utils.py

# ##### BEGIN GPL LICENSE BLOCK #####
#
#  This program is free software; you can redistribute it and/or
#  modify it under the terms of the GNU General Public License
#  as published by the Free Software Foundation; either version 2
#  of the License, or (at your option) any later version.
#
#  This program is distributed in the hope that it will be useful,
#  but WITHOUT ANY WARRANTY; without even the implied warranty of
#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
#  GNU General Public License for more details.
#
#  You should have received a copy of the GNU General Public License
#  along with this program; if not, write to the Free Software Foundation,
#  Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
#
# ##### END GPL LICENSE BLOCK #####


if "bpy" in locals():
    from importlib import reload

    paths = reload(paths)
    rerequests = reload(rerequests)

else:
    from blenderkit import paths, rerequests

import bpy
from mathutils import Vector
import json
import os
import sys
import shutil
import logging
import traceback
import inspect

bk_logger = logging.getLogger('blenderkit')

ABOVE_NORMAL_PRIORITY_CLASS = 0x00008000
BELOW_NORMAL_PRIORITY_CLASS = 0x00004000
HIGH_PRIORITY_CLASS = 0x00000080
IDLE_PRIORITY_CLASS = 0x00000040
NORMAL_PRIORITY_CLASS = 0x00000020
REALTIME_PRIORITY_CLASS = 0x00000100


def get_process_flags():
    flags = BELOW_NORMAL_PRIORITY_CLASS
    if sys.platform != 'win32':  # TODO test this on windows
        flags = 0
    return flags


def activate(ob):
    bpy.ops.object.select_all(action='DESELECT')
    ob.select_set(True)
    bpy.context.view_layer.objects.active = ob


def selection_get():
    aob = bpy.context.view_layer.objects.active
    selobs = bpy.context.view_layer.objects.selected[:]
    return (aob, selobs)


def selection_set(sel):
    bpy.ops.object.select_all(action='DESELECT')
    bpy.context.view_layer.objects.active = sel[0]
    for ob in sel[1]:
        ob.select_set(True)


def get_active_model():
    if bpy.context.view_layer.objects.active is not None:
        ob = bpy.context.view_layer.objects.active
        while ob.parent is not None:
            ob = ob.parent
        return ob
    return None


def get_active_HDR():
    scene = bpy.context.scene
    ui_props = scene.blenderkitUI
    image = ui_props.hdr_upload_image
    return image


def get_selected_models():
    '''
    Detect all hierarchies that contain asset data from selection. Only parents that have actual ['asset data'] get returned
    Returns
    list of objects containing asset data.

    '''
    obs = bpy.context.selected_objects[:]
    done = {}
    parents = []
    for ob in obs:
        if ob not in done:
            while ob.parent is not None and ob not in done and ob.blenderkit.asset_base_id == '' and ob.instance_collection is None:
                done[ob] = True
                ob = ob.parent

            if ob not in parents and ob not in done:
                if ob.blenderkit.name != '' or ob.instance_collection is not None:
                    parents.append(ob)
            done[ob] = True

    # if no blenderkit - like objects were found, use the original selection.
    if len(parents) == 0:
        parents = obs
    return parents


def get_selected_replace_adepts():
    '''
    Detect all hierarchies that contain either asset data from selection, or selected objects themselves.
    Returns
    list of objects for replacement.

    '''
    obs = bpy.context.selected_objects[:]
    done = {}
    parents = []
    for selected_ob in obs:
        ob = selected_ob
        if ob not in done:
            while ob.parent is not None and ob not in done and ob.blenderkit.asset_base_id == '' and ob.instance_collection is None:
                done[ob] = True
                # print('step,',ob.name)
                ob = ob.parent

            # print('fin', ob.name)
            if ob not in parents and ob not in done:
                if ob.blenderkit.name != '' or ob.instance_collection is not None:
                    parents.append(ob)

            done[ob] = True
    # print(parents)
    # if no blenderkit - like objects were found, use the original selection.
    if len(parents) == 0:
        parents = obs
    pprint('replace adepts')
    pprint(str(parents))
    return parents


def get_search_props():
    scene = bpy.context.scene
    if scene is None:
        return;
    uiprops = scene.blenderkitUI
    props = None
    if uiprops.asset_type == 'MODEL':
        if not hasattr(scene, 'blenderkit_models'):
            return;
        props = scene.blenderkit_models
    if uiprops.asset_type == 'SCENE':
        if not hasattr(scene, 'blenderkit_scene'):
            return;
        props = scene.blenderkit_scene
    if uiprops.asset_type == 'HDR':
        if not hasattr(scene, 'blenderkit_HDR'):
            return;
        props = scene.blenderkit_HDR
    if uiprops.asset_type == 'MATERIAL':
        if not hasattr(scene, 'blenderkit_mat'):
            return;
        props = scene.blenderkit_mat

    if uiprops.asset_type == 'TEXTURE':
        if not hasattr(scene, 'blenderkit_tex'):
            return;
        # props = scene.blenderkit_tex

    if uiprops.asset_type == 'BRUSH':
        if not hasattr(scene, 'blenderkit_brush'):
            return;
        props = scene.blenderkit_brush
    return props


def get_active_asset():
    scene = bpy.context.scene
    ui_props = scene.blenderkitUI
    if ui_props.asset_type == 'MODEL':
        if bpy.context.view_layer.objects.active is not None:
            ob = get_active_model()
            return ob
    if ui_props.asset_type == 'SCENE':
        return bpy.context.scene
    if ui_props.asset_type == 'HDR':
        return get_active_HDR()
    elif ui_props.asset_type == 'MATERIAL':
        if bpy.context.view_layer.objects.active is not None and bpy.context.active_object.active_material is not None:
            return bpy.context.active_object.active_material
    elif ui_props.asset_type == 'TEXTURE':
        return None
    elif ui_props.asset_type == 'BRUSH':
        b = get_active_brush()
        if b is not None:
            return b
    return None


def get_upload_props():
    scene = bpy.context.scene
    ui_props = scene.blenderkitUI
    if ui_props.asset_type == 'MODEL':
        if bpy.context.view_layer.objects.active is not None:
            ob = get_active_model()
            return ob.blenderkit
    if ui_props.asset_type == 'SCENE':
        s = bpy.context.scene
        return s.blenderkit
    if ui_props.asset_type == 'HDR':

        hdr = ui_props.hdr_upload_image#bpy.data.images.get(ui_props.hdr_upload_image)
        if not hdr:
            return None
        return hdr.blenderkit
    elif ui_props.asset_type == 'MATERIAL':
        if bpy.context.view_layer.objects.active is not None and bpy.context.active_object.active_material is not None:
            return bpy.context.active_object.active_material.blenderkit
    elif ui_props.asset_type == 'TEXTURE':
        return None
    elif ui_props.asset_type == 'BRUSH':
        b = get_active_brush()
        if b is not None:
            return b.blenderkit
    return None


def previmg_name(index, fullsize=False):
    if not fullsize:
        return '.bkit_preview_' + str(index).zfill(3)
    else:
        return '.bkit_preview_full_' + str(index).zfill(3)


def get_active_brush():
    context = bpy.context
    brush = None
    if context.sculpt_object:
        brush = context.tool_settings.sculpt.brush
    elif context.image_paint_object:  # could be just else, but for future possible more types...
        brush = context.tool_settings.image_paint.brush
    return brush


def load_prefs():
    user_preferences = bpy.context.preferences.addons['blenderkit'].preferences
    # if user_preferences.api_key == '':
    fpath = paths.BLENDERKIT_SETTINGS_FILENAME
    if os.path.exists(fpath):
        try:
            with open(fpath, 'r') as s:
                prefs = json.load(s)
                user_preferences.api_key = prefs.get('API_key', '')
                user_preferences.global_dir = prefs.get('global_dir', paths.default_global_dict())
                user_preferences.api_key_refresh = prefs.get('API_key_refresh', '')
        except Exception as e:
            print('failed to read addon preferences.')
            print(e)


def save_prefs(self, context):
    # first check context, so we don't do this on registration or blender startup
    if not bpy.app.background:  # (hasattr kills blender)
        user_preferences = bpy.context.preferences.addons['blenderkit'].preferences
        # we test the api key for length, so not a random accidentally typed sequence gets saved.
        lk = len(user_preferences.api_key)
        if 0 < lk < 25:
            # reset the api key in case the user writes some nonsense, e.g. a search string instead of the Key
            user_preferences.api_key = ''
            props = get_search_props()
            props.report = 'Login failed. Please paste a correct API Key.'

        prefs = {
            'API_key': user_preferences.api_key,
            'API_key_refresh': user_preferences.api_key_refresh,
            'global_dir': user_preferences.global_dir,
        }
        try:
            fpath = paths.BLENDERKIT_SETTINGS_FILENAME
            if not os.path.exists(paths._presets):
                os.makedirs(paths._presets)
            f = open(fpath, 'w')
            with open(fpath, 'w') as s:
                json.dump(prefs, s)
        except Exception as e:
            print(e)


def uploadable_asset_poll():
    '''returns true if active asset type can be uploaded'''
    ui_props = bpy.context.scene.blenderkitUI
    if ui_props.asset_type == 'MODEL':
        return bpy.context.view_layer.objects.active is not None
    if ui_props.asset_type == 'MATERIAL':
        return bpy.context.view_layer.objects.active is not None and bpy.context.active_object.active_material is not None
    if ui_props.asset_type == 'HDR':
        return ui_props.hdr_upload_image is not None
    return True

def get_hidden_texture(tpath, bdata_name, force_reload=False):
    i = get_hidden_image(tpath, bdata_name, force_reload=force_reload)
    bdata_name = f".{bdata_name}"
    t = bpy.data.textures.get(bdata_name)
    if t is None:
        t = bpy.data.textures.new('.test', 'IMAGE')
    if t.image != i:
        t.image = i
    return t


def get_hidden_image(tpath, bdata_name, force_reload=False, colorspace = 'sRGB'):
    if bdata_name[0] == '.':
        hidden_name = bdata_name
    else:
        hidden_name = '.%s' % bdata_name
    img = bpy.data.images.get(hidden_name)

    if tpath.startswith('//'):
        tpath = bpy.path.abspath(tpath)

    if img == None or (img.filepath != tpath):
        if tpath.startswith('//'):
            tpath = bpy.path.abspath(tpath)
        if not os.path.exists(tpath) or os.path.isdir(tpath):
            tpath = paths.get_addon_thumbnail_path('thumbnail_notready.jpg')

        if img is None:
            img = bpy.data.images.load(tpath)
            img.name = hidden_name
        else:
            if img.filepath != tpath:
                if img.packed_file is not None:
                    img.unpack(method='USE_ORIGINAL')

                img.filepath = tpath
                img.reload()
        img.colorspace_settings.name = colorspace
    elif force_reload:
        if img.packed_file is not None:
            img.unpack(method='USE_ORIGINAL')
        img.reload()
        img.colorspace_settings.name = colorspace
    return img


def get_thumbnail(name):
    p = paths.get_addon_thumbnail_path(name)
    name = '.%s' % name
    img = bpy.data.images.get(name)
    if img == None:
        img = bpy.data.images.load(p)
        img.colorspace_settings.name = 'sRGB'
        img.name = name
        img.name = name

    return img


def files_size_to_text(size):
    fsmb = size // (1024 * 1024)
    fskb = size % 1024
    if fsmb == 0:
        return f'{fskb}KB'
    else:
        return f'{fsmb}MB {fskb}KB'


def get_brush_props(context):
    brush = get_active_brush()
    if brush is not None:
        return brush.blenderkit
    return None


def p(text, text1='', text2='', text3='', text4='', text5='', level = 'DEBUG'):
    '''debug printing depending on blender's debug value'''

    if 1:#bpy.app.debug_value != 0:
        # print('-----BKit debug-----\n')
        # traceback.print_stack()
        texts = [text1,text2,text3,text4,text5]
        text = str(text)
        for t in texts:
            if t!= '':
                text += ' ' + str(t)

        bk_logger.debug(text)
        # print('---------------------\n')


def copy_asset(fp1, fp2):
    '''synchronizes the asset between folders, including it's texture subdirectories'''
    if 1:
        bk_logger.debug('copy asset')
        bk_logger.debug(fp1, fp2)
        if not os.path.exists(fp2):
            shutil.copyfile(fp1, fp2)
            bk_logger.debug('copied')
        source_dir = os.path.dirname(fp1)
        target_dir = os.path.dirname(fp2)
        for subdir in os.scandir(source_dir):
            if not subdir.is_dir():
                continue
            target_subdir = os.path.join(target_dir, subdir.name)
            if os.path.exists(target_subdir):
                continue
            bk_logger.debug(subdir, target_subdir)
            shutil.copytree(subdir, target_subdir)
            bk_logger.debug('copied')

    # except Exception as e:
    #     print('BlenderKit failed to copy asset')
    #     print(fp1, fp2)
    #     print(e)


def pprint(data, data1=None, data2=None, data3=None, data4=None):
    '''pretty print jsons'''
    p(json.dumps(data, indent=4, sort_keys=True))


def get_hierarchy(ob):
    '''get all objects in a tree'''
    obs = []
    doobs = [ob]
    # pprint('get hierarchy')
    pprint(ob.name)
    while len(doobs) > 0:
        o = doobs.pop()
        doobs.extend(o.children)
        obs.append(o)
    return obs


def select_hierarchy(ob, state=True):
    obs = get_hierarchy(ob)
    for ob in obs:
        ob.select_set(state)
    return obs


def delete_hierarchy(ob):
    obs = get_hierarchy(ob)
    bpy.ops.object.delete({"selected_objects": obs})


def get_bounds_snappable(obs, use_modifiers=False):
    # progress('getting bounds of object(s)')
    parent = obs[0]
    while parent.parent is not None:
        parent = parent.parent
    maxx = maxy = maxz = -10000000
    minx = miny = minz = 10000000

    s = bpy.context.scene

    obcount = 0  # calculates the mesh obs. Good for non-mesh objects
    matrix_parent = parent.matrix_world
    for ob in obs:
        # bb=ob.bound_box
        mw = ob.matrix_world
        subp = ob.parent
        # while parent.parent is not None:
        #     mw =

        if ob.type == 'MESH' or ob.type == 'CURVE':
            # If to_mesh() works we can use it on curves and any other ob type almost.
            # disabled to_mesh for 2.8 by now, not wanting to use dependency graph yet.
            depsgraph = bpy.context.evaluated_depsgraph_get()

            object_eval = ob.evaluated_get(depsgraph)
            if ob.type == 'CURVE':
                mesh = object_eval.to_mesh()
            else:
                mesh = object_eval.data

            # to_mesh(context.depsgraph, apply_modifiers=self.applyModifiers, calc_undeformed=False)
            obcount += 1
            if mesh is not None:
                for c in mesh.vertices:
                    coord = c.co
                    parent_coord = matrix_parent.inverted() @ mw @ Vector(
                        (coord[0], coord[1], coord[2]))  # copy this when it works below.
                    minx = min(minx, parent_coord.x)
                    miny = min(miny, parent_coord.y)
                    minz = min(minz, parent_coord.z)
                    maxx = max(maxx, parent_coord.x)
                    maxy = max(maxy, parent_coord.y)
                    maxz = max(maxz, parent_coord.z)
                # bpy.data.meshes.remove(mesh)
            if ob.type == 'CURVE':
                object_eval.to_mesh_clear()

    if obcount == 0:
        minx, miny, minz, maxx, maxy, maxz = 0, 0, 0, 0, 0, 0

    minx *= parent.scale.x
    maxx *= parent.scale.x
    miny *= parent.scale.y
    maxy *= parent.scale.y
    minz *= parent.scale.z
    maxz *= parent.scale.z

    return minx, miny, minz, maxx, maxy, maxz


def get_bounds_worldspace(obs, use_modifiers=False):
    # progress('getting bounds of object(s)')
    s = bpy.context.scene
    maxx = maxy = maxz = -10000000
    minx = miny = minz = 10000000
    obcount = 0  # calculates the mesh obs. Good for non-mesh objects
    for ob in obs:
        # bb=ob.bound_box
        mw = ob.matrix_world
        if ob.type == 'MESH' or ob.type == 'CURVE':
            depsgraph = bpy.context.evaluated_depsgraph_get()
            ob_eval = ob.evaluated_get(depsgraph)
            mesh = ob_eval.to_mesh()
            obcount += 1
            if mesh is not None:
                for c in mesh.vertices:
                    coord = c.co
                    world_coord = mw @ Vector((coord[0], coord[1], coord[2]))
                    minx = min(minx, world_coord.x)
                    miny = min(miny, world_coord.y)
                    minz = min(minz, world_coord.z)
                    maxx = max(maxx, world_coord.x)
                    maxy = max(maxy, world_coord.y)
                    maxz = max(maxz, world_coord.z)
            ob_eval.to_mesh_clear()

    if obcount == 0:
        minx, miny, minz, maxx, maxy, maxz = 0, 0, 0, 0, 0, 0
    return minx, miny, minz, maxx, maxy, maxz


def is_linked_asset(ob):
    return ob.get('asset_data') and ob.instance_collection != None


def get_dimensions(obs):
    minx, miny, minz, maxx, maxy, maxz = get_bounds_snappable(obs)
    bbmin = Vector((minx, miny, minz))
    bbmax = Vector((maxx, maxy, maxz))
    dim = Vector((maxx - minx, maxy - miny, maxz - minz))
    return dim, bbmin, bbmax


def requests_post_thread(url, json, headers):
    r = rerequests.post(url, json=json, verify=True, headers=headers)


def get_headers(api_key):
    headers = {
        "accept": "application/json",
    }
    if api_key != '':
        headers["Authorization"] = "Bearer %s" % api_key
    return headers


def scale_2d(v, s, p):
    '''scale a 2d vector with a pivot'''
    return (p[0] + s[0] * (v[0] - p[0]), p[1] + s[1] * (v[1] - p[1]))


def scale_uvs(ob, scale=1.0, pivot=Vector((.5, .5))):
    mesh = ob.data
    if len(mesh.uv_layers) > 0:
        uv = mesh.uv_layers[mesh.uv_layers.active_index]

        # Scale a UV map iterating over its coordinates to a given scale and with a pivot point
        for uvindex in range(len(uv.data)):
            uv.data[uvindex].uv = scale_2d(uv.data[uvindex].uv, scale, pivot)


# map uv cubic and switch of auto tex space and set it to 1,1,1
def automap(target_object=None, target_slot=None, tex_size=1, bg_exception=False, just_scale=False):
    s = bpy.context.scene
    mat_props = s.blenderkit_mat
    if mat_props.automap:
        tob = bpy.data.objects[target_object]
        # only automap mesh models
        if tob.type == 'MESH' and len(tob.data.polygons) > 0:
            # check polycount for a rare case where no polys are in editmesh
            actob = bpy.context.active_object
            bpy.context.view_layer.objects.active = tob

            # auto tex space
            if tob.data.use_auto_texspace:
                tob.data.use_auto_texspace = False

            if not just_scale:
                tob.data.texspace_size = (1, 1, 1)

            if 'automap' not in tob.data.uv_layers:
                bpy.ops.mesh.uv_texture_add()
                uvl = tob.data.uv_layers[-1]
                uvl.name = 'automap'

            # TODO limit this to active material
            # tob.data.uv_textures['automap'].active = True

            scale = tob.scale.copy()

            if target_slot is not None:
                tob.active_material_index = target_slot
            bpy.ops.object.mode_set(mode='EDIT')
            bpy.ops.mesh.select_all(action='DESELECT')

            # this exception is just for a 2.8 background thunmbnailer crash, can be removed when material slot select works...
            if bg_exception:
                bpy.ops.mesh.select_all(action='SELECT')
            else:
                bpy.ops.object.material_slot_select()

            scale = (scale.x + scale.y + scale.z) / 3.0
            if not just_scale:
                bpy.ops.uv.cube_project(
                    cube_size=scale * 2.0 / (tex_size),
                    correct_aspect=False)  # it's * 2.0 because blender can't tell size of a unit cube :)

            bpy.ops.object.editmode_toggle()
            tob.data.uv_layers.active = tob.data.uv_layers['automap']
            tob.data.uv_layers["automap"].active_render = True
            # this by now works only for thumbnail preview, but should be extended to work on arbitrary objects.
            # by now, it takes the basic uv map = 1 meter. also, it now doeasn't respect more materials on one object,
            # it just scales whole UV.
            if just_scale:
                scale_uvs(tob, scale=Vector((1 / tex_size, 1 / tex_size)))
            bpy.context.view_layer.objects.active = actob


def name_update():
    scene = bpy.context.scene
    ui_props = scene.blenderkitUI

    props = get_upload_props()
    if props.name_old != props.name:
        props.name_changed = True
        props.name_old = props.name
        nname = props.name.strip()
        nname = nname.replace('_', ' ')

        if nname.isupper():
            nname = nname.lower()
        nname = nname[0].upper() + nname[1:]
        props.name = nname
        # here we need to fix the name for blender data = ' or " give problems in path evaluation down the road.
    fname = props.name
    fname = fname.replace('\'', '')
    fname = fname.replace('\"', '')
    asset = get_active_asset()
    if ui_props.asset_type != 'HDR':
        # Here we actually rename assets datablocks, but don't do that with HDR's and possibly with others
        asset.name = fname


def get_param(asset_data, parameter_name):
    if not asset_data.get('parameters'):
        # this can appear in older version files.
        return None

    for p in asset_data['parameters']:
        if p.get('parameterType') == parameter_name:
            return p['value']
    return None



def params_to_dict(params):
    params_dict = {}
    for p in params:
        params_dict[p['parameterType']] = p['value']
    return params_dict


def dict_to_params(inputs, parameters=None):
    if parameters == None:
        parameters = []
    for k in inputs.keys():
        if type(inputs[k]) == list:
            strlist = ""
            for idx, s in enumerate(inputs[k]):
                strlist += s
                if idx < len(inputs[k]) - 1:
                    strlist += ','

            value = "%s" % strlist
        elif type(inputs[k]) != bool:
            value = inputs[k]
        else:
            value = str(inputs[k])
        parameters.append(
            {
                "parameterType": k,
                "value": value
            })
    return parameters

def update_tags(self, context):
    props = self

    commasep = props.tags.split(',')
    ntags = []
    for tag in commasep:
        if len(tag) > 19:
            short_tags = tag.split(' ')
            for short_tag in short_tags:
                if len(short_tag) > 19:
                    short_tag = short_tag[:18]
                ntags.append(short_tag)
        else:
            ntags.append(tag)
    if len(ntags) == 1:
        ntags = ntags[0].split(' ')
    ns = ''
    for t in ntags:
        if t != '':
            ns += t + ','
    ns = ns[:-1]
    if props.tags != ns:
        props.tags = ns

def user_logged_in():
    a = bpy.context.window_manager.get('bkit profile')
    if a is not None:
        return True
    return False


def profile_is_validator():
    a = bpy.context.window_manager.get('bkit profile')
    if a is not None and a['user'].get('exmenu'):
        return True
    return False


def guard_from_crash():
    '''Blender tends to crash when trying to run some functions with the addon going through unregistration process.'''
    if bpy.context.preferences.addons.get('blenderkit') is None:
        return False;
    if bpy.context.preferences.addons['blenderkit'].preferences is None:
        return False;
    return True


def get_largest_area(area_type='VIEW_3D'):
    maxsurf = 0
    maxa = None
    maxw = None
    region = None
    for w in bpy.data.window_managers[0].windows:
        for a in w.screen.areas:
            if a.type == area_type:
                asurf = a.width * a.height
                if asurf > maxsurf:
                    maxa = a
                    maxw = w
                    maxsurf = asurf

                    for r in a.regions:
                        if r.type == 'WINDOW':
                            region = r
    global active_area_pointer, active_window_pointer, active_region_pointer
    active_window_pointer = maxw.as_pointer()
    active_area_pointer = maxa.as_pointer()
    active_region_pointer = region.as_pointer()
    return maxw, maxa, region


def get_fake_context(context, area_type='VIEW_3D'):
    C_dict = {}  # context.copy() #context.copy was a source of problems - incompatibility with addons that also define context
    C_dict.update(region='WINDOW')

    try:
        context = context.copy()
        # print('bk context copied successfully')
    except Exception as e:
        print(e)
        print('BlenderKit: context.copy() failed. Can be a colliding addon.')
        context = {}

    if context.get('area') is None or context.get('area').type != area_type:
        w, a, r = get_largest_area(area_type=area_type)
        if w:
            # sometimes there is no area of the requested type. Let's face it, some people use Blender without 3d view.
            override = {'window': w, 'screen': w.screen, 'area': a, 'region': r}
            C_dict.update(override)
        # print(w,a,r)
    return C_dict


def label_multiline(layout, text='', icon='NONE', width=-1):
    ''' draw a ui label, but try to split it in multiple lines.'''
    if text.strip() == '':
        return
    lines = text.split('\n')
    if width > 0:
        threshold = int(width / 5.5)
    else:
        threshold = 35
    maxlines = 8
    li = 0
    for l in lines:
        while len(l) > threshold:
            i = l.rfind(' ', 0, threshold)
            if i < 1:
                i = threshold
            l1 = l[:i]
            layout.label(text=l1, icon=icon)
            icon = 'NONE'
            l = l[i:].lstrip()
            li += 1
            if li > maxlines:
                break;
        if li > maxlines:
            break;
        layout.label(text=l, icon=icon)
        icon = 'NONE'