mirror of
https://github.com/ynput/ayon-core.git
synced 2026-01-01 16:34:53 +01:00
fixed redshift support
This commit is contained in:
Ondrej Samohel
parent
f572d97fe0
commit
022862d632
2 changed files with 190 additions and 87 deletions
|
|
@ -59,6 +59,7 @@ R_LAYER_TOKEN = re.compile(
|
|||
r'.*%l.*|.*<layer>.*|.*<renderlayer>.*', re.IGNORECASE)
|
||||
R_AOV_TOKEN = re.compile(r'.*%a.*|.*<aov>.*|.*<renderpass>.*', re.IGNORECASE)
|
||||
R_SUBSTITUTE_AOV_TOKEN = re.compile(r'%a|<aov>|<renderpass>', re.IGNORECASE)
|
||||
R_REMOVE_AOV_TOKEN = re.compile(r'_%a|_<aov>|_<renderpass>', re.IGNORECASE)
|
||||
R_SUBSTITUTE_LAYER_TOKEN = re.compile(
|
||||
r'%l|<layer>|<renderlayer>', re.IGNORECASE)
|
||||
R_SUBSTITUTE_CAMERA_TOKEN = re.compile(r'%c|<camera>', re.IGNORECASE)
|
||||
|
|
@ -160,6 +161,7 @@ class CollectMayaRender(pyblish.api.ContextPlugin):
|
|||
# return all expected files for all cameras and aovs in given
|
||||
# frame range
|
||||
exp_files = ExpectedFiles().get(renderer, layer_name)
|
||||
assert exp_files, ("no file names were generated, this is bug")
|
||||
|
||||
# if we want to attach render to subset, check if we have AOV's
|
||||
# in expectedFiles. If so, raise error as we cannot attach AOV
|
||||
|
|
@ -173,16 +175,32 @@ class CollectMayaRender(pyblish.api.ContextPlugin):
|
|||
full_exp_files = []
|
||||
aov_dict = {}
|
||||
|
||||
for aov, files in exp_files[0].items():
|
||||
# we either get AOVs or just list of files. List of files can
|
||||
# mean two things - there are no AOVs enabled or multipass EXR
|
||||
# is produced. In either case we treat those as `beauty`.
|
||||
if isinstance(exp_files[0], dict):
|
||||
for aov, files in exp_files[0].items():
|
||||
full_paths = []
|
||||
for ef in files:
|
||||
full_path = os.path.join(workspace, "renders", ef)
|
||||
full_path = full_path.replace("\\", "/")
|
||||
full_paths.append(full_path)
|
||||
aov_dict[aov] = full_paths
|
||||
else:
|
||||
full_paths = []
|
||||
for ef in files:
|
||||
for ef in exp_files:
|
||||
full_path = os.path.join(workspace, "renders", ef)
|
||||
full_path = full_path.replace("\\", "/")
|
||||
full_paths.append(full_path)
|
||||
aov_dict[aov] = full_paths
|
||||
aov_dict["beauty"] = full_paths
|
||||
|
||||
full_exp_files.append(aov_dict)
|
||||
|
||||
from pprint import pprint
|
||||
print("=" * 40)
|
||||
pprint(full_exp_files)
|
||||
print("=" * 40)
|
||||
|
||||
self.log.info("collecting layer: {}".format(layer_name))
|
||||
# Get layer specific settings, might be overrides
|
||||
data = {
|
||||
|
|
@ -363,7 +381,15 @@ class AExpectedFiles:
|
|||
def get_aovs(self):
|
||||
pass
|
||||
|
||||
def get_files(self):
|
||||
def get_renderer_prefix(self):
|
||||
try:
|
||||
file_prefix = cmds.getAttr(ImagePrefixes[self.renderer])
|
||||
except KeyError:
|
||||
raise UnsupportedRendererException(
|
||||
"Unsupported renderer {}".format(self.renderer))
|
||||
return file_prefix
|
||||
|
||||
def _get_layer_data(self):
|
||||
# ______________________________________________
|
||||
# ____________________/ ____________________________________________/
|
||||
# 1 - get scene name /__________________/
|
||||
|
|
@ -381,11 +407,7 @@ class AExpectedFiles:
|
|||
# __________________/ ______________________________________________/
|
||||
# 3 - image prefix /__________________/
|
||||
# __________________/
|
||||
try:
|
||||
file_prefix = cmds.getAttr(ImagePrefixes[renderer])
|
||||
except KeyError:
|
||||
raise UnsupportedRendererException(
|
||||
"Unsupported renderer {}".format(renderer))
|
||||
file_prefix = self.get_renderer_prefix()
|
||||
|
||||
if not file_prefix:
|
||||
raise RuntimeError("Image prefix not set")
|
||||
|
|
@ -397,6 +419,9 @@ class AExpectedFiles:
|
|||
# 4 - get renderable cameras_____________/
|
||||
# __________________/
|
||||
|
||||
# if we have <camera> token in prefix path we'll expect output for
|
||||
# every renderable camera in layer.
|
||||
|
||||
renderable_cameras = self.get_renderable_cameras()
|
||||
# ________________________________________________
|
||||
# __________________/ ______________________________________________/
|
||||
|
|
@ -404,11 +429,11 @@ class AExpectedFiles:
|
|||
# __________________/
|
||||
|
||||
enabled_aovs = self.get_aovs()
|
||||
from pprint import pprint
|
||||
print("-" * 40)
|
||||
pprint(enabled_aovs)
|
||||
print("-" * 40)
|
||||
|
||||
# if we have <camera> token in prefix path we'll expect output for
|
||||
# every renderable camera in layer.
|
||||
|
||||
expected_files = []
|
||||
layer_name = self.layer
|
||||
if self.layer.startswith("rs_"):
|
||||
layer_name = self.layer[3:]
|
||||
|
|
@ -417,62 +442,104 @@ class AExpectedFiles:
|
|||
frame_step = int(self.get_render_attribute('byFrameStep'))
|
||||
padding = int(self.get_render_attribute('extensionPadding'))
|
||||
|
||||
resolved_path = file_prefix
|
||||
if enabled_aovs:
|
||||
aov_file_list = {}
|
||||
for aov in enabled_aovs:
|
||||
for cam in renderable_cameras:
|
||||
scene_data = {
|
||||
"frameStart": start_frame,
|
||||
"frameEnd": end_frame,
|
||||
"frameStep": frame_step,
|
||||
"padding": padding,
|
||||
"cameras": renderable_cameras,
|
||||
"sceneName": scene_name,
|
||||
"layerName": layer_name,
|
||||
"renderer": renderer,
|
||||
"defaultExt": default_ext,
|
||||
"filePrefix": file_prefix,
|
||||
"enabledAOVs": enabled_aovs
|
||||
}
|
||||
return scene_data
|
||||
|
||||
mappings = (
|
||||
(R_SUBSTITUTE_SCENE_TOKEN, scene_name),
|
||||
(R_SUBSTITUTE_LAYER_TOKEN, layer_name),
|
||||
(R_SUBSTITUTE_CAMERA_TOKEN, cam),
|
||||
(R_SUBSTITUTE_AOV_TOKEN, aov[0])
|
||||
)
|
||||
def _generate_single_file_sequence(self, layer_data):
|
||||
expected_files = []
|
||||
file_prefix = layer_data["filePrefix"]
|
||||
for cam in layer_data["cameras"]:
|
||||
mappings = (
|
||||
(R_SUBSTITUTE_SCENE_TOKEN, layer_data["sceneName"]),
|
||||
(R_SUBSTITUTE_LAYER_TOKEN, layer_data["layerName"]),
|
||||
(R_SUBSTITUTE_CAMERA_TOKEN, cam),
|
||||
# this is required to remove unfilled aov token, for example
|
||||
# in Redshift
|
||||
(R_REMOVE_AOV_TOKEN, "")
|
||||
)
|
||||
|
||||
for regex, value in mappings:
|
||||
file_prefix = re.sub(regex, value, file_prefix)
|
||||
for regex, value in mappings:
|
||||
file_prefix = re.sub(regex, value, file_prefix)
|
||||
|
||||
aov_files = []
|
||||
for frame in range(
|
||||
int(start_frame),
|
||||
int(end_frame) + 1,
|
||||
int(frame_step)):
|
||||
aov_files.append(
|
||||
'{}.{}.{}'.format(
|
||||
file_prefix,
|
||||
str(frame).rjust(padding, "0"),
|
||||
aov[1]))
|
||||
for frame in range(
|
||||
int(layer_data["frameStart"]),
|
||||
int(layer_data["frameEnd"]) + 1,
|
||||
int(layer_data["frameStep"])):
|
||||
expected_files.append(
|
||||
'{}.{}.{}'.format(file_prefix,
|
||||
str(frame).rjust(
|
||||
layer_data["padding"], "0"),
|
||||
layer_data["defaultExt"]))
|
||||
return expected_files
|
||||
|
||||
# if we have more then one renderable camera, append
|
||||
# camera name to AOV to allow per camera AOVs.
|
||||
aov_name = aov[0]
|
||||
if len(renderable_cameras) > 1:
|
||||
aov_name = "{}_{}".format(aov[0], cam)
|
||||
def _generate_aov_file_sequences(self, layer_data):
|
||||
expected_files = []
|
||||
aov_file_list = {}
|
||||
file_prefix = layer_data["filePrefix"]
|
||||
for aov in layer_data["enabledAOVs"]:
|
||||
for cam in layer_data["cameras"]:
|
||||
|
||||
aov_file_list[aov_name] = aov_files
|
||||
file_prefix = resolved_path
|
||||
|
||||
expected_files.append(aov_file_list)
|
||||
else:
|
||||
for cam in renderable_cameras:
|
||||
mappings = (
|
||||
(R_SUBSTITUTE_SCENE_TOKEN, scene_name),
|
||||
(R_SUBSTITUTE_LAYER_TOKEN, layer_name),
|
||||
(R_SUBSTITUTE_CAMERA_TOKEN, cam)
|
||||
(R_SUBSTITUTE_SCENE_TOKEN, layer_data["sceneName"]),
|
||||
(R_SUBSTITUTE_LAYER_TOKEN, layer_data["layerName"]),
|
||||
(R_SUBSTITUTE_CAMERA_TOKEN, cam),
|
||||
(R_SUBSTITUTE_AOV_TOKEN, aov[0])
|
||||
)
|
||||
|
||||
for regex, value in mappings:
|
||||
file_prefix = re.sub(regex, value, file_prefix)
|
||||
|
||||
aov_files = []
|
||||
for frame in range(
|
||||
int(start_frame),
|
||||
int(end_frame) + 1,
|
||||
int(frame_step)):
|
||||
expected_files.append(
|
||||
'{}.{}.{}'.format(file_prefix,
|
||||
str(frame).rjust(padding, "0"),
|
||||
default_ext))
|
||||
int(layer_data["frameStart"]),
|
||||
int(layer_data["frameEnd"]) + 1,
|
||||
int(layer_data["frameStep"])):
|
||||
aov_files.append(
|
||||
'{}.{}.{}'.format(
|
||||
file_prefix,
|
||||
str(frame).rjust(layer_data["padding"], "0"),
|
||||
aov[1]))
|
||||
|
||||
# if we have more then one renderable camera, append
|
||||
# camera name to AOV to allow per camera AOVs.
|
||||
aov_name = aov[0]
|
||||
if len(layer_data["cameras"]) > 1:
|
||||
aov_name = "{}_{}".format(aov[0], cam)
|
||||
|
||||
aov_file_list[aov_name] = aov_files
|
||||
file_prefix = layer_data["filePrefix"]
|
||||
|
||||
expected_files.append(aov_file_list)
|
||||
return expected_files
|
||||
|
||||
def get_files(self):
|
||||
"""
|
||||
This method will return list of expected files.
|
||||
|
||||
It will translate render token strings ('<RenderPass>', etc.) to
|
||||
their values. This task is tricky as every renderer deals with this
|
||||
differently. It depends on `get_aovs()` abstract method implemented
|
||||
for every supported renderer.
|
||||
"""
|
||||
layer_data = self._get_layer_data()
|
||||
|
||||
expected_files = []
|
||||
if layer_data.get("enabledAOVs"):
|
||||
expected_files = self._generate_aov_file_sequences(layer_data)
|
||||
else:
|
||||
expected_files = self._generate_single_file_sequence(layer_data)
|
||||
|
||||
return expected_files
|
||||
|
||||
|
|
@ -656,13 +723,53 @@ class ExpectedFilesVray(AExpectedFiles):
|
|||
|
||||
class ExpectedFilesRedshift(AExpectedFiles):
|
||||
|
||||
# mapping redshift extension dropdown values to strings
|
||||
ext_mapping = ['iff', 'exr', 'tif', 'png', 'tga', 'jpg']
|
||||
|
||||
def __init__(self, layer):
|
||||
super(ExpectedFilesRedshift, self).__init__(layer)
|
||||
self.renderer = 'redshift'
|
||||
|
||||
def get_renderer_prefix(self):
|
||||
prefix = super(ExpectedFilesRedshift, self).get_renderer_prefix()
|
||||
prefix = "{}_<aov>".format(prefix)
|
||||
return prefix
|
||||
|
||||
def get_files(self):
|
||||
expected_files = super(ExpectedFilesRedshift, self).get_files()
|
||||
|
||||
# we need to add one sequence for plain beauty if AOVs are enabled.
|
||||
# as redshift output beauty without 'beauty' in filename.
|
||||
|
||||
layer_data = self._get_layer_data()
|
||||
if layer_data.get("enabledAOVs"):
|
||||
expected_files[0][u"beauty"] = self._generate_single_file_sequence(layer_data) # noqa: E501
|
||||
|
||||
return expected_files
|
||||
|
||||
def get_aovs(self):
|
||||
enabled_aovs = []
|
||||
default_ext = cmds.getAttr('defaultRenderGlobals.imfPluginKey')
|
||||
|
||||
try:
|
||||
if self.maya_is_true(
|
||||
cmds.getAttr("redshiftOptions.exrForceMultilayer")):
|
||||
# AOVs are merged in mutli-channel file
|
||||
print("*" * 40)
|
||||
print(cmds.getAttr("redshiftOptions.exrForceMultilayer"))
|
||||
print("*" * 40)
|
||||
return enabled_aovs
|
||||
except ValueError:
|
||||
# this occurs when Render Setting windows was not opened yet. In
|
||||
# such case there are no Arnold options created so query for AOVs
|
||||
# will fail. We terminate here as there are no AOVs specified then.
|
||||
# This state will most probably fail later on some Validator
|
||||
# anyway.
|
||||
print("+" * 40)
|
||||
return enabled_aovs
|
||||
|
||||
default_ext = self.ext_mapping[
|
||||
cmds.getAttr('redshiftOptions.imageFormat')
|
||||
]
|
||||
rs_aovs = [n for n in cmds.ls(type='RedshiftAOV')]
|
||||
|
||||
# todo: find out how to detect multichannel exr for redshift
|
||||
|
|
@ -674,7 +781,6 @@ class ExpectedFilesRedshift(AExpectedFiles):
|
|||
enabled = self.maya_is_true(override)
|
||||
|
||||
if enabled:
|
||||
# todo: find how redshift set format for AOVs
|
||||
enabled_aovs.append(
|
||||
(
|
||||
cmds.getAttr('%s.name' % aov),
|
||||
|
|
|
|||
|
|
@ -48,6 +48,13 @@ class ValidateRenderSettings(pyblish.api.InstancePlugin):
|
|||
'redshift': 'defaultRenderGlobals.imageFilePrefix'
|
||||
}
|
||||
|
||||
ImagePrefixTokens = {
|
||||
|
||||
'arnold': 'maya/<Scene>/<RenderLayer>/<RenderLayer>_<RenderPass',
|
||||
'redshift': 'maya/<Scene>/<RenderLayer>/<RenderLayer>',
|
||||
'vray': 'maya/<scene>/<Layer>/<Layer>'
|
||||
}
|
||||
|
||||
R_AOV_TOKEN = re.compile(
|
||||
r'%a|<aov>|<renderpass>', re.IGNORECASE)
|
||||
R_LAYER_TOKEN = re.compile(
|
||||
|
|
@ -99,26 +106,31 @@ class ValidateRenderSettings(pyblish.api.InstancePlugin):
|
|||
"doesn't have: '<renderlayer>' or "
|
||||
"'<layer>' token".format(prefix))
|
||||
|
||||
if not re.search(cls.R_AOV_TOKEN, prefix):
|
||||
invalid = True
|
||||
cls.log.error("Wrong image prefix [ {} ] - "
|
||||
"doesn't have: '<renderpass>' or "
|
||||
"'<aov>' token".format(prefix))
|
||||
|
||||
if len(cameras) > 1:
|
||||
if not re.search(cls.R_CAMERA_TOKEN, prefix):
|
||||
invalid = True
|
||||
cls.log.error("Wrong image prefix [ {} ] - "
|
||||
"doesn't have: '<camera>' token".format(prefix))
|
||||
|
||||
# renderer specific checks
|
||||
if renderer == "vray":
|
||||
if prefix.lower() != cls.VRAY_PREFIX.lower():
|
||||
cls.log.warning("warning: prefix differs from "
|
||||
"recommended {}".format(cls.VRAY_PREFIX))
|
||||
# no vray checks implemented yet
|
||||
pass
|
||||
elif renderer == "redshift":
|
||||
# no redshift check implemented yet
|
||||
pass
|
||||
else:
|
||||
if prefix.lower() != cls.DEFAULT_PREFIX.lower():
|
||||
cls.log.warning("warning: prefix differs from "
|
||||
"recommended {}".format(cls.DEFAULT_PREFIX))
|
||||
if not re.search(cls.R_AOV_TOKEN, prefix):
|
||||
invalid = True
|
||||
cls.log.error("Wrong image prefix [ {} ] - "
|
||||
"doesn't have: '<renderpass>' or "
|
||||
"token".format(prefix))
|
||||
|
||||
# prefix check
|
||||
if prefix.lower() != cls.ImagePrefixTokens[renderer].lower():
|
||||
cls.log.warning("warning: prefix differs from "
|
||||
"recommended {}".format(
|
||||
cls.ImagePrefixTokens[renderer]))
|
||||
|
||||
if padding != cls.DEFAULT_PADDING:
|
||||
invalid = True
|
||||
|
|
@ -127,21 +139,6 @@ class ValidateRenderSettings(pyblish.api.InstancePlugin):
|
|||
|
||||
return invalid
|
||||
|
||||
@classmethod
|
||||
def get_prefix(cls, renderer):
|
||||
prefix = cls.RENDERER_PREFIX.get(renderer, cls.DEFAULT_PREFIX)
|
||||
# maya.cmds and pymel.core return only default project directory and
|
||||
# not the current one but only default.
|
||||
output_path = os.path.join(
|
||||
mel.eval("workspace -q -rd;"), pm.workspace.fileRules["images"]
|
||||
)
|
||||
# Workfile paths can be configured to have host name in file path.
|
||||
# In this case we want to avoid duplicate folder names.
|
||||
if "maya" in output_path.lower():
|
||||
prefix = prefix.replace("maya/", "")
|
||||
|
||||
return prefix
|
||||
|
||||
@classmethod
|
||||
def repair(cls, instance):
|
||||
|
||||
|
|
@ -156,7 +153,7 @@ class ValidateRenderSettings(pyblish.api.InstancePlugin):
|
|||
node = render_attrs["node"]
|
||||
prefix_attr = render_attrs["prefix"]
|
||||
|
||||
fname_prefix = cls.get_prefix(renderer)
|
||||
fname_prefix = cls.ImagePrefixTokens[renderer]
|
||||
cmds.setAttr("{}.{}".format(node, prefix_attr),
|
||||
fname_prefix, type="string")
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue