"""
Contains convenience functions to create ready-made PlotRenderer
and PlotFrame instances of various types.
"""

from numpy import array, ndarray, transpose, cos, sin

# Local relative imports
from .abstract_data_source import AbstractDataSource
from .array_data_source import ArrayDataSource
from .axis import PlotAxis
from .barplot import BarPlot
from .data_range_1d import DataRange1D
from .grid import PlotGrid
from .linear_mapper import LinearMapper
from .scatterplot import ScatterPlot
from .polar_mapper import PolarMapper
from .lineplot import LinePlot
from .polar_line_renderer import PolarLineRenderer

def _create_data_sources(data, index_sort="none"):
    """
    Returns datasources for index and value based on the inputs.  Assumes that
    the index data is unsorted unless otherwise specified.
    """
    if (type(data) == ndarray) or (len(data) == 2):
        index, value = data
        if type(index) in (list, tuple, ndarray):
            index = ArrayDataSource(array(index), sort_order=index_sort)
        elif not isinstance(index, AbstractDataSource):
            raise RuntimeError("Need an array or list of values or a DataSource, got %s instead." % type(index))

        if type(value) in (list, tuple, ndarray):
            value = ArrayDataSource(array(value))
        elif not isinstance(value, AbstractDataSource):
            raise RuntimeError("Need an array or list of values or a DataSource, got %s instead." % type(index))

        return index, value
    else:
        raise RuntimeError("Unable to create datasources.")


def create_scatter_plot(data=[], index_bounds=None, value_bounds=None,
                        orientation="h", color="green", marker="square",
                        marker_size=4,
                        bgcolor="transparent", outline_color="black",
                        border_visible=True,
                        add_grid=False, add_axis=False,
                        index_sort="none"):
    """
    Creates a ScatterPlot from a single Nx2 data array or a tuple of
    two length-N 1-D arrays.  The data must be sorted on the index if any
    reverse-mapping tools are to be used.

    Pre-existing "index" and "value" datasources can be passed in.
    """

    index, value = _create_data_sources(data)

    if index_bounds is not None:
        index_range = DataRange1D(low=index_bounds[0], high=index_bounds[1])
    else:
        index_range = DataRange1D()
    index_range.add(index)
    index_mapper = LinearMapper(range=index_range)

    if value_bounds is not None:
        value_range = DataRange1D(low=value_bounds[0], high=value_bounds[1])
    else:
        value_range = DataRange1D()
    value_range.add(value)
    value_mapper = LinearMapper(range=value_range)

    plot = ScatterPlot(index=index, value=value,
                         index_mapper=index_mapper,
                         value_mapper=value_mapper,
                         orientation=orientation,
                         marker=marker,
                         marker_size=marker_size,
                         color=color,
                         bgcolor=bgcolor,
                         outline_color=outline_color,
                         border_visible=border_visible,)

    if add_grid:
        add_default_grids(plot, orientation)
    if add_axis:
        add_default_axes(plot, orientation)
    return plot


def create_line_plot(data=[], index_bounds=None, value_bounds=None,
                     orientation="h", color="red", width=1.0,
                     dash="solid", value_mapper_class=LinearMapper,
                     bgcolor="transparent", border_visible=False,
                     add_grid=False, add_axis=False,
                     index_sort="none"):

    index, value = _create_data_sources(data, index_sort)

    if index_bounds is not None:
        index_range = DataRange1D(low=index_bounds[0], high=index_bounds[1])
    else:
        index_range = DataRange1D()
    index_range.add(index)
    index_mapper = LinearMapper(range=index_range)

    if value_bounds is not None:
        value_range = DataRange1D(low=value_bounds[0], high=value_bounds[1])
    else:
        value_range = DataRange1D()
    value_range.add(value)
    value_mapper = value_mapper_class(range=value_range)

    plot = LinePlot(index=index, value=value,
                    index_mapper = index_mapper,
                    value_mapper = value_mapper,
                    orientation = orientation,
                    color = color,
                    bgcolor = bgcolor,
                    line_width = width,
                    line_style = dash,
                    border_visible=border_visible)

    if add_grid:
        add_default_grids(plot, orientation)
    if add_axis:
        add_default_axes(plot, orientation)
    return plot


def create_bar_plot(data=[], index_bounds=None, value_bounds=None,
                     orientation="h", color="red", bar_width=10.0,
                     value_mapper_class=LinearMapper,
                     line_color="black",
                     fill_color="red", line_width=1,
                     bgcolor="transparent", border_visible=False,
                     antialias=True,
                     add_grid=False, add_axis=False):

    index, value = _create_data_sources(data)

    if index_bounds is not None:
        index_range = DataRange1D(low=index_bounds[0], high=index_bounds[1])
    else:
        index_range = DataRange1D()
    index_range.add(index)
    index_mapper = LinearMapper(range=index_range)

    if value_bounds is not None:
        value_range = DataRange1D(low=value_bounds[0], high=value_bounds[1])
    else:
        value_range = DataRange1D()
    value_range.add(value)
    value_mapper = value_mapper_class(range=value_range)

    # Create the plot
    plot = BarPlot(index=index,
                    value=value,
                    value_mapper=value_mapper,
                    index_mapper=index_mapper,
                    orientation=orientation,
                    line_color=line_color,
                    fill_color=fill_color,
                    line_width=line_width,
                    bar_width=bar_width,
                    antialias=antialias,)

    if add_grid:
        add_default_grids(plot, orientation)
    if add_axis:
        add_default_axes(plot, orientation)
    return plot


def create_polar_plot(data, orientation='h', color='black', width=1.0,
                      dash="solid", grid="dot", value_mapper_class=PolarMapper):
    if (type(data) != ndarray) and (len(data) == 2):
        data = transpose(array(data))

    r_data, t_data = transpose(data)
    index_data= r_data*cos(t_data)
    value_data= r_data*sin(t_data)

    index = ArrayDataSource(index_data, sort_order='ascending')
    # Typically the value data is unsorted
    value = ArrayDataSource(value_data)

    index_range = DataRange1D()
    index_range.add(index)
    index_mapper = PolarMapper(range=index_range)

    value_range = DataRange1D()
    value_range.add(value)
    value_mapper = value_mapper_class(range=value_range)

    plot = PolarLineRenderer(index=index, value=value,
                    index_mapper = index_mapper,
                    value_mapper = value_mapper,
                    orientation = orientation,
                    color = color,
                    line_width = width,
                    line_style = dash,
                    grid_style = grid)

    return plot


def add_default_axes(plot, orientation="normal", vtitle="", htitle="",
                     axis_class=PlotAxis):
    """
    Creates left and bottom axes for a plot.  Assumes that the index is
    horizontal and value is vertical by default; set *orientation* to
    something other than "normal" if they are flipped.
    """
    if orientation in ("normal", "h"):
        v_mapper = plot.value_mapper
        h_mapper = plot.index_mapper
    else:
        v_mapper = plot.index_mapper
        h_mapper = plot.value_mapper

    left = axis_class(orientation='left',
                    title= vtitle,
                    mapper=v_mapper,
                    component=plot)

    bottom = axis_class(orientation='bottom',
                      title= htitle,
                      mapper=h_mapper,
                      component=plot)

    plot.underlays.append(left)
    plot.underlays.append(bottom)
    return left, bottom


def add_default_grids(plot, orientation="normal"):
    """
    Creates horizontal and vertical gridlines for a plot.  Assumes that the
    index is horizontal and value is vertical by default; set orientation to
    something other than "normal" if they are flipped.
    """
    if orientation in ("normal", "h"):
        v_mapper = plot.index_mapper
        h_mapper = plot.value_mapper
    else:
        v_mapper = plot.value_mapper
        h_mapper = plot.index_mapper

    vgrid = PlotGrid(mapper=v_mapper, orientation='vertical',
                     component=plot,
                     line_color="lightgray", line_style="dot")

    hgrid = PlotGrid(mapper=h_mapper, orientation='horizontal',
                     component=plot,
                     line_color="lightgray", line_style="dot")

    plot.underlays.append(vgrid)
    plot.underlays.append(hgrid)
    return hgrid, vgrid



# EOF