parse a text file to retrieve keyframes

http://forums.cgsociety.org/showthread.php?p=6473887#post6473887

I stumbled on this at CgTalk and though I would share it here, as it is something that we encounter now and then. Almost all programs as the option to output a text file with raw data of keyframes. But it can be tedious if we don’t know how to parse a text file and get it into maya or softimage.

I will only cover how to parse the file here as it’s mostly a python exercise.

The content of the text file looks something like this

13.232 12.134 11.134
14.234 67.423 22.634
13.232 10.001 7.346
13.232 12.134 4.126

So each column is X Y Z and each row is a frame. We now need to open this file and read in the text

raw_keyframes_file = open('text.txt')
text = raw_keyframes_file.read()
raw_keyframes_file.close()
print text

Once you have it as a list we can start to split the text so that we get each frame. Python sees a linebreak as \n, so we can use that to split the text into frames (since each line is a frame).

raw_keyframes_file = open('text.txt')
frames = raw_keyframes_file.read().split('\n')
raw_keyframes_file.close()
print frames

And we can now also split each frame in into channels. And we should remove empty rows as they might mess up the animation (if each frame was on every other row instead if each row).

raw_keyframes_file = open('text.txt')
frames = raw_keyframes_file.read().split('\n')
raw_keyframes_file.close()
keyframes = [] #empty to be filled with keyframes
for frame in frames:
    ''' removed empty rows '''
    if not frame == '':
        keyframes.append(frame)
x = []
y = []
z = []
for channel in keyframes:
    ''' use the space to split the keyfame into channels '''
    chan = channel.split(' ')
    x.append(chan[0])
    y.append(chan[2])
    z.append(chan[4])
print x, y, x

We have now successfully gathered all information that we need. But the code isn’t very pretty, and we would be better off if we could reuse this. So……… **loud music*** let’s make it into a class.

class MyTools(object):
    def parse_keys(self, textfile):
        raw_keyframes_file = open(textfile)
        frames = raw_keyframes_file.read().split('\n')
        raw_keyframes_file.close()
        keys = MyTools()
        keys.x = []
        keys.y = []
        keys.z = []
        keyframe = []
        for frame in frames:
            if not frame == '':
                keyframe.append(frame)
        for channel in keyframe:
            chan = channel.split(' ')
            keys.x.append(chan[0])
            keys.y.append(chan[2])
            keys.z.append(chan[4])
        return keys

It’s almost the same code, with the difference that we now have it wrapped up in a class. And we now have a attribute called ‘keys’ which also has x, y, z in it. So if you save out the file in a folder called ‘tools’ and call it ‘__init__.py’. If you read the previous ‘lesson’ we can now import this into other scripts.

import sys
sys.path.append('/server/tools/python')
from sanders3d import tools

tools = tools.MyTools() #first we need to instance the class
keys = tools.parse_keys('raw_keyframe.txt') #now point the text file into it

#print out the channels
print keys.x
print keys.y
print keys.z

Did that work? I hope so