source: sasview/src/sas/sascalc/dataloader/file_reader_base_class.py @ b6f9c9f

"""
This is the base file reader class most file readers should inherit from.
All generic functionality required for a file loader/reader is built into this
class
"""

import os
import sys
import math
import logging
from abc import abstractmethod

import numpy as np
from .loader_exceptions import NoKnownLoaderException, FileContentsException,\
    DataReaderException, DefaultReaderException
from .data_info import Data1D, Data2D, DataInfo, plottable_1D, plottable_2D,\
    combine_data_info_with_plottable

logger = logging.getLogger(__name__)

if sys.version_info[0] < 3:
    def decode(s):
        return s
else:
    def decode(s):
        return s.decode() if isinstance(s, bytes) else s

# Data 1D fields for iterative purposes
FIELDS_1D = ('x', 'y', 'dx', 'dy', 'dxl', 'dxw')
# Data 2D fields for iterative purposes
FIELDS_2D = ('data', 'qx_data', 'qy_data', 'q_data', 'err_data',
                 'dqx_data', 'dqy_data', 'mask')


class FileReader(object):
    # String to describe the type of data this reader can load
    type_name = "ASCII"
    # Wildcards to display
    type = ["Text files (*.txt|*.TXT)"]
    # List of allowed extensions
    ext = ['.txt']
    # Bypass extension check and try to load anyway
    allow_all = False
    # Able to import the unit converter
    has_converter = True
    # Default value of zero
    _ZERO = 1e-16

    def __init__(self):
        # List of Data1D and Data2D objects to be sent back to data_loader
        self.output = []
        # Current plottable_(1D/2D) object being loaded in
        self.current_dataset = None
        # Current DataInfo object being loaded in
        self.current_datainfo = None
        # File path sent to reader
        self.filepath = None
        # Open file handle
        self.f_open = None

    def read(self, filepath):
        """
        Basic file reader

        :param filepath: The full or relative path to a file to be loaded
        """
        self.filepath = filepath
        if os.path.isfile(filepath):
            basename, extension = os.path.splitext(os.path.basename(filepath))
            self.extension = extension.lower()
            # If the file type is not allowed, return nothing
            if self.extension in self.ext or self.allow_all:
                # Try to load the file, but raise an error if unable to.
                try:
                    self.f_open = open(filepath, 'rb')
                    self.get_file_contents()

                except DataReaderException as e:
                    self.handle_error_message(e.message)
                except OSError as e:
                    # If the file cannot be opened
                    msg = "Unable to open file: {}\n".format(filepath)
                    msg += e.message
                    self.handle_error_message(msg)
                finally:
                    # Close the file handle if it is open
                    if not self.f_open.closed:
                        self.f_open.close()
                    if len(self.output) > 0:
                        # Sort the data that's been loaded
                        self.sort_one_d_data()
                        self.sort_two_d_data()
        else:
            msg = "Unable to find file at: {}\n".format(filepath)
            msg += "Please check your file path and try again."
            self.handle_error_message(msg)

        # Return a list of parsed entries that data_loader can manage
        final_data = self.output
        self.reset_state()
        return final_data

    def reset_state(self):
        """
        Resets the class state to a base case when loading a new data file so previous
        data files do not appear a second time
        """
        self.current_datainfo = None
        self.current_dataset = None
        self.filepath = None
        self.ind = None
        self.output = []

    def nextline(self):
        """
        Returns the next line in the file as a string.
        """
        #return self.f_open.readline()
        return decode(self.f_open.readline())

    def nextlines(self):
        """
        Returns the next line in the file as a string.
        """
        for line in self.f_open:
            #yield line
            yield decode(line)

    def readall(self):
        """
        Returns the entire file as a string.
        """
        #return self.f_open.read()
        return decode(self.f_open.read())

    def handle_error_message(self, msg):
        """
        Generic error handler to add an error to the current datainfo to
        propagate the error up the error chain.
        :param msg: Error message
        """
        if len(self.output) > 0:
            self.output[-1].errors.append(msg)
        elif isinstance(self.current_datainfo, DataInfo):
            self.current_datainfo.errors.append(msg)
        else:
            logger.warning(msg)

    def send_to_output(self):
        """
        Helper that automatically combines the info and set and then appends it
        to output
        """
        data_obj = combine_data_info_with_plottable(self.current_dataset,
                                                    self.current_datainfo)
        self.output.append(data_obj)

    def sort_one_d_data(self):
        """
        Sort 1D data along the X axis for consistency
        """
        for data in self.output:
            if isinstance(data, Data1D):
                # Normalize the units for
                data.x_unit = self.format_unit(data.x_unit)
                data.y_unit = self.format_unit(data.y_unit)
                # Sort data by increasing x and remove 1st point
                ind = np.lexsort((data.y, data.x))
                data.x = self._reorder_1d_array(data.x, ind)
                data.y = self._reorder_1d_array(data.y, ind)
                if data.dx is not None:
                    if len(data.dx) == 0:
                        data.dx = None
                        continue
                    data.dx = self._reorder_1d_array(data.dx, ind)
                if data.dxl is not None:
                    data.dxl = self._reorder_1d_array(data.dxl, ind)
                if data.dxw is not None:
                    data.dxw = self._reorder_1d_array(data.dxw, ind)
                if data.dy is not None:
                    if len(data.dy) == 0:
                        data.dy = None
                        continue
                    data.dy = self._reorder_1d_array(data.dy, ind)
                if data.lam is not None:
                    data.lam = self._reorder_1d_array(data.lam, ind)
                if data.dlam is not None:
                    data.dlam = self._reorder_1d_array(data.dlam, ind)
                data = self._remove_nans_in_data(data)
                if len(data.x) > 0:
                    data.xmin = np.min(data.x)
                    data.xmax = np.max(data.x)
                    data.ymin = np.min(data.y)
                    data.ymax = np.max(data.y)

    @staticmethod
    def _reorder_1d_array(array, ind):
        """
        Reorders a 1D array based on the indices passed as ind
        :param array: Array to be reordered
        :param ind: Indices used to reorder array
        :return: reordered array
        """
        array = np.asarray(array, dtype=np.float64)
        return array[ind]

    @staticmethod
    def _remove_nans_in_data(data):
        """
        Remove data points where nan is loaded
        :param data: 1D or 2D data object
        :return: data with nan points removed
        """
        if isinstance(data, Data1D):
            fields = FIELDS_1D
        elif isinstance(data, Data2D):
            fields = FIELDS_2D
        else:
            return data
        # Make array of good points - all others will be removed
        good = np.isfinite(getattr(data, fields[0]))
        for name in fields[1:]:
            array = getattr(data, name)
            if array is not None:
                # Update good points only if not already changed
                good &= np.isfinite(array)
        if not np.all(good):
            for name in fields:
                array = getattr(data, name)
                if array is not None:
                    setattr(data, name, array[good])
        return data

    def sort_two_d_data(self):
        for dataset in self.output:
            if isinstance(dataset, Data2D):
                # Normalize the units for
                dataset.x_unit = self.format_unit(dataset.Q_unit)
                dataset.y_unit = self.format_unit(dataset.I_unit)
                dataset.data = dataset.data.astype(np.float64)
                dataset.qx_data = dataset.qx_data.astype(np.float64)
                dataset.xmin = np.min(dataset.qx_data)
                dataset.xmax = np.max(dataset.qx_data)
                dataset.qy_data = dataset.qy_data.astype(np.float64)
                dataset.ymin = np.min(dataset.qy_data)
                dataset.ymax = np.max(dataset.qy_data)
                dataset.q_data = np.sqrt(dataset.qx_data * dataset.qx_data
                                         + dataset.qy_data * dataset.qy_data)
                if dataset.err_data is not None:
                    dataset.err_data = dataset.err_data.astype(np.float64)
                if dataset.dqx_data is not None:
                    dataset.dqx_data = dataset.dqx_data.astype(np.float64)
                if dataset.dqy_data is not None:
                    dataset.dqy_data = dataset.dqy_data.astype(np.float64)
                if dataset.mask is not None:
                    dataset.mask = dataset.mask.astype(dtype=bool)

                if len(dataset.data.shape) == 2:
                    n_rows, n_cols = dataset.data.shape
                    dataset.y_bins = dataset.qy_data[0::int(n_cols)]
                    dataset.x_bins = dataset.qx_data[:int(n_cols)]
                dataset.data = dataset.data.flatten()
                dataset = self._remove_nans_in_data(dataset)
                if len(dataset.data) > 0:
                    dataset.xmin = np.min(dataset.qx_data)
                    dataset.xmax = np.max(dataset.qx_data)
                    dataset.ymin = np.min(dataset.qy_data)
                    dataset.ymax = np.max(dataset.qx_data)

    def format_unit(self, unit=None):
        """
        Format units a common way
        :param unit:
        :return:
        """
        if unit:
            split = unit.split("/")
            if len(split) == 1:
                return unit
            elif split[0] == '1':
                return "{0}^".format(split[1]) + "{-1}"
            else:
                return "{0}*{1}^".format(split[0], split[1]) + "{-1}"

    def set_all_to_none(self):
        """
        Set all mutable values to None for error handling purposes
        """
        self.current_dataset = None
        self.current_datainfo = None
        self.output = []

    def data_cleanup(self):
        """
        Clean up the data sets and refresh everything
        :return: None
        """
        self.remove_empty_q_values()
        self.send_to_output()  # Combine datasets with DataInfo
        self.current_datainfo = DataInfo()  # Reset DataInfo

    def remove_empty_q_values(self):
        """
        Remove any point where Q == 0
        """
        if isinstance(self.current_dataset, plottable_1D):
            # Booleans for resolutions
            has_error_dx = self.current_dataset.dx is not None
            has_error_dxl = self.current_dataset.dxl is not None
            has_error_dxw = self.current_dataset.dxw is not None
            has_error_dy = self.current_dataset.dy is not None
            # Create arrays of zeros for non-existent resolutions
            if has_error_dxw and not has_error_dxl:
                array_size = self.current_dataset.dxw.size - 1
                self.current_dataset.dxl = np.append(self.current_dataset.dxl,
                                                    np.zeros([array_size]))
                has_error_dxl = True
            elif has_error_dxl and not has_error_dxw:
                array_size = self.current_dataset.dxl.size - 1
                self.current_dataset.dxw = np.append(self.current_dataset.dxw,
                                                    np.zeros([array_size]))
                has_error_dxw = True
            elif not has_error_dxl and not has_error_dxw and not has_error_dx:
                array_size = self.current_dataset.x.size - 1
                self.current_dataset.dx = np.append(self.current_dataset.dx,
                                                    np.zeros([array_size]))
                has_error_dx = True
            if not has_error_dy:
                array_size = self.current_dataset.y.size - 1
                self.current_dataset.dy = np.append(self.current_dataset.dy,
                                                    np.zeros([array_size]))
                has_error_dy = True

            # Remove points where q = 0
            x = self.current_dataset.x
            self.current_dataset.x = self.current_dataset.x[x != 0]
            self.current_dataset.y = self.current_dataset.y[x != 0]
            if has_error_dy:
                self.current_dataset.dy = self.current_dataset.dy[x != 0]
            if has_error_dx:
                self.current_dataset.dx = self.current_dataset.dx[x != 0]
            if has_error_dxl:
                self.current_dataset.dxl = self.current_dataset.dxl[x != 0]
            if has_error_dxw:
                self.current_dataset.dxw = self.current_dataset.dxw[x != 0]
        elif isinstance(self.current_dataset, plottable_2D):
            has_error_dqx = self.current_dataset.dqx_data is not None
            has_error_dqy = self.current_dataset.dqy_data is not None
            has_error_dy = self.current_dataset.err_data is not None
            has_mask = self.current_dataset.mask is not None
            x = self.current_dataset.qx_data
            self.current_dataset.data = self.current_dataset.data[x != 0]
            self.current_dataset.qx_data = self.current_dataset.qx_data[x != 0]
            self.current_dataset.qy_data = self.current_dataset.qy_data[x != 0]
            self.current_dataset.q_data = np.sqrt(
                np.square(self.current_dataset.qx_data) + np.square(
                    self.current_dataset.qy_data))
            if has_error_dy:
                self.current_dataset.err_data = self.current_dataset.err_data[x != 0]
            if has_error_dqx:
                self.current_dataset.dqx_data = self.current_dataset.dqx_data[x != 0]
            if has_error_dqy:
                self.current_dataset.dqy_data = self.current_dataset.dqy_data[x != 0]
            if has_mask:
                self.current_dataset.mask = self.current_dataset.mask[x != 0]

    def reset_data_list(self, no_lines=0):
        """
        Reset the plottable_1D object
        """
        # Initialize data sets with arrays the maximum possible size
        x = np.zeros(no_lines)
        y = np.zeros(no_lines)
        dx = np.zeros(no_lines)
        dy = np.zeros(no_lines)
        self.current_dataset = plottable_1D(x, y, dx, dy)

    @staticmethod
    def splitline(line):
        """
        Splits a line into pieces based on common delimiters
        :param line: A single line of text
        :return: list of values
        """
        # Initial try for CSV (split on ,)
        toks = line.split(',')
        # Now try SCSV (split on ;)
        if len(toks) < 2:
            toks = line.split(';')
        # Now go for whitespace
        if len(toks) < 2:
            toks = line.split()
        return toks

    @abstractmethod
    def get_file_contents(self):
        """
        Reader specific class to access the contents of the file
        All reader classes that inherit from FileReader must implement
        """
        pass