python-ccdproc 1.3.0-2 / usr / lib / python2.7 / dist-packages / ccdproc / ccddata.py

# Licensed under a 3-clause BSD style license - see LICENSE.rst

"""This module implements the base CCDData class."""

from __future__ import (absolute_import, division, print_function,
                        unicode_literals)

import copy
import numbers
import weakref
from collections import OrderedDict

import numpy as np

from astropy.nddata import (NDDataArray, StdDevUncertainty, NDUncertainty,
                            MissingDataAssociationException)
from astropy.io import fits, registry
from astropy import units as u
from astropy import log
from astropy.utils import minversion
from astropy.wcs import WCS

_ASTROPY_LT_1_2 = not minversion("astropy", "1.2")
_ASTROPY_LT_1_3 = not minversion("astropy", "1.3")
_ASTROPY_GT_2_0 = minversion("astropy", "2.0")

# FIXME: Remove the content of the following "if" as soon as astropy 1.1 isn't
# supported anymore. This is just a temporary workaround to fix the memory leak
# described in https://github.com/astropy/astropy/issues/4825
if _ASTROPY_LT_1_2:

    class ParentNDDataDescriptor(object):
        def __get__(self, obj, objtype=None):
            message = "uncertainty is not associated with an NDData object."
            try:
                if obj._parent_nddata is None:
                    raise MissingDataAssociationException(message)
                else:
                    # The NDData is saved as weak reference so we must call it
                    # to get the object the reference points to.
                    if isinstance(obj._parent_nddata, weakref.ref):
                        return obj._parent_nddata()
                    log.info("parent_nddata should be a weakref to an "
                             "NDData object.")
                    return obj._parent_nddata
            except AttributeError:
                raise MissingDataAssociationException(message)

        def __set__(self, obj, value):
            if value is not None and not isinstance(value, weakref.ref):
                # Save a weak reference on the uncertainty that points to this
                # instance of NDData. Direct references should NOT be used:
                # https://github.com/astropy/astropy/pull/4799#discussion_r61236832
                value = weakref.ref(value)
            obj._parent_nddata = value

    # Use the descriptor as parent_nddata property. This only affects
    # instances created after importing this module.
    StdDevUncertainty.parent_nddata = ParentNDDataDescriptor()


__all__ = ['CCDData', 'fits_ccddata_reader', 'fits_ccddata_writer']


# Global value which can turn on/off the unit requirements when creating a
# CCDData. Should be used with care because several functions actually break
# if the unit is None!
_config_ccd_requires_unit = True

if not _ASTROPY_LT_1_2:
    from astropy.utils.decorators import sharedmethod

    def _arithmetic(op):
        """Decorator factory which temporarly disables the need for a unit when
        creating a new CCDData instance. The final result must have a unit.

        Parameters
        ----------
        op : function
            The function to apply. Supported are:

            - ``np.add``
            - ``np.subtract``
            - ``np.multiply``
            - ``np.true_divide``

        Notes
        -----
        Should only be used on CCDData ``add``, ``subtract``, ``divide`` or
        ``multiply`` because only these methods from NDArithmeticMixin are
        overwritten.
        """
        def decorator(func):
            def inner(self, operand, operand2=None, **kwargs):
                global _config_ccd_requires_unit
                _config_ccd_requires_unit = False
                result = self._prepare_then_do_arithmetic(op, operand,
                                                          operand2, **kwargs)
                # Wrap it again as CCDData so it checks the final unit.
                _config_ccd_requires_unit = True
                return result.__class__(result)
            inner.__doc__ = ("See `astropy.nddata.NDArithmeticMixin.{}`."
                             "".format(func.__name__))
            return sharedmethod(inner)
        return decorator


class CCDData(NDDataArray):
    """A class describing basic CCD data.

    The CCDData class is based on the NDData object and includes a data array,
    uncertainty frame, mask frame, meta data, units, and WCS information for a
    single CCD image.

    Parameters
    -----------
    data : `~ccdproc.CCDData`-like or `numpy.ndarray`-like
        The actual data contained in this `~ccdproc.CCDData` object.
        Note that the data will always be saved by *reference*, so you should
        make a copy of the ``data`` before passing it in if that's the desired
        behavior.

    uncertainty : `~astropy.nddata.StdDevUncertainty`, `numpy.ndarray` or \
            None, optional
        Uncertainties on the data.
        Default is ``None``.

    mask : `numpy.ndarray` or None, optional
        Mask for the data, given as a boolean Numpy array with a shape
        matching that of the data. The values must be `False` where
        the data is *valid* and `True` when it is not (like Numpy
        masked arrays). If ``data`` is a numpy masked array, providing
        ``mask`` here will causes the mask from the masked array to be
        ignored.
        Default is ``None``.

    flags : `numpy.ndarray` or `~astropy.nddata.FlagCollection` or None, \
            optional
        Flags giving information about each pixel. These can be specified
        either as a Numpy array of any type with a shape matching that of the
        data, or as a `~astropy.nddata.FlagCollection` instance which has a
        shape matching that of the data.
        Default is ``None``.

    wcs : `~astropy.wcs.WCS` or None, optional
        WCS-object containing the world coordinate system for the data.
        Default is ``None``.

    meta : dict-like object or None, optional
        Metadata for this object. "Metadata" here means all information that
        is included with this object but not part of any other attribute
        of this particular object, e.g. creation date, unique identifier,
        simulation parameters, exposure time, telescope name, etc.

    unit : `~astropy.units.Unit` or str, optional
        The units of the data.
        Default is ``None``.

        .. warning::

            If the unit is ``None`` or not otherwise specified it will raise a
            ``ValueError``

    Raises
    ------
    ValueError
        If the ``uncertainty`` or ``mask`` inputs cannot be broadcast (e.g.,
        match shape) onto ``data``.

    Methods
    -------
    read(\\*args, \\**kwargs)
        ``Classmethod`` to create an CCDData instance based on a ``FITS`` file.
        This method uses :func:`fits_ccddata_reader` with the provided
        parameters.
    write(\\*args, \\**kwargs)
        Writes the contents of the CCDData instance into a new ``FITS`` file.
        This method uses :func:`fits_ccddata_writer` with the provided
        parameters.

    Notes
    -----
    `~ccdproc.CCDData` objects can be easily converted to a regular
     Numpy array using `numpy.asarray`.

    For example::

        >>> from ccdproc import CCDData
        >>> import numpy as np
        >>> x = CCDData([1,2,3], unit='adu')
        >>> np.asarray(x)
        array([1, 2, 3])

    This is useful, for example, when plotting a 2D image using
    matplotlib.

        >>> from ccdproc import CCDData
        >>> from matplotlib import pyplot as plt   # doctest: +SKIP
        >>> x = CCDData([[1,2,3], [4,5,6]], unit='adu')
        >>> plt.imshow(x)   # doctest: +SKIP

    """
    def __init__(self, *args, **kwd):
        if 'meta' not in kwd:
            kwd['meta'] = kwd.pop('header', None)
        if 'header' in kwd:
            raise ValueError("can't have both header and meta.")

        super(CCDData, self).__init__(*args, **kwd)

        # Check if a unit is set. This can be temporarly disabled by the
        # _CCDDataUnit contextmanager.
        if _config_ccd_requires_unit and self.unit is None:
            raise ValueError("a unit for CCDData must be specified.")

    @property
    def data(self):
        return self._data

    @data.setter
    def data(self, value):
        self._data = value

    @property
    def wcs(self):
        return self._wcs

    @wcs.setter
    def wcs(self, value):
        self._wcs = value

    @property
    def unit(self):
        return self._unit

    @unit.setter
    def unit(self, value):
        self._unit = u.Unit(value)

    @property
    def header(self):
        return self._meta

    @header.setter
    def header(self, value):
        self.meta = value

    @property
    def meta(self):
        return self._meta

    @meta.setter
    def meta(self, value):
        if value is None:
            self._meta = OrderedDict()
        else:
            if hasattr(value, 'keys'):
                self._meta = value
            else:
                raise TypeError(
                    'the meta attribute of CCDData must be dict-like.')

    @property
    def uncertainty(self):
        return self._uncertainty

    @uncertainty.setter
    def uncertainty(self, value):
        if value is not None:
            if isinstance(value, NDUncertainty):
                if getattr(value, '_parent_nddata', None) is not None:
                    value = value.__class__(value, copy=False)
                self._uncertainty = value
            elif isinstance(value, np.ndarray):
                if value.shape != self.shape:
                    raise ValueError("uncertainty must have same shape as "
                                     "data.")
                self._uncertainty = StdDevUncertainty(value)
                log.info("array provided for uncertainty; assuming it is a "
                         "StdDevUncertainty.")
            else:
                raise TypeError("uncertainty must be an instance of a "
                                "NDUncertainty object or a numpy array.")
            self._uncertainty.parent_nddata = self
        else:
            self._uncertainty = value

    def to_hdu(self, hdu_mask='MASK', hdu_uncertainty='UNCERT',
               hdu_flags=None, wcs_relax=True):
        """Creates an HDUList object from a CCDData object.

        Parameters
        ----------
        hdu_mask, hdu_uncertainty, hdu_flags : str or None, optional
            If it is a string append this attribute to the HDUList as
            `~astropy.io.fits.ImageHDU` with the string as extension name.
            Flags are not supported at this time. If ``None`` this attribute
            is not appended.
            Default is ``'MASK'`` for mask, ``'UNCERT'`` for uncertainty and
            ``None`` for flags.

        wcs_relax : bool
            Value of the ``relax`` parameter to use in converting the WCS to a
            FITS header using `~astropy.wcs.WCS.to_header`. The common
            ``CTYPE`` ``RA---TAN-SIP`` and ``DEC--TAN-SIP`` requires
            ``relax=True`` for the ``-SIP`` part of the ``CTYPE`` to be
            preserved.

        Raises
        -------
        ValueError
            - If ``self.mask`` is set but not a `numpy.ndarray`.
            - If ``self.uncertainty`` is set but not a
              `~astropy.nddata.StdDevUncertainty`.
            - If ``self.uncertainty`` is set but has another unit then
              ``self.data``.

        NotImplementedError
            Saving flags is not supported.

        Returns
        -------
        hdulist : `~astropy.io.fits.HDUList`
        """
        if isinstance(self.header, fits.Header):
            # Copy here so that we can modify the HDU header by adding WCS
            # information without changing the header of the CCDData object.
            header = self.header.copy()
        else:
            # Because _insert_in_metadata_fits_safe is written as a method
            # we need to create a dummy CCDData instance to hold the FITS
            # header we are constructing. This probably indicates that
            # _insert_in_metadata_fits_safe should be rewritten in a more
            # sensible way...
            dummy_ccd = CCDData([1], meta=fits.Header(), unit="adu")
            for k, v in self.header.items():
                dummy_ccd._insert_in_metadata_fits_safe(k, v)
            header = dummy_ccd.header
        if self.unit is not u.dimensionless_unscaled:
            header['bunit'] = self.unit.to_string()
        if self.wcs:
            # Simply extending the FITS header with the WCS can lead to
            # duplicates of the WCS keywords; iterating over the WCS
            # header should be safer.
            #
            # Turns out if I had read the io.fits.Header.extend docs more
            # carefully, I would have realized that the keywords exist to
            # avoid duplicates and preserve, as much as possible, the
            # structure of the commentary cards.
            #
            # Note that until astropy/astropy#3967 is closed, the extend
            # will fail if there are comment cards in the WCS header but
            # not header.
            wcs_header = self.wcs.to_header(relax=wcs_relax)
            header.extend(wcs_header, useblanks=False, update=True)
        hdus = [fits.PrimaryHDU(self.data, header)]

        if hdu_mask and self.mask is not None:
            # Always assuming that the mask is a np.ndarray (check that it has
            # a 'shape').
            if not hasattr(self.mask, 'shape'):
                raise ValueError('only a numpy.ndarray mask can be saved.')

            # Convert boolean mask to uint since io.fits cannot handle bool.
            hduMask = fits.ImageHDU(self.mask.astype(np.uint8), name=hdu_mask)
            hdus.append(hduMask)

        if hdu_uncertainty and self.uncertainty is not None:
            # We need to save some kind of information which uncertainty was
            # used so that loading the HDUList can infer the uncertainty type.
            # No idea how this can be done so only allow StdDevUncertainty.
            if self.uncertainty.__class__.__name__ != 'StdDevUncertainty':
                raise ValueError('only StdDevUncertainty can be saved.')

            # Assuming uncertainty is an StdDevUncertainty save just the array
            # this might be problematic if the Uncertainty has a unit differing
            # from the data so abort for different units. This is important for
            # astropy > 1.2
            if (hasattr(self.uncertainty, 'unit') and
                    self.uncertainty.unit is not None and
                    self.uncertainty.unit != self.unit):
                raise ValueError('saving uncertainties with a unit differing'
                                 'from the data unit is not supported.')

            hduUncert = fits.ImageHDU(self.uncertainty.array,
                                      name=hdu_uncertainty)
            hdus.append(hduUncert)

        if hdu_flags and self.flags:
            raise NotImplementedError('adding the flags to a HDU is not '
                                      'supported at this time.')

        hdulist = fits.HDUList(hdus)

        return hdulist

    def copy(self):
        """
        Return a copy of the CCDData object.
        """
        try:
            return self.__class__(self, copy=True)
        except TypeError:
            new = self.__class__(copy.deepcopy(self))
        return new

    def _ccddata_arithmetic(self, other, operation, scale_uncertainty=False):
        """
        Perform the common parts of arithmetic operations on CCDData objects.

        This should only be called when ``other`` is a Quantity or a number.
        """
        # THE "1 *" IS NECESSARY to get the right result, at least in
        # astropy-0.4dev. Using the np.multiply, etc, methods with a Unit
        # and a Quantity is currently broken, but it works with two Quantity
        # arguments.
        if isinstance(other, u.Quantity):
            if (operation.__name__ in ['add', 'subtract'] and
                    self.unit != other.unit):
                # For addition and subtraction we need to convert the unit
                # to the same unit otherwise operating on the values alone will
                # give wrong results (#291)
                other_value = other.to(self.unit).value
            else:
                other_value = other.value
        elif isinstance(other, numbers.Number):
            other_value = other
        else:
            raise TypeError("cannot do arithmetic with type '{0}' "
                            "and 'CCDData'".format(type(other)))

        result_unit = operation(1 * self.unit, other).unit
        result_data = operation(self.data, other_value)

        if self.uncertainty:
            result_uncertainty = self.uncertainty.array
            if scale_uncertainty:
                result_uncertainty = operation(result_uncertainty, other_value)
            result_uncertainty = StdDevUncertainty(result_uncertainty)
        else:
            result_uncertainty = None

        new_mask = copy.deepcopy(self.mask)
        new_meta = copy.deepcopy(self.meta)
        new_wcs = copy.deepcopy(self.wcs)
        result = CCDData(result_data, unit=result_unit, mask=new_mask,
                         uncertainty=result_uncertainty,
                         meta=new_meta, wcs=new_wcs)
        return result

    def multiply(self, other, compare_wcs='first_found'):
        if isinstance(other, CCDData):
            if compare_wcs is None or compare_wcs == 'first_found':
                tmp_wcs_1, tmp_wcs_2 = self.wcs, other.wcs
                self.wcs, other.wcs = None, None

                # Determine the WCS of the result
                if compare_wcs is None:
                    result_wcs = None
                else:
                    result_wcs = tmp_wcs_1 if tmp_wcs_1 else tmp_wcs_2

                result = super(CCDData, self).multiply(other)
                result.wcs = result_wcs
                self.wcs, other.wcs = tmp_wcs_1, tmp_wcs_2
                return result
            else:
                if hasattr(self, '_arithmetics_wcs'):
                    return super(CCDData, self).multiply(
                        other, compare_wcs=compare_wcs)
                else:
                    raise ImportError("wcs_compare functionality requires "
                                      "astropy 1.2 or greater.")

        return self._ccddata_arithmetic(other, np.multiply,
                                        scale_uncertainty=True)

    def divide(self, other, compare_wcs='first_found'):
        if isinstance(other, CCDData):
            if compare_wcs is None or compare_wcs == 'first_found':
                tmp_wcs_1, tmp_wcs_2 = self.wcs, other.wcs
                self.wcs, other.wcs = None, None

                # Determine the WCS of the result
                if compare_wcs is None:
                    result_wcs = None
                else:
                    result_wcs = tmp_wcs_1 if tmp_wcs_1 else tmp_wcs_2

                result = super(CCDData, self).divide(other)
                result.wcs = result_wcs
                self.wcs, other.wcs = tmp_wcs_1, tmp_wcs_2
                return result
            else:
                if hasattr(self, '_arithmetics_wcs'):
                    return super(CCDData, self).divide(
                        other, compare_wcs=compare_wcs)
                else:
                    raise ImportError("wcs_compare functionality requires "
                                      "astropy 1.2 or greater.")

        return self._ccddata_arithmetic(other, np.divide,
                                        scale_uncertainty=True)

    def add(self, other, compare_wcs='first_found'):
        if isinstance(other, CCDData):
            if compare_wcs is None or compare_wcs == 'first_found':
                tmp_wcs_1, tmp_wcs_2 = self.wcs, other.wcs
                self.wcs, other.wcs = None, None

                # Determine the WCS of the result
                if compare_wcs is None:
                    result_wcs = None
                else:
                    result_wcs = tmp_wcs_1 if tmp_wcs_1 else tmp_wcs_2

                result = super(CCDData, self).add(other)
                result.wcs = result_wcs
                self.wcs, other.wcs = tmp_wcs_1, tmp_wcs_2
                return result
            else:
                if hasattr(self, '_arithmetics_wcs'):
                    return super(CCDData, self).add(
                        other, compare_wcs=compare_wcs)
                else:
                    raise ImportError("wcs_compare functionality requires "
                                      "astropy 1.2 or greater.")

        return self._ccddata_arithmetic(other, np.add,
                                        scale_uncertainty=False)

    def subtract(self, other, compare_wcs='first_found'):
        if isinstance(other, CCDData):
            if compare_wcs is None or compare_wcs == 'first_found':
                tmp_wcs_1, tmp_wcs_2 = self.wcs, other.wcs
                self.wcs, other.wcs = None, None

                # Determine the WCS of the result
                if compare_wcs is None:
                    result_wcs = None
                else:
                    result_wcs = tmp_wcs_1 if tmp_wcs_1 else tmp_wcs_2

                result = super(CCDData, self).subtract(other)
                result.wcs = result_wcs
                self.wcs, other.wcs = tmp_wcs_1, tmp_wcs_2
                return result

            else:
                if hasattr(self, '_arithmetics_wcs'):
                    return super(CCDData, self).subtract(
                        other, compare_wcs=compare_wcs)
                else:
                    raise ImportError("wcs_compare functionality requires "
                                      "astropy 1.2 or greater.")

        return self._ccddata_arithmetic(other, np.subtract,
                                        scale_uncertainty=False)

    # Use NDDataArithmetic methods if astropy version is 1.2 or greater
    if not _ASTROPY_LT_1_2:
        del add, subtract, divide, multiply, _ccddata_arithmetic

        add = _arithmetic(np.add)(NDDataArray.add)
        subtract = _arithmetic(np.subtract)(NDDataArray.subtract)
        multiply = _arithmetic(np.multiply)(NDDataArray.multiply)
        divide = _arithmetic(np.true_divide)(NDDataArray.divide)

    def _insert_in_metadata_fits_safe(self, key, value):
        """
        Insert key/value pair into metadata in a way that FITS can serialize.

        Parameters
        ----------
        key : str
            Key to be inserted in dictionary.

        value : str or None
            Value to be inserted.

        Notes
        -----
        This addresses a shortcoming of the FITS standard. There are length
        restrictions on both the ``key`` (8 characters) and ``value`` (72
        characters) in the FITS standard. There is a convention for handline
        long keywords and a convention for handling long values, but the
        two conventions cannot be used at the same time.

        Autologging in `ccdproc` frequently creates keywords/values with this
        combination. The workaround is to use a shortened name for the keyword.
        """
        from .core import _short_names

        if key in _short_names and isinstance(self.meta, fits.Header):
            # This keyword was (hopefully) added by autologging but the
            # combination of it and its value not FITS-compliant in two
            # ways: the keyword name may be more than 8 characters and
            # the value may be too long. FITS cannot handle both of
            # those problems at once, so this fixes one of those
            # problems...
            # Shorten, sort of...
            short_name = _short_names[key]
            self.meta['HIERARCH {0}'.format(key.upper())] = (
                short_name, "Shortened name for ccdproc command")
            self.meta[short_name] = value
        else:
            self.meta[key] = value


# This needs to be importable by the tests...
_KEEP_THESE_KEYWORDS_IN_HEADER = [
    'JD-OBS',
    'MJD-OBS',
    'DATE-OBS'
]


def _generate_wcs_and_update_header(hdr):
    """
    Generate a WCS object from a header and remove the WCS-specific
    keywords from the header.
    Parameters
    ----------
    hdr : astropy.io.fits.header or other dict-like
    Returns
    -------
    new_header, wcs
    """

    # Try constructing a WCS object.
    try:
        wcs = WCS(hdr)
    except Exception as exc:
        # Normally WCS only raises Warnings and doesn't fail but in rare
        # cases (malformed header) it could fail...
        log.info('An exception happened while extracting WCS informations from '
                 'the Header.\n{}: {}'.format(type(exc).__name__, str(exc)))
        return hdr, None
    # Test for success by checking to see if the wcs ctype has a non-empty
    # value, return None for wcs if ctype is empty.
    if not wcs.wcs.ctype[0]:
        return hdr, None

    new_hdr = hdr.copy()
    # If the keywords below are in the header they are also added to WCS.
    # It seems like they should *not* be removed from the header, though.

    wcs_header = wcs.to_header(relax=True)
    for k in wcs_header:
        if k not in _KEEP_THESE_KEYWORDS_IN_HEADER:
            try:
                new_hdr.remove(k)
            except KeyError:
                pass
    return new_hdr, wcs


def fits_ccddata_reader(filename, hdu=0, unit=None, hdu_uncertainty='UNCERT',
                        hdu_mask='MASK', hdu_flags=None, **kwd):
    """
    Generate a CCDData object from a FITS file.

    Parameters
    ----------
    filename : str
        Name of fits file.

    hdu : int, optional
        FITS extension from which CCDData should be initialized. If zero and
        and no data in the primary extension, it will search for the first
        extension with data. The header will be added to the primary header.
        Default is ``0``.

    unit : `~astropy.units.Unit`, optional
        Units of the image data. If this argument is provided and there is a
        unit for the image in the FITS header (the keyword ``BUNIT`` is used
        as the unit, if present), this argument is used for the unit.
        Default is ``None``.

    hdu_uncertainty : str or None, optional
        FITS extension from which the uncertainty should be initialized. If the
        extension does not exist the uncertainty of the CCDData is ``None``.
        Default is ``'UNCERT'``.

    hdu_mask : str or None, optional
        FITS extension from which the mask should be initialized. If the
        extension does not exist the mask of the CCDData is ``None``.
        Default is ``'MASK'``.

    hdu_flags : str or None, optional
        Currently not implemented.
        Default is ``None``.

    kwd :
        Any additional keyword parameters are passed through to the FITS reader
        in :mod:`astropy.io.fits`; see Notes for additional discussion.

    Notes
    -----
    FITS files that contained scaled data (e.g. unsigned integer images) will
    be scaled and the keywords used to manage scaled data in
    :mod:`astropy.io.fits` are disabled.
    """
    unsupport_open_keywords = {
        'do_not_scale_image_data': ('Image data must be scaled to perform '
                                    'ccdproc operations.'),
        'scale_back': 'Scale information is not preserved.'
    }
    for key, msg in unsupport_open_keywords.items():
        if key in kwd:
            prefix = 'unsupported keyword: {0}.'.format(key)
            raise TypeError(' '.join([prefix, msg]))
    with fits.open(filename, **kwd) as hdus:
        hdr = hdus[hdu].header

        if hdu_uncertainty is not None and hdu_uncertainty in hdus:
            uncertainty = StdDevUncertainty(hdus[hdu_uncertainty].data)
        else:
            uncertainty = None

        if hdu_mask is not None and hdu_mask in hdus:
            # Mask is saved as uint but we want it to be boolean.
            mask = hdus[hdu_mask].data.astype(np.bool_)
        else:
            mask = None

        if hdu_flags is not None and hdu_flags in hdus:
            raise NotImplementedError('loading flags is currently not '
                                      'supported.')

        # search for the first instance with data if
        # the primary header is empty.
        if hdu == 0 and hdus[hdu].data is None:
            for i in range(len(hdus)):
                if hdus.fileinfo(i)['datSpan'] > 0:
                    hdu = i
                    comb_hdr = hdus[hdu].header.copy()
                    comb_hdr.extend(hdr, unique=True)
                    hdr = comb_hdr
                    log.info("first HDU with data is extension "
                             "{0}.".format(hdu))
                    break

        if 'bunit' in hdr:
            fits_unit_string = hdr['bunit']
            # patch to handle FITS files using ADU for the unit instead of the
            # standard version of 'adu'
            if fits_unit_string.strip().lower() == 'adu':
                fits_unit_string = fits_unit_string.lower()
        else:
            fits_unit_string = None

        if unit is not None and fits_unit_string:
            log.info("using the unit {0} passed to the FITS reader instead of "
                     "the unit {1} in the FITS file.".format(unit,
                                                             fits_unit_string))

        use_unit = unit or fits_unit_string
        hdr, wcs = _generate_wcs_and_update_header(hdr)
        ccd_data = CCDData(hdus[hdu].data, meta=hdr, unit=use_unit,
                           mask=mask, uncertainty=uncertainty, wcs=wcs)

    return ccd_data


def fits_ccddata_writer(ccd_data, filename, hdu_mask='MASK',
                        hdu_uncertainty='UNCERT', hdu_flags=None, **kwd):
    """
    Write CCDData object to FITS file.

    Parameters
    ----------
    filename : str
        Name of file.

    hdu_mask, hdu_uncertainty, hdu_flags : str or None, optional
        If it is a string append this attribute to the HDUList as
        `~astropy.io.fits.ImageHDU` with the string as extension name.
        Flags are not supported at this time. If ``None`` this attribute
        is not appended.
        Default is ``'MASK'`` for mask, ``'UNCERT'`` for uncertainty and
        ``None`` for flags.

    kwd :
        All additional keywords are passed to :py:mod:`astropy.io.fits`

    Raises
    -------
    ValueError
        - If ``self.mask`` is set but not a `numpy.ndarray`.
        - If ``self.uncertainty`` is set but not a
          `~astropy.nddata.StdDevUncertainty`.
        - If ``self.uncertainty`` is set but has another unit then
          ``self.data``.

    NotImplementedError
        Saving flags is not supported.
    """
    hdu = ccd_data.to_hdu(hdu_mask=hdu_mask, hdu_uncertainty=hdu_uncertainty,
                          hdu_flags=hdu_flags)
    hdu.writeto(filename, **kwd)

# This should be be a tuple to ensure it isn't inadvertently changed
# elsewhere.
_recognized_fits_file_extensions = ('fit', 'fits', 'fts')

if _ASTROPY_LT_1_3:
    def is_fits(origin, filepath, fileobj, *args, **kwargs):
        """
        Wrapper around astropy.io.fits.connect.is_fits that handles the extra
        extension.

        Can be removed if fts is added to astropy.io as a recognized FITS
        extension.
        """
        if ((filepath is not None) and
                filepath.lower().endswith(('.fts', '.fts.gz'))):

            return True

        else:
            return fits.connect.is_fits(origin, filepath, fileobj,
                                        *args, **kwargs)
else:
    is_fits = fits.connect.is_fits

if _ASTROPY_LT_1_3:
    registry.register_reader('fits', CCDData, fits_ccddata_reader)
    registry.register_writer('fits', CCDData, fits_ccddata_writer)
    registry.register_identifier('fits', CCDData, is_fits)
else:
    with registry.delay_doc_updates(CCDData):
        registry.register_reader('fits', CCDData, fits_ccddata_reader)
        registry.register_writer('fits', CCDData, fits_ccddata_writer)
        registry.register_identifier('fits', CCDData, is_fits)

try:
    CCDData.read.__doc__ = fits_ccddata_reader.__doc__
except AttributeError:
    CCDData.read.__func__.__doc__ = fits_ccddata_reader.__doc__

try:
    CCDData.write.__doc__ = fits_ccddata_writer.__doc__
except AttributeError:
    CCDData.write.__func__.__doc__ = fits_ccddata_writer.__doc__

# CCDData moved to astropy core so we just import them from there (overwriting)
# the classes defined here.
if _ASTROPY_GT_2_0:
    from astropy.nddata import fits_ccddata_reader, fits_ccddata_writer, CCDData