fitToWindow: true,		// option to fit preview to window size if the
                                // image is too big
showInfo:    true		// option to show or hide image info

To fit the preview to the window size, I’ve replaced the call to _resize_container_image_box() with _adjust_image_size().

Here is the definition of the new function:

function _adjust_image_size(objImagePreloader) {
    // get image size
    var imgWidth = objImagePreloader.width;
    var imgHeight = objImagePreloader.height;

    var arrayPageSize = ___getPageSize();
    // calculate proportions
    var imageProportion = imgWidth / imgHeight;
    var winProportion = arrayPageSize[2] / arrayPageSize[3];
    if (imageProportion > winProportion) {
        // calculate max width base on page width
        var maxWidth = arrayPageSize[2] - (settings.containerBorderSize * 2) - (arrayPageSize[2] / 10);
        var maxHeight = Math.round(maxWidth / imageProportion);
    } else {
        // calculate max height base on page height
        var maxHeight = arrayPageSize[3] - (settings.containerBorderSize * 2) - (arrayPageSize[3] / 10) - 40;
        var maxWidth = Math.round(maxHeight * imageProportion);
    }
    if (imgWidth > maxWidth || imgHeight > maxHeight) {
        imgWidth = maxWidth;
        imgHeight = maxHeight;
    }
    $('#lightbox-image').attr('width', imgWidth);
    $('#lightbox-image').attr('height', imgHeight);
    _resize_container_image_box(imgWidth, imgHeight);
};

Finally, to optionally hide or show the image info, just canged the following line:

if ( settings.imageArray.length > 1 ) {

to:

if ( settings.showInfo && (settings.imageArray.length > 1) ) {

Download modified code here – jQuery lightBox 0.5 AlferSoft Mod

Here you will find some python scripts that will help you to move the Blobstore files from the old application to the new one, and correct all the references in the new database. As always, use at your own risk, don’t try to do anything without reading and understanding the scripts, otherwise you may loose your data permanently.

Some considerations:

• We did used these scripts in our very own application, so I can say it works. However, I had to edit the original scripts to make them more “generic”, remove my database references and names, etc.
• I will try to describe the steps of the migration, but please don’t try to do anything before reading the whole article first.
• This is not by any means intended to be a fully automatic migration, in fact, the steps needs to be manually executed one by one.
• The scripts won’t migrate big files. We have migrated files up to 40Mb without problems, but if you have files bigger than that, it will probably fail. For example we tried a 300Mb file unsuccessfully, the reason is that the file needs to be downloaded from the old app to the new, and there is a 1 minute timeout for the app requests, so if the download takes more than that it will fail.

How does the script works

You will need to put the script in your applications, both the new (HRD) and the old one. It basically exposes four URLs and you will need to access two of them from the HRD application.
By doing this you will download all the blob files from the old application to the new one, the script will create a model that keeps a map between the old and new references.
The final step is to migrate the old references to the new one in the HRD databases.

Prerequisites

• All the references to blob files in your models must be of type blobstore.BlobReferenceProperty, if not you will need to adjust them first.
• Migrate the databases from the old app to the new HRD app using the standard Google migration method.

Steps

• Add the URLs to your application

def main():
    application = webapp.WSGIApplication(
          [
           # ... Your app URLs here ...
           # TODO - REMOVE THIS AFTER MIGRATION
           ('/mig/__getblob/(.+)', GetBlob),
           ('/mig/__getblobkeys/?', GetBlobKeys),
           ('/mig/__migrateblobs/?', MigrateBlobs),
           ('/mig/__migratereferences/?', MigrateBlobReferences),
          ], debug=True)
    run_wsgi_app(application)

if __name__ == '__main__':
  main()

• Then add the migration code

####
# TEMPORARY BLOBSTORE MIGRATION CODE
####

MODELS = (
    # add your (model, blobinfo_field_name) tuples here
    ("myModel1", "blobinfofield"),
    ("myModel2", "BlobInfo"),
)

class mig_BlobMig(db.Model):
    fname = db.StringProperty()
    blobinfo = blobstore.BlobReferenceProperty()
    origkey = db.StringProperty()

class MigrateBlobReferences(blobstore_handlers.BlobstoreDownloadHandler):
    '''Migrate blob info references'''
    def get(self):
        import models
        tb = dict([(str(blob.origkey), blob.blobinfo) for blob in mig_BlobMig.all()])
        for modelname, field in MODELS:
            to_put = []
            skipped = 0
            for obj in db.GqlQuery('SELECT * FROM %s'%modelname):
                oldkey = getattr(obj, field)
                if oldkey:
                    oldkey = str(oldkey.key())
                if oldkey in tb:
                    setattr(obj, field, tb[oldkey])
                    to_put.append(obj)
                else:
                    skipped += 1
            db.put(to_put)
            self.response.out.write("[%s] - Migrated %d references!<br/>Skipped: %d<br/>\n"%(modelname, len(to_put), skipped))

class MigrateBlobs(blobstore_handlers.BlobstoreDownloadHandler):
    '''Migrate blobs from myoldapp to mynewapp'''
    def get(self):
        import time
        from google.appengine.api import files, urlfetch
        start_time = time.time()
        timed_out = False
        migrated_keys = set([str(blob.origkey) for blob in mig_BlobMig.all()])
        all_blobs = eval(urlfetch.fetch("https://myoldapp.appspot.com/mig/__getblobkeys", deadline=15.0).content)
        all_keys = set(all_blobs.keys())
        missing_keys = list(all_keys-migrated_keys)
        if not missing_keys:
            self.response.out.write("Nothing to migrate!")
            return
        # Download 20 blobs per round
        download_keys = missing_keys[:20]
        MAXSIZE = (10*1024*1024) # 10MB
        for origkey in download_keys:
            blob = all_blobs[origkey]
            url = "https://myoldapp.appspot.com/mig/__getblob/%s"%urllib.quote_plus(origkey)
            blob_path = files.blobstore.create(mime_type=blob["content_type"], _blobinfo_uploaded_filename=blob["filename"])
            fsize = int(blob["size"])
            with files.open(blob_path, 'a') as f:
                for first_byte in range(0, fsize, MAXSIZE):
                    last_byte = (first_byte+MAXSIZE-1)
                    if last_byte>=fsize: last_byte=(fsize-1)
                    bytes_range = "bytes=%d-%d"%(first_byte,last_byte)
                    logging.info("Downloading [%s] range [%s] key=[%s]"%(blob["filename"], bytes_range, origkey))
                    res = urlfetch.fetch(url, deadline=35.0, headers={"Range": bytes_range})
                    f.write(res.content)
                    del res.content
                    del res
                    if (time.time()-start_time > 40.0):
                        timed_out = True
                        break
            if timed_out:
                self.response.out.write("Stopped due to time limit!<br/>")
                break
            files.finalize(blob_path)
            blob_key = files.blobstore.get_blob_key(blob_path)
            blob_info = blobstore.BlobInfo.get(blob_key)
            mig_BlobMig(fname=blob["filename"], blobinfo=blob_info, origkey=origkey).put()
            logging.info("Successfully downloaded [%s]!!!"%(blob["filename"]))
            self.response.out.write("Migrated %s<br/>"%blob["filename"])
        self.response.out.write("<br/>SUCCESS!")

class GetBlobKeys(blobstore_handlers.BlobstoreDownloadHandler):
    '''Retrieve all Blob Keys, as a JSON string'''
    def get(self):
        from simplejson.encoder import JSONEncoder
        blobs = {}
        for blob_info in blobstore.BlobInfo.all():
            blobs[str(blob_info.key())] = {
             "filename":str(blob_info.filename),
             "content_type":str(blob_info.content_type),
             "size":int(blob_info.size),
             "key":str(blob_info.key()),
            }
        self.response.headers["Content-type"] = "application/json"
        self.response.out.write(JSONEncoder().encode(blobs))

class GetBlob(blobstore_handlers.BlobstoreDownloadHandler):
    '''Download a blob given its key'''
    def get(self, blobkey):
        blobkey = str(urllib.unquote(blobkey))
        blob_info = blobstore.BlobInfo.get(blobkey)
        if not blob_info:
            self.error(404)
            return
        self.response.headers["Content-type"] = str(blob_info.content_type)
        self.send_blob(blob_info, save_as=True)

####
# END OF TEMPORARY BLOBSTORE MIGRATION CODE
####

• Modify the MODELS variable adding the models and blob info fields that you need to update
• Replace myoldapp URLs with the actual URL of your old application
• Synchronize the code on both apps
• Open a browser and point it to http://mynewapp.appspot.com/mig/__migrateblobs. Of course, mynewapp must be replaced with your actual HRD application name. This will download the blobs from the old app to the new one. Because there is a timeout limitation you will need to call it many times. WARNING: You must make only one request at time, never call it from two browsers / tabs at the same time or you will mess the migration database!!!. So this step must be carefully executed, one request at time and wait for it to finish before calling the next one. Repeat this until you see the message Nothing to migrate!. This step can be very slow depending on how many files you have in the blobstore and the size of them, it might take a long time to get all the files downloaded. Be patient!
• Once all the files are copied to the new app, point the browser to http://mynewapp.appspot.com/mig/__migratereferences. Should be quick, and it needs to be executed only one time. However doesn’t harm to call it more than once.
• The last step, remove the migration code and URLs and synchronize the code again. You can also remove the mig_BlobMig database.

Good migration.

Too easy? Ok, a real ninja hacker solution would had been taking apart the photo frame, put a Linux on the nice ARC processor through the JTAG interface and then plug some Wi-Fi dongle and do Everything directly on the DPF itself. But that’s just dreaming, we don’t have experience nor time for that.

Now you may ask, why didn’t you plug a monitor on the machine and just display whatever you want directly on the monitor? and the first reason is, we don’t have spare LCD monitors for this purpose (in fact we have many CRT monitors that nobody uses, but that requires just too much space). The second reason is that we wanted to do something more exiting, like hacker things .

So we setup everything: we made the scripts to generate the images, installed Ubuntu Linux 10.4 on the Itona and actually we borrowed another 512mb (flash) hard disk form another Itona and put it in the secondary IDE to get a RAID configuration with 1GB of space, just enough for Ubuntu and our script. Then realized of a small problem, as soon as you plug the DPF on the USB port, it enters in “transfer mode” and stops displaying the images. No matter if you eject, unmount, safe remove, shut down the computer, etc. the photo frame wont show the pictures until you physically unplug the USB cable (cut off the +5V). Here is were our ingenious circuit enters in action. A USB interrupter controlled by the parallel port (I guess serial RS-232 would work fine also using one of the signal pins).
After googling, I’ve found a circuit that seemed usable for this purpose. Just a couple transistors and a few resistors is everything we need.

Crop the board and put it in a case…



Finally, the whole thing mounted:
The Itona hidden behind the desk…

And the DPF showing the reports!!

For the circuit you will need:

• 1k resistor (2)
• 10k resistor (2)
• PNP transistor
• NPN transistor
• USB connectors, 1 to plug into the computer, the other whatever is better for your DPF
• DB-25 male connector (if you decided to do it via parallel too)

IMPORTANT: following image was borrowed from http://forum.allaboutcircuits.com/showthread.php?t=38544&page=2

I’ve connected the control-in to the PIN #2 of the DB-25, and the PIN #20 to the ground.
To turn the parallel port signal on/off I’ve made the following program in C (para.c):

#include 
#include 
#include 
#include 
#include 
#include 
#include 

#include


static int pfd;

int main(int argc, char* argv[])
{
	unsigned char data=0;

	if (argc != 3) {
		printf("Usage:\n\tpara [port] [val]\n\nport: parallel port e.g. /dev/parport0\nval:  1 to activate all data bits in parallel port, or\n      0 to disable all data bits in parallel port\n\n");
		return 1;
	}
	if (argv[2][0] == '1') {
		data=0xff;
	}
	pfd = open(argv[1], O_RDWR);
	if (pfd < 0) {
		perror("Failed to open port");
		exit(0);
	}
	if ((ioctl(pfd, PPEXCL) < 0) || (ioctl(pfd, PPCLAIM) < 0)) {
		perror("Failed to lock port");
		close(pfd);
		exit(0);
	}
	printf("Setting %s data bits to: %.2X\n", argv[1], data);
	ioctl(pfd, PPWDATA, &data);
	close(pfd);
	return 0;
}

in line 674 add:
if mail in ("xxx@yyy.com", ): continue

then in line 770 add:
if email in ("xxx@yyy.com", ): return

So the final Controller.py will look something like:

....

    def checkPending(self):
        '''Check for users pending to be added'''

        if self.msn is None:
            return False

        if self.addBuddy is None:
            self.addBuddy = dialog.AddBuddy(self)

        users = self.msn.checkPending()
        if len(users) > 0:
            for mail in users:
                if mail in ("xxx@yyy.com", ): continue
                nick = self.msn.getUserDisplayName(mail)
                self.addBuddy.append(nick, mail)
        return False

....

    def addNotification(self, msnp, command, tid, params, email, nick):
        '''this method is called when a user adds you'''

        if self.addBuddy is None:
            self.addBuddy = dialog.AddBuddy(self)
        if email in ("xxx@yyy.com", ): return
        self.addBuddy.append(nick, email)

Useful for sites that does not shows the upload progress like youtube.

This is my first add-on, so if you find something wrong please let me know.
I’ve submitted it to AMO but since it takes time to get released to the general public, I’ve decided to put it here in our blog if you want to give it a try.

Download it here, enjoy!

Tested in Linux, MacOS X and Windows, this small add-on is 100% JavaScript, no dlls no native XPCOMs. I’ve read in some forums that once upon a time (2004) Firefox already included an upload progress indicator but since then till today is broken. Today other modern browsers like Chrome already provides an upload percent indicator.

Edit: version 0.3 now displaying a small progress bar in Firefox’s status-bar. Check it out!

To create two bridged virtual serial ports use the following command:

socat -d -d pty,raw,echo=0 pty,raw,echo=0


The output will show you which are the virtual ports (or pseudo terminals) created, e.g.:

2010/02/19 16:16:33 socat[9662] N PTY is /dev/pts/3
2010/02/19 16:16:33 socat[9662] N PTY is /dev/pts/4
2010/02/19 16:16:33 socat[9662] N starting data transfer loop with FDs [3,3] and [5,5]


Note: if you are using Ubuntu and you do not have this command, try:

sudo apt-get install socat


Image source

The easiest way I found is dumping the binaries into text files using hexdump and then comparing them with your favourite program (diff, Meld, etc.). E.g.:

hexdump -C a.bin >a.txt
hexdump -C b.bin >b.txt
diff a.txt b.txt


Image source

from struct import unpack
from datetime import datetime

class FLVReader(dict):
    """
    Reads metadata from FLV files
    """

    # Tag types
    AUDIO = 8
    VIDEO = 9
    META = 18
    UNDEFINED = 0

    def __init__(self, filename):
        """
        Pass the filename of an flv file and it will return a dictionary of meta
        data.
        """
        # Lock on to the file
        self.file = open(filename, 'rb')
        self.signature = self.file.read(3)
        assert self.signature == 'FLV', 'Not an flv file'
        self.version = self.readbyte()
        self.typeFlags = self.readbyte()
        self.dataOffset = self.readint()
        extraDataLen = self.dataOffset - self.file.tell()
        self.extraData = self.file.read(extraDataLen)
        self.readtag()

    def readtag(self):
        unknown = self.readint()
        tagType = self.readbyte()
        dataSize = self.read24bit()
        timeStamp = self.read24bit()
        unknown = self.readint()
        if tagType == self.AUDIO:
            print "Can't handle audio tags yet"
        elif tagType == self.VIDEO:
            print "Can't handle video tags yet"
        elif tagType == self.META:
            endpos = self.file.tell() + dataSize
            self.event = self.readAMFData()
            metaData = self.readAMFData()
            # We got the meta data.
            # Our job is done.
            # We are complete
            self.update(metaData)
        elif tagType == self.UNDEFINED:
            print "Can't handle undefined tags yet"

    def readint(self):
      data = self.file.read(4)
      return unpack('>I', data)[0]

    def readshort(self):
      data = self.file.read(2)
      return unpack('>H', data)[0]

    def readbyte(self):
      data = self.file.read(1)
      return unpack('B', data)[0]

    def read24bit(self):
      b1, b2, b3 = unpack('3B', self.file.read(3))
      return (b1 << 16) + (b2 << 8 ) + b3

    def readAMFData(self, dataType=None):
        if dataType is None:
            dataType = self.readbyte()
        funcs = {
            0: self.readAMFDouble,
            1: self.readAMFBoolean,
            2: self.readAMFString,
            3: self.readAMFObject,
            8: self.readAMFMixedArray,
           10: self.readAMFArray,
           11: self.readAMFDate
        }
        func = funcs[dataType]
        if callable(func):
            return func()

    def readAMFDouble(self):
        return unpack('>d', self.file.read(8))[0]

    def readAMFBoolean(self):
        return self.readbyte() == 1

    def readAMFString(self):
        size = self.readshort()
        return self.file.read(size)

    def readAMFObject(self):
        result = {}
        data = True
        while data:
            size = self.readshort()
            key = self.file.read(size)
            dataType = self.readbyte()
            if not key and dataType == 9:
                break
            data = self.readAMFData(dataType)
            result[key] = data
        return result

    def readAMFMixedArray(self):
        size = self.readint()
        result = {}
        i = 0
        while i < size:
            key = self.readAMFString()
            dataType = self.readbyte()
            if not key and dataType == 9:
                break
            result[key] = self.readAMFData(dataType)
            i += 1
        return result

    def readAMFArray(self):
        size = self.readint()
        result = []
        i = 0
        while i < size:
            result.append(self.readAMFData())
            i += 1
        return result

    def readAMFDate(self):
        date = self.readAMFDouble()/1000
        localoffset = self.readshort()
        return datetime.fromtimestamp(date)

if __name__ == '__main__':
    import sys
    from pprint import pprint
    if len(sys.argv) == 1:
        print 'Usage: %s filename [filename]...' % sys.argv[0]
        print 'Where filename is a .flv file'
        print 'eg. %s myfile.flv' % sys.argv[0]
    for fn in sys.argv[1:]:
        x = FLVReader(fn)
        pprint(x)

Download this code snippet from here

<html>
<body>
<div style="color: blue;">
	<table>
	<tbody>
		<tr>
			<td>BLUE TEXT</td>
		</tr>
	</tbody>
	</table>
<div>
</body>
</html>

In the below example I would expect to see the “BLUE TEXT” in blue, but is not (at least in FireFox). So what the hell is wrong with my code? probably many things but because I’m not interested in following/reading the W3C recommendations for my uncle’s home page, here is one possible solution: Add the transitional DOCTYPE to your document, or the DOCTYPE that adequate better to your page (read more here). E.g.:

<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">

So, now you know. Never forget the DOCTYPE it is very important… Why? read this… It has to do with poor standards, browsers and all that crap.

Simplified DES

This encryption algorithm is not secure at all, in fact it was made for educational purposes. It uses a 10-bit key that can be quickly broken by a brute force attack. But, who knows, it may be good for someone interested in some quick encryption where security is not that important, or for embedded devices were resources are limited, or just for lazy students.

So, here is my version of the S-DES algorithm in plain C. Differently from the other version in the web my implementation is more byte-oriented while the other (probably more didactic) is converting the input to a string with ’0′s and ’1′s and working over it. Also the other implementation uses some C++ functions while mine is written in plain C.

Download it and take a look to the code (it is less than 300 lines) — released under BSD license.

To build it with gcc type:

gcc sdes.c -lm -o sdes

Note that I’m using math library only because of the pow() function that is used only in debug mode.

And here is a copy of a document explaining the algorithm.

Image source