source: sasview/src/sas/dataloader/readers/cansas_reader.py @ 6b0fc72

"""
    CanSAS data reader - new recursive cansas_version.
"""
############################################################################
#This software was developed by the University of Tennessee as part of the
#Distributed Data Analysis of Neutron Scattering Experiments (DANSE)
#project funded by the US National Science Foundation.
#If you use DANSE applications to do scientific research that leads to
#publication, we ask that you acknowledge the use of the software with the
#following sentence:
#This work benefited from DANSE software developed under NSF award DMR-0520547.
#copyright 2008,2009 University of Tennessee
#############################################################################

import logging
import numpy
import os
import sys
import datetime
import inspect
# For saving individual sections of data
from sas.dataloader.data_info import Data1D
from sas.dataloader.data_info import Collimation
from sas.dataloader.data_info import TransmissionSpectrum
from sas.dataloader.data_info import Detector
from sas.dataloader.data_info import Process
from sas.dataloader.data_info import Aperture
# Both imports used. Do not remove either.
from xml.dom.minidom import parseString
import sas.dataloader.readers.xml_reader as xml_reader
from sas.dataloader.readers.xml_reader import XMLreader
from sas.dataloader.readers.cansas_constants import CansasConstants

_ZERO = 1e-16
PREPROCESS = "xmlpreprocess"
ENCODING = "encoding"
RUN_NAME_DEFAULT = "None"
HAS_CONVERTER = True
try:
    from sas.data_util.nxsunit import Converter
except ImportError:
    HAS_CONVERTER = False

CONSTANTS = CansasConstants()
CANSAS_FORMAT = CONSTANTS.format
CANSAS_NS = CONSTANTS.names
ALLOW_ALL = True

# DO NOT REMOVE
# Called by outside packages:
#    sas.perspectives.invariant.invariant_state
#    sas.perspectives.fitting.pagestate
def get_content(location, node):
    """
    Get the first instance of the content of a xpath location.

    :param location: xpath location
    :param node: node to start at

    :return: Element, or None
    """
    nodes = node.xpath(location,
                       namespaces={'ns': CANSAS_NS.get("1.0").get("ns")})

    if len(nodes) > 0:
        return nodes[0]
    else:
        return None


# DO NOT REMOVE
# Called by outside packages:
#    sas.perspectives.fitting.pagestate
def write_node(doc, parent, name, value, attr=None):
    """
    :param doc: document DOM
    :param parent: parent node
    :param name: tag of the element
    :param value: value of the child text node
    :param attr: attribute dictionary

    :return: True if something was appended, otherwise False
    """
    if attr is None:
        attr = {}
    if value is not None:
        node = doc.createElement(name)
        node.appendChild(doc.createTextNode(str(value)))
        for item in attr:
            node.setAttribute(item, attr[item])
        parent.appendChild(node)
        return True
    return False


class Reader(XMLreader):
    """
    Class to load cansas 1D XML files

    :Dependencies:
        The CanSAS reader requires PyXML 0.8.4 or later.
    """
    ##CanSAS version - defaults to version 1.0
    cansas_version = "1.0"
    base_ns = "{cansas1d/1.0}"

    logging = []
    errors = []

    type_name = "canSAS"
    ## Wildcards
    type = ["XML files (*.xml)|*.xml", "SasView Save Files (*.svs)|*.svs"]
    ## List of allowed extensions
    ext = ['.xml', '.XML', '.svs', '.SVS']

    ## Flag to bypass extension check
    allow_all = True


    def __init__(self):
        ## List of errors
        self.errors = []
        self.encoding = None


    def is_cansas(self, ext="xml"):
        """
        Checks to see if the xml file is a CanSAS file

        :param ext: The file extension of the data file
        """
        if self.validate_xml():
            name = "{http://www.w3.org/2001/XMLSchema-instance}schemaLocation"
            value = self.xmlroot.get(name)
            if CANSAS_NS.get(self.cansas_version).get("ns") == \
                    value.rsplit(" ")[0]:
                return True
        if ext == "svs":
            return True
        return False


    def load_file_and_schema(self, xml_file):
        """
        Loads the file and associates a schema, if a known schema exists

        :param xml_file: The xml file path sent to Reader.read
        """
        base_name = xml_reader.__file__
        base_name = base_name.replace("\\", "/")
        base = base_name.split("/sas/")[0]

        # Load in xml file and get the cansas version from the header
        self.set_xml_file(xml_file)
        self.cansas_version = self.xmlroot.get("version", "1.0")

        # Generic values for the cansas file based on the version
        cansas_defaults = CANSAS_NS.get(self.cansas_version, "1.0")
        schema_path = "{0}/sas/dataloader/readers/schema/{1}".format\
                (base, cansas_defaults.get("schema")).replace("\\", "/")

        # Link a schema to the XML file.
        self.set_schema(schema_path)
        return cansas_defaults


    def read(self, xml_file):
        """
        Validate and read in an xml_file file in the canSAS format.

        :param xml_file: A canSAS file path in proper XML format
        """

        # output - Final list of Data1D objects
        output = []
        # ns - Namespace hierarchy for current xml_file object
        ns_list = []

        # Check that the file exists
        if os.path.isfile(xml_file):
            basename = os.path.basename(xml_file)
            _, extension = os.path.splitext(basename)
            # If the file type is not allowed, return nothing
            if extension in self.ext or self.allow_all:
                # Get the file location of
                cansas_defaults = self.load_file_and_schema(xml_file)

                # Try to load the file, but raise an error if unable to.
                # Check the file matches the XML schema
                try:
                    if self.is_cansas(extension):
                        # Get each SASentry from XML file and add it to a list.
                        entry_list = self.xmlroot.xpath(
                                '/ns:SASroot/ns:SASentry',
                                namespaces={'ns': cansas_defaults.get("ns")})
                        ns_list.append("SASentry")

                        # If multiple files, modify the name for each is unique
                        increment = 0
                        # Parse each SASentry item
                        for entry in entry_list:
                            # Define a new Data1D object with zeroes for
                            # x_vals and y_vals
                            data1d = Data1D(numpy.empty(0), numpy.empty(0),
                                            numpy.empty(0), numpy.empty(0))
                            data1d.dxl = numpy.empty(0)
                            data1d.dxw = numpy.empty(0)

                            # If more than one SASentry, increment each in order
                            name = basename
                            if len(entry_list) - 1 > 0:
                                name += "_{0}".format(increment)
                                increment += 1

                            # Set the Data1D name and then parse the entry.
                            # The entry is appended to a list of entry values
                            data1d.filename = name
                            data1d.meta_data["loader"] = "CanSAS 1D"

                            # Get all preprocessing events and encoding
                            self.set_processing_instructions()
                            data1d.meta_data[PREPROCESS] = \
                                    self.processing_instructions

                            # Parse the XML file
                            return_value, extras = \
                                self._parse_entry(entry, ns_list, data1d)
                            del extras[:]

                            return_value = self._final_cleanup(return_value)
                            output.append(return_value)
                    else:
                        output.append("Invalid XML at: {0}".format(\
                                                    self.find_invalid_xml()))
                except:
                    # If the file does not match the schema, raise this error
                    raise RuntimeError, "%s cannot be read" % xml_file
                return output
        # Return a list of parsed entries that dataloader can manage
        return None


    def _final_cleanup(self, data1d):
        """
        Final cleanup of the Data1D object to be sure it has all the
        appropriate information needed for perspectives

        :param data1d: Data1D object that has been populated
        """
        # Final cleanup
        # Remove empty nodes, verify array sizes are correct
        for error in self.errors:
            data1d.errors.append(error)
        del self.errors[:]
        numpy.trim_zeros(data1d.x)
        numpy.trim_zeros(data1d.y)
        numpy.trim_zeros(data1d.dy)
        size_dx = data1d.dx.size
        size_dxl = data1d.dxl.size
        size_dxw = data1d.dxw.size
        if size_dxl == 0 and size_dxw == 0:
            data1d.dxl = None
            data1d.dxw = None
            numpy.trim_zeros(data1d.dx)
        elif size_dx == 0:
            data1d.dx = None
            size_dx = size_dxl
            numpy.trim_zeros(data1d.dxl)
            numpy.trim_zeros(data1d.dxw)
        return data1d

    def _create_unique_key(self, dictionary, name, numb=0):
        """
        Create a unique key value for any dictionary to prevent overwriting
        Recurses until a unique key value is found.

        :param dictionary: A dictionary with any number of entries
        :param name: The index of the item to be added to dictionary
        :param numb: The number to be appended to the name, starts at 0
        """
        if dictionary.get(name) is not None:
            numb += 1
            name = name.split("_")[0]
            name += "_{0}".format(numb)
            name = self._create_unique_key(dictionary, name, numb)
        return name


    def _unit_conversion(self, node, new_current_level, data1d, \
                                                tagname, node_value):
        """
        A unit converter method used to convert the data included in the file
        to the default units listed in data_info

        :param new_current_level: cansas_constants level as returned by
            iterate_namespace
        :param attr: The attributes of the node
        :param data1d: Where the values will be saved
        :param node_value: The value of the current dom node
        """
        attr = node.attrib
        value_unit = ''
        if 'unit' in attr and new_current_level.get('unit') is not None:
            try:
                local_unit = attr['unit']
                if isinstance(node_value, float) is False:
                    exec("node_value = float({0})".format(node_value))
                default_unit = None
                unitname = new_current_level.get("unit")
                exec "default_unit = data1d.{0}".format(unitname)
                if local_unit is not None and default_unit is not None and \
                        local_unit.lower() != default_unit.lower() \
                        and local_unit.lower() != "none":
                    if HAS_CONVERTER == True:
                        ## Check local units - bad units raise KeyError
                        data_conv_q = Converter(local_unit)
                        value_unit = default_unit
                        i_string = "node_value = data_conv_q"
                        i_string += "(node_value, units=data1d.{0})"
                        exec i_string.format(unitname)
                    else:
                        value_unit = local_unit
                        err_msg = "Unit converter is not available.\n"
                        self.errors.append(err_msg)
                else:
                    value_unit = local_unit
            except KeyError:
                err_msg = "CanSAS reader: unexpected "
                err_msg += "\"{0}\" unit [{1}]; "
                err_msg = err_msg.format(tagname, local_unit)
                intermediate = "err_msg += " + \
                            "\"expecting [{1}]\"" + \
                            ".format(data1d.{0})"
                exec intermediate.format(unitname, "{0}", "{1}")
                self.errors.append(err_msg)
                value_unit = local_unit
            except:
                print sys.exc_info()
                err_msg = "CanSAS reader: unknown error converting "
                err_msg += "\"{0}\" unit [{1}]"
                err_msg = err_msg.format(tagname, local_unit)
                self.errors.append(err_msg)
                value_unit = local_unit
        elif 'unit' in attr:
            value_unit = attr['unit']
        node_value = "float({0})".format(node_value)
        return node_value, value_unit


    def _check_for_empty_data(self, data1d):
        """
        Creates an empty data set if no data is passed to the reader

        :param data1d: presumably a Data1D object
        """
        if data1d == None:
            self.errors = []
            x_vals = numpy.empty(0)
            y_vals = numpy.empty(0)
            dx_vals = numpy.empty(0)
            dy_vals = numpy.empty(0)
            dxl = numpy.empty(0)
            dxw = numpy.empty(0)
            data1d = Data1D(x_vals, y_vals, dx_vals, dy_vals)
            data1d.dxl = dxl
            data1d.dxw = dxw
        return data1d

    def _handle_special_cases(self, tagname, data1d, children):
        """
        Handle cases where the data type in Data1D is a dictionary or list

        :param tagname: XML tagname in use
        :param data1d: The original Data1D object
        :param children: Child nodes of node
        :param node: existing node with tag name 'tagname'
        """
        if tagname == "SASdetector":
            data1d = Detector()
        elif tagname == "SAScollimation":
            data1d = Collimation()
        elif tagname == "SAStransmission_spectrum":
            data1d = TransmissionSpectrum()
        elif tagname == "SASprocess":
            data1d = Process()
            for child in children:
                if child.tag.replace(self.base_ns, "") == "term":
                    term_attr = {}
                    for attr in child.keys():
                        term_attr[attr] = \
                            ' '.join(child.get(attr).split())
                    if child.text is not None:
                        term_attr['value'] = \
                            ' '.join(child.text.split())
                    data1d.term.append(term_attr)
        elif tagname == "aperture":
            data1d = Aperture()
        if tagname == "Idata" and children is not None:
            data1d = self._check_for_empty_resolution(data1d, children)
        return data1d

    def _check_for_empty_resolution(self, data1d, children):
        """
        A method to check all resolution data sets are the same size as I and Q
        """
        dql_exists = False
        dqw_exists = False
        dq_exists = False
        di_exists = False
        for child in children:
            tag = child.tag.replace(self.base_ns, "")
            if tag == "dQl":
                dql_exists = True
            if tag == "dQw":
                dqw_exists = True
            if tag == "Qdev":
                dq_exists = True
            if tag == "Idev":
                di_exists = True
        if dqw_exists and dql_exists == False:
            data1d.dxl = numpy.append(data1d.dxl, 0.0)
        elif dql_exists and dqw_exists == False:
            data1d.dxw = numpy.append(data1d.dxw, 0.0)
        elif dql_exists == False and dqw_exists == False \
                                            and dq_exists == False:
            data1d.dx = numpy.append(data1d.dx, 0.0)
        if di_exists == False:
            data1d.dy = numpy.append(data1d.dy, 0.0)
        return data1d

    def _restore_original_case(self,
                               tagname_original,
                               tagname,
                               save_data1d,
                               data1d):
        """
        Save the special case data to the appropriate location and restore
        the original Data1D object

        :param tagname_original: Unmodified tagname for the node
        :param tagname: modified tagname for the node
        :param save_data1d: The original Data1D object
        :param data1d: If a special case was handled, an object of that type
        """
        if tagname_original == "SASdetector":
            save_data1d.detector.append(data1d)
        elif tagname_original == "SAScollimation":
            save_data1d.collimation.append(data1d)
        elif tagname == "SAStransmission_spectrum":
            save_data1d.trans_spectrum.append(data1d)
        elif tagname_original == "SASprocess":
            save_data1d.process.append(data1d)
        elif tagname_original == "aperture":
            save_data1d.aperture.append(data1d)
        else:
            save_data1d = data1d
        return save_data1d

    def _handle_attributes(self, node, data1d, cs_values, tagname):
        """
        Process all of the attributes for a node
        """
        attr = node.attrib
        if attr is not None:
            for key in node.keys():
                try:
                    node_value, unit = self._get_node_value(node, cs_values, \
                                                   data1d, tagname)
                    cansas_attrib = \
                        cs_values.current_level.get("attributes").get(key)
                    attrib_variable = cansas_attrib.get("variable")
                    if key == 'unit' and unit != '':
                        attrib_value = unit
                    else:
                        attrib_value = node.attrib[key]
                    store_attr = attrib_variable.format("data1d", \
                                                    attrib_value, key)
                    exec store_attr
                except AttributeError:
                    pass
        return data1d

    def _get_node_value(self, node, cs_values, data1d, tagname):
        """
        Get the value of a node and any applicable units

        :param node: The XML node to get the value of
        :param cs_values: A CansasConstants.CurrentLevel object
        :param attr: The node attributes
        :param dataid: The working object to be modified
        :param tagname: The tagname of the node
        """
        #Get the text from the node and convert all whitespace to spaces
        units = ''
        node_value = node.text
        if node_value == "":
            node_value = None
        if node_value is not None:
            node_value = ' '.join(node_value.split())

        # If the value is a float, compile with units.
        if cs_values.ns_datatype == "float":
            # If an empty value is given, set as zero.
            if node_value is None or node_value.isspace() \
                                    or node_value.lower() == "nan":
                node_value = "0.0"
            #Convert the value to the base units
            node_value, units = self._unit_conversion(node, \
                        cs_values.current_level, data1d, tagname, node_value)

        # If the value is a timestamp, convert to a datetime object
        elif cs_values.ns_datatype == "timestamp":
            if node_value is None or node_value.isspace():
                pass
            else:
                try:
                    node_value = \
                        datetime.datetime.fromtimestamp(node_value)
                except ValueError:
                    node_value = None
        return node_value, units

    def _parse_entry(self, dom, names=None, data1d=None, extras=None):
        """
        Parse a SASEntry - new recursive method for parsing the dom of
            the CanSAS data format. This will allow multiple data files
            and extra nodes to be read in simultaneously.

        :param dom: dom object with a namespace base of names
        :param names: A list of element names that lead up to the dom object
        :param data1d: The data1d object that will be modified
        :param extras: Any values that should go into meta_data when data1d
            is not a Data1D object
        """

        if extras is None:
            extras = []
        if names is None or names == []:
            names = ["SASentry"]

        data1d = self._check_for_empty_data(data1d)

        self.base_ns = "{0}{1}{2}".format("{", \
                            CANSAS_NS.get(self.cansas_version).get("ns"), "}")
        tagname = ''
        tagname_original = ''

        # Go through each child in the parent element
        for node in dom:
            try:
                # Get the element name and set the current names level
                tagname = node.tag.replace(self.base_ns, "")
                tagname_original = tagname
                if tagname == "fitting_plug_in" or tagname == "pr_inversion" or\
                    tagname == "invariant":
                    continue
                names.append(tagname)
                children = node.getchildren()
                if len(children) == 0:
                    children = None
                save_data1d = data1d

                # Look for special cases
                data1d = self._handle_special_cases(tagname, data1d, children)

                # Get where to store content
                cs_values = CONSTANTS.iterate_namespace(names)
                # If the element is a child element, recurse
                if children is not None:
                    # Returned value is new Data1D object with all previous and
                    # new values in it.
                    data1d, extras = self._parse_entry(node,
                                                       names, data1d, extras)

                #Get the information from the node
                node_value, _ = self._get_node_value(node, cs_values, \
                                                            data1d, tagname)

                # If appending to a dictionary (meta_data | run_name)
                # make sure the key is unique
                if cs_values.ns_variable == "{0}.meta_data[\"{2}\"] = \"{1}\"":
                    # If we are within a Process, Detector, Collimation or
                    # Aperture instance, pull out old data1d
                    tagname = self._create_unique_key(data1d.meta_data, \
                                                      tagname, 0)
                    if isinstance(data1d, Data1D) == False:
                        store_me = cs_values.ns_variable.format("data1d", \
                                                            node_value, tagname)
                        extras.append(store_me)
                        cs_values.ns_variable = None
                if cs_values.ns_variable == "{0}.run_name[\"{2}\"] = \"{1}\"":
                    tagname = self._create_unique_key(data1d.run_name, \
                                                      tagname, 0)

                # Check for Data1D object and any extra commands to save
                if isinstance(data1d, Data1D):
                    for item in extras:
                        exec item
                # Don't bother saving empty information unless it is a float
                if cs_values.ns_variable is not None and \
                            node_value is not None and \
                            node_value.isspace() == False:
                    # Format a string and then execute it.
                    store_me = cs_values.ns_variable.format("data1d", \
                                                            node_value, tagname)
                    exec store_me
                # Get attributes and process them
                data1d = self._handle_attributes(node, data1d, cs_values, \
                                                 tagname)

            except TypeError:
                pass
            except Exception as excep:
                exc_type, exc_obj, exc_tb = sys.exc_info()
                fname = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]
                print(excep, exc_type, fname, exc_tb.tb_lineno, \
                      tagname, exc_obj)
            finally:
                # Save special cases in original data1d object
                # then restore the data1d
                save_data1d = self._restore_original_case(tagname_original, \
                                                tagname, save_data1d, data1d)
                if tagname_original == "fitting_plug_in" or \
                    tagname_original == "invariant" or \
                    tagname_original == "pr_inversion":
                    pass
                else:
                    data1d = save_data1d
                    # Remove tagname from names to restore original base
                    names.remove(tagname_original)
        return data1d, extras

    def _get_pi_string(self):
        """
        Creates the processing instructions header for writing to file
        """
        pis = self.return_processing_instructions()
        if len(pis) > 0:
            pi_tree = self.create_tree(pis[0])
            i = 1
            for i in range(1, len(pis) - 1):
                pi_tree = self.append(pis[i], pi_tree)
            pi_string = self.to_string(pi_tree)
        else:
            pi_string = ""
        return pi_string

    def _create_main_node(self):
        """
        Creates the primary xml header used when writing to file
        """
        xsi = "http://www.w3.org/2001/XMLSchema-instance"
        version = self.cansas_version
        n_s = CANSAS_NS.get(version).get("ns")
        if version == "1.1":
            url = "http://www.cansas.org/formats/1.1/"
        else:
            url = "http://svn.smallangles.net/svn/canSAS/1dwg/trunk/"
        schema_location = "{0} {1}cansas1d.xsd".format(n_s, url)
        attrib = {"{" + xsi + "}schemaLocation" : schema_location,
                  "version" : version}
        nsmap = {'xsi' : xsi, None: n_s}

        main_node = self.create_element("{" + n_s + "}SASroot",
                                        attrib=attrib, nsmap=nsmap)
        return main_node

    def _write_run_names(self, datainfo, entry_node):
        """
        Writes the run names to the XML file

        :param datainfo: The Data1D object the information is coming from
        :param entry_node: lxml node ElementTree object to be appended to
        """
        if datainfo.run == None or datainfo.run == []:
            datainfo.run.append(RUN_NAME_DEFAULT)
            datainfo.run_name[RUN_NAME_DEFAULT] = RUN_NAME_DEFAULT
        for item in datainfo.run:
            runname = {}
            if item in datainfo.run_name and \
            len(str(datainfo.run_name[item])) > 1:
                runname = {'name': datainfo.run_name[item]}
            self.write_node(entry_node, "Run", item, runname)

    def _write_data(self, datainfo, entry_node):
        """
        Writes the I and Q data to the XML file

        :param datainfo: The Data1D object the information is coming from
        :param entry_node: lxml node ElementTree object to be appended to
        """
        node = self.create_element("SASdata")
        self.append(node, entry_node)

        for i in range(len(datainfo.x)):
            point = self.create_element("Idata")
            node.append(point)
            self.write_node(point, "Q", datainfo.x[i],
                            {'unit': datainfo.x_unit})
            if len(datainfo.y) >= i:
                self.write_node(point, "I", datainfo.y[i],
                                {'unit': datainfo.y_unit})
            if datainfo.dy != None and len(datainfo.dy) > i:
                self.write_node(point, "Idev", datainfo.dy[i],
                                {'unit': datainfo.y_unit})
            if datainfo.dx != None and len(datainfo.dx) > i:
                self.write_node(point, "Qdev", datainfo.dx[i],
                                {'unit': datainfo.x_unit})
            if datainfo.dxw != None and len(datainfo.dxw) > i:
                self.write_node(point, "dQw", datainfo.dxw[i],
                                {'unit': datainfo.x_unit})
            if datainfo.dxl != None and len(datainfo.dxl) > i:
                self.write_node(point, "dQl", datainfo.dxl[i],
                                {'unit': datainfo.x_unit})

    def _write_trans_spectrum(self, datainfo, entry_node):
        """
        Writes the transmission spectrum data to the XML file

        :param datainfo: The Data1D object the information is coming from
        :param entry_node: lxml node ElementTree object to be appended to
        """
        for i in range(len(datainfo.trans_spectrum)):
            spectrum = datainfo.trans_spectrum[i]
            node = self.create_element("SAStransmission_spectrum",
                                       {"name" : spectrum.name})
            self.append(node, entry_node)
            if isinstance(spectrum.timestamp, datetime.datetime):
                node.setAttribute("timestamp", spectrum.timestamp)
            for i in range(len(spectrum.wavelength)):
                point = self.create_element("Tdata")
                node.append(point)
                self.write_node(point, "Lambda", spectrum.wavelength[i],
                                {'unit': spectrum.wavelength_unit})
                self.write_node(point, "T", spectrum.transmission[i],
                                {'unit': spectrum.transmission_unit})
                if spectrum.transmission_deviation != None \
                and len(spectrum.transmission_deviation) >= i:
                    self.write_node(point, "Tdev",
                                    spectrum.transmission_deviation[i],
                                    {'unit':
                                     spectrum.transmission_deviation_unit})

    def _write_sample_info(self, datainfo, entry_node):
        """
        Writes the sample information to the XML file

        :param datainfo: The Data1D object the information is coming from
        :param entry_node: lxml node ElementTree object to be appended to
        """
        sample = self.create_element("SASsample")
        if datainfo.sample.name is not None:
            self.write_attribute(sample, "name",
                                 str(datainfo.sample.name))
        self.append(sample, entry_node)
        self.write_node(sample, "ID", str(datainfo.sample.ID))
        self.write_node(sample, "thickness", datainfo.sample.thickness,
                        {"unit": datainfo.sample.thickness_unit})
        self.write_node(sample, "transmission", datainfo.sample.transmission)
        self.write_node(sample, "temperature", datainfo.sample.temperature,
                        {"unit": datainfo.sample.temperature_unit})

        pos = self.create_element("position")
        written = self.write_node(pos,
                                  "x",
                                  datainfo.sample.position.x,
                                  {"unit": datainfo.sample.position_unit})
        written = written | self.write_node( \
            pos, "y", datainfo.sample.position.y,
            {"unit": datainfo.sample.position_unit})
        written = written | self.write_node( \
            pos, "z", datainfo.sample.position.z,
            {"unit": datainfo.sample.position_unit})
        if written == True:
            self.append(pos, sample)

        ori = self.create_element("orientation")
        written = self.write_node(ori, "roll",
                                  datainfo.sample.orientation.x,
                                  {"unit": datainfo.sample.orientation_unit})
        written = written | self.write_node( \
            ori, "pitch", datainfo.sample.orientation.y,
            {"unit": datainfo.sample.orientation_unit})
        written = written | self.write_node( \
            ori, "yaw", datainfo.sample.orientation.z,
            {"unit": datainfo.sample.orientation_unit})
        if written == True:
            self.append(ori, sample)

        for item in datainfo.sample.details:
            self.write_node(sample, "details", item)

    def _write_instrument(self, datainfo, entry_node):
        """
        Writes the instrumental information to the XML file

        :param datainfo: The Data1D object the information is coming from
        :param entry_node: lxml node ElementTree object to be appended to
        """
        instr = self.create_element("SASinstrument")
        self.append(instr, entry_node)
        self.write_node(instr, "name", datainfo.instrument)
        return instr

    def _write_source(self, datainfo, instr):
        """
        Writes the source information to the XML file

        :param datainfo: The Data1D object the information is coming from
        :param instr: instrument node  to be appended to
        """
        source = self.create_element("SASsource")
        if datainfo.source.name is not None:
            self.write_attribute(source, "name",
                                 str(datainfo.source.name))
        self.append(source, instr)
        if datainfo.source.radiation == None or datainfo.source.radiation == '':
            datainfo.source.radiation = "neutron"
        self.write_node(source, "radiation", datainfo.source.radiation)

        size = self.create_element("beam_size")
        if datainfo.source.beam_size_name is not None:
            self.write_attribute(size, "name",
                                 str(datainfo.source.beam_size_name))
        written = self.write_node( \
            size, "x", datainfo.source.beam_size.x,
            {"unit": datainfo.source.beam_size_unit})
        written = written | self.write_node( \
            size, "y", datainfo.source.beam_size.y,
            {"unit": datainfo.source.beam_size_unit})
        written = written | self.write_node( \
            size, "z", datainfo.source.beam_size.z,
            {"unit": datainfo.source.beam_size_unit})
        if written == True:
            self.append(size, source)

        self.write_node(source, "beam_shape", datainfo.source.beam_shape)
        self.write_node(source, "wavelength",
                        datainfo.source.wavelength,
                        {"unit": datainfo.source.wavelength_unit})
        self.write_node(source, "wavelength_min",
                        datainfo.source.wavelength_min,
                        {"unit": datainfo.source.wavelength_min_unit})
        self.write_node(source, "wavelength_max",
                        datainfo.source.wavelength_max,
                        {"unit": datainfo.source.wavelength_max_unit})
        self.write_node(source, "wavelength_spread",
                        datainfo.source.wavelength_spread,
                        {"unit": datainfo.source.wavelength_spread_unit})

    def _write_collimation(self, datainfo, instr):
        """
        Writes the collimation information to the XML file

        :param datainfo: The Data1D object the information is coming from
        :param instr: lxml node ElementTree object to be appended to
        """
        if datainfo.collimation == [] or datainfo.collimation == None:
            coll = Collimation()
            datainfo.collimation.append(coll)
        for item in datainfo.collimation:
            coll = self.create_element("SAScollimation")
            if item.name is not None:
                self.write_attribute(coll, "name", str(item.name))
            self.append(coll, instr)

            self.write_node(coll, "length", item.length,
                            {"unit": item.length_unit})

            for aperture in item.aperture:
                apert = self.create_element("aperture")
                if aperture.name is not None:
                    self.write_attribute(apert, "name", str(aperture.name))
                if aperture.type is not None:
                    self.write_attribute(apert, "type", str(aperture.type))
                self.append(apert, coll)

                size = self.create_element("size")
                if aperture.size_name is not None:
                    self.write_attribute(size, "name",
                                         str(aperture.size_name))
                written = self.write_node(size, "x", aperture.size.x,
                                          {"unit": aperture.size_unit})
                written = written | self.write_node( \
                    size, "y", aperture.size.y,
                    {"unit": aperture.size_unit})
                written = written | self.write_node( \
                    size, "z", aperture.size.z,
                    {"unit": aperture.size_unit})
                if written == True:
                    self.append(size, apert)

                self.write_node(apert, "distance", aperture.distance,
                                {"unit": aperture.distance_unit})


    def _write_detectors(self, datainfo, instr):
        """
        Writes the detector information to the XML file

        :param datainfo: The Data1D object the information is coming from
        :param inst: lxml instrument node to be appended to
        """
        if datainfo.detector == None or datainfo.detector == []:
            det = Detector()
            det.name = ""
            datainfo.detector.append(det)

        for item in datainfo.detector:
            det = self.create_element("SASdetector")
            written = self.write_node(det, "name", item.name)
            written = written | self.write_node(det, "SDD", item.distance,
                                                {"unit": item.distance_unit})
            if written == True:
                self.append(det, instr)

            off = self.create_element("offset")
            written = self.write_node(off, "x", item.offset.x,
                                      {"unit": item.offset_unit})
            written = written | self.write_node(off, "y", item.offset.y,
                                                {"unit": item.offset_unit})
            written = written | self.write_node(off, "z", item.offset.z,
                                                {"unit": item.offset_unit})
            if written == True:
                self.append(off, det)

            ori = self.create_element("orientation")
            written = self.write_node(ori, "roll", item.orientation.x,
                                      {"unit": item.orientation_unit})
            written = written | self.write_node(ori, "pitch",
                                                item.orientation.y,
                                                {"unit": item.orientation_unit})
            written = written | self.write_node(ori, "yaw",
                                                item.orientation.z,
                                                {"unit": item.orientation_unit})
            if written == True:
                self.append(ori, det)

            center = self.create_element("beam_center")
            written = self.write_node(center, "x", item.beam_center.x,
                                      {"unit": item.beam_center_unit})
            written = written | self.write_node(center, "y",
                                                item.beam_center.y,
                                                {"unit": item.beam_center_unit})
            written = written | self.write_node(center, "z",
                                                item.beam_center.z,
                                                {"unit": item.beam_center_unit})
            if written == True:
                self.append(center, det)

            pix = self.create_element("pixel_size")
            written = self.write_node(pix, "x", item.pixel_size.x,
                                      {"unit": item.pixel_size_unit})
            written = written | self.write_node(pix, "y", item.pixel_size.y,
                                                {"unit": item.pixel_size_unit})
            written = written | self.write_node(pix, "z", item.pixel_size.z,
                                                {"unit": item.pixel_size_unit})
            written = written | self.write_node(det, "slit_length",
                                                item.slit_length,
                                                {"unit": item.slit_length_unit})
            if written == True:
                self.append(pix, det)

    def _write_process_notes(self, datainfo, entry_node):
        """
        Writes the process notes to the XML file

        :param datainfo: The Data1D object the information is coming from
        :param entry_node: lxml node ElementTree object to be appended to

        """
        for item in datainfo.process:
            node = self.create_element("SASprocess")
            self.append(node, entry_node)
            self.write_node(node, "name", item.name)
            self.write_node(node, "date", item.date)
            self.write_node(node, "description", item.description)
            for term in item.term:
                value = term['value']
                del term['value']
                self.write_node(node, "term", value, term)
            for note in item.notes:
                self.write_node(node, "SASprocessnote", note)
            if len(item.notes) == 0:
                self.write_node(node, "SASprocessnote", "")

    def _write_notes(self, datainfo, entry_node):
        """
        Writes the notes to the XML file and creates an empty note if none
        exist

        :param datainfo: The Data1D object the information is coming from
        :param entry_node: lxml node ElementTree object to be appended to

        """
        if len(datainfo.notes) == 0:
            node = self.create_element("SASnote")
            self.append(node, entry_node)
        else:
            for item in datainfo.notes:
                node = self.create_element("SASnote")
                self.write_text(node, item)
                self.append(node, entry_node)

    def _check_origin(self, entry_node, doc):
        """
        Return the document, and the SASentry node associated with
        the data we just wrote.
        If the calling function was not the cansas reader, return a minidom
        object rather than an lxml object.

        :param entry_node: lxml node ElementTree object to be appended to
        :param doc: entire xml tree
        """
        frm = inspect.stack()[1]
        mod_name = frm[1].replace("\\", "/").replace(".pyc", "")
        mod_name = mod_name.replace(".py", "")
        mod = mod_name.split("sas/")
        mod_name = mod[1]
        if mod_name != "dataloader/readers/cansas_reader":
            string = self.to_string(doc, pretty_print=False)
            doc = parseString(string)
            node_name = entry_node.tag
            node_list = doc.getElementsByTagName(node_name)
            entry_node = node_list.item(0)
        return entry_node

    def _to_xml_doc(self, datainfo):
        """
        Create an XML document to contain the content of a Data1D

        :param datainfo: Data1D object
        """
        if not issubclass(datainfo.__class__, Data1D):
            raise RuntimeError, "The cansas writer expects a Data1D instance"

        # Get PIs and create root element
        pi_string = self._get_pi_string()

        # Define namespaces and create SASroot object
        main_node = self._create_main_node()

        # Create ElementTree, append SASroot and apply processing instructions
        base_string = pi_string + self.to_string(main_node)
        base_element = self.create_element_from_string(base_string)
        doc = self.create_tree(base_element)

        # Create SASentry Element
        entry_node = self.create_element("SASentry")
        root = doc.getroot()
        root.append(entry_node)

        # Add Title to SASentry
        self.write_node(entry_node, "Title", datainfo.title)

        # Add Run to SASentry
        self._write_run_names(datainfo, entry_node)

        # Add Data info to SASEntry
        self._write_data(datainfo, entry_node)

        # Transmission Spectrum Info
        self._write_trans_spectrum(datainfo, entry_node)

        # Sample info
        self._write_sample_info(datainfo, entry_node)

        # Instrument info
        instr = self._write_instrument(datainfo, entry_node)

        #   Source
        self._write_source(datainfo, instr)

        #   Collimation
        self._write_collimation(datainfo, instr)

        #   Detectors
        self._write_detectors(datainfo, instr)

        # Processes info
        self._write_process_notes(datainfo, entry_node)

        # Note info
        self._write_notes(datainfo, entry_node)

        # Return the document, and the SASentry node associated with
        #      the data we just wrote
        # If the calling function was not the cansas reader, return a minidom
        #      object rather than an lxml object.
        entry_node = self._check_origin(entry_node, doc)

        return doc, entry_node

    def write_node(self, parent, name, value, attr=None):
        """
        :param doc: document DOM
        :param parent: parent node
        :param name: tag of the element
        :param value: value of the child text node
        :param attr: attribute dictionary

        :return: True if something was appended, otherwise False
        """
        if value is not None:
            parent = self.ebuilder(parent, name, value, attr)
            return True
        return False

    def write(self, filename, datainfo):
        """
        Write the content of a Data1D as a CanSAS XML file

        :param filename: name of the file to write
        :param datainfo: Data1D object
        """
        # Create XML document
        doc, _ = self._to_xml_doc(datainfo)
        # Write the file
        file_ref = open(filename, 'w')
        if self.encoding == None:
            self.encoding = "UTF-8"
        doc.write(file_ref, encoding=self.encoding,
                  pretty_print=True, xml_declaration=True)
        file_ref.close()

    # DO NOT REMOVE - used in saving and loading panel states.
    def _store_float(self, location, node, variable, storage, optional=True):
        """
        Get the content of a xpath location and store
        the result. Check that the units are compatible
        with the destination. The value is expected to
        be a float.

        The xpath location might or might not exist.
        If it does not exist, nothing is done

        :param location: xpath location to fetch
        :param node: node to read the data from
        :param variable: name of the data member to store it in [string]
        :param storage: data object that has the 'variable' data member
        :param optional: if True, no exception will be raised
            if unit conversion can't be done

        :raise ValueError: raised when the units are not recognized
        """
        entry = get_content(location, node)
        try:
            value = float(entry.text)
        except:
            value = None

        if value is not None:
            # If the entry has units, check to see that they are
            # compatible with what we currently have in the data object
            units = entry.get('unit')
            if units is not None:
                toks = variable.split('.')
                local_unit = None
                exec "local_unit = storage.%s_unit" % toks[0]
                if local_unit != None and units.lower() != local_unit.lower():
                    if HAS_CONVERTER == True:
                        try:
                            conv = Converter(units)
                            exec "storage.%s = %g" % \
                                (variable, conv(value, units=local_unit))
                        except:
                            _, exc_value, _ = sys.exc_info()
                            err_mess = "CanSAS reader: could not convert"
                            err_mess += " %s unit [%s]; expecting [%s]\n  %s" \
                                % (variable, units, local_unit, exc_value)
                            self.errors.append(err_mess)
                            if optional:
                                logging.info(err_mess)
                            else:
                                raise ValueError, err_mess
                    else:
                        err_mess = "CanSAS reader: unrecognized %s unit [%s];"\
                        % (variable, units)
                        err_mess += " expecting [%s]" % local_unit
                        self.errors.append(err_mess)
                        if optional:
                            logging.info(err_mess)
                        else:
                            raise ValueError, err_mess
                else:
                    exec "storage.%s = value" % variable
            else:
                exec "storage.%s = value" % variable

    # DO NOT REMOVE - used in saving and loading panel states.
    def _store_content(self, location, node, variable, storage):
        """
        Get the content of a xpath location and store
        the result. The value is treated as a string.

        The xpath location might or might not exist.
        If it does not exist, nothing is done

        :param location: xpath location to fetch
        :param node: node to read the data from
        :param variable: name of the data member to store it in [string]
        :param storage: data object that has the 'variable' data member

        :return: return a list of errors
        """
        entry = get_content(location, node)
        if entry is not None and entry.text is not None:
            exec "storage.%s = entry.text.strip()" % variable