import os,pickle,csv

import Tkinter as tk

import Tkconstants, tkFileDialog, tkMessageBox

import numpy as np

root = tk.Tk()

options = {}

options['title'] = 'Select the tif and dmp file folder'

options['parent'] = root

folder_path = tkFileDialog.askdirectory(**options)

for thisfile in os.listdir(folder_path):

if thisfile.endswith(".dmp"):

print thisfile

try:

pickle_path = os.path.join(folder_path, thisfile)

#Fix the fact that it was done originally in Windows...

pickle_file = open(pickle_path, 'rb')

text = pickle_file.read().replace('\r\n', '\n')

pickle_file.close()

pickle_file = open(pickle_path, 'wb')

pickle_file.write(text)

pickle_file.close()

pickle_file = open(pickle_path,'r')

pick = pickle.load(pickle_file)

ff_translation_info_rowshift = pick['full_frame_ncc']['row_shifts']

ff_translation_info_colshift = pick['full_frame_ncc']['column_shifts']

strip_translation_info = pick['sequence_interval_data_list']

firsttime=True

pickle_file.close()

matches=list()

#Find all of the tifs that match this dmp file, and record them.

for thatfile in os.listdir(folder_path):

if thatfile.endswith(".tif") and thatfile.startswith(thisfile[0:-4]):

matches.append(thatfile)

print "Matches transform file: "+thatfile

minmaxpix = np.empty([1,2])

for frame in strip_translation_info:

for frame_contents in frame:

ref_pixels = frame_contents['slow_axis_pixels_in_current_frame_interpolated']

minmaxpix = np.append(minmaxpix,[[ref_pixels[0], ref_pixels[-1]]], axis=0)

minmaxpix=minmaxpix[1:,:]

#print minmaxpix[:,1].min()

#print minmaxpix[:,0].max()

topmostrow = minmaxpix[:,0].max()

bottommostrow= minmaxpix[:,1].min()

#print np.array([pick['strip_cropping_ROI_2'][-1]]).shape

# The first row is the crop ROI.

for i in range(0,len(matches)):

np.savetxt(os.path.join(folder_path, matches[i][0:-4])+ "_transforms.csv" , np.array([pick['strip_cropping_ROI_2'][i]]), delimiter=",", newline="\n", fmt="%f")

for frame in strip_translation_info:

if len(frame) > 0:

print "************************ Frame " +str(frame[0]['frame_index']+1) + "************************"

frame_ind = frame[0]['frame_index']

slow_axis_pixels=np.zeros([1])

all_col_shifts=np.zeros([1])

all_row_shifts=np.zeros([1])

for frame_contents in frame:

slow_axis_pixels = np.append(slow_axis_pixels,frame_contents['slow_axis_pixels_in_reference_frame'])

ff_row_shift = ff_translation_info_rowshift[frame_ind]

ff_col_shift = ff_translation_info_colshift[frame_ind]

#First set the relative shifts

row_shift = (np.subtract(frame_contents['slow_axis_pixels_in_reference_frame'], \

frame_contents['slow_axis_pixels_in_current_frame_interpolated']))

col_shift = (frame_contents['fast_axis_pixels_in_reference_frame_interpolated'])

#These will contain all of the motion, not the relative motion between the aligned frames-

#So then subtract the full frame row shift

row_shift = np.add(row_shift, ff_row_shift)

col_shift = np.add(col_shift, ff_col_shift)

all_col_shifts = np.append(all_col_shifts,col_shift)

all_row_shifts = np.append(all_row_shifts,row_shift)

slow_axis_pixels = slow_axis_pixels[1:]

all_col_shifts = all_col_shifts[1:]

all_row_shifts = all_row_shifts[1:]

for matchedfile in matches:

transhandle = open( os.path.join(folder_path, matchedfile[0:-4]) + "_transforms.csv", 'a')

np.savetxt(transhandle, np.vstack( (slow_axis_pixels, all_col_shifts, all_row_shifts) ),delimiter=',',fmt='%f')

transhandle.close()

except(ValueError, RuntimeError):

tkMessageBox.showwarning("DMP failed to process.","Failed to process DMP ("+thisfile+")! This file may be corrupted. Re-process the DMP, or contact your local RFC.")

root.destroy()