# This script gathers available sentinel files from the database and checks the coverage in space and time for this data
# stack.

import os
import warnings
from collections import Counter, OrderedDict
from datetime import datetime
import shutil

import fiona
import numpy as np
from shapely.geometry import shape, mapping, box
from shapely.ops import cascaded_union

from . import image as image
from doris.doris_stack.functions.load_shape_unzip import extract_kml_preview, shape_im_kml, load_shape
from doris.doris_stack.main_code.dorisparameters import DorisParameters
from doris.doris_stack.functions.burst_metadata import center_shape_from_res
from doris.doris_stack.main_code.jobs import Jobs


class StackData(object):
    # This function holds information for a full datastack of sentinel data and is used to select relevant images and
    # bursts based on a given area of interest.

    def __init__(self, track_dir, shape_dat, buffer=0.02, start_date='2014-04-01', end_date='', polarisation='vh', path='', db_type=1, precise_dir=''):
        # Initialize variables:

        Jobs.id = 0

        # Datastack folder where all data from the datastack is stored (database,shapes,burst data, res files and results)
        # You should have write acces to this folder!
        self.path = path
        self.unzip_path = ''

        # Search path, shape, buffer and polarisation for this datastack. Currently only one polarisation is implemented. If
        # needed this can be extended later.
        self.search_path = []
        self.start_date = []
        self.end_date = []
        self.master_date = []
        self.shape = []
        self.shape_filename = shape_dat
        self.buffer = []
        self.polarisation = ''
        self.precise_orbits = ''

        # All images included in this datastack. Including lists of shapes and acquistion dates
        self.images = []
        self.image_files = []
        self.image_shapes = []
        self.image_dates = []

        # The shapes and names of the swaths
        self.swath_names = list()
        self.swath_shapes = list()

        # The resulting dates, with underlying bursts. This information can be used to create res files and images
        # of individual bursts. Variable data includes a structure using (dates > swaths > bursts).
        self.dates = list()
        self.datastack = dict()
        self.concatenated = dict()
        self.coordinates = dict()
        self.burst_names = list()
        self.burst_shapes = list()
        self.burst_centers = list()

        # Some important characteristics of the dataset of bursts. (matrices with names on y and dates on x axis)
        self.burst_no = 0
        self.burst_availability = []
        self.burst_lon = []
        self.burst_lat = []
        self.burst_baselines = []

        # Temporary variable to store images which are not yet checked for shape or dates.
        self.image_dump = []

        # parallel computing:
        doris_parameters = DorisParameters(os.path.dirname(self.path))
        self.doris_parameters = doris_parameters
        self.nr_of_jobs = doris_parameters.nr_of_jobs
        self.parallel = doris_parameters.parallel
        self.function_path = doris_parameters.function_path

        ####################################################################

        # This function initializes the datastack using a search path, start/end dates and a buffer shape. The start and
        # end date should have the format yyyy-mm-dd and the shape should either be a shapefile or a list of coordinate
        # pairs. [[lat,lon][lat,lon] enz. ]. Minimum input is a folder which contains the .SAFE folders (a track folder)

        if not track_dir or not os.path.exists(track_dir):
            warnings.warn('This function needs an existing path as input!')
            return

        self.search_files(track_dir)

        if shape:
            self.create_shape(shape_dat,buffer)

        if end_date:
            self.end_date = np.datetime64(end_date).astype('datetime64[s]') + np.timedelta64(1, 'D').astype('timedelta64[s]')
        else:
            self.end_date = np.datetime64('now').astype('datetime64[s]')
        self.start_date = np.datetime64(start_date).astype('datetime64[s]')

        if isinstance(polarisation, str):
            polarisation = [polarisation]
            for i in polarisation:
                if not i in ['hh','vv','hv','vh']:
                    warnings.warn('This polarisation does not exist for sentinel data.')
                    return
        self.polarisation = polarisation
        self.search_path = track_dir

        if not path:
            warnings.warn('You did not specify an output path. Please do so later on using the add_path function')
        else:
            self.add_path(path)

        if precise_dir:
            if os.path.exists(precise_dir):
                self.precise_orbits = precise_dir
            else:
                print('Precise orbit path does not exist')

    def add_path(self,path):
        # This function adds the output path.

        if os.path.isdir(path):
            self.path = path
        elif os.path.isdir(os.path.dirname(path[:-1])):
            os.mkdir(path)
            self.path = path
        else:
            warnings.warn('Neither the directory itself nor the parents directory exists. Choose another path.')

    def search_files(self, track_dir):
        # This function searches for files within a certain folder. This folder should contain images from the same
        # track. These images are added to the variable image dump.
        images = list()

        top_dir = next(os.walk(track_dir))

        for data in top_dir[2]:
            if data.endswith('.zip'):
                images.append(os.path.join(track_dir, data))
        for data in top_dir[1]:
            if data.endswith('.SAFE'):
                images.append(os.path.join(track_dir, data))
            else:  # Likely images are stored in a folder.
                sec_dir = next(os.walk(os.path.join(track_dir, data)))

                for dat in sec_dir[1]:
                    if dat.endswith('.SAFE'):
                        images.append(os.path.join(track_dir, data, dat))
                for dat in sec_dir[2]:
                    if dat.endswith('.zip'):
                        images.append(os.path.join(track_dir, data, dat))

        if images:
            images = sorted(images)
        else:
            print('No images found in track_dir = ' + str(track_dir) + ', switching to archive folder')

            track_name = self.doris_parameters.direction + '_t' + self.doris_parameters.track.zfill(3)
            track_dir = os.path.join(track_dir, track_name)

            print('Searching in folder ' + track_dir)

            top_dirs = next(os.walk(track_dir))[1]

            if len(top_dirs) == 1:
                track_dir = os.path.join(track_dir, top_dirs[0])
            elif len(top_dirs) > 1:
                for top_dir in top_dirs:
                    user_input = input("Do you want to use folder " + top_dir + " as resource folder? (yes/no)").lower()
                    if user_input in ['yes', 'y']:
                        track_dir = os.path.join(track_dir, top_dir)

            dir_new = next(os.walk(track_dir))
            for dat in dir_new[1]:
                dat_files = next(os.walk(os.path.join(track_dir, dat)))[2]
                for data in dat_files:
                    if data.endswith('.zip'):
                        images.append(os.path.join(track_dir, dat, data))

        if images:
            images = sorted(images)
        else:
            warnings.warn('No images found! Please choose another data folder. Track_dir = ' + str(track_dir))

        base = []  # Remove double hits because data is already unzipped.
        for i in images: # Drop all .zip files which are unpacked already.
            if i.endswith('.SAFE.zip'):
                base.append(os.path.basename(i[:-9]))
            elif i.endswith('.zip'):
                base.append(os.path.basename(i[:-4]))
            elif i.endswith('.SAFE'):
                base.append(os.path.basename(i[:-5]))
        b, id = np.unique(base, return_index=True)

        rem = []
        for i in range(len(base)):
            if i in id:
                self.image_dump.append(images[i])
            else:
                rem.append(images[i])

        if rem:
            print('removed the following zip files from stack:')
            for r in rem:
                print(r)
            print('It is advised to work with zipfiles instead of unpacked data. This saves diskspace and will ')

    def create_shape(self,shape_dat,buffer=0.02):
        # This function creates a shape to make a selection of usable bursts later on. Buffer around shape is in
        # degrees.

        self.shape = load_shape(shape_dat, buffer)
        self.buffer = buffer

    def check_new_images(self, master):
        # This function checks which images are already processed, and which are not. If certain dates are already
        # processed they are removed from the list. You have to specify the master date, otherwise the script will not
        # know how many burst are expected per date.

        # Which dates are available?
        image_dates = [im.astype('datetime64[D]') for im in self.image_dates]
        # What is the master date
        date = np.datetime64(master).astype('datetime64[D]')

        date_folders = [d for d in next(os.walk(self.path))[1] if len(d) == 8]
        rm_id = []

        if date_folders:
            dates = [np.datetime64(d[0:4] + '-' + d[4:6] + '-' + d[6:8]) for d in date_folders]

            if date in dates:
                date_folder = date_folders[np.where(dates == date)[0][0]]

                # Check existing files in master folder
                swaths = dict()
                swath_folders = next(os.walk(os.path.join(self.path, date_folder)))[1]
                if len(swath_folders) == 0:
                    print('No swaths in master folder')
                    return

                for swath in swath_folders:
                    self.swath_names.append(swath)
                    swaths[swath] = sorted(next(os.walk(os.path.join(self.path, date_folder, swath)))[1])

                # Now check if the burst also in slave folders exist....
                for folder, d in zip(date_folders, dates):
                    # Check existing files in master folder
                    try:
                        swath_folders = next(os.walk(os.path.join(self.path, folder)))[1]
                        if not set(swath_folders) == set(swaths.keys()):
                            raise LookupError('Amount of swaths is not the same for ' + folder)

                        for swath in swath_folders:
                            bursts = sorted(next(os.walk(os.path.join(self.path, date_folder, swath)))[1])
                            if not set(bursts) == set(swaths[swath]):
                                raise LookupError('Amount of bursts is not the same for ' + folder)

                            if d == date:
                                # If the master is already processed we have to create the list of center and coverage of
                                # bursts.
                                # TODO make this robust for the case no seperate input data folders are created.
                                res_files = []
                                for burst in bursts:
                                    dat_file = [r for r in next(os.walk(os.path.join(self.path, date_folder, swath, burst)))[2]
                                                if (r.startswith('slave_iw') and len(r) < 25)]
                                    res_file = os.path.join(self.path, date_folder, swath, burst, 'slave.res')
                                    res_files.append(res_file)

                                    center, coverage = center_shape_from_res(resfile=res_file)
                                    # Assign coverage, center coordinates and burst name.
                                    self.burst_shapes.append(coverage)
                                    self.burst_centers.append(center)
                                    burst_num = os.path.basename(dat_file[0])[17:-4]
                                    self.burst_names.append(swath + '_burst_' + burst_num)
                                    self.burst_no += 1

                        # If all bursts are the same these files are not processed.
                        for id in np.where(image_dates == d)[0][::-1]:
                            del self.image_dates[id]
                            del image_dates[id]
                            del self.images[id]
                            del self.image_files[id]

                    except LookupError as error:
                        print(error)

    def select_image(self,start_date='',end_date='', dest_folder=''):
        # This function selects usable images based on .kml files and dates

        if not dest_folder:
            dest_folder = os.path.join(self.path, 'kml')
        if not os.path.exists(dest_folder):
            os.makedirs(dest_folder)

        if not self.shape:
            warnings.warn('There is no shape loaded to select images. Please use the create_shape function to do so.')

        if start_date:
            self.start_date = np.datetime64(start_date).astype('datetime64[s]')
        if end_date:
            self.end_date = np.datetime64(end_date).astype('datetime64[s]') + np.timedelta64(1, 'D').astype('timedelta64[s]')

        # First select images based on dates and check if polygons intersect.
        for i in self.image_dump:
            d = os.path.basename(i)[17:32]
            acq_time = np.datetime64(d[0:4] + '-' + d[4:6] + '-' + d[6:11] + ':' + d[11:13] + ':' + d[13:] + '-0000')
            if acq_time >= self.start_date and acq_time <= self.end_date:
                im = image.ImageMeta(path=i)
                kml, png = extract_kml_preview(i, dir=dest_folder, png=False, overwrite=True)
                succes = shape_im_kml(self.shape, kml)

                if succes:
                    self.images.append(im)
                    self.image_dates.append(acq_time)
                    self.image_files.append(os.path.basename(i))
        self.dates = sorted(list(set([d.astype('datetime64[D]') for d in self.image_dates])))
        for date in self.dates:
            print(date)

    def select_burst(self, date=''):
        # This function selects the usefull bursts at one epoch (user defined or automatically selected) and searches
        # usefull burst at other dates. This function uses the extend_burst function, which is intended to search for
        # bursts at other dates. This function can be run later on to update the datastack.

        if self.burst_names:
            print('Master data is already loaded from earlier processed data.')
            return

        image_dates = [im.astype('datetime64[D]') for im in self.image_dates]
        # First select which date will be the master
        if date:
            date = np.datetime64(date).astype('datetime64[D]')
            # date = deepcopy(image_dates[min(abs(self.image_dates-date))])
        else:  # if no date is specified
            date = Counter(image_dates).most_common(1)[0][0]

        # Load the metadata for this date for the bursts and swaths
        for date in image_dates:
            print(date)
        print(date)
        image_id = np.where(image_dates == date)[0]
        for i in image_id:
            print('processing data: ' + self.images[i].unzip_path)
            self.images[i].meta_swath(precise_folder=self.precise_orbits)

        # Order the selected images by acquisition time.
        image_id = [x for (y,x) in sorted(zip([self.image_dates[i] for i in image_id],image_id))]
        date = date.astype(datetime).strftime('%Y-%m-%d')
        self.master_date = date
        self.datastack[date] = dict()

        swath_nums = ['1', '2', '3']

        for p in self.polarisation:
            for swath in swath_nums:
                swath_id = 'iw' + swath + '-slc-' + p
                swath_name = 'swath_' + swath
                burst_no = 1
                data = []

                for i in image_id:
                    # Check which swath should be selected.
                    if not self.images[i].swaths_data:
                        continue

                    swath_names = [os.path.basename(data) for data in self.images[i].swaths_data]
                    swath_no = [no for no in range(len(swath_names)) if swath_id in swath_names[no]]
                    if not swath_no:  # If there is no data for this swath
                        continue

                    swath_no = swath_no[0]

                    for burst in self.images[i].swaths[swath_no].bursts:
                        if burst.burst_coverage.intersects(self.shape):

                            # Check if we do not have a copy of former bursts...
                            if len(self.burst_centers) != 0:
                                dist_center = [np.sqrt((burst.burst_center[0] - center[0])**2 +
                                               (burst.burst_center[1] - center[1])**2) for center in self.burst_centers]
                                if min(dist_center) < 0.02:
                                    print('Possible duplicate burst deleted')
                                    continue

                            # If there are bursts in this swath, which intersect, add burst to list.
                            if swath_name not in self.datastack[date].keys(): # If there are burst and no burst list exists
                                self.datastack[date][swath_name] = dict()
                                if swath_name not in self.swath_names:
                                    self.swath_names.append(swath_name)

                            # Assign burst to data stack
                            burst_name = 'burst_' + str(burst_no)
                            burst.new_burst_num = burst_no
                            self.datastack[date][swath_name][burst_name] = burst

                            # Create .res file data
                            burst.meta_burst(swath_meta=self.images[i].swaths[swath_no].metadata)

                            # Assign coverage, center coordinates and burst name.
                            self.burst_shapes.append(burst.burst_coverage)
                            self.burst_centers.append(burst.burst_center)
                            self.burst_names.append(swath_name + '_' + burst_name)
                            burst_no += 1

                            # Finally add also to the number of bursts from the image
                            self.images[i].burst_no += 1
                            self.burst_no += 1

    def extend_burst(self):
        # Searches for burst at dates other than the dates that are already available. This means that if there are
        # images added for dates which are already indexed, this data will not be used!

        image_dates = [d.astype('datetime64[D]') for d in self.image_dates]

        for date in self.dates:
            # Append data to datastack variable
            date_str = date.astype(datetime).strftime('%Y-%m-%d')
            self.datastack[date_str] = dict()
            data = []

            # Load the metadata for this date for the bursts and swaths
            image_id = np.where(image_dates == date)[0]
            if len(image_id) == 0:
                continue

            for i in image_id:
                if date_str == self.master_date:
                    continue
                print('processing data: ' + self.images[i].unzip_path)
                self.images[i].meta_swath(precise_folder=self.precise_orbits)

            for swath in self.swath_names:
                # Add swath to datastack
                self.datastack[date_str][swath] = OrderedDict()

                for i in image_id:
                    # Select correct swath in image
                    swath_id = swath[-1] + '-slc-' + self.polarisation[0]

                    swath_names = [os.path.basename(data) for data in self.images[i].swaths_data]
                    swath_no = [no for no in range(len(swath_names)) if swath_id in swath_names[no]]
                    if not swath_no:  # If there is no data for this swath
                        continue

                    swath_no = swath_no[0]

                    for burst in self.images[i].swaths[swath_no].bursts:
                        x_dist = np.array([xy[0] - burst.burst_center[0] for xy in self.burst_centers])
                        y_dist = np.array([xy[1] - burst.burst_center[1] for xy in self.burst_centers])

                        dist = np.sqrt(x_dist**2 + y_dist**2)
                        burst_id = np.argmin(dist)

                        if dist[burst_id] < 0.1:
                            # Assign burst to data stack
                            burst.new_burst_num = int(self.burst_names[burst_id][14:])
                            self.datastack[date_str][swath][self.burst_names[burst_id][8:]] = burst

                            # Create .res file data
                            burst.meta_burst(swath_meta=self.images[i].swaths[swath_no].metadata)
                        else:
                            print('No corresponding burst found! Closest is ' + str(dist[burst_id]) + ' ' +
                                  date_str + ' ' + swath + ' ' + self.burst_names[burst_id])

        # Remove all unnecessary dates from stack
        for dat_key in self.datastack.keys():
            if not self.datastack[dat_key]:
                del self.datastack[dat_key]

    def remove_incomplete_images(self):
        # This function removes all the images with less than maximum bursts. This will make a stack more consistent.

        for key in self.datastack.keys():
            burst_no = 0
            for key_swath in self.datastack[key].keys():
                burst_no += len(self.datastack[key][key_swath])
            if burst_no != self.burst_no:
                self.datastack.pop(key)
                print('Number of burst for ' + key + ' is ' + str(burst_no) + ' instead of ' + str(self.burst_no) +
                      ' and is removed from the datastack.')

    def define_burst_coordinates(self,slaves=False):
        # This function defines the exact coordinates in pixels of every burst based on the lower left corner of the first
        # burst image. In this way the total overlap of these bursts can easily be monitored. Results are written to the
        # coordinates variable

        if slaves is True:
            dates = list(self.datastack.keys())
        else:
            dates = [self.master_date]

        self.coordinates = OrderedDict()

        for date in dates:
            self.coordinates[date] = OrderedDict()
            self.coordinates[date]['shapes'] = []
            ref = False

            min_line = 1; min_pixel = 1
            max_line = 1; max_pixel = 1
            for swath in self.datastack[date].keys():

                self.coordinates[date][swath] = OrderedDict()
                self.coordinates[date][swath]['corners'] = np.zeros([len(list(self.datastack[date][swath].keys())), 4, 2],dtype='int')

                b = 0
                for burst in sorted(self.datastack[date][swath].keys(), key = lambda x: int(x[6:])):

                    if ref is False:
                        self.coordinates[date]['ref_az_time'] = self.datastack[date][swath][burst].processes['readfiles']['First_pixel_azimuth_time (UTC)']
                        self.coordinates[date]['ref_range_time'] = self.datastack[date][swath][burst].processes['readfiles']['Range_time_to_first_pixel (2way) (ms)']
                        ref = True

                    az_first = self.datastack[date][swath][burst].processes['readfiles']['First_pixel_azimuth_time (UTC)']
                    az_samp = self.datastack[date][swath][burst].processes['readfiles']['Pulse_Repetition_Frequency (computed, Hz)']
                    first_line = int(self.datastack[date][swath][burst].processes['crop']['First_line (w.r.t. original_image)'])
                    last_line = int(self.datastack[date][swath][burst].processes['crop']['Last_line (w.r.t. original_image)'])

                    range_first = self.datastack[date][swath][burst].processes['readfiles']['Range_time_to_first_pixel (2way) (ms)']
                    range_samp = self.datastack[date][swath][burst].processes['readfiles']['Range_sampling_rate (computed, MHz)']
                    first_pixel = int(self.datastack[date][swath][burst].processes['crop']['First_pixel (w.r.t. original_image)'])
                    last_pixel = int(self.datastack[date][swath][burst].processes['crop']['Last_pixel (w.r.t. original_image)'])

                    no_lines = int(self.datastack[date][swath][burst].processes['readfiles']['Number_of_lines_original'])
                    no_pixels = int(self.datastack[date][swath][burst].processes['readfiles']['Number_of_pixels_original'])

                    # Calculate difference w.r.t. reference point.
                    range_time_diff = (float(range_first) - float(self.coordinates[date]['ref_range_time']))
                    pixel_diff = int(round((float(range_samp) * 1e3) * range_time_diff))

                    az_time1 = datetime.strptime(az_first, '%Y-%b-%d %H:%M:%S.%f')
                    az_time2 = datetime.strptime(self.coordinates[date]['ref_az_time'], '%Y-%b-%d %H:%M:%S.%f')
                    az_time_diff = (az_time1 - az_time2).total_seconds()
                    line_diff = int(round(float(az_samp) * az_time_diff))

                    # Calculate final corner coordinates.
                    ll = np.array([line_diff + first_line, pixel_diff + first_pixel], ndmin=2)
                    ul = np.array([line_diff + last_line, pixel_diff + first_pixel], ndmin=2)
                    ur = np.array([line_diff + last_line, pixel_diff + last_pixel], ndmin=2)
                    lr = np.array([line_diff + first_line, pixel_diff + last_pixel], ndmin=2)

                    self.coordinates[date][swath]['corners'][b,:,:] = np.vstack([ll,ul,ur,lr])
                    b += 1

                    # Check for max/min line/pixel to prevent negative pixel numbers.
                    min_line = min(1 + line_diff, min_line)
                    max_line = max(line_diff + no_lines, max_line)
                    min_pixel = min(1 + pixel_diff, min_pixel)
                    max_pixel = max(pixel_diff + no_pixels, max_pixel)

            if min_line < 1:
                max_line = max_line + (1 - min_line)
            if min_pixel < 1:
                max_pixel = max_pixel + (1 - min_pixel)

            for swath in self.datastack[date].keys():
                # If one of the lines or pixels is lower then 1, correct all coordinates to have positive coordinates for all
                # bursts.
                if min_line < 1:
                    self.coordinates[date][swath]['corners'][:,:,0] = self.coordinates[date][swath]['corners'][:,:,0] + (1 - min_line)
                if min_pixel < 1:
                    self.coordinates[date][swath]['corners'][:,:,1] = self.coordinates[date][swath]['corners'][:,:,1] + (1 - min_pixel)

                for burst in range(len(self.datastack[date][swath].keys())):
                    shape_c = self.coordinates[date][swath]['corners'][burst,:,:]
                    shape = box(shape_c[0,1],shape_c[0,0],shape_c[2,1],shape_c[2,0])
                    self.coordinates[date]['shapes'].append(shape)

                    # Finally add information to the .res file if already loaded
                    if self.datastack[date][swath]['burst_' + str(burst+1)].processes['readfiles']:

                        read = self.datastack[date][swath]['burst_' + str(burst+1)].processes['readfiles']
                        read['First_line (w.r.t. output_image)'] = str(shape_c[0,0])
                        read['Last_line (w.r.t. output_image)'] = str(shape_c[2,0])
                        read['First_pixel (w.r.t. output_image)'] = str(shape_c[0,1])
                        read['Last_pixel (w.r.t. output_image)'] = str(shape_c[2,1])
                        read['Number_of_pixels_output_image'] = str(max_pixel)
                        read['Number_of_lines_output_image'] = str(max_line)
                        self.datastack[date][swath]['burst_' + str(burst+1)].processes['readfiles'] = read
                    else:
                        print('No resfile available, so information is not added to resfile')

    def write_stack(self,write_path='',no_data=False):
        # This function writes the full datastack to a given folder using the dates / swaths / bursts setup. This
        # also generates the res readfiles data.
        if write_path and os.path.exists(write_path):
            self.path = write_path
        if (not write_path or not os.path.exists(write_path)) and not self.path:
            warnings.warn('Please specify a path that exists to write the data')
            return

        write_jobs = []
        burst_num = []

        for date in self.datastack.keys():

            date_basic = date.translate(None,'-')
            date_path = os.path.join(self.path, date_basic)
            if not os.path.exists(date_path):
                os.mkdir(date_path)

            for swath in self.datastack[date].keys():

                swath_path = os.path.join(date_path, swath)
                if not os.path.exists(swath_path):
                    os.mkdir(swath_path)

                for burst in self.datastack[date][swath].keys():
                    burst_path = os.path.join(date_path, swath, burst)
                    if not os.path.exists(burst_path):
                        os.mkdir(burst_path)

                    # Finally write the bursts with their res files and precise orbits
                    xml = self.datastack[date][swath][burst].swath_xml
                    data = self.datastack[date][swath][burst].swath_data
                    image_no = str(self.datastack[date][swath][burst].burst_num)
                    stack_no = burst[6:]
                    xml_base = os.path.basename(xml)
                    res_name = os.path.join(burst_path, 'slave.res')
                    outdata = os.path.join(burst_path, 'slave_iw_' + xml_base[6] + '_burst_' + stack_no + '.raw')

                    self.datastack[date][swath][burst].write(res_name)
                    if not os.path.exists(res_name) or not os.path.exists(outdata):

                        write_jobs.append(sys.executable + ' ' + self.function_path + 'sentinel_dump_data_function.py ' + data + ' ' + res_name + ' ' + outdata)
                        burst_num.append(stack_no + '_' + xml_base[6] + '_' + xml_base[15:23])

        # Burst are sorted in such a way that mainly read from different data files sorted by burst then swath then date.
        ids = sorted(range(len(burst_num)), key=lambda x: burst_num[x])

        jobList1 = []
        for id_val in ids:
            jobList1.append({"path": self.path, "command": write_jobs[id_val]})
            if not self.parallel:
                os.chdir(self.path)
                # Resample
                os.system(write_jobs[id_val])
        if self.parallel:
            jobs = Jobs(self.nr_of_jobs, self.doris_parameters)
            jobs.run(jobList1)

    def unpack_image(self, dest_folder=''):

        if not dest_folder:
            dest_folder = os.path.join(self.path, 'slc_data_files')
            self.unzip_path = dest_folder
        if not os.path.exists(dest_folder):
            os.mkdir(dest_folder)

        jobList1 = []

        # This program unpacks the images which are needed for processing. If unpacking fails, they are removed..
        for imagefile in self.images:

            zipped_folder = imagefile.zip_path
            if zipped_folder.endswith('.SAFE.zip'):
                imagefile.unzip_path = os.path.join(dest_folder, os.path.basename(zipped_folder[:-9] + '.SAFE'))
            elif zipped_folder.endswith('.zip'):
                imagefile.unzip_path = os.path.join(dest_folder, os.path.basename(zipped_folder[:-4] + '.SAFE'))
            shapefile = self.shape_filename
            pol = self.polarisation[0]
            overwrite = False
            command1 = (sys.executable + ' ' + self.function_path + 'load_shape_unzip.py ' + zipped_folder + ' ' + dest_folder +
                        ' ' + shapefile + ' ' + pol + ' ' + str(overwrite))
            jobList1.append({"path": self.path, "command": command1})
            if not self.parallel:
                os.chdir(self.path)
                # Resample
                os.system(command1)

        if self.parallel:
            jobs = Jobs(self.nr_of_jobs, self.doris_parameters)
            jobs.run(jobList1)

    def del_unpacked_image(self):
        # This program unpacks the images which are needed for processing.
        for image_dat in self.images:
            shutil.rmtree(image_dat.unzip_path)

    def swath_coverage(self):
        # Create a convex hull for the different swaths.

        for swath in self.swath_names:
            # Assign coverage of swath using convex hull.
            burst_id = [i for i in range(len(self.burst_names)) if swath in self.burst_names[i]]
            swath_shape = cascaded_union([self.burst_shapes[i] for i in burst_id])
            self.swath_shapes.append(swath_shape)

    def write_shapes(self,coverage=True,images=True,swaths=True,bursts=True):
        # This function writes shapefiles of the area of interest, images and bursts.

        if not coverage and not images and not bursts:
            warnings.warn('Select at least one of shape types')
            return

        shapes = list(); shape_names = list(); shape_files = list()
        if coverage and self.shape:
            shape_files.append(self.path + 'area_of_interest.shp')
            if self.shape.type == 'MultiPolygon':
                shapes.append([sh for sh in self.shape])
                shape_names.append([('coverage_with_buffer_of_' + str(self.buffer) + '_degrees_' + str(i)) for i in range(len(shapes[0]))])
            elif self.shape.type == 'Polygon':
                shape_names.append(['coverage_with_buffer_of_' + str(self.buffer) + '_degrees'])
                shapes.append([self.shape])
        if images and self.image_shapes:
            shape_files.append(self.path + 'image_coverage.shp')
            shape_names.append([date.astype(datetime).strftime('%Y-%m-%dT%H:%M:%S') for date in self.image_dates])
            shapes.append(self.image_shapes)
        if swaths and self.swath_shapes:
            shape_files.append(self.path + 'swath_coverage.shp')
            shape_names.append(self.swath_names)
            shapes.append(self.swath_shapes)
        if bursts and self.burst_shapes:
            shape_files.append(self.path + 'burst_coverage.shp')
            shape_names.append(self.burst_names)
            shapes.append(self.burst_shapes)

        shape_setup = {
            'geometry': 'Polygon',
            'properties': {'name': 'str'},
        }

        for i in range(len(shape_files)):
            with fiona.open(shape_files[i], 'w', 'ESRI Shapefile', shape_setup) as sh:
                for n in range(len(shapes[i])):
                    sh.write({
                        'geometry': mapping(shapes[i][n]),
                        'properties': {'name': shape_names[i][n]},
                    })