
You should check for the latest versions of all software
curl -O -L http://effbot.org/media/downloads/Imaging-1.1.7.tar.gz
tar -xzf Imaging-1.1.7.tar.gz
cd Imaging-1.1.7
python setup.py install --user
pip install xattr --user- Install xattrpip install biplist --user- Install biplistgit clone https://github.com/kghose/Chhobi2.git- Get Chhobi2
You can now start Chhobi by going into the download directory and typing
python guichhobi.py
Click on the command window (the input box with white background) and press 'h' to get the manual.
The user manual is accessed by running the program with the -h option python guichhobi.py -h or pressing the 'h' key in the command window.
You can find the manual as part of the docstring of the main file (guichhobi.py)
I think iPhoto is bloated, and any photo organizer concept I could come up with, seemed to unnecessarily duplicate many standard OS functions, such as file organization and metadata searching. Several powerful commandline tools like exiftool, mdfind and convert already exist to manipulate images and image metadata. So I harked back to my Unix days and thought, why not go retro? Why not go back to the GUI as a wrapper around powerful and fast commandline tools? The use of an actual command line to perform all the actions was, of course, not strictly necessary, but if you are going retro, why not go all the way?
Chhobi is made available for download in the hope that it may be of some use. No claim is made that the code works and is free of bugs, but I would appreciate bug reports (iamkaushik.ghose@gmail.com: When you email me remove the first three letters)
http://www.gnu.org/licenses/gpl.html
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
The captions (exiftool: Caption-Abstract) and keywords (exiftool keywords) you enter are stored in the standard photo metadata itself. This means that they will show up in Finder, show up in Inspector, show up in spotlight and can be read by any exif compliant tool. They will survive transfers across file systems and operating systems. This is the beauty of standardized metadata. This is all possible because of Phil Harvy's wonderful exiftool command line program and its super awesome interactive mode, which allows you to perform multiple queries without having to call the program repeatedly.
Exiftool, however, does not handle captions and keywords for videos. I do not know if there is a standard for such user defined metadata for videos. I first toyed with the idea of using sidecar files, but this left a very bad taste in my mouth. I finally settled on using extended attributes to save the captions (com.apple.metadata:kMDItemDescription), keywords (com.apple.metadata:kMDItemKeywords) and even the thumbnail (chhobi2:thumbnail). This has the bonus that mdfind (and therefore Spotlight) indexes the caption and keywords.
Chhobi stores two pieces of information - your photo root and its own geometry - in a simple text configuration file called chhobi2.cfg in your home directory.
Chhobi runs exiftool in safe mode, which means your original images are always kept and captions and keywords are added in a copy of the original.
- ( ) Log buffer that you can pull up as a window (like for help)
- (x) Fix logical bug related to preview and rotations (need to do some pre computations related to final image aspect)
- (x) Handle rotations
- ( ) Bugfix: some kind of bug involving keyword removal
- (x) Have a quickshow window that shows a larger preview if we pause on a picture for a bit.
- (x) Add command to add photos to Flickr
- (x) Flickr upload should push to command window.
- ( ) Flickr privacy options during upload
- ( ) Fix '=' vs '==' issue
- ( ) Increase the video types handled
- ( ) Command to force regenerate thumbnail
- (x) Handle video metadata
- ( ) Expand search parsing
- (x) Selections/collections
- (x) Resize and zip collection to send via mail
- ( ) Export queries as smart folders
- (x) (silently generate thumbnail) Indicate when preview thumbnail is not available - autogenerate one
- (x) Command history
- ( ) Commands to change appearance
- (x) Increase status bar time
- (x) Show status when changing browser panes
- ( ) Write how to do searches
- ( ) Add a commandline option to reset geometry
xattr -wx com.apple.metadata:kMDItemKeywords <hexdumped plist> File
- you need to use -wx because the data needs to be in binary format
- You need to preface the tag with
com.apple.metadata:for Spotlight to index it (and for it to showup on Finder info) - The plist needs to be in binary format
- The hexdump is in the format that
xxd -pgives which is straight hex and is what is obtained by binascii.b2a_hex( - Setting a plain text string by doing
xattr -x blah blahwill not work - the xattr will not be recognized by spotlight
Interestingly, all this can be done from within Python, without going to the xattr, mdls and other tools for binary plists
You just need the xattr and biplist modules.
To read keywords you need to do
biplist.readPlistFromString(xattr.getxattr('TestData/2013-06-29/MVI_0843.AVI', 'com.apple.metadata:kMDItemKeywords'))
which returns the keywords as a list.
You can set them by doing
pl = ['Cat','hat','mat']
xattr.setxattr('TestData/2013-06-29/MVI_0843.AVI', 'com.apple.metadata:kMDItemKeywords', biplist.writePlistToString(pl))
Exif tool can read video metadata but not write it. Chhobi simply writes comments and keywords in extended metadata, which is also indexed by spotlight, so this nicely solves our problem with video metadata.
py2applet --make-setup ../Chhobi2/guichhobi.py ../Chhobi2/icon_sm.pgm ../Chhobi2/chhobi.icns python setup.py py2app -e sip,sympy
"""
This is a setup.py script generated by py2applet
Usage:
python setup.py py2app
"""
from setuptools import setup
APP = ['../Chhobi2/guichhobi.py']
DATA_FILES = ['../Chhobi2/icon_sm.pgm']
OPTIONS = {'argv_emulation': False,
'iconfile': '/Users/kghose/Source/Chhobi2/chhobi.icns'}
setup(
app=APP,
data_files=DATA_FILES,
options={'py2app': OPTIONS},
setup_requires=['py2app'],
)
python setup.py py2app -e sip,pandas,matplotlib,sympy,scipy --matplotlib-backends -
gcc -I/Library/Frameworks/Python.framework/Versions/2.7/include/python2.7 -I/Library/Frameworks/Python.framework/Versions/2.7/include/python2.7 -fno-strict-aliasing -fno-common -dynamic -isysroot /Developer/SDKs/MacOSX10.6.sdk -arch i386 -arch x86_64 -g -O2 -DNDEBUG -g -O3 -L/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/config -ldl -framework CoreFoundation -lpython2.7 -u _PyMac_Error Python.framework/Versions/2.7/Python test.c -o testme
- When using Popen, do not us PIPE for production sized data. It freezes. Use files instead to read data from process output [pipe1].
- A Tkinter app, when launched from the Terminal will not grab focus (focus stays in Terminal)
- Get rid of 'frame' around text box : set
highlightthickness=0
RawQuery format: http://developer.apple.com/library/mac/#documentation/Carbon/Conceptual/SpotlightQuery/Concepts/QueryFormat.html#//apple_ref/doc/uid/TP40001849-CJBEJBHH
My initial idea was to have no GUI elements at all, instead using the finder to select and view the images. In a preliminary version I had a thread that polled the finder for its current selection periodically and used that to drive the interface. This however had a major drawbacks - the applescript (run as a subprocess through oascript) that I used loaded the Finder and the load increased with the size of the selection. The GUI remained responsive since I used a spearate thread for the polling, but my fans started going off and Finder would start consuming CPU when the application was just sitting there. I decided instead to build in a tree view into the application and use Finder for previews only instead.
Smart Folders are typically created by opening up a finder window and entering a search criterion. The search is very powerful, allowing you to search on very detailed file properties, including (relevant to us) things like f-numbers, shutter speeds, ISOs etc etc. You can also save the search for later access.
When a smart folder is saved as a plist file which has an extension called .savedSearch. Finder is programmed to interpret a .plist file with a .savedSearch extension as a set of commands to list the results of the search.
Saved Searches were probably not meant to be generated programatically - their format is a little obscure. If you load up a saved search you will note that the main business seems to occur in the tags 'RawQuery' and 'RawQueryDict'. 'RawQuery' carries the search string that is passed to mdfind to execute the search and return results. 'RawQueryDict' carries the sub tag ''SearchScopes' which is a list of tags that tell Finder which folders to search within, with 'kMDQueryScopeComputer' refering to the local computer and folders referred to in the usual POSIX notation.
The 'RawQuery' tag under 'RawQueryDict' appears to be cosmetic.
mdimport -X will give a list of tags that the RawQuery (and mdfind) will operate on mdimport -X 2> attributes.txt (Yes, the output is through stderr, go figure)
[Searchable attributes][http://developer.apple.com/library/mac/#documentation/Carbon/Reference/MetadataAttributesRef/Reference/CommonAttrs.html#//apple_ref/doc/uid/TP40001694-SW1]
Python's plistlib module offers a very elegant interface to reading and writing plists as Python dictionaries. Chhobi uses a template saved search (created by doing a blank search and saving it) to create a default savedSearch file scaffold. Then the ['RawQuery'] tag and the ['RawQueryDict']['SearchScopes'] are updated with the current search parameters.
import Image, base64
im = Image.open('chhobi-icon.png')
imsm = im.resize((150,150))
str = imsm.tostring(encoder_name='raw')
str64 = base64.b64encode(str)