message
stringlengths 13
484
| diff
stringlengths 38
4.63k
|
|---|---|
Minor improvements in torch approximator
using call instead of forward for the network
removed useless requires_grad_(False)
closes | @@ -98,7 +98,7 @@ class TorchApproximator(Serializable):
if not self._use_cuda:
torch_args = [torch.from_numpy(x) if isinstance(x, np.ndarray) else x
for x in args]
- val = self.network.forward(*torch_args, **kwargs)
+ val = self.network(*torch_args, **kwargs)
if output_tensor:
return val
@@ -109,8 +109,7 @@ class TorchApproximator(Serializable):
else:
torch_args = [torch.from_numpy(x).cuda()
if isinstance(x, np.ndarray) else x.cuda() for x in args]
- val = self.network.forward(*torch_args,
- **kwargs)
+ val = self.network(*torch_args, **kwargs)
if output_tensor:
return val
@@ -259,7 +258,7 @@ class TorchApproximator(Serializable):
else:
output_type = y_hat.dtype
- y = [y_i.clone().detach().requires_grad_(False).type(output_type) for y_i
+ y = [y_i.clone().detach().type(output_type) for y_i
in torch_args[-self._n_fit_targets:]]
if self._use_cuda:
|
Add SCRIPTS_ROOT to configuration.example.py
Fixes by adding the new variable to the example configuration. | @@ -154,6 +154,10 @@ PREFER_IPV4 = False
# this setting is derived from the installed location.
# REPORTS_ROOT = '/opt/netbox/netbox/reports'
+# The file path where custom scripts will be stored. A trailing slash is not needed. Note that the default value of
+# this setting is derived from the installed location.
+# SCRIPTS_ROOT = '/opt/netbox/netbox/scripts'
+
# By default, NetBox will store session data in the database. Alternatively, a file path can be specified here to use
# local file storage instead. (This can be useful for enabling authentication on a standby instance with read-only
# database access.) Note that the user as which NetBox runs must have read and write permissions to this path.
|
DOC: updated Constellation documentation
Updated the Constellation documentation to reflect the updated string output. | @@ -44,4 +44,6 @@ to create a Constellation of real-time solar wind data.
This last command will show that four :py:class:`~pysat._instrument.Instrument`
objects match the desired :py:attr:`platform` and :py:attr:`tag` criteria.
The :py:attr:`~pysat._instrument.Instrument.name` values are: :py:data:`epam`,
-:py:data:`mag`, :py:data:`sis`, and :py:data:`swepam`.
+:py:data:`mag`, :py:data:`sis`, and :py:data:`swepam`. The only
+:py:class:`~pysat._instrument.Instrument` defining attributes that are not
+unique are the :py:attr:`~pysat._instrument.Instrument.name` values.
|
TypeRepo.env_assoc: use T.defer_root_node instead of T.root_node
TN: | @@ -2880,12 +2880,10 @@ class TypeRepo(object):
EnvAssoc type, used to add associations of key and value to the lexical
environments, via the add_to_env primitive.
"""
- assert T.root_node
-
class EnvAssoc(StructType):
_fields = [
('key', UserField(type=Symbol)),
- ('val', UserField(type=T.root_node)),
+ ('val', UserField(type=self.defer_root_node)),
]
return EnvAssoc
|
Pulling out 3.2.2 version of Robot Framework from Github Actions
Latest builds are failing in somewhat unusual places. I want to see what
is the consitancy of runs for version 4.x, 5.x, and 6.x. | @@ -9,7 +9,7 @@ jobs:
strategy:
matrix:
python-version: [3.7, 3.9, pypy-3.7]
- rf-version: [3.2.2, 4.1.3, 5.0.1]
+ rf-version: [4.1.3, 5.0.1, 6.0.1]
steps:
- uses: actions/checkout@v3
|
Refactor "delete part category" dialog
Translations
Simplification | {% load i18n %}
{% block pre_form_content %}
-{% trans 'Are you sure you want to delete category' %} <strong>{{ category.name }}</strong>?
+
+<div class='alert alert-block alert-danger'>
+ {% trans "Are you sure you want to delete this part category?" %}
+</div>
{% if category.children.all|length > 0 %}
-<p>{% blocktrans with count=category.children.all|length%}This category contains {{count}} child categories{% endblocktrans %}.<br>
-{% trans 'If this category is deleted, these child categories will be moved to the' %}
+<div class='alert alert-block alert-warning'>
+ {% blocktrans with n=category.children.all|length %}This category contains {{ n }} child categories{% endblocktrans %}.<br>
{% if category.parent %}
- <strong>{{ category.parent.name }}</strong> {% trans 'category' %}.
+ {% blocktrans with category=category.parent.name %}If this category is deleted, these child categories will be moved to {{ category }}{% endblocktrans %}.
{% else %}
- {% trans 'top level Parts category' %}.
+ {% trans "If this category is deleted, these child categories will be moved to the top level part category" %}.
{% endif %}
-</p>
-
-<ul class='list-group'>
-{% for cat in category.children.all %}
- <li class='list-group-item'><strong>{{ cat.name }}</strong> - <em>{{ cat.description }}</em></li>
-{% endfor %}
-</ul>
+</div>
{% endif %}
{% if category.parts.all|length > 0 %}
-<p>{% blocktrans with count=category.parts.all|length %}This category contains {{count}} parts{% endblocktrans %}.<br>
+<div class='alert alert-block alert-warning'>
+ {% blocktrans with n=category.parts.all|length %}This category contains {{ n }} parts{% endblocktrans %}.<br>
{% if category.parent %}
- {% blocktrans with path=category.parent.pathstring %}If this category is deleted, these parts will be moved to the parent category {{path}}{% endblocktrans %}
+ {% blocktrans with category=category.parent.name %}If this category is deleted, these parts will be moved to {{ category }}{% endblocktrans %}.
{% else %}
- {% trans 'If this category is deleted, these parts will be moved to the top-level category Teile' %}
+ {% trans "If this category is deleted, these parts will be moved to the top level part category" %}.
{% endif %}
-</p>
-<ul class='list-group'>
-{% for part in category.parts.all %}
- <li class='list-group-item'><strong>{{ part.full_name }}</strong> - <em>{{ part.description }}</em></li>
-{% endfor %}
-</ul>
+</div>
{% endif %}
{% endblock %}
\ No newline at end of file
|
fix pipeline definition doc to reference solid
Summary: pipeline def was referencing `apply_op_three` instead of `apply_op`
Test Plan: none
Reviewers: max, sashank | @@ -114,8 +114,8 @@ def adder_resource(init_context):
pipeline_def = PipelineDefinition(
name='basic',
- solid_defs=[return_one, apply_op_three],
- dependencies={'apply_op_three': {'num': DependencyDefinition('return_one')}},
+ solid_defs=[return_one, apply_op],
+ dependencies={'apply_op': {'num': DependencyDefinition('return_one')}},
mode_defs=[add_mode],
preset_defs=[add_three_preset],
)
|
X-API-Key and Authorization: Bearer headers.
* MS teams sends Authorization: Bearer header. To use X-API-Key header with
teams (webhook) we need to look for X-API-Key if Authorization header
doesn't have Key. (this was briefly discussed in alerta/alerta-contrib#280)
* use startswith('Key ') instead of find('Key '). | @@ -24,7 +24,7 @@ def permission(scope=None):
def wrapped(*args, **kwargs):
# API Key (Authorization: Key <key>)
- if 'Authorization' in request.headers:
+ if 'Authorization' in request.headers and request.headers['Authorization'].startswith('Key '):
auth_header = request.headers['Authorization']
m = re.match(r'Key (\S+)', auth_header)
key = m.group(1) if m else None
|
Ctypes: Fix missing adaptations for C target type conversion in list
operations. | @@ -23,7 +23,8 @@ Right now only the creation is done here. But more should be added later on.
from .CodeHelpers import (
decideConversionCheckNeeded,
generateChildExpressionsCode,
- generateExpressionCode
+ generateExpressionCode,
+ withObjectCodeTemporaryAssignment
)
from .ErrorCodes import getErrorExitBoolCode, getErrorExitCode
from .PythonAPICodes import generateCAPIObjectCode
@@ -31,11 +32,13 @@ from .PythonAPICodes import generateCAPIObjectCode
def generateListCreationCode(to_name, expression, emit, context):
elements = expression.getElements()
-
assert elements
element_name = context.allocateTempName("list_element")
+ with withObjectCodeTemporaryAssignment(to_name, "list_extend_result", expression, emit, context) \
+ as result_name:
+
for count, element in enumerate(elements):
generateExpressionCode(
to_name = element_name,
@@ -48,12 +51,12 @@ def generateListCreationCode(to_name, expression, emit, context):
if count == 0:
emit(
"%s = PyList_New( %d );" % (
- to_name,
+ result_name,
len(elements)
)
)
- context.addCleanupTempName(to_name)
+ context.addCleanupTempName(result_name)
if not context.needsCleanup(element_name):
emit("Py_INCREF( %s );" % element_name)
@@ -62,7 +65,7 @@ def generateListCreationCode(to_name, expression, emit, context):
emit(
"PyList_SET_ITEM( %s, %d, %s );" % (
- to_name,
+ result_name,
count,
element_name
)
@@ -115,22 +118,25 @@ def generateListOperationExtendCode(to_name, expression, emit, context):
)
emit("assert( PyList_Check( %s ) );" % list_arg_name)
+
+ with withObjectCodeTemporaryAssignment(to_name, "list_extend_result", expression, emit, context) \
+ as result_name:
emit(
"%s = _PyList_Extend( (PyListObject *)%s, %s );" % (
- to_name,
+ result_name,
list_arg_name,
value_arg_name
)
)
getErrorExitCode(
- check_name = to_name,
+ check_name = result_name,
release_names = (list_arg_name, value_arg_name),
emit = emit,
context = context
)
- context.addCleanupTempName(to_name)
+ context.addCleanupTempName(result_name)
def generateListOperationPopCode(to_name, expression, emit, context):
@@ -140,23 +146,28 @@ def generateListOperationPopCode(to_name, expression, emit, context):
context = context
)
- # TODO: Have a dedicated helper instead, this could be more efficient.
+
emit("assert( PyList_Check( %s ) );" % list_arg_name)
+
+ with withObjectCodeTemporaryAssignment(to_name, "list_extend_result", expression, emit, context) \
+ as result_name:
+
+ # TODO: Have a dedicated helper instead, this could be more efficient.
emit(
'%s = PyObject_CallMethod( %s, (char *)"pop", NULL );' % (
- to_name,
+ result_name,
list_arg_name
)
)
getErrorExitCode(
- check_name = to_name,
+ check_name = result_name,
release_name = list_arg_name,
emit = emit,
context = context
)
- context.addCleanupTempName(to_name)
+ context.addCleanupTempName(result_name)
def generateBuiltinListCode(to_name, expression, emit, context):
|
Change the default source for the workflow fixture
When passing source=None, the workflow object will create a DummySource
in a newly created system tmpdir. Use the source_dir fixture instead, to
make sure unit tests reuse the pytest tmpdir rather than creating new
ones in /tmp. | @@ -14,6 +14,7 @@ import requests
import requests.exceptions
from atomic_reactor.constants import DOCKERFILE_FILENAME
from atomic_reactor.dirs import RootBuildDir
+from atomic_reactor.source import DummySource
from tests.constants import LOCALHOST_REGISTRY_HTTP, DOCKER0_REGISTRY_HTTP, TEST_IMAGE
from tests.util import uuid_value
@@ -66,8 +67,11 @@ def user_params(monkeypatch):
@pytest.fixture
-def workflow(build_dir, user_params):
- return DockerBuildWorkflow(RootBuildDir(build_dir), source=None)
+def workflow(build_dir, source_dir, user_params):
+ return DockerBuildWorkflow(
+ build_dir=RootBuildDir(build_dir),
+ source=DummySource(None, None, workdir=source_dir),
+ )
@pytest.mark.optionalhook
|
Fix GPT2 impl partially
TODO: still need to add `until` everywhere | @@ -17,19 +17,22 @@ class GPT2LM(LM):
return cls(device=args.get("device", "cpu"))
def generate(self, context, max_gen_length, truncate=True):
- context_tensor = torch.tensor([self.tokenizer.encode(context.strip())], dtype=torch.long).to(self.device)
+ # when too long to fit in context, truncate from the left
+ context_tensor = torch.tensor([self.tokenizer.encode(context.strip())[max_gen_length - 1024:]], dtype=torch.long).to(self.device)
res = self.gpt2.generate(
context_tensor,
+ # TODO: change to have until rather than using eos_token_id
eos_token_id=self.tokenizer.eos_token_id,
do_sample=False,
max_length=self.num_tokens(context) + max_gen_length,
)
# chop off the prompt and the final eos token
- return self.tokenizer.decode(res[0][len(context[0]):-1]).strip()
+ return self.tokenizer.decode(res[0][min(1024 - max_gen_length, len(context_tensor[0])):-1]).strip()
def loglikelihood(self, context, continuation, truncate=True):
- inp = torch.tensor([self.tokenizer.encode(context + continuation)], dtype=torch.long).to(self.device)
+ # when too long to fit in context, truncate from the left
+ inp = torch.tensor([self.tokenizer.encode(context + continuation)[-1024:]], dtype=torch.long).to(self.device)
ctxlen = len(self.tokenizer.encode(context.strip()))
cont_toks = inp[:, ctxlen:] # [batch, seq]
|
added pull request link to changelog
added entry to contributors list | ### Features
- Added warning to nodes selector if nothing was matched ([#2115](https://github.com/fishtown-analytics/dbt/issues/2115), [#2343](https://github.com/fishtown-analytics/dbt/pull/2343))
- Suport column descriptions for BigQuery models ([#2335](https://github.com/fishtown-analytics/dbt/issues/2335), [#2402](https://github.com/fishtown-analytics/dbt/pull/2402))
-- Added BigQuery option maximum_bytes_billed to set an upper limit for query costs ([#2346](https://github.com/fishtown-analytics/dbt/issues/2346))
+- Added BigQuery option maximum_bytes_billed to set an upper limit for query costs ([#2346](https://github.com/fishtown-analytics/dbt/issues/2346), [#2427](https://github.com/fishtown-analytics/dbt/pull/2427))
### Fixes
- When tracking is disabled due to errors, do not reset the invocation ID ([#2398](https://github.com/fishtown-analytics/dbt/issues/2398), [#2400](https://github.com/fishtown-analytics/dbt/pull/2400))
@@ -20,6 +20,7 @@ Contributors:
- [@azhard](https://github.com/azhard) [#2413](https://github.com/fishtown-analytics/dbt/pull/2413)
- [@mikaelene](https://github.com/mikaelene) [#2414](https://github.com/fishtown-analytics/dbt/pull/2414)
- [@raalsky](https://github.com/Raalsky) ([#2343](https://github.com/fishtown-analytics/dbt/pull/2343))
+ - [@haukeduden](https://github.com/haukeduden) ([#2427](https://github.com/fishtown-analytics/dbt/pull/2427))
## dbt 0.17.0b1 (May 5, 2020)
|
Create a mapping: kind constant -> ASTNodeType subclass
TN: | @@ -406,6 +406,13 @@ class CompileCtx(object):
:type: dict[langkit.compiled_types.ASTNodeType, int]
"""
+ self.kind_constant_to_node = {}
+ """
+ Reverse mapping for `node_kind_constants`.
+
+ :type: dict[int, langkit.compiled_types.ASTNodeType]
+ """
+
self._struct_types = []
"""
List of all plain struct types.
@@ -1710,6 +1717,7 @@ class CompileCtx(object):
start=1
):
self.node_kind_constants[astnode] = i
+ self.kind_constant_to_node[i] = astnode
def expose_public_api_types(self, astnode):
"""
|
Quality: Avoid temporary file from isorted
* This causes flicker to git and can be left over if interrupted
and it's not in the ignore list for git. | @@ -335,6 +335,7 @@ def _cleanupImportSortOrder(filename):
isort_call
+ [
"-q", # quiet, but stdout is still garbage
+ "--overwrite-in-place", # avoid using another temp file, this is already on one.
"-ot", # Order imports by type in addition to alphabetically
"-m3", # "vert-hanging"
"-tc", # Trailing commas
|
Update decisiontree.cu
remove printf from stateless | @@ -187,7 +187,6 @@ void decisionTreeClassifierPredict(const ML::cumlHandle &handle,
bool verbose) {
std::shared_ptr<DecisionTreeClassifier<float>> dt_classifier =
std::make_shared<DecisionTreeClassifier<float>>();
- dt_classifier->print(tree->root);
dt_classifier->predict(handle, tree, rows, n_rows, n_cols, predictions,
verbose);
}
|
Update requirements.txt after testing.
pandas.read_excel requires xlrd>=0.9.0 | -pytest==3.4.2
-pandas==0.23.1
-numpy==1.14.1
-setuptools==38.5.2
+pytest>=3.4.2
+pandas>=0.23.1
+numpy>=1.14.1
+setuptools>=38.5.2
sphinxcontrib-fulltoc==1.2.0
scikit-learn==0.19.1
+xlrd>=0.9.0
\ No newline at end of file
|
Fix shaft elements plot and add slenderness ratio parameter
This piece of code was missing in the first commit. It's a list with elements lenght, built just like "nodes_o_d" and "nodes_i_d", and has the same objective, provide data for the rotor plot. | @@ -259,6 +259,10 @@ class Rotor(object):
nodes_o_d.append(df_shaft["o_d"].iloc[-1])
self.nodes_o_d = nodes_o_d
+ nodes_le = list(df_shaft.groupby("n_l")["L"].min())
+ nodes_le.append(df_shaft["L"].iloc[-1])
+ self.nodes_le = nodes_le
+
self.nodes = list(range(len(self.nodes_pos)))
self.elements_length = [sh_el.L for sh_el in self.shaft_elements]
self.L = nodes_pos[-1]
|
UserField: document the "is_public" constructor argument
TN: | @@ -1041,6 +1041,9 @@ class UserField(AbstractField):
:type type: CompiledType
:type doc: str
+
+ :param bool is_public: Whether this field is public in the generated
+ APIs.
"""
super(UserField, self).__init__(repr, doc, type)
self._is_public = is_public
|
Update ctc.py
setting the --ctc_conf ignore_nan_grad= true like and suggested at | @@ -23,7 +23,7 @@ class CTC(torch.nn.Module):
dropout_rate: float = 0.0,
ctc_type: str = "builtin",
reduce: bool = True,
- ignore_nan_grad: bool = False,
+ ignore_nan_grad: bool = True,
):
assert check_argument_types()
super().__init__()
|
help text for container_format, disk_format
Updated the container_format and disk_format in v1 help text.
Closes-Bug: | @@ -28,8 +28,9 @@ from glanceclient.common import utils
from glanceclient import exc
import glanceclient.v1.images
-CONTAINER_FORMATS = 'Acceptable formats: ami, ari, aki, bare, and ovf.'
-DISK_FORMATS = ('Acceptable formats: ami, ari, aki, vhd, vmdk, raw, '
+CONTAINER_FORMATS = ('Acceptable formats: ami, ari, aki, bare, ovf, ova,'
+ 'docker.')
+DISK_FORMATS = ('Acceptable formats: ami, ari, aki, vhd, vdhx, vmdk, raw, '
'qcow2, vdi, iso, and ploop.')
DATA_FIELDS = ('location', 'copy_from', 'file')
|
Update contrib.py
update concat docs | @@ -243,7 +243,7 @@ def concat(arr, dim=0):
Example::
jt.concat([jt.array([[1],[2]]), jt.array([[2],[2]])], dim=1)
- # return [[1],[2],[2],[2]]
+ # return jt.Var([[1,2],[2,2]],dtype=int32)
'''
if not isinstance(arr, Sequence):
raise TypeError("concat arr needs to be a tuple or list")
|
validate: support obs repository
Otherwise, installation on SuSe fails.
Fixes: | - name: validate ceph_repository
fail:
- msg: "ceph_repository must be either 'community', 'rhcs', 'dev', 'custom' or 'uca'"
+ msg: "ceph_repository must be either 'community', 'rhcs', 'obs', 'dev', 'custom' or 'uca'"
when:
- ceph_origin == 'repository'
- - ceph_repository not in ['community', 'rhcs', 'dev', 'custom', 'uca']
+ - ceph_repository not in ['community', 'rhcs', 'obs', 'dev', 'custom', 'uca']
- name: validate ceph_repository_community
fail:
|
Update southern-california-earthquakes.yaml
Adding more tutorials | Name: Southern California Earthquake Data
Description: This dataset contains ground motion velocity and acceleration
seismic waveforms recorded by the Southern California Seismic Network (SCSN)
- and archived at the Southern California Earthquake Data Center (SCEDC).
+ and archived at the Southern California Earthquake Data Center (SCEDC). A
+ Distributed Acousting Sensing (DAS) dataset is included.
Documentation: https://scedc.caltech.edu/data/cloud.html
Contact: [email protected]
ManagedBy: "[Southern California Earthquake Data Center](https://scedc.caltech.edu)"
@@ -49,7 +50,17 @@ DataAtWork:
- Title: 'SeisNoise.jl GPU Computing Tutorial - Another example of accessing data s3://scedc-pds for ambient noise cross-correlation'
URL: https://nextjournal.com/thclements/seisnoisejl-gpu-computing-tutorial
AuthorName: Tim Clements
- AuthorURL: https://tclements.github.io/
+ - Title: 'Using Amazon API Gateway and Lambda to window and decimate waveforms'
+ URL: https://github.com/SCEDC/cloud/tree/master/pds-lambda-apis
+ AuthorName: Shang-Lin Chen
+ AuthorURL: https://scedc.caltech.edu
+ Services:
+ - Lambda
+ - Title: 'Jupyter Notebook tutorial on Ridgecrest DAS dataset'
+ URL: https://github.com/SCEDC/cloud/blob/master/pds_ridgecrest_das.ipynb
+ AuthorName: Jiuxun Yin
+
+
Publications:
- Title: Southern California Earthquake Data Now Available in the AWS Cloud
URL: https://pubs.geoscienceworld.org/ssa/srl/article-abstract/doi/10.1785/0220210039/600883/Southern-California-Earthquake-Data-Now-Available
|
Update agilent34410A.py
fixed spelling typo | @@ -45,5 +45,5 @@ class Agilent34410A(Instrument):
def __init__(self, adapter, delay=0.02, **kwargs):
super(Agilent34410A, self).__init__(
- adapter, "HP/Agilent/Keysight 34410A Multimiter", **kwargs
+ adapter, "HP/Agilent/Keysight 34410A Multimeter", **kwargs
)
|
Overload Concat.__repr__
TN: | @@ -643,3 +643,6 @@ class Concat(AbstractExpression):
return CallExpr('Concat_Result', 'Concat', array_1.type,
[array_1, array_2],
abstract_expr=self)
+
+ def __repr__(self):
+ return '<Concat>'
|
[bugfix] use page.site.data_repository when creating a _WbDataPage
use page.site.data_repository() instead of default Site().data_repository()
when creating a _WbDataPage. | @@ -943,7 +943,7 @@ class _WbDataPage(_WbRepresentation):
:param page: page containing the data
:param site: The Wikibase site
"""
- site = site or Site().data_repository()
+ site = site or page.site.data_repository()
specifics = type(self)._get_type_specifics(site)
_WbDataPage._validate(page, specifics['data_site'],
specifics['ending'], specifics['label'])
|
[commands] Handle nick mentions in HelpFormatter
Modifies the help formatter to handle nicknamed bot users for
mentions in clean_prefix | @@ -181,12 +181,12 @@ class HelpFormatter:
@property
def clean_prefix(self):
"""The cleaned up invoke prefix. i.e. mentions are ``@name`` instead of ``<@id>``."""
- user = self.context.bot.user
+ user = self.context.guild.me if self.context.guild else self.context.bot.user
# this breaks if the prefix mention is not the bot itself but I
# consider this to be an *incredibly* strange use case. I'd rather go
# for this common use case rather than waste performance for the
# odd one.
- return self.context.prefix.replace(user.mention, '@' + user.name)
+ return self.context.prefix.replace(user.mention, '@' + user.display_name)
def get_command_signature(self):
"""Retrieves the signature portion of the help page."""
|
Fix, need to reset return release mode when generating outline codes.
* This could lead to crashes when this kind of function was executed
in a return handler of a tried block. | @@ -779,10 +779,10 @@ def generateFunctionOutlineCode(to_name, expression, emit, context):
context = PythonFunctionOutlineContext(parent=context, outline=expression)
# Need to set return target, to assign to_name from.
- old_return_release_mode = context.getReturnReleaseMode()
return_target = context.allocateLabel("outline_result")
old_return_target = context.setReturnTarget(return_target)
+ old_return_release_mode = context.setReturnReleaseMode(False)
# TODO: Put the return value name as that to_name.c_type too.
|
Allow iterable params to column
Add type checking in QueryBuilder column methods
After apply this patch, list and tuple is allow to column (fixed | @@ -719,6 +719,9 @@ class QueryBuilder(Selectable, Term):
if self._insert_table is None:
raise AttributeError("'Query' object has no attribute '%s'" % "insert")
+ if isinstance(terms[0], (list, tuple)):
+ terms = terms[0]
+
for term in terms:
if isinstance(term, str):
term = Field(term, table=self._insert_table)
|
tools: Include `test_server_url.py` to be checked by mypy.
This commit adds `test_server_url.py` to the
`type_consistent_testfiles` list to check for type consistency
with mypy. | @@ -78,7 +78,8 @@ repo_python_files['tests'] = []
# Added incrementally as newer test files are type-annotated.
type_consistent_testfiles = [
- "test_run.py", "test_core.py", "test_emoji_data.py", "test_helper.py"
+ "test_run.py", "test_core.py", "test_emoji_data.py", "test_helper.py",
+ "test_server_url.py"
]
for file_path in python_files:
|
Add group permission check to Get BI dashboard
HG--
branch : feature/microservices | @@ -172,6 +172,7 @@ class BIAPI(API):
:return:
"""
user = self.handler.current_user
+ groups = user.groups.values_list("id", flat=True)
d = Dashboard.objects.filter(id=id).first()
if not d:
return None
@@ -181,8 +182,10 @@ class BIAPI(API):
for i in d.access:
if i.user == user and i.level >= access_level:
return d
+ elif i.group and i.group.id in groups and i.level >= access_level:
+ return d
# No access
- return None
+ raise APIError("User have no permission to access dashboard")
@executor("query")
@api
|
Fix, properly construct the path of extension modules.
* This was at least a problem on MacOS, it should be a fix for Linux
too. | @@ -632,7 +632,8 @@ static PyObject *loadModule(PyObject *module_name, struct Nuitka_MetaPathBasedLo
char filename[MAXPATHLEN + 1];
strcpy(filename, getBinaryDirectoryHostEncoded());
- filename[strlen(filename)] = SEP;
+ char const sep_str[2] = {SEP, 0};
+ strcat(filename, sep_str);
copyModulenameAsPath(filename + strlen(filename), entry->name);
strcat(filename, ".so");
|
Tests: Fixups for the distutils test runner
* Need to be able to both execute "runner" and "runner.exe" which
occurs for CPython too.
* Was not really doing the output of Nuitka when it said so.
* More fixups | @@ -87,10 +87,21 @@ for filename in sorted(os.listdir(".")):
'pip install "%s"' % (os.path.join(dist_dir, os.listdir(dist_dir)[0]))
)
+ runner_binary = os.path.join(
+ venv.getVirtualenvDir(),
+ "bin" if os.name != "nt" else "scripts",
+ "runner",
+ )
+
+ if os.path.exists(runner_binary):
# Need to call CPython binary for Windows.
process = subprocess.Popen(
args=[
- sys.executable,
+ os.path.join(
+ venv.getVirtualenvDir(),
+ "bin" if os.name != "nt" else "scripts",
+ "python",
+ ),
os.path.join(
venv.getVirtualenvDir(),
"bin" if os.name != "nt" else "scripts",
@@ -100,6 +111,14 @@ for filename in sorted(os.listdir(".")):
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
)
+ else:
+ assert os.path.exists(runner_binary + ".exe")
+
+ process = subprocess.Popen(
+ args=[runner_binary + ".exe"],
+ stdout=subprocess.PIPE,
+ stderr=subprocess.PIPE,
+ )
stdout_cpython, stderr_cpython = process.communicate()
exit_cpython = process.returncode
@@ -136,26 +155,41 @@ for filename in sorted(os.listdir(".")):
'pip install "%s"' % (os.path.join(dist_dir, os.listdir(dist_dir)[0]))
)
+ runner_binary = os.path.join(
+ venv.getVirtualenvDir(),
+ "bin" if os.name != "nt" else "scripts",
+ "runner",
+ )
+
+ if os.path.exists(runner_binary):
process = subprocess.Popen(
args=[
- sys.executable,
os.path.join(
venv.getVirtualenvDir(),
"bin" if os.name != "nt" else "scripts",
- "runner",
+ "python",
),
+ runner_binary,
],
stdout=subprocess.PIPE,
stderr=subprocess.PIPE,
)
+ else:
+ assert os.path.exists(runner_binary + ".exe")
+
+ process = subprocess.Popen(
+ args=[runner_binary + ".exe"],
+ stdout=subprocess.PIPE,
+ stderr=subprocess.PIPE,
+ )
stdout_nuitka, stderr_nuitka = process.communicate()
exit_nuitka = process.returncode
print("STDOUT Nuitka:")
- print(stdout_cpython)
+ print(stdout_nuitka)
print("STDERR Nuitka:")
- print(stderr_cpython)
+ print(stderr_nuitka)
assert exit_nuitka == 0, exit_nuitka
print("EXIT was OK.")
|
Fix date filters.
Handle lists of operations. Filter[date][eq]=day becomes [filter[date][gt]=day-1, filter[date][lt]=day+1]. | @@ -541,10 +541,12 @@ class ListFilterMixin(FilterMixin):
if filters:
for key, field_names in filters.iteritems():
for field_name, data in field_names.iteritems():
+ operations = data if isinstance(data, list) else [data]
+ for operation in operations:
if isinstance(queryset, list):
- queryset = self.get_filtered_queryset(field_name, data, queryset)
+ queryset = self.get_filtered_queryset(field_name, operation, queryset)
else:
- queryset = self.filter_by_field(queryset, field_name=field_name, operation=data)
+ queryset = self.filter_by_field(queryset, field_name, operation)
return queryset
def filter_by_field(self, queryset, field_name, operation):
@@ -581,7 +583,7 @@ class ListFilterMixin(FilterMixin):
if operation['value'] not in (list(), tuple()):
operation['source_field_name'] = '_contributors__guids___id'
operation['op'] = 'iexact'
- if operation['source_field_name'] == 'kind':
+ if field_name == 'kind':
operation['source_field_name'] = 'is_file'
# The value should be boolean
operation['value'] = operation['value'] == 'file'
|
Squashed commit of the following:
commit
Author: William Jones
Date: Mon Jul 8 14:34:18 2019 +0100
fix order of function parameters
commit
Author: William Jones
Date: Mon Jul 8 14:31:37 2019 +0100
add boolean for atomic writes | @@ -115,7 +115,7 @@ def create_dir(path):
def save_df_as_csv(df, path, filename,
- comment="", prepend_info=True,**kwargs, atomic=True):
+ comment="", prepend_info=True, atomic=True, **kwargs):
""" Save dataframe to a CSV file.
Parameters
|
Update README.md
Clean up quick start | @@ -22,18 +22,20 @@ See more on the [hypergan youtube](https://www.youtube.com/channel/UCU33XvBbMnS8
* [Quick start](#quick-start)
* [Requirements](#requirements)
* [Install](#install)
- * [Troubleshooting](#troubleshooting)
* [Train](#train)
- * [Development Mode](#development-mode)
* [The pip package hypergan](#the-pip-package-hypergan)
* [Training](#training)
* [Sampling](#sampling)
+ * [Additional Arguments](#additional-arguments)
* [Running on CPU](#running-on-cpu)
+ * [Troubleshooting](#troubleshooting)
+ * [Development Mode](#development-mode)
* [API](#api)
* [Using a trained hypergan model](#using-a-trained-hypergan-model)
* [Training a gan](#training-a-gan)
* [Examples](#examples)
* [Tutorials](#tutorials)
+
* [Datasets](#datasets)
* [Creating a Dataset](#creating-a-dataset)
* [Downloadable Datasets](#downloadable-datasets)
@@ -84,12 +86,6 @@ GPU: Nvidia, GTX 1080+ recommended
3. Join the community
* Once you've made something cool, be sure to share it on the Discord \([https://discord.gg/t4WWBPF](https://discord.gg/t4WWBPF)\).
-### Troubleshooting
-
-Make sure that your cuda, nvidia drivers, pillow, pytorch, and pytorch vision are the latest version.
-
-Check the discord for help.
-
### Create a new model
```bash
@@ -134,7 +130,7 @@ See all configuration templates with `--list-templates` or `-l`.
By default hypergan will not save training samples to disk. To change this, use `--save_samples`.
-### Arguments
+### Additional Arguments
To see a detailed list, run
@@ -142,6 +138,12 @@ To see a detailed list, run
hypergan -h
```
+### Troubleshooting
+
+Make sure that your cuda, nvidia drivers, pillow, pytorch, and pytorch vision are the latest version.
+
+Check the discord for help.
+
### Running on CPU
You can switch the backend with:
@@ -152,7 +154,6 @@ You can switch the backend with:
Don't train on CPU! It's too slow.
-
## API
```python
|
install ceph-mds packages on SUSE/openSUSE
install packages on SUSE/openSUSE distributions, using the
same logic as on RedHat-based distributions
Fixes | register: result
until: result is succeeded
-- name: install redhat ceph-mds package
+- name: install ceph-mds package on redhat or suse
package:
name: "ceph-mds"
state: "{{ (upgrade_ceph_packages|bool) | ternary('latest','present') }}"
until: result is succeeded
when:
- mds_group_name in group_names
- - ansible_os_family == 'RedHat'
+ - ansible_os_family in ['Suse', 'RedHat']
- name: create mds keyring
command: ceph --cluster {{ cluster }} --name client.bootstrap-mds --keyring /var/lib/ceph/bootstrap-mds/{{ cluster }}.keyring auth get-or-create mds.{{ mds_name }} osd 'allow rwx' mds 'allow' mon 'allow profile mds' -o /var/lib/ceph/mds/{{ cluster }}-{{ mds_name }}/keyring
|
Infraction Tests: Small fixes
Remove unnecessary space from placeholder
Rename `has_active_infraction` to `get_active_infraction` | @@ -17,11 +17,11 @@ class TruncationTests(unittest.IsolatedAsyncioTestCase):
self.guild = MockGuild(id=4567)
self.ctx = MockContext(bot=self.bot, author=self.user, guild=self.guild)
- @patch("bot.cogs.moderation.utils.has_active_infraction")
+ @patch("bot.cogs.moderation.utils.get_active_infraction")
@patch("bot.cogs.moderation.utils.post_infraction")
- async def test_apply_ban_reason_truncation(self, post_infraction_mock, has_active_mock):
+ async def test_apply_ban_reason_truncation(self, post_infraction_mock, get_active_mock):
"""Should truncate reason for `ctx.guild.ban`."""
- has_active_mock.return_value = False
+ get_active_mock.return_value = 'foo'
post_infraction_mock.return_value = {"foo": "bar"}
self.cog.apply_infraction = AsyncMock()
|
Make test_leaky_relu_inplace_with_neg_slope device-generic and skipIfRocm.
Summary:
Pull Request resolved:
Fixes
Test Plan: Imported from OSS | @@ -4108,18 +4108,6 @@ for shape in [(1,), ()]:
with self.assertRaisesRegex(RuntimeError, "must implement the backward"):
BadBw.apply(inp).sum().backward()
- def test_leaky_relu_inplace_with_neg_slope(self):
- for device in torch.testing.get_all_device_types():
- a = torch.tensor([-1., 1.], device=device, requires_grad=True)
- b = torch.nn.functional.leaky_relu_(a.clone(), -2)
- with self.assertRaisesRegex(RuntimeError, "call out-of-place version"):
- b.backward(torch.ones(2, device=device))
-
- a = torch.tensor([-1., 1.], device=device, requires_grad=True)
- b = torch.nn.functional.rrelu_(a.clone(), -5.0, 1.0)
- with self.assertRaisesRegex(RuntimeError, "call out-of-place version"):
- b.backward(torch.ones(2, device=device))
-
def test_custom_function_local_inplace(self):
class MyFn(torch.autograd.Function):
@staticmethod
@@ -5704,6 +5692,18 @@ class TestAutogradDeviceType(TestCase):
grad_cudnn, = torch.autograd.grad(loss_cudnn, log_probs, grad_out)
self.assertEqual(grad_cudnn, grad_native, prec=1e-4)
+ @skipCUDAIfRocm
+ def test_leaky_relu_inplace_with_neg_slope(self, device):
+ a = torch.tensor([-1., 1.], device=device, requires_grad=True)
+ b = torch.nn.functional.leaky_relu_(a.clone(), -2)
+ with self.assertRaisesRegex(RuntimeError, "call out-of-place version"):
+ b.backward(torch.ones(2, device=device))
+
+ a = torch.tensor([-1., 1.], device=device, requires_grad=True)
+ b = torch.nn.functional.rrelu_(a.clone(), -5.0, 1.0)
+ with self.assertRaisesRegex(RuntimeError, "call out-of-place version"):
+ b.backward(torch.ones(2, device=device))
+
@onlyCUDA
def test_free_unneeded_tensor(self, device):
x = torch.randn(2, 3, 10, 10, device=device, requires_grad=True)
|
Simplify column and info copying for table.
In particular, since parent_table and indices are never copied
internally, there is no need to save and reset them.
Also, as it was, the code set indices to an empty list even when not
present initially, thus changing a column being copied. | @@ -74,26 +74,15 @@ def col_copy(col, copy_indices=True):
if isinstance(col, BaseColumn):
return col.copy()
- # The new column should have None for the parent_table ref. If the
- # original parent_table weakref there at the point of copying then it
- # generates an infinite recursion. Instead temporarily remove the weakref
- # on the original column and restore after the copy in an exception-safe
- # manner.
-
- parent_table = col.info.parent_table
- indices = col.info.indices
- col.info.parent_table = None
- col.info.indices = []
-
- try:
newcol = col.copy() if hasattr(col, 'copy') else deepcopy(col)
newcol.info = col.info
- newcol.info.indices = deepcopy(indices or []) if copy_indices else []
+ if copy_indices and col.info.indices:
+ newcol.info.indices = deepcopy(col.info.indices)
for index in newcol.info.indices:
index.replace_col(col, newcol)
- finally:
- col.info.parent_table = parent_table
- col.info.indices = indices
+ else:
+ # TODO: bit strange that this is needed. Why not the default?
+ newcol.info.indices = []
return newcol
|
allow runway tests to exit 0 without signaling failure
This will check the exit code of tests so any that have called
sys.exit(0) (e.g. yamllint) won't trigger Runway to count it as failed. | @@ -78,6 +78,8 @@ class Test(BaseCommand): # pylint: disable=too-few-public-methods
# tool it is wrapping.
if not isinstance(err, SystemExit):
traceback.print_exc()
+ elif err.code == 0:
+ continue # Tests calling sys.exit(0) don't indicate failure
LOGGER.error('Test failed: %s', test.name)
if test.required:
LOGGER.error('Failed test was required, the remaining '
|
Fix test_delete_create_pvc_same_name
Fixed teardown mechanism using teardown_factory | @@ -11,28 +11,6 @@ from tests import helpers
logger = logging.getLogger(__name__)
[email protected]()
-def resources(request):
- """
- Delete the pvc resources and validate pv deletion created during the test
-
- Returns:
- list: empty list of pvcs
- """
- pvcs = []
-
- def finalizer():
- for instance in pvcs:
- instance.delete()
- instance.ocp.wait_for_delete(
- instance.name
- )
- assert helpers.validate_pv_delete(instance.backed_pv)
-
- request.addfinalizer(finalizer)
- return pvcs
-
-
@tier2
class TestDeleteCreatePVCSameName(ManageTest):
"""
@@ -56,28 +34,23 @@ class TestDeleteCreatePVCSameName(ManageTest):
]
)
def test_delete_create_pvc_same_name(
- self, interface, pvc_factory, resources
+ self, interface, pvc_factory, teardown_factory
):
"""
Delete PVC and create a new PVC with same name
"""
# Create a PVC
- pvcs = resources
pvc_obj1 = pvc_factory(
- interface=interface, status=constants.STATUS_BOUND
- )
- pvcs.append(pvc_obj1)
-
- # Check PV is Bound
- pv_obj1 = pvc_obj1.backed_pv_obj
- helpers.wait_for_resource_state(
- resource=pv_obj1, state=constants.STATUS_BOUND
+ interface=interface,
+ access_mode=constants.ACCESS_MODE_RWO,
+ status=constants.STATUS_BOUND
)
# Delete the PVC
logger.info(f"Deleting PVC {pvc_obj1.name}")
pvc_obj1.delete()
pvc_obj1.ocp.wait_for_delete(pvc_obj1.name)
+ logger.info(f"Deleted PVC {pvc_obj1.name}")
# Create a new PVC with same name
logger.info(f"Creating new PVC with same name {pvc_obj1.name}")
@@ -87,8 +60,8 @@ class TestDeleteCreatePVCSameName(ManageTest):
namespace=pvc_obj1.project.namespace,
do_reload=False
)
- assert pvc_obj2, "Failed to create PVC"
- pvcs.append(pvc_obj2)
+
+ teardown_factory(pvc_obj2)
# Check the new PVC and PV are Bound
helpers.wait_for_resource_state(
|
qt: Link dbus-1 not dbus-1d in Debug builds on Windows
Neither the Conan or the vanilla CMake-build version of the dbus
package use a d-suffix for the debug version of the library on
Windows.
Fixes | @@ -420,6 +420,10 @@ class QtConan(ConanFile):
" if (enable_precompiled_headers) {\n if (is_win) {",
" if (enable_precompiled_headers) {\n if (false) {"
)
+ tools.replace_in_file(os.path.join(self.source_folder, "qt5", "qtbase", "configure.json"),
+ "-ldbus-1d",
+ "-ldbus-1"
+ )
def _make_program(self):
if self._is_msvc:
|
Fixup some documentation references for new grain.equals function
versionadded
code-block notation | @@ -763,12 +763,15 @@ def equals(key, value):
'''
Used to make sure the minion's grain key/value matches.
- Returns ``True`` if matches otherwise ``False``
+ Returns ``True`` if matches otherwise ``False``.
+
+ .. versionadded:: Nitrogen
CLI Example:
- salt '*' grains.equals fqdn <expected_fqdn>
+ .. code-block:: bash
+ salt '*' grains.equals fqdn <expected_fqdn>
salt '*' grains.equals systemd:version 219
'''
return str(value) == str(get(key))
|
sql: Improve description of i_s.engines
It was improved to no longer show dummy data: | @@ -80,7 +80,7 @@ NULL.
### ENGINES table
-The ENGINES table provides information about storage engines. But it contains dummy data only. In the production environment, use the TiKV engine for TiDB.
+The ENGINES table provides information about storage engines. For compatibility, TiDB will always describe InnoDB as the only supported engine.
### EVENTS table
|
reference/tools: add a brief explanation for a default value in br doc
* sql: fix sequence docs
* Revert "sql: fix sequence docs"
This reverts commit
* reference/tools: add a brief explanation for a default value in br doc | @@ -154,7 +154,7 @@ Each of the above three sub-commands might still include the following three sub
To back up the cluster data, use the `br backup` command. You can add the `full` or `table` sub-command to specify the scope of your backup operation: the whole cluster or a single table.
-If the backup time might exceed the [`tikv_gc_life_time`](/reference/garbage-collection/configuration.md#tikv_gc_life_time) configuration which is `10m0s` by default, increase the value of this configuration.
+If the backup time might exceed the [`tikv_gc_life_time`](/reference/garbage-collection/configuration.md#tikv_gc_life_time) configuration which is `10m0s` by default (`10m0s` means 10 minutes), increase the value of this configuration.
For example, set `tikv_gc_life_time` to `720h`:
|
Only apply clip_values if available
Also removed some unnecessary `.detach()` calls. | @@ -134,7 +134,8 @@ class FeatureAdversariesPyTorch(EvasionAttack):
if self.random_start:
# Starting at a uniformly random point
adv = adv + torch.empty_like(adv).uniform_(-self.delta, self.delta)
- adv = torch.clamp(adv, *self.estimator.clip_values).detach()
+ if self.estimator.clip_values is not None:
+ adv = torch.clamp(adv, *self.estimator.clip_values)
if self.optimizer is None:
# run a plain-vanilla PGD
@@ -145,13 +146,12 @@ class FeatureAdversariesPyTorch(EvasionAttack):
loss.backward()
# pgd step
- perturbation = -adv.grad.data.sign() * self.step_size
- adv = adv + perturbation
- perturbation = torch.clamp(adv.data - x.data, -self.delta, self.delta)
- adv = torch.autograd.Variable(
- torch.clamp(x.data + perturbation, self.estimator.clip_values[0], self.estimator.clip_values[1]),
- requires_grad=True,
- )
+ adv.data = adv - adv.grad.detach().sign() * self.step_size
+ perturbation = torch.clamp(adv.detach() - x.detach(), -self.delta, self.delta)
+ adv.data = x.detach() + perturbation
+ if self.estimator.clip_values is not None:
+ adv.data = torch.clamp(adv.detach(), *self.estimator.clip_values)
+ adv.grad.zero_()
else:
# optimize soft constraint problem with chosen optimizer
opt = self.optimizer(params=[adv], **self._optimizer_kwargs) # type: ignore
@@ -159,7 +159,7 @@ class FeatureAdversariesPyTorch(EvasionAttack):
for _ in trange(self.max_iter, desc="Feature Adversaries PyTorch", disable=not self.verbose):
adv.requires_grad = True
# clip adversarial within L-Inf delta-ball
- adv.data = x.detach() + torch.clamp(adv.detach() - x.detach(), -self.delta, self.delta).detach()
+ adv.data = x.detach() + torch.clamp(adv.detach() - x.detach(), -self.delta, self.delta)
def closure():
if torch.is_grad_enabled():
@@ -172,7 +172,7 @@ class FeatureAdversariesPyTorch(EvasionAttack):
opt.step(closure)
if self.estimator.clip_values is not None:
- adv.data = torch.clamp(adv.detach(), *self.estimator.clip_values).detach()
+ adv.data = torch.clamp(adv.detach(), *self.estimator.clip_values)
return adv.detach().cpu()
def generate(self, x: np.ndarray, y: Optional[np.ndarray] = None, **kwargs) -> np.ndarray:
|
[IMPR] Reduce code complexity of pwb.main for further improvements
Introduce a new function find_filename to look for the given filename
and return its path. The code was moved from main().
This change will be used for further improvements. | @@ -168,26 +168,22 @@ except RuntimeError:
sys.exit(1)
-def main():
- """Command line entry point."""
- global filename
- if not filename:
- return False
+def find_filename(filename):
+ """Search for the filename in the given script paths."""
+ from pywikibot import config
- file_package = None
- argvu = pwb.argvu[1:]
- if not os.path.exists(filename):
script_paths = ['scripts',
'scripts.maintenance',
'scripts.userscripts']
- from pywikibot import config
+
if config.user_script_paths:
if isinstance(config.user_script_paths, (tuple, list)):
script_paths = config.user_script_paths + script_paths
else:
warn("'user_script_paths' must be a list or tuple,\n"
'found: {0}. Ignoring this setting.'
- ''.format(type(config.user_script_paths)))
+ .format(type(config.user_script_paths)))
+
for file_package in script_paths:
paths = file_package.split('.') + [filename]
testpath = os.path.join(_pwb_dir, *paths)
@@ -215,6 +211,22 @@ def main():
.format(' is' if len(similar_scripts) == 1
else 's are'))
print('\t' + '.py\n\t'.join(similar_scripts) + '.py')
+ return None
+ return filename
+
+
+def main():
+ """Command line entry point."""
+ global filename
+ if not filename:
+ return False
+
+ file_package = None
+ argvu = pwb.argvu[1:]
+
+ if not os.path.exists(filename):
+ filename = find_filename(filename)
+ if filename is None:
return True
# When both pwb.py and the filename to run are within the current
|
remove burn valid check
since this has its own disposition options no need to apply twice | @@ -1063,7 +1063,7 @@ class MediaProcessor:
def burnSubtitleFilter(self, inputfile, subtitle_streams, swl, valid_external_subs=None):
if self.settings.burn_subtitles:
- filtered_subtitle_streams = [x for x in subtitle_streams if self.validLanguage(x.metadata.get('language'), swl) and self.validDisposition(x.metadata.get('language'), x.dispostr, self.settings.ignored_subtitle_dispositions)]
+ filtered_subtitle_streams = [x for x in subtitle_streams if self.validLanguage(x.metadata.get('language'), swl)]
filtered_subtitle_streams = sorted(filtered_subtitle_streams, key=lambda x: swl.index(x.metadata.get('language')) if x.metadata.get('language') in swl else 999)
sub_candidates = []
if len(filtered_subtitle_streams) > 0:
|
Clean up change_parent matrix calculation
Causes less updates. | @@ -66,15 +66,16 @@ class Presentation(Matrices, Element, Generic[S]):
"""Change the parent and update the item's matrix so the item visualy
remains in the same place."""
old_parent = self.parent
- self.parent = new_parent
+ if new_parent is old_parent:
+ return
+ self.parent = new_parent
+ m = self.matrix
if old_parent:
- m = old_parent.matrix_i2c
- self.matrix.set(*self.matrix.multiply(m))
-
+ m = m * old_parent.matrix_i2c
if new_parent:
- m = new_parent.matrix_i2c.inverse()
- self.matrix.set(*self.matrix.multiply(m))
+ m = m * new_parent.matrix_i2c.inverse()
+ self.matrix.set(*m)
def load(self, name, value):
if name == "matrix":
@@ -120,7 +121,7 @@ class Presentation(Matrices, Element, Generic[S]):
if new_parent:
new_parent.matrix_i2c.add_handler(self._on_matrix_changed)
- self._on_matrix_changed(self.matrix, self.matrix.tuple())
+ self._on_matrix_changed(None, ())
def _on_matrix_changed(self, matrix, old_value):
if self.parent:
|
[BYOC][MergeComposite] if root->args[i] isn't a CallNode, then Donwcast<Call> will check fail
we needn't execute L131 "call_map->Set(arg, new_arg)", because when arg
is CallNode and root->args[i] is not CallNode, new_arg will be a null
pointer. There is no point in caching null pointer. | @@ -121,7 +121,7 @@ class MergeCompositeWrapper : public ExprMutator {
Array<Expr> new_args;
for (const auto& arg : pattern->args) {
Expr new_arg;
- if (arg->IsInstance<CallNode>()) {
+ if (arg->IsInstance<CallNode>() && root->args[i]->IsInstance<CallNode>()) {
new_arg =
ExtractPattern(Downcast<Call>(arg), Downcast<Call>(root->args[i]), var_map, call_map);
// if we've already processed this call node, return the previous result
|
Fix typo in contribution guide
If you're only exposure to using... -> If your only exposure to using... | @@ -81,7 +81,7 @@ Following that we'll tell you about how you can test your changes locally and th
# If you're using shells other than bash you'll need to use
pip install -e ".[test,examples,doc]"
```
- * If you're only exposure to using pip is `pip install package_name` then this might be a bit confusing.
+ * If your only exposure to using pip is `pip install package_name` then this might be a bit confusing.
* If we type `pip install -e .` (notice the 'dot'), this tells `pip` to install a package located here, in this directory, `.`.
The `-e` flag indicates that it should be editable, meaning you will not have to run `pip install .` every time you make a change and want to try it.
* Finally the `[test,examples,doc]` tells `pip` that there's some extra optional dependencies that we want to install.
|
Prevent file called `1` from being created during build
* Fix issue with shell redirection
A file called `1` is no longer created
* Remove trailing whitespace | @@ -83,12 +83,12 @@ endif
# Fall back to "-march=native" if the compiler doesn't support either of those.
ifeq ($(filter -march=%,$(CXXFLAGS)),)
- FAIL_A :=$(shell cp /dev/null a.cpp; $(CXX) -march=sandybridge -c a.cpp 2>1 || echo FAIL; rm -f a.cpp a.o)
+ FAIL_A :=$(shell cp /dev/null a.cpp; $(CXX) -march=sandybridge -c a.cpp 2>&1 || echo FAIL; rm -f a.cpp a.o)
ifeq ($(FAIL_A),)
CXXFLAGS += -march=sandybridge
else
# g++-4.8.x accepts "-march=corei7-avx" but not "-march=sandybridge "
- FAIL_B :=$(shell cp /dev/null a.cpp; $(CXX) -march=corei7-avx -c a.cpp 2>1 || echo FAIL; rm -f a.cpp a.o)
+ FAIL_B :=$(shell cp /dev/null a.cpp; $(CXX) -march=corei7-avx -c a.cpp 2>&1 || echo FAIL; rm -f a.cpp a.o)
ifeq ($(FAIL_B),)
CXXFLAGS += -march=corei7-avx
else
|
Add one more `# noqa` to unblock
Missed one. | import sys
from types import MappingProxyType
-from typing import ( # noqa: Y027
+from typing import ( # noqa: Y027,Y038
AbstractSet as Set,
AsyncGenerator as AsyncGenerator,
AsyncIterable as AsyncIterable,
|
latex can't recognize path seperator \\ in windows
Modify path before transport it to system command | import os
import hashlib
+from pathlib import Path
+
from manimlib.constants import TEX_TEXT_TO_REPLACE
from manimlib.constants import TEX_USE_CTEX
import manimlib.constants as consts
@@ -39,12 +41,15 @@ def generate_tex_file(expression, template_tex_file_body):
def tex_to_dvi(tex_file):
result = tex_file.replace(".tex", ".dvi" if not TEX_USE_CTEX else ".xdv")
+ result = Path(result).as_posix()
+ tex_file = Path(tex_file).as_posix()
+ tex_dir = Path(consts.TEX_DIR).as_posix()
if not os.path.exists(result):
commands = [
"latex",
"-interaction=batchmode",
"-halt-on-error",
- "-output-directory=\"{}\"".format(consts.TEX_DIR),
+ "-output-directory=\"{}\"".format(tex_dir),
"\"{}\"".format(tex_file),
">",
os.devnull
@@ -53,7 +58,7 @@ def tex_to_dvi(tex_file):
"-no-pdf",
"-interaction=batchmode",
"-halt-on-error",
- "-output-directory=\"{}\"".format(consts.TEX_DIR),
+ "-output-directory=\"{}\"".format(tex_dir),
"\"{}\"".format(tex_file),
">",
os.devnull
@@ -76,6 +81,8 @@ def dvi_to_svg(dvi_file, regen_if_exists=False):
where in the dvi
"""
result = dvi_file.replace(".dvi" if not TEX_USE_CTEX else ".xdv", ".svg")
+ result = Path(result).as_posix()
+ dvi_file = Path(dvi_file).as_posix()
if not os.path.exists(result):
commands = [
"dvisvgm",
|
changelog!
[skip-ci] | @@ -7,7 +7,12 @@ Changelog
v0.4.2
======
-Released: |today|
+Released: August 4, 2017
+
+New Features
+------------
+
+ * Packets with partial body lengths can now be parsed. For now, these packets are converted to have definite lengths instead. (#95) (#208)
Bugs Fixed
----------
@@ -15,6 +20,9 @@ Bugs Fixed
* PGPKey.decrypt was mistakenly using message.issuers instead of message.encrypters when determining whether or not the key was eligible
to attempt decrypting the message (#183)
* Fixed an issue with parsing some cleartext messages (#184)
+ * Fixed signing already-encrypted messages (encrypt-then-sign) (#186) (#191)
+ * PGP*.from_blob now correctly raises an exception if given zero-length input (#199) (#200)
+ * Fixed an issue where PGPKey.decrypt would fail with an arcane traceback if the key is passphrase-protected and not unlocked. (#204)
v0.4.1
======
|
Fix MapR dependency on mysql on RHEL
MapR is missing mysql-java-connector and that makes it necessary to have
subscription enable on RHEL7
Story: | @@ -75,10 +75,14 @@ validators:
- libtirpc
- libvisual
- libxslt
+ - mariadb
+ - mariadb-server
+ - mariadb-libs
- mesa-dri-drivers
- mesa-libGL
- mesa-libGLU
- mesa-private-llvm
+ - mysql-connector-java
- nmap-ncat
- numactl
- openjpeg-libs
|
qt send tab: when clicking "Max", show tooltip explaining max amt
follow | @@ -41,7 +41,7 @@ from typing import Optional, TYPE_CHECKING, Sequence, List, Union
import eth_abi
from PyQt5.QtGui import QPixmap, QKeySequence, QIcon, QCursor, QFont
-from PyQt5.QtCore import Qt, QRect, QStringListModel, QSize, pyqtSignal
+from PyQt5.QtCore import Qt, QRect, QStringListModel, QSize, pyqtSignal, QPoint
from PyQt5.QtWidgets import (QMessageBox, QComboBox, QSystemTrayIcon, QTabWidget,
QMenuBar, QFileDialog, QCheckBox, QLabel,
QVBoxLayout, QGridLayout, QLineEdit,
@@ -1478,6 +1478,18 @@ class ElectrumWindow(QMainWindow, MessageBoxMixin, Logger):
amount_after_all_fees = amount - x_fee_amount
self.amount_e.setAmount(amount_after_all_fees)
+ # show tooltip explaining max amount
+ mining_fee = tx.get_fee()
+ mining_fee_str = self.format_amount_and_units(mining_fee)
+ msg = _("Mining fee: {} (can be adjusted on next screen)").format(mining_fee_str)
+ if x_fee_amount:
+ twofactor_fee_str = self.format_amount_and_units(x_fee_amount)
+ msg += "\n" + _("2fa fee: {} (for the next batch of transactions)").format(twofactor_fee_str)
+ frozen_bal = self.get_frozen_balance_str()
+ if frozen_bal:
+ msg += "\n" + _("Some coins are frozen: {} (can be unfrozen in the Addresses or in the Coins tab)").format(frozen_bal)
+ QToolTip.showText(self.max_button.mapToGlobal(QPoint(0, 0)), msg)
+
def get_contact_payto(self, key):
_type, label = self.contacts.get(key)
return label + ' <' + key + '>' if _type == 'address' else key
@@ -1677,6 +1689,21 @@ class ElectrumWindow(QMainWindow, MessageBoxMixin, Logger):
"""
return self.utxo_list.get_spend_list()
+ def get_text_not_enough_funds_mentioning_frozen(self) -> str:
+ text = _("Not enough funds")
+ frozen_str = self.get_frozen_balance_str()
+ if frozen_str:
+ text += " ({} {})".format(
+ frozen_str, _("are frozen")
+ )
+ return text
+
+ def get_frozen_balance_str(self) -> Optional[str]:
+ frozen_bal = sum(self.wallet.get_frozen_balance())
+ if not frozen_bal:
+ return None
+ return self.format_amount_and_units(frozen_bal)
+
def pay_onchain_dialog(
self, inputs: Sequence[PartialTxInput],
outputs: List[PartialTxOutput], *,
|
fix: Now PyScript can import in the script
path. | @@ -23,10 +23,11 @@ class PyScript(Task):
def execute_action(self, **params):
root = str(Path(self.path).parent)
- task_func = self.get_task_func()
-
sys.path.append(root)
+
+ task_func = self.get_task_func()
output = task_func(**params)
+
sys.path.remove(root)
return output
|
Add tox environ to test without external utilities available.
Break PATH to simulate not having imagemagick, git, etc. installed. | @@ -10,6 +10,9 @@ setenv =
# Use per-testenv coverage files to prevent contention when parallel
# tests (using `tox -p`)
COVERAGE_FILE=.coverage.{envname}
+ # To test in environment without external utitilities like imagemagick and git installed,
+ # break PATH in noutils environment(s).
+ noutils: PATH=/dev/null
deps =
pytest>=6
pytest-click
|
CMSIS-DAPv2 USB match class catches UnicodeDecodeError.
This error is raised by certain versions of STLinkV2 that are known to
have a corrupted interface name. | @@ -268,8 +268,15 @@ class HasCmsisDapv2Interface(object):
try:
def match_cmsis_dap_interface_name(desc):
+ try:
interface_name = usb.util.get_string(desc.device, desc.iInterface)
return (interface_name is not None) and ("CMSIS-DAP" in interface_name)
+ except UnicodeDecodeError:
+ # This exception can be raised if the device has a corrupted interface name.
+ # Certain versions of STLinkV2 are known to have this problem. If we can't
+ # read the interface name, there's no way to tell if it's a CMSIS-DAPv2
+ # interface.
+ return False
config = dev.get_active_configuration()
cmsis_dap_interface = usb.util.find_descriptor(config, custom_match=match_cmsis_dap_interface_name)
|
Fix small typo in setup.py
Fixed small typo in setup.py | @@ -279,7 +279,7 @@ class build_deps(Command):
# This is not perfect solution as build does not depend on any of
# the auto-generated code and auto-generated files will not be
# included in this copy. If we want to use auto-generated files,
- # we need to find a batter way to do this.
+ # we need to find a better way to do this.
# More information can be found in conversation thread of PR #5772
self.copy_tree('torch/csrc', 'torch/lib/include/torch/csrc/')
|
Fix / update data path assertion
forgot to push this final change before i merged the last commit | @@ -696,11 +696,11 @@ class NeoXArgs(*BASE_CLASSES):
"in args "
# assert that if one of train/test/valid_data_path are provided, all should be
- assert_error_mess = "One of train/valid/test data_path is not provided\n"
- assert_error_mess += "\n".join(
- [f"{name}_data_path:{data_path}," for name, data_path in [self.train_data_path,
- self.valid_data_path,
- self.test_data_path]])
+ assert_error_mess = "One or more of train/valid/test data_path are not provided:\n\t"
+ assert_error_mess += "\n\t".join(
+ [f"{name}_data_path: {data_path}," for name, data_path in [['train', self.train_data_path],
+ ['valid', self.valid_data_path],
+ ['test', self.test_data_path]]]
assert any(has_separate_path) == all(has_separate_path), assert_error_mess
return True
|
Fixed gdbserver response to 'vCont?' command.
Search-replace error introduced during PEP8 rename. | @@ -760,7 +760,7 @@ class GDBServer(threading.Thread):
# v_cont capabilities query.
elif b'Cont?' == cmd:
- return self.create_rsp_packet(b"v_cont;c;C;s;S;t")
+ return self.create_rsp_packet(b"vCont;c;C;s;S;t")
# v_cont, thread action command.
elif cmd.startswith(b'Cont'):
|
communication: handle attachment case for both filename and fileid
some incoming files could have either filename or fileid for whatever reason. this handles both those cases (at least for now). | @@ -279,12 +279,16 @@ def prepare_to_notify(doc, print_html=None, print_format=None, attachments=None)
if isinstance(a, string_types):
# is it a filename?
try:
- # keep this for error handling
- _file = frappe.get_doc("File", a)
+ # check for both filename and file id
+ file_id = frappe.db.get_list('File', or_filters={'file_name': a, 'name': a}, limit=1)
+ if not file_id:
+ frappe.throw(_("Unable to find attachment {0}").format(a))
+ file_id = file_id[0]['name']
+ _file = frappe.get_doc("File", file_id)
_file.get_content()
# these attachments will be attached on-demand
# and won't be stored in the message
- doc.attachments.append({"fid": a})
+ doc.attachments.append({"fid": file_id})
except IOError:
frappe.throw(_("Unable to find attachment {0}").format(a))
else:
|
Bugfix htmlsafe function
Somehow the unicode characters intended were in fact the ascii
versions which negated the point of htmlsafe as it left the output
unsafe. This wasn't directly noticeable (as was the intention), so to
help avoid this in the future I've used the explicit forms. | @@ -52,10 +52,13 @@ def loads(object_: str, app: Optional["Quart"] = None, **kwargs: Any) -> Any:
def htmlsafe_dumps(object_: Any, **kwargs: Any) -> str:
- # Note in the below the ascii characters are replaced with a
- # unicode similar version.
- result = dumps(object_, **kwargs).replace("<", "<").replace(">", ">")
- return result.replace("&", "&").replace("'", "'")
+ return (
+ dumps(object_, **kwargs)
+ .replace("<", "\\u003c")
+ .replace(">", "\\u003e")
+ .replace("&", "\\u0026")
+ .replace("'", "\\u0027")
+ )
def tojson_filter(object_: Any, **kwargs: Any) -> Markup:
|
travis.yml: fix pytorch version compatibility bug
`conda install torchvision` will installs the latest one, and
it will update the pytorch to 0.3.0, which has some compatibility bugs.
Fix the torchvision verion to 0.1.8 for now. | @@ -20,7 +20,7 @@ before_install:
- pip install -r requirements.txt
# dependencies for im2text
- pip install Pillow
- - conda install torchvision
+ - conda install torchvision==0.1.8
# dependencies for speech2text
- sudo apt-get install -y sox libsox-dev libsox-fmt-all
- pip install librosa git+https://github.com/pytorch/audio
|
Fix checkFormat fail for null input
Patched in the function checkFormat itself,
instead of the code which calls it | @@ -344,6 +344,7 @@ def formatSubclassOf(fmt, cls, ontology, visited):
def checkFormat(actualFile, inputFormats, ontology):
# type: (Union[Dict[Text, Any], List, Text], Union[List[Text], Text], Graph) -> None
for af in aslist(actualFile):
+ if not af: continue
if "format" not in af:
raise validate.ValidationException(u"Missing required 'format' for File %s" % af)
for inpf in aslist(inputFormats):
@@ -551,7 +552,7 @@ class Process(object):
if self.formatgraph:
for i in self.tool["inputs"]:
d = shortname(i["id"])
- if d in builder.job and i.get("format") and builder.job[d]:
+ if d in builder.job and i.get("format"):
checkFormat(builder.job[d], builder.do_eval(i["format"]), self.formatgraph)
builder.bindings.extend(builder.bind_input(self.inputs_record_schema, builder.job))
|
linked-list: Implement extra-credit tests
add optional tests for users to implement __len__() and __iter__() | @@ -44,6 +44,24 @@ class LinkedListTests(unittest.TestCase):
self.assertEqual(50, self.list.pop())
self.assertEqual(30, self.list.shift())
+ @unittest.skip("extra-credit")
+ def test_length(self):
+ self.list.push(10)
+ self.list.push(20)
+ self.assertEqual(2, len(self.list))
+ self.list.shift()
+ self.assertEqual(1, len(self.list))
+ self.list.pop()
+ self.assertEqual(0, len(self.list))
+
+ @unittest.skip("extra-credit")
+ def test_iterator(self):
+ self.list.push(10)
+ self.list.push(20)
+ iterator = iter(self.list)
+ self.assertEqual(10, next(iterator))
+ self.assertEqual(20, next(iterator))
+
if __name__ == '__main__':
unittest.main()
|
Swap order of coverage actions in source
It would be nice to get the html report even if the coverage percent has
dropped below the failing threshold. | @@ -42,8 +42,8 @@ coverage:
python3 -m coverage --version
python3 -m coverage run --timid --branch -m pytest ./tests/whitebox/integration
python3 -m coverage run --timid --branch -a -m pytest ./tests/whitebox/monkey_patching/test_keyboard_interrupt.py
- python3 -m coverage report -m --fail-under=86 --show-missing --include="./src/*"
python3 -m coverage html --include="./src/*"
+ python3 -m coverage report -m --fail-under=86 --show-missing --include="./src/*"
keyboard-interrupt-test:
py.test-3 ${PYTEST_OPTS} ./tests/whitebox/monkey_patching/test_keyboard_interrupt.py
|
clipping values, and replacing transposes with permute
Without clipping, any values larger than 255 will be replaced with
int(v mod 256) by byte(), which results in high freq noise in image. | @@ -90,6 +90,6 @@ def save_image(tensor, filename, nrow=8, padding=2,
tensor = tensor.cpu()
grid = make_grid(tensor, nrow=nrow, padding=padding,
normalize=normalize, range=range, scale_each=scale_each)
- ndarr = grid.mul(255).byte().transpose(0, 2).transpose(0, 1).numpy()
+ ndarr = grid.mul(255).clip(0, 255).byte().permute(1, 2, 0).numpy()
im = Image.fromarray(ndarr)
im.save(filename)
|
Have CaseBlock leave date_opened unset unless explicitly passed in
Letting the default behavior kick in where date_opened automatically
is set to date_modified | @@ -35,8 +35,7 @@ class CaseBlock(object):
self.update = copy.copy(update) if update else {}
now = datetime.utcnow()
self.date_modified = date_modified or now
- self.date_opened = (now if create and date_opened is CaseBlock.undefined
- else date_opened)
+ self.date_opened = date_opened
self.case_type = "" if create and case_type is CaseBlock.undefined else case_type
self.case_name = "" if create and case_name is CaseBlock.undefined else case_name
|
Action is now recorded as event:PutEvents (fully qualified)
We no longer need to amend how the Action is reported in the describe_event_bus policy | @@ -235,7 +235,7 @@ class EventsBackend(BaseBackend):
'Sid': statement_id,
'Effect': 'Allow',
'Principal': {'AWS': 'arn:aws:iam::{0}:root'.format(data['principal'])},
- 'Action': 'events:{0}'.format(data['action']),
+ 'Action': data['action'],
'Resource': arn
})
return {
|
fix: return 0 if no users are found
if no users are present in the system, get_total_users returns None
instead of 0 causing comparison issues | @@ -841,11 +841,11 @@ def user_query(doctype, txt, searchfield, start, page_len, filters):
def get_total_users():
"""Returns total no. of system users"""
- return frappe.db.sql('''SELECT SUM(`simultaneous_sessions`)
+ return cint(frappe.db.sql('''SELECT SUM(`simultaneous_sessions`)
FROM `tabUser`
WHERE `enabled` = 1
AND `user_type` = 'System User'
- AND `name` NOT IN ({})'''.format(", ".join(["%s"]*len(STANDARD_USERS))), STANDARD_USERS)[0][0]
+ AND `name` NOT IN ({})'''.format(", ".join(["%s"]*len(STANDARD_USERS))), STANDARD_USERS)[0][0])
def get_system_users(exclude_users=None, limit=None):
if not exclude_users:
|
fix: Make infer dummy entity join key idempotent
fix: Infer dummy entity join key once | @@ -156,7 +156,11 @@ def update_feature_views_with_inferred_features_and_entities(
)
# Infer a dummy entity column for entityless feature views.
- if len(fv.entities) == 1 and fv.entities[0] == DUMMY_ENTITY_NAME:
+ if (
+ len(fv.entities) == 1
+ and fv.entities[0] == DUMMY_ENTITY_NAME
+ and not fv.entity_columns
+ ):
fv.entity_columns.append(Field(name=DUMMY_ENTITY_ID, dtype=String))
# Run inference for entity columns if there are fewer entity fields than expected.
|
Remove comment that is in conflict with the code
From git blame, this hasn't been the case for at least 2 years | @@ -43,8 +43,6 @@ domain_specific = [
url(r'^logo.png', logo, name='logo'),
url(r'^apps/', include('corehq.apps.app_manager.urls')),
url(r'^api/', include('corehq.apps.api.urls')),
- # the receiver needs to accept posts at an endpoint that might
- # not have a slash, so don't include it at the root urlconf
url(r'^receiver/', include('corehq.apps.receiverwrapper.urls')),
url(r'^settings/', include(settings_domain_specific)),
url(r'^enterprise/', include(accounting_domain_specific)),
|
Modify 'Click to expand' detail margin
Decrease the margin-top where a <detail> panel follows a paragraph, and increase the margin-bottom. | @@ -313,6 +313,10 @@ h1, h2, .rst-content .toctree-wrapper p.caption, h3, h4, h5, h6, legend {
line-height: 1.5;
}
+.wy-body-for-nav p + details {
+ margin: -10px 0 15px;
+}
+
.wy-body-for-nav blockquote {
margin: 1em 0;
padding-left: 1em;
|
Convert strs to bytes before decoding in tests
Fixes pypy3 test failure
Resolves | import base64
+import six
from globus_sdk.authorizers import BasicAuthorizer
from tests.framework import CapturedIOTestCase
@@ -24,7 +25,7 @@ class BasicAuthorizerTests(CapturedIOTestCase):
# confirm value
self.assertEqual(header_dict["Authorization"][:6], "Basic ")
decoded = base64.b64decode(
- header_dict["Authorization"][6:]).decode('utf-8')
+ six.b(header_dict["Authorization"][6:])).decode('utf-8')
self.assertEqual(decoded, "{0}:{1}".format(self.username,
self.password))
@@ -38,7 +39,7 @@ class BasicAuthorizerTests(CapturedIOTestCase):
# confirm values
self.assertEqual(header_dict["Authorization"][:6], "Basic ")
decoded = base64.b64decode(
- header_dict["Authorization"][6:]).decode('utf-8')
+ six.b(header_dict["Authorization"][6:])).decode('utf-8')
self.assertEqual(decoded, "{0}:{1}".format(self.username,
self.password))
self.assertEqual(header_dict["Header"], "value")
|
fix bug that gallery is none in generated envs
The gallery is created in __post_init__, but it's not called for a
generated environment. So update code to create a _gallery cache, and
update with gallery_raw.
a small fix that remove vhd secret, as it may block troubleshooting on
vhd path. | @@ -2,7 +2,7 @@ import json
import logging
import os
import re
-from dataclasses import dataclass, field
+from dataclasses import InitVar, dataclass, field
from datetime import datetime
from functools import lru_cache
from pathlib import Path
@@ -151,30 +151,32 @@ class AzureNodeSchema:
# for gallery image, which need to accept terms
purchase_plan: Optional[AzureVmPurchasePlanSchema] = None
- def __post_init__(self, *args: Any, **kwargs: Any) -> None:
- add_secret(self.vhd)
+ _gallery: InitVar[Optional[AzureVmGallerySchema]] = None
+ @property
+ def gallery(self) -> Optional[AzureVmGallerySchema]:
+ # this is a safe guard and prevent mypy error on typing
+ if not hasattr(self, "_gallery"):
+ self._gallery: Optional[AzureVmGallerySchema] = None
+ gallery: Optional[AzureVmGallerySchema] = self._gallery
+ if not gallery:
if isinstance(self.gallery_raw, dict):
- self.gallery: Optional[
- AzureVmGallerySchema
- ] = AzureVmGallerySchema.schema().load( # type: ignore
+ gallery = AzureVmGallerySchema.schema().load( # type: ignore
self.gallery_raw
)
# this step makes gallery_raw is validated, and filter out any unwanted
# content.
- self.gallery_raw = self.gallery.to_dict() # type: ignore
+ self.gallery_raw = gallery.to_dict() # type: ignore
elif self.gallery_raw:
assert isinstance(
self.gallery_raw, str
), f"actual: {type(self.gallery_raw)}"
- gallery = re.split(r"[:\s]+", self.gallery_raw.strip())
+ gallery_strings = re.split(r"[:\s]+", self.gallery_raw.strip())
- if len(gallery) == 4:
- self.gallery = AzureVmGallerySchema(
- gallery[0], gallery[1], gallery[2], gallery[3]
- )
+ if len(gallery_strings) == 4:
+ gallery = AzureVmGallerySchema(*gallery_strings)
# gallery_raw is used
- self.gallery_raw = self.gallery.to_dict() # type: ignore
+ self.gallery_raw = gallery.to_dict() # type: ignore
else:
raise LisaException(
f"Invalid value for the provided gallery "
@@ -183,12 +185,22 @@ class AzureNodeSchema:
f"'<Publisher> <Offer> <Sku> <Version>' "
f"or '<Publisher>:<Offer>:<Sku>:<Version>'"
)
+ self._gallery = gallery
+ return gallery
+
+ @gallery.setter
+ def gallery(self, value: Optional[AzureVmGallerySchema]) -> None:
+ self._status = value
+ if value is None:
+ self.gallery_raw = None
+ else:
+ self.gallery_raw = value.to_dict() # type: ignore
def get_image_name(self) -> str:
result = ""
if self.vhd:
result = self.vhd
- elif self.gallery_raw:
+ elif self.gallery:
assert isinstance(
self.gallery_raw, dict
), f"actual type: {type(self.gallery_raw)}"
|
bypassConvert adjustments
Logging now does not show options which are never created or used if bypass convert is trigger
optionsonly mode also reflects bypassConvert | @@ -86,10 +86,17 @@ class MediaProcessor:
options = None
preopts = None
postopts = None
+ outputfile = None
+ ripped_subs = []
+ downloaded_subs = []
info = info or self.isValidSource(inputfile)
if info:
+ if self.canBypassConvert(inputfile, info):
+ outputfile = inputfile
+ self.log.info("Bypassing conversion and setting outputfile to inputfile.")
+ else:
try:
options, preopts, postopts, ripsubopts, downloaded_subs = self.generateOptions(inputfile, info=info, original=original)
except:
@@ -114,13 +121,6 @@ class MediaProcessor:
self.log.exception("Unable to log options.")
ripped_subs = self.ripSubs(inputfile, ripsubopts)
-
- if self.canBypassConvert(inputfile, info, options):
- outputfile = inputfile
- for sub in downloaded_subs:
- self.log.debug("Removing downloaded sub %s for media that is not being converted." % (sub))
- self.removeFile(sub)
- else:
try:
outputfile, inputfile = self.convert(options, preopts, postopts, reportProgress)
except:
@@ -257,7 +257,11 @@ class MediaProcessor:
# Generate a JSON formatter dataset with the input and output information and ffmpeg command for a theoretical conversion
def jsonDump(self, inputfile, original=None):
dump = {}
- dump["input"] = self.generateSourceDict(inputfile)
+ dump["input"], info = self.generateSourceDict(inputfile)
+ if self.canBypassConvert(inputfile, info):
+ dump["output"] = dump["input"]
+ dump["output"]["bypassConvert"] = True
+ else:
dump["output"], dump["preopts"], dump["postopts"], dump["ripsubopts"], dump["downloadedsubs"] = self.generateOptions(inputfile, original)
parsed = self.converter.parse_options(dump["output"])
input_dir, filename, input_extension = self.parseFile(inputfile)
@@ -287,7 +291,7 @@ class MediaProcessor:
output.update(probe.toJson)
else:
output['error'] = "Invalid input, unable to read"
- return output
+ return output, probe
# Pass over audio and subtitle streams to ensure the language properties are safe, return any adjustments made to SWL/AWL if relax is enabled
def safeLanguage(self, info):
@@ -1101,7 +1105,7 @@ class MediaProcessor:
return True
return False
- def canBypassConvert(self, inputfile, info, options):
+ def canBypassConvert(self, inputfile, info):
# Process same extensions
if self.settings.output_extension == self.parseFile(inputfile)[2]:
if not self.settings.force_convert and not self.settings.process_same_extensions:
@@ -1110,13 +1114,6 @@ class MediaProcessor:
elif info.format.metadata.get('encoder', '').startswith('sma') and not self.settings.force_convert:
self.log.info("Input and output extensions match and the file appears to have already been processed by SMA, enable force-convert to override [force-convert: %s]." % self.settings.force_convert)
return True
- '''
- elif not self.settings.force_convert and len([x for x in [options['video']] + [x for x in options['audio']] + [x for x in options['subtitle']] if x['codec'] != 'copy']) == 0:
- self.log.info("Input and output extensions match and every codec is copy, this file probably doesn't need conversion, returning [force-convert: %s]." % self.settings.force_convert)
- return True
- elif self.settings.force_convert:
- self.log.info("Input and output extensions match and every codec is copy, this file probably doesn't need conversion, but conversion being forced [force-convert: %s]." % self.settings.force_convert)
- '''
return False
# Encode a new file based on selected options, built in naming conflict resolution
|
handle existing phenotype terms
log better | @@ -26,13 +26,16 @@ def update_variants(adapter, case_obj, old_variants):
for variant in old_variants:
new_variant = Variant.objects(variant_id=variant['variant_id']).first()
if new_variant is None:
- logger.warning("missing variant: %s", variant['variant_id'])
+ logger.warn("missing variant: %s", variant['variant_id'])
continue
if variant.get('manual_rank'):
+ logger.info("updating manual rank: %s", variant['manual_rank'])
new_variant.manual_rank = variant['manual_rank']
if variant.get('sanger_ordered') is True:
+ logger.info("marking sanger order")
new_variant.sanger_ordered = True
if new_variant not in case_obj.suspects:
+ logger.info("adding suspect: %s", variant['variant_id'])
new_variant.suspects.append(new_variant)
yield new_variant
@@ -118,19 +121,19 @@ def update_case(adapter, case_obj, exported_data):
if phenotype_terms:
logger.info("Updating phenotype terms")
existing_terms = case_obj.phenotype_terms
- if len(existing_terms) == 0:
- for term in phenotype_terms:
- hpo_obj = adapter.hpo_term(term)
+ existing_ids = set(term.phenotype_id for term in existing_terms)
+ for hpo_id in phenotype_terms:
+ if hpo_id in existing_ids:
+ logger.info("term already added: %s", hpo_id)
+ else:
+ hpo_obj = adapter.hpo_term(hpo_id)
if hpo_obj:
- new_term = PhenotypeTerm(phenotype_id=term,
+ new_term = PhenotypeTerm(phenotype_id=hpo_id,
feature=hpo_obj.description)
- logger.info("Adding term %s", term)
- existing_terms.append(new_term)
- else:
- logger.info("Could not find term %s", term)
- case_obj.phenotype_terms = existing_terms
+ logger.info("Adding term %s", hpo_id)
+ case_obj.phenotype_terms.append(new_term)
else:
- logger.info("Case was already phenotyped")
+ logger.info("Could not find term %s", hpo_id)
# Update the phenotype groups
# If the case has a phenotype group, skip to add the groups
@@ -189,6 +192,7 @@ def update_cases(context, exported_cases):
new_events = update_events(adapter, case_obj,
exported_info['events'])
for new_event in new_events:
+ logger.info("adding new event: %s", new_event.verb)
new_event.save()
case_obj.is_migrated = True
case_obj.save()
|
Update README.md
Simple, minor typo in the Readme. | <img height="120px" align="right" src="/static/mythril.png"/>
-Mythril is a security analysis tool for Ethereum smart contracts. I uses concolic analysis, taint analysis and control flow checking to detect a variety of security vulnerabilities. The analysis is based on [laser-ethereum](https://github.com/b-mueller/laser-ethereum), a symbolic execution library for EVM bytecode.
+Mythril is a security analysis tool for Ethereum smart contracts. It uses concolic analysis, taint analysis and control flow checking to detect a variety of security vulnerabilities. The analysis is based on [laser-ethereum](https://github.com/b-mueller/laser-ethereum), a symbolic execution library for EVM bytecode.
## Installation and setup
|
docs: tutorials: models: slr: Document that hyperparameters live in model config
Related: | @@ -62,6 +62,9 @@ Anything that a user might want to tweak about a models behavior should go in
the ``Config`` class for the model. The naming convention is ``TheName`` +
``Model`` + ``Config``.
+Hyperparameters for a model should live inside the model's config. Ideally at
+the top level and not nested within another structure.
+
Our model has three configurable properties.
- ``features``. A list of :py:class:`Feature <dffml.feature.Feature>` objects
|
Updated acknowledgements file to change old username
I recently moved from to and this commit fixes that in the acknowledgements file | @@ -57,7 +57,7 @@ Documenters
- Matt Caldwell (@mattcaldwell)
- berdario (@berdario)
- Cory Taylor (@coryandrewtaylor)
-- James C. (@Jammy4312)
+- James C. (@JamesMCo)
- Ally Weir (@allyjweir)
- Steven Loria (@sloria)
- Patrick Abeya (@wombat2k)
|
Menu : Fix search widget focus styling on Mac
By default, Qt draws an additional focus box that we don't want. | @@ -201,6 +201,7 @@ class Menu( GafferUI.Widget ) :
self.__searchMenu = _Menu( self._qtWidget(), "" )
self.__searchMenu.aboutToShow.connect( Gaffer.WeakMethod( self.__searchMenuShow ) )
self.__searchLine = QtWidgets.QLineEdit()
+ self.__searchLine.setAttribute( QtCore.Qt.WA_MacShowFocusRect, False )
self.__searchLine.textEdited.connect( Gaffer.WeakMethod( self.__updateSearchMenu ) )
self.__searchLine.returnPressed.connect( Gaffer.WeakMethod( self.__searchReturnPressed ) )
self.__searchLine.setObjectName( "search" )
|
Move start of consensus items after Genesis
Move the start of the consensus network items to after genesis, as the
validator is not ready to satisfy any requests on that interconnect. | @@ -404,14 +404,14 @@ class Validator:
def start(self):
self._component_dispatcher.start()
self._component_service.start()
- self._consensus_dispatcher.start()
- self._consensus_service.start()
if self._genesis_controller.requires_genesis():
self._genesis_controller.start(self._start)
else:
self._start()
def _start(self):
+ self._consensus_dispatcher.start()
+ self._consensus_service.start()
self._network_dispatcher.start()
self._network_service.start()
|
adding the Privacy, Security and OSINT Show
adding the Privacy, Security and OSINT Show | - [Talos Takes](https://talosintelligence.com/podcasts/shows/talos_takes)
- [Open Source Security Podcast](https://opensourcesecurity.io/category/podcast/)
- [Crypto-Gram Security Podcast](http://crypto-gram.libsyn.com/)
+- [The Privacy, Security and OSINT Show](https://inteltechniques.com/podcast.html)
|
Update elf_coinminer.txt
Detection update due to PA Unit 42 information. | # Reference: https://isc.sans.edu/forums/diary/Crypto+Mining+Is+More+Popular+Than+Ever/24050
# Reference: https://www.alibabacloud.com/blog/jbossminer-mining-malware-analysis_593804
# Reference: https://blog.talosintelligence.com/2018/08/rocke-champion-of-monero-miners.html
+# Reference: https://unit42.paloaltonetworks.com/malware-used-by-rocke-group-evolves-to-evade-detection-by-cloud-security-products/
3g2upl4pq6kufc4m.tk
+a.ssvs.space
+aybc.so
blockbitcoin.com
d3goboxon32grk2l.tk
d20blzxlz9ydha.cloudfront.net
dazqc4f140wtl.cloudfront.net
+dwn.rundll32.ml
enjoytopic.tk
realtimenews.tk
sydwzl.cn
|
fix typos
fix some typos | @@ -21,7 +21,8 @@ TAG=0.2.1
kubectl apply -f ./install/$TAG/kfserving.yaml
```
By default, you can create InferenceService instances in any namespace which has no label with `control-plane` as key.
-You can aslo configure KFServing to make InferenceService instances only work in the namespace which has label pair `serving.kubeflow.org/inferenceservice: enabled`. To enable this mode, you need add `env` field as below to `manager` container of statefulset `kfserving-controller-manager`.
+You can also configure KFServing to make InferenceService instances only work in the namespace which has label pair `serving.kubeflow.org/inferenceservice: enabled`. To enable this mode, you need to add `env` field as stated below to `manager` container of statefulset `kfserving-controller-manager`.
+
```
env:
- name: ENABLE_WEBHOOK_NAMESPACE_SELECTOR
|
operations.format: don't regen all fetchables (and chksums) for fetch ops
If they're provided to the call. This fixes `pmaint digest` complaining
about missing digests when trying to fetch/digest distfiles for new
ebuilds. | @@ -135,8 +135,9 @@ class operations(_operations_mod.base):
if not isinstance(fetchables, (tuple, list)):
fetchables = [fetchables]
ret = []
+ fetcher = self._fetch_kls(self.domain, self.pkg, fetchables, self._find_fetcher())
for fetchable in fetchables:
- ret.append(self._fetch_op.fetch_one(fetchable, self._get_observer(observer)))
+ ret.append(fetcher.fetch_one(fetchable, self._get_observer(observer)))
return all(ret)
return self._fetch_op.fetch_all(self._get_observer(observer))
|
devstack: enable flow based tunnels for sfc
ODL SFC requires flow based tunnels to operate correctly. This is
enabled through a specific property set as an external id in the
OVS database. | @@ -336,6 +336,9 @@ function bind_opendaylight_controller {
other_config:provider_mappings=$ODL_PROVIDER_MAPPINGS
fi
sudo ovs-vsctl set Open_vSwitch $ovstbl other_config:local_ip=$ODL_LOCAL_IP
+ if [[ ",$ODL_NETVIRT_KARAF_FEATURE," =~ ",odl-netvirt-sfc," ]]; then
+ sudo ovs-vsctl set Open_vSwitch $ovstbl external_ids:of-tunnel=true
+ fi
# for pseudo agent port binding
if [ "$ODL_PORT_BINDING_CONTROLLER" == "pseudo-agentdb-binding" ]; then
ODL_OVS_HOSTCONFIGS_OPTIONS=${ODL_OVS_HOSTCONFIGS_OPTIONS:---debug --noovs_dpdk}
|
Bugfix in PC_Miner.py
Fix a missing ) makes crash in app | @@ -222,7 +222,7 @@ class Client:
return (NODE_ADDRESS, NODE_PORT)
elif "message" in response:
- pretty_print(f"Warning: {response['message']}"
+ pretty_print(f"Warning: {response['message']}")
+ f", retrying in {retry_count*2}s",
"warning", "net0")
|
Fix formulation of eigenvectors
Add the missing imaginary unit in the section 'Create modulated structure' | @@ -1141,7 +1141,7 @@ eigenvectors with amplitudes and phase factors as
```{math}
\frac{A} { \sqrt{N_\mathrm{a}m_j} } \operatorname{Re} \left[ \exp(i\phi)
-\mathbf{e}_j \exp( \mathbf{q} \cdot \mathbf{r}_{jl} ) \right],
+\mathbf{e}_j \exp( i \mathbf{q} \cdot \mathbf{r}_{jl} ) \right],
```
where {math}`A` is the amplitude, {math}`\phi` is the phase,
|
Update `AsyncHTTPClient` to latest version `1.11.1`
Version `1.11.1` of `AsyncHTTPClient` only support Swift 5.4+ | "maintainer": "[email protected]",
"compatibility": [
{
- "version": "5.0",
- "commit": "a72c5adce3986ff6b5092ae0464a8f2675087860"
+ "version": "5.4",
+ "commit": "794dc9d42720af97cedd395e8cd2add9173ffd9a"
}
],
"platforms": [
|
update download location for BLAST 2.2.26 for Travis CI
Travis builds were failing because NCBI have removed the 2.2.26
obsolete executables from their FTP site. The Travis config file
now points at VBI at Virginia Tech's mirror | @@ -19,7 +19,7 @@ script:
# application dependencies: BLAST+, legacy BLAST, MUMMER
before_install:
- cd $HOME
- - wget ftp://ftp.ncbi.nlm.nih.gov/blast/executables/legacy.NOTSUPPORTED/2.2.26/blast-2.2.26-x64-linux.tar.gz
+ - wget http://mirrors.vbi.vt.edu/mirrors/ftp.ncbi.nih.gov/blast/executables/blast+/2.2.26/blast-2.2.26-x64-linux.tar.gz
- tar -zxvf blast-2.2.26-x64-linux.tar.gz
- export PATH=$HOME/blast-2.2.26/bin:$PATH
- cd $TRAVIS_BUILD_DIR
|
[drivers.SAFE.quicklook] generalized to allow unpacked + tar.gz scenes
until now this method worked on zipped archives only | @@ -1290,15 +1290,14 @@ class SAFE(ID):
if format != 'kmz':
raise RuntimeError('currently only kmz is supported as format')
kml_name = self.findfiles('map-overlay.kml')[0]
- kml_membername = kml_name.replace(self.scene, '').strip(r'\/')
- png = self.findfiles('quick-look.png')[0]
- png_membername = png.replace(self.scene, '').strip(r'\/')
- with zf.ZipFile(self.scene, 'r') as archive:
+ png_name = self.findfiles('quick-look.png')[0]
with zf.ZipFile(outname, 'w') as out:
- kml = archive.open(kml_membername).read().decode('utf-8')
+ with self.getFileObj(kml_name) as kml_in:
+ kml = kml_in.getvalue().decode('utf-8')
kml = kml.replace('Sentinel-1 Map Overlay', self.outname_base())
out.writestr('doc.kml', data=kml)
- out.writestr('quick-look.png', data=archive.open(png_membername).read())
+ with self.getFileObj(png_name) as png_in:
+ out.writestr('quick-look.png', data=png_in.getvalue())
def scanMetadata(self):
with self.getFileObj(self.findfiles('manifest.safe')[0]) as input:
|
adapted code for q0 off-sweetspot
the changes regard only the functions handling the noise | @@ -684,7 +684,7 @@ def distort_amplitude(fitted_stepresponse_ty,amp,tlist_new,sim_step_new):
def shift_due_to_fluxbias_q0(fluxlutman,amp_final,fluxbias_q0):
if not fluxlutman.czd_double_sided():
- omega_0 = fluxlutman.calc_amp_to_freq(0,'01')
+ omega_0 = compute_sweetspot_frequency(fluxlutman.q_polycoeffs_freq_01_det(),fluxlutman.q_freq_01())
f_pulse = fluxlutman.calc_amp_to_freq(amp_final,'01')
f_pulse = np.clip(f_pulse,a_min=None,a_max=omega_0) # necessary otherwise the sqrt below gives nan
@@ -701,7 +701,7 @@ def shift_due_to_fluxbias_q0(fluxlutman,amp_final,fluxbias_q0):
amp_B = amp_final[half_length:]
- omega_0 = fluxlutman.calc_amp_to_freq(0,'01')
+ omega_0 = compute_sweetspot_frequency(fluxlutman.q_polycoeffs_freq_01_det(),fluxlutman.q_freq_01())
f_pulse_A = fluxlutman.calc_amp_to_freq(amp_A,'01')
f_pulse_A = np.clip(f_pulse_A,a_min=None,a_max=omega_0)
@@ -748,8 +748,8 @@ def return_jump_operators(noise_parameters_CZ, f_pulse_final, fluxlutman):
# time-dependent jump operators on q0
if T2_q0_amplitude_dependent[0] != -1:
- f_pulse_final = np.clip(f_pulse_final,a_min=None,a_max=fluxlutman.q_freq_01())
- sensitivity = calc_sensitivity(f_pulse_final,fluxlutman.q_freq_01())
+ f_pulse_final = np.clip(f_pulse_final,a_min=None,a_max=compute_sweetspot_frequency(fluxlutman.q_polycoeffs_freq_01_det(),fluxlutman.q_freq_01()))
+ sensitivity = calc_sensitivity(f_pulse_final,compute_sweetspot_frequency(fluxlutman.q_polycoeffs_freq_01_det(),fluxlutman.q_freq_01()))
for i in range(len(sensitivity)):
if sensitivity[i] < 1e-1:
sensitivity[i] = 1e-1
@@ -1258,6 +1258,8 @@ def sensitivity_to_fluxoffsets(U_final_vec,input_to_parallelize,t_final,w_q0,w_q
print('infid_vec =',infid_vec.tolist())
+def compute_sweetspot_frequency(polycoeff,freq_at_0_amp):
+ return polycoeff[1]**2/2/polycoeff[0]-polycoeff[2]+freq_at_0_amp
def repeated_CZs_decay_curves(U_superop_average,t_final,w_q0,w_q1,alpha_q0):
|
Add float types to load_items to support configuration parameters of float type
Complement missing float types when loading configuration group parameters
so that parameters of type float can be attached into the trove instance successfully
Story:
Task: 42508 | @@ -166,6 +166,8 @@ class Configuration(object):
item.configuration_value = bool(int(item.configuration_value))
elif rule.data_type == 'integer':
item.configuration_value = int(item.configuration_value)
+ elif rule.data_type == 'float':
+ item.configuration_value = float(item.configuration_value)
else:
item.configuration_value = str(item.configuration_value)
return config_items
|
Update README.md
Add new install steps for conjure-up | @@ -22,10 +22,7 @@ This is a minimal Kubernetes cluster comprised of the following components and f
Installation has been automated via [conjure-up](http://conjure-up.io/):
- sudo apt-add-repository ppa:juju/stable
- sudo apt-add-repository ppa:conjure-up/next
- sudo apt update
- sudo apt install conjure-up
+ sudo snap install conjure-up --classic
conjure-up canonical-kubernetes
Conjure will prompt you for deployment options (AWS, GCE, Azure, etc.) and credentials.
|
Update knowledge_graph.py
The l in "bool" is lost. | @@ -27,7 +27,7 @@ class KnowledgeGraphDataset(DGLBuiltinDataset):
-----------
name: str
Name can be 'FB15k-237', 'FB15k' or 'wn18'.
- reverse: boo
+ reverse: bool
Whether add reverse edges. Default: True.
raw_dir : str
Raw file directory to download/contains the input data directory.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.