gemini - kennedy.gemi.dev

💾 Archived View for clemat.is › saccophore › library › ezines › textfiles › ezines › EUROHACKER › IS… captured on 2022-01-08 at 15:37:09.
-=-=-=-=-=-=-
<html>

<head>

<title> EuroHacker Magazine </title>

<link rel="stylesheet" type="text/css" href="style.css" />

</head>

<body>

<table style="width: 100%;">

<tr class="page_top_thingy">

<td width="10%"><a style="color: #000000;" href="a13.html">Previous</a></td>

<td align="center"><a style="color: #FFFFFF;" href="index.html">EuroHacker Magazine, issue #1</a></td>

<td width="10%"><a style="color: #000000;" href="a15.html">Next</a></td>

</tr>

</table>

<hr>
<h1>Pr0ngallery</h1>

<p align="center"><b>Written by:</b> guess who?</p>

<p> A while ago I needed to sample a pretty vast file collection,
belonging to a dude I know on the Net. The files were large movie files
and so I would have to download all of them just to check out if I liked
'em. This annoyed me so I thought about what to do. The idea I came up
with was using mplayer to rip stills out of each file, so I could easily
check out if it was worth downloading. In hindsight I could have
downloaded chunks of each file using HTTP's Range header and then done
mplayer -idx on them but I don't know if that would work plus that's an
unnecessary waste of bandwidth. This solution only requires me to run
the script on the dude's box. </p>

<p> It's worked so far, but it wouldn't surprise me if it craps out.
YMMV. Also, this is my first and only Python script at this point, so
don't harass me. </p>

<div style="height: 80%; width: 80%; overflow: scroll;">

<pre>

#!/usr/bin/python

import os, sys, commands, re, string, glob

def usage():
	print &#039;pr0ngallery - A movie still gallery generator&#039;
	print &#039;Usage: pr0ngallery &lt;files&gt;&#039;

def make_stills(filename, dir, interval, num_stills):

	os.chdir(dir)

	i=1
	
	while i &lt;= num_stills:
	
		# grab a few frames
		# I&#039;ve found that to avoid black frames, cluttered frames and other crap, generally you need to grab the fourth frame
		# -ss seek to second
		# -frames number of frames to grab (you need to increment the desired number by one, oddly enough)
		# -vo video out
		# -ao audio out
		# -z png compression level
		os.system(&#039;mplayer -ss %d -frames 5 -vo png -z 1 -ao null &quot;%s&quot; 2&gt;/dev/null &gt;/dev/null&#039; % (i*interval, filename))

		#print &quot;num_stills: %d&quot; % num_stills
		
		print &#039;Creating still %d for &quot;%s&quot;&#039; % (i, filename)
		
		try:
		    try:
			os.rename(&#039;00000004.png&#039;, &#039;still_%06d.png&#039; % i) # may fail, I&#039;ve found
		    except OSError:
			print &#039;Error, skipping&#039;
			continue
		finally:
		    # delete all the crap
		    d=glob.glob(&#039;*.png&#039;)
		
		    for curr in d:
			if not re.compile(&#039;^still.+{body}amp;#039;).match(curr):
			    os.unlink(curr)
		
		    i+=1

	os.chdir(&#039;..&#039;)

def get_movie_info(filename):
	buffer=commands.getoutput(&#039;mplayer -vo null -ao null -frames 0 -identify &quot;%s&quot;&#039; % filename)

	m=re.compile(r&#039;.*?ID_VIDEO_FORMAT=([^\n]+).*?ID_VIDEO_BITRATE=([^\n]+).*?ID_VIDEO_WIDTH=([^\n]+).*?ID_VIDEO_HEIGHT=([^\n]+).*?ID_VIDEO_FPS=([^\n]+).*?ID_VIDEO_ASPECT=([^\n]+).*?ID_AUDIO_CODEC=([^\n]+).*?ID_AUDIO_FORMAT=([^\n]+).*?ID_AUDIO_BITRATE=([^\n]+).*?ID_AUDIO_RATE=([^\n]+).*?ID_AUDIO_NCH=([^\n]+).*?ID_LENGTH=([^\n]+)&#039;, re.S).match(buffer)
	
	if not m:
		print &#039;Unexpected error!&#039;
		sys.exit(-1)

	t=m.groups()

	info={&#039;vformat&#039;: t[0], &#039;vbitrate&#039;: int(t[1]), &#039;vwidth&#039;: int(t[2]), &#039;vheight&#039;: int(t[3]), &#039;vfps&#039;: float(t[4]), &#039;vaspect&#039;: float(t[5]), &#039;acodec&#039;: t[6], &#039;aformat&#039;: int(t[7]), &#039;abitrate&#039;: int(t[8]), &#039;arate&#039;: int(t[9]), &#039;anch&#039;: int(t[10]), &#039;length&#039;: int(t[11])}

	return info

if __name__ == &#039;__main__&#039;:

	# TODO: this shouldn&#039;t be hardcoded
	base_dir=&#039;gallery&#039;

	if len(sys.argv) &lt; 2:
		usage()
		sys.exit(-1)
		
	# get curr dir
	launch_dir=os.getcwd()

	# make the dir where the galleries will be made
	# abort if it already exists
	if (os.access(base_dir, os.F_OK)):
		print &#039;&quot;%s&quot; already exists, aborting&#039; % base_dir
		sys.exit(-1)
	
	# make the dir
	os.mkdir(base_dir)
	
	# change into it
	os.chdir(base_dir)

	i=1

	# loop thru all the filenames given to us
	for filename in sys.argv[1:]:
	
		if not os.path.isabs(filename):
		    filename=&#039;%s/%s&#039; % (launch_dir, filename)

		num_stills=200 # TODO: shouldn&#039;t be hardcoded
	
		movie_info=get_movie_info(filename)
		
		#print movie_info
		
		if movie_info[&#039;length&#039;] == 0:
		    print &#039;Zero-length movie, skipping...&#039;
		    continue
		elif movie_info[&#039;length&#039;] &lt;= 10:
		    num_stills=1
		# determine number of stills we need
		# there should be at least 10 seconds between stills
		elif float(movie_info[&#039;length&#039;])/float(num_stills) &lt; 10:
		    num_stills=int(float(movie_info[&#039;length&#039;])/10)
		    
		interval=int(float(movie_info[&#039;length&#039;])/num_stills)
		
		print &#039;Making %d stills for movie &quot;%s&quot;&#039; % (num_stills, filename)
		print &#039;Using an interval of %d seconds for the stills&#039; % interval
		print &#039;Length of movie: %d seconds&#039; % movie_info[&#039;length&#039;]
		
		dir=&quot;stills_%09d&quot; % i
		
		os.mkdir(&#039;%s&#039; % dir)
		
		make_stills(filename, dir, interval, num_stills)
		
		i+=1

</pre>

</div>

<hr>

<small>Copyright 2005, EuroHacker Magazine</small>
</body>

</html>