microsoft/ptvs

Python Tools for Visual Studio

Python tools for Visual Studio

Python support for Visual Studio is developed and managed here.

• Visit our product overview and downloads page, read our documentation and watch our (old) YouTube tutorials.
• Feel free to file issues or ask questions on our issue tracker, or use the Send Feedback functionality in Visual Studio to report a problem or suggest a feature.
• We welcome code contributions - see Contributing for information. Documentation contributions may be made at the visualstudio-docs repository.
• For all other documentation, please see the Wiki

Visual Studio 2017 and Visual Studio 2019 installation

The Python development workload is available as part of Visual Studio Community, Professional and Enterprise. To install, run the normal VS installer and select either the Python development workload or the Data science and analytical applications workload.

Visual Studio 2015 and earlier

Prior to Visual Studio 2017, Python support was released as a standalone extension. We are no longer actively developing these versions, but if you are unable to upgrade to Visual Studio 2017 yet, you can use these older installers to get Python support.

PTVS 2.2.6 is the latest available version for Visual Studio 2015. You can also find it in the Visual Studio 2015 installer by selecting "Custom" and looking under the "Programming Languages" category.

PTVS 2.2.2 is the last release for Visual Studio 2013 users.

Thank you for using and supporting Python Tools for Visual Studio.

ptvs's Issues

Python Environment window items should have a context menu

Would be nice to have a right click context menu that lists what's in the drop down. Makes it easier to discover "remove".

PyPy doesn't show installed packages in virtual environment

It looks like PyPy's virtual environments use a bin folder rather than Scripts, and so we don't integrate properly in the UI.

Stepping sometimes requires you to step twice, or ends up in Python's native implementation

Pre-requisites:

• Mixed Mode Debugging enabled
• Debug Python 2.7 environment

Repro script:

import sys

sys.path.insert("d:\\src")   # Step over this line
sys.path.insert("d:\\python_src")

When you step over this line, the first time nothing happens. The second time, I end up in the C implementation of PyMain -- in native code. I've seen double stepping in other instances as well where the second step actually succeeds and ends up on the next line of python code, but I can't find a reliable repro for that, so I'm not sure what triggers it to succeed on the second step sometimes vs. ending up in the implementation of Python on the second step other times.

Tons of exception spam in the output window

Pre-requisites:

• Mixed mode debugging enabled
• Debug Python 2.7 environment

You can repro this with any script. You can just have a one-line script with import sys and put a breakpoint on the first line. By the time the breakpoint has hit, there are already hundreds of lines of exceptions in the output window, making it hard to get any useful information from the debug window if you are using OutputDebugString or otherwise trying to read the contents of the debug window.

Short sample output:

Short snippet:
exceptions.AttributeError
exceptions.TypeError
exceptions.AttributeError
exceptions.AttributeError
exceptions.AttributeError
exceptions.TypeError
exceptions.AttributeError
exceptions.AttributeError
exceptions.TypeError
exceptions.TypeError
exceptions.TypeError
exceptions.TypeError
exceptions.TypeError
exceptions.TypeError
exceptions.AttributeError
exceptions.AttributeError
'python_d.exe' (Python): Loaded 'D:\python_src\Python-2.7.8\lib\sysconfig.py'. Module was built without symbols.
exceptions.AttributeError
exceptions.TypeError
exceptions.TypeError
exceptions.TypeError
exceptions.AttributeError
'python_d.exe' (Python): Loaded 'D:\python_src\Python-2.7.8\lib\re.py'. Module was built without symbols.
'python_d.exe' (Python): Loaded 'D:\python_src\Python-2.7.8\lib\sre_compile.py'. Module was built without symbols.
'python_d.exe' (Python): Loaded 'D:\python_src\Python-2.7.8\lib\sre_parse.py'. Module was built without symbols.
'python_d.exe' (Python): Loaded 'D:\python_src\Python-2.7.8\lib\sre_constants.py'. Module was built without symbols.

Over the normal course of running my application, I counted that ~3,000 out of ~3,500 lines were just these exception messages. It would be nice if these could be silenced. If the issue is by design, then maybe there's a way to silence it only for exceptions that are thrown by the interpreter?

Refresh DB is slow / not working

Carrying on from this issue on codeplex:

https://pytools.codeplex.com/workitem/3017

I had a wealth of information to provide to back up what I was saying, unfortunately codeplex decided to take all comments and discussions offline right around that time, and I never got around to posting them when it came back online (if it is back online).

The gist of what I was getting at with regards to the issue was that it was taking upwards of 30 minutes for the code completion database to finish on a system that is capable of 2000MB/s data transfers, with 12 cores (24 with hyperthreading) and 64GB of RAM to analyze the data with (hint: it's the same time even on a slow system, it seems). The end result was that it mostly provided the same method signatures and auto-completion as PyCharm was able to glean in 2-3 minutes of analysis, except that often times the PTVS data was truncated in the display (and PyCharm was not).

Meanwhile, in an IPython session I am able to get more complete method signatures (many of which neither PTVS or PyCharm appear to pick up), and help easily available through either the console or a notebook - so the data is certainly there, it's just not being analyzed, and what is being analyzed is taking too long (especially when regular virtualenv creation is taken into consideration) and being underutilized (truncated).

Migrate Doc Page From CodePlex: NumPy-and-SciPy-for-Net

https://github.com/Microsoft/PTVS/wiki/NumPy-and-SciPy-for-Net

Code Snippet for def should let you tab over to the pass

Adding a new function snippet adds
def function(args):
pass

We let you tab over the name and args to replace them. We should also let you tab over pass.

Can't debug unit tests on VS2015 RC

VS 2015 22822.1 D14REL
PTVS 2.2 CL1457733_VS2015_22820

Create python app
Add item, unit test

    def test_A(self):
        a = 0
        b = 10
        pass

Put a breakpoint on a = 0.
Debug test

Occasionally, I get this access violation exception (debugger breaks).

But most of the time, it runs without debugging. You don't see the "2nd attach" happening at all. This is worse than with CTP 6 where it would sometimes work as expected (debug and stop at that breakpoint), sometimes get the access violation.

I tried C# and C++ (native) unit test debugging, and haven't had seen any issue yet.

System.AccessViolationException was unhandled
Message: An unhandled exception of type 'System.AccessViolationException' occurred in System.ServiceModel.dll
Additional information: Attempted to read or write protected memory. This is often an indication that other memory is corrupt.

Migrate Doc Page From CodePlex: Pyvot

Move page from CodePlex to https://github.com/Microsoft/PTVS/wiki/Pyvot

Deadlock opening VS 2015

Got a deadlock opening a file with VS 2015 with the following two stacks:

UI thread:

WindowsBase.dll!System.Windows.Threading.DispatcherSynchronizationContext.Wait(System.IntPtr[] waitHandles, bool waitAll, int millisecondsTimeout)
mscorlib.dll!System.Threading.SynchronizationContext.InvokeWaitMethodHelper(System.Threading.SynchronizationContext syncContext, System.IntPtr[] waitHandles, bool waitAll, int millisecondsTimeout)
[Native to Managed Transition]
mscorlib.dll!System.Lazy<Microsoft.PythonTools.Intellisense.CommentTaskProvider>.LazyInitValue()
mscorlib.dll!System.Lazy<Microsoft.PythonTools.Intellisense.CommentTaskProvider>.Value.get()
Microsoft.PythonTools.dll!Microsoft.PythonTools.Intellisense.VsProjectAnalyzer.ConnectErrorList(Microsoft.PythonTools.Analysis.IProjectEntry projEntry, Microsoft.VisualStudio.Text.ITextBuffer buffer)
Microsoft.PythonTools.dll!Microsoft.PythonTools.Intellisense.VsProjectAnalyzer.MonitorTextBuffer(Microsoft.VisualStudio.Text.Editor.ITextView textView, Microsoft.VisualStudio.Text.ITextBuffer buffer)
Microsoft.PythonTools.dll!Microsoft.PythonTools.Intellisense.IntellisenseControllerProvider.TryCreateIntellisenseController(Microsoft.VisualStudio.Text.Editor.ITextView textView, System.Collections.Generic.IList<Microsoft.VisualStudio.Text.ITextBuffer> subjectBuffers)
Microsoft.VisualStudio.Platform.VSEditor.dll!Microsoft.VisualStudio.Language.Intellisense.Implementation.IntellisenseManagerConnectionListener.SubjectBuffersConnected.AnonymousMethod__3(Microsoft.VisualStudio.Language.Intellisense.IIntellisenseControllerProvider provider)

Worker thread:

mscorlib.dll!System.Lazy<Microsoft.PythonTools.Intellisense.CommentTaskProvider>.LazyInitValue()
mscorlib.dll!System.Lazy<Microsoft.PythonTools.Intellisense.CommentTaskProvider>.Value.get()
Microsoft.PythonTools.dll!Microsoft.PythonTools.Intellisense.VsProjectAnalyzer.UpdateErrorsAndWarnings(Microsoft.PythonTools.Analysis.IProjectEntry entry, Microsoft.VisualStudio.Text.ITextSnapshot snapshot, Microsoft.PythonTools.Parsing.CollectingErrorSink errorSink, System.Collections.Generic.List<Microsoft.PythonTools.Intellisense.TaskProviderItem> commentTasks)
Microsoft.PythonTools.dll!Microsoft.PythonTools.Intellisense.VsProjectAnalyzer.ParseFile(Microsoft.PythonTools.Analysis.IProjectEntry entry, string filename, System.IO.Stream content, Microsoft.PythonTools.Parsing.Severity indentationSeverity)
Microsoft.PythonTools.dll!Microsoft.PythonTools.Intellisense.ParseQueue.EnqueueFile.AnonymousMethod__0()
Microsoft.PythonTools.dll!Microsoft.PythonTools.Intellisense.ParseQueue.EnqueWorker.AnonymousMethod__0(object dummy)
mscorlib.dll!System.Threading.QueueUserWorkItemCallback.WaitCallback_Context(object state)
mscorlib.dll!System.Threading.ExecutionContext.RunInternal(System.Threading.ExecutionContext executionContext, System.Threading.ContextCallback callback, object state, bool preserveSyncCtx)
mscorlib.dll!System.Threading.ExecutionContext.Run(System.Threading.ExecutionContext executionContext, System.Threading.ContextCallback callback, object state, bool preserveSyncCtx)
mscorlib.dll!System.Threading.QueueUserWorkItemCallback.System.Threading.IThreadPoolWorkItem.ExecuteWorkItem()
mscorlib.dll!System.Threading.ThreadPoolWorkQueue.Dispatch()
mscorlib.dll!System.Threading._ThreadPoolWaitCallback.PerformWaitCallback()

Refactor rename in doc string raises a null reference

class abc(object):
"""doc string"""
def init(self, value):
self.value = value
def x(self):
"""doc string"""
pass

Do refactor rename in either of the doc strings, a null reference exception is raised.
Instead we should give you the "Please select a symbol to be renamed." dialog

Launching PyPy interpreter crashes Visual Studio

I am using Visual Studio 2015 CTP with PTVS 2.2 Beta, PyPy3 2.4.

Going to Python Environments and calling the pypy interpreter crashes VS.

Migrate Doc Page From CodePlex: Build-Instructions-for-PTVS

https://github.com/Microsoft/PTVS/wiki/Build-Instructions-for-PTVS

Migrate Doc Page From CodePlex: Code-Formatting

https://github.com/Microsoft/PTVS/wiki/Code-Formatting

[Python view] doesn't work in VS 2015

When expanding [Python View] nodes in debugger, there are no children, and in debug builds there's an assertion failure:

GetChildren called on an unsupported DkmEvaluationResult.
at Microsoft.PythonTools.DkmDebugger.ExpressionEvaluator.GetChildren(DkmEvaluationResult result, DkmWorkList workList, Int32 initialRequestSize, DkmInspectionContext inspectionContext, DkmCompletionRoutine1 completionRoutine) in C:\TFS\xTVS\Python_Main\Python\Product\Debugger\DkmDebugger\ExpressionEvaluator.cs:line 516 at Microsoft.PythonTools.DkmDebugger.LocalComponent.Microsoft.VisualStudio.Debugger.ComponentInterfaces.IDkmLanguageExpressionEvaluator.GetChildren(DkmEvaluationResult result, DkmWorkList workList, Int32 initialRequestSize, DkmInspectionContext inspectionContext, DkmCompletionRoutine1 completionRoutine) in C:\TFS\xTVS\Python_Main\Python\Product\Debugger\DkmDebugger\LocalComponent.cs:line 423
at Microsoft.VisualStudio.Debugger.EntryPoint.IDkmLanguageExpressionEvaluator_GetChildren(IntPtr pvClassInfo, IntPtr Result, IntPtr WorkList, Int32 InitialRequestSize, IntPtr InspectionContext, IntPtr CompletionRoutine)

Deadlock while starting debugging

If the engine attaches while getting a breakpoint hit count, we will deadlock.

Location of lock on UI thread:

        int IDebugBoundBreakpoint2.GetHitCount(out uint pdwHitCount) {
            pdwHitCount = (uint)_breakpoint.GetHitCountAsync().GetAwaiter().GetResult();
            return VSConstants.S_OK;
        }

Stack trace of the worker thread causing the lock:

Microsoft.PythonTools.dll!Microsoft.VisualStudioTools.UIThread.Invoke(System.Action action) Line 66 C#
Microsoft.PythonTools.dll!Microsoft.PythonTools.Repl.BasePythonReplEvaluator.EnsureConnected() Line 127 C#
Microsoft.PythonTools.dll!Microsoft.PythonTools.Repl.PythonDebugProcessReplEvaluator.PythonDebugProcessReplEvaluator(System.IServiceProvider serviceProvider, Microsoft.PythonTools.Debugger.PythonProcess process, Microsoft.PythonTools.PythonToolsService pyService, Microsoft.PythonTools.Debugger.IThreadIdMapper threadIdMapper) Line 534 C#
Microsoft.PythonTools.dll!Microsoft.PythonTools.Repl.PythonDebugReplEvaluator.AttachProcess(Microsoft.PythonTools.Debugger.PythonProcess process, Microsoft.PythonTools.Debugger.IThreadIdMapper threadIdMapper) Line 460   C#
Microsoft.PythonTools.dll!Microsoft.PythonTools.Repl.PythonDebugReplEvaluator.OnEngineAttached(object sender, Microsoft.PythonTools.Debugger.DebugEngine.AD7EngineEventArgs e) Line 430 C#
Microsoft.PythonTools.Debugger.dll!Microsoft.PythonTools.Debugger.DebugEngine.AD7Engine.SendLoadComplete(Microsoft.PythonTools.Debugger.DebugEngine.AD7Thread thread) Line 372  C#
Microsoft.PythonTools.Debugger.dll!Microsoft.PythonTools.Debugger.DebugEngine.AD7Engine.OnProcessLoaded(object sender, Microsoft.PythonTools.Debugger.ThreadEventArgs e) Line 1262  C#
Microsoft.PythonTools.Debugger.dll!Microsoft.PythonTools.Debugger.PythonProcess.HandleProcessLoad(System.IO.Stream stream) Line 796 C#
Microsoft.PythonTools.Debugger.dll!Microsoft.PythonTools.Debugger.PythonProcess.DebugEventThread() Line 448 C#

It looks like we need to marshal the LoadComplete event back to the UI thread sooner without blocking.

NullReferenceException in datetime analysis

Got the following exception while analysing test.support.init.py:1472 from CPython 3.5. Can't see any obvious way to fix it other than checking _type for null. The null is apparently coming from the days property in datetime.py:474, which apparently has not assigned a type.

Application: devenv.exe
Framework Version: v4.0.30319
Description: The process was terminated due to an unhandled exception.
Exception Info: System.NullReferenceException
Stack:
   at Microsoft.PythonTools.Analysis.Values.BuiltinNamespace`1[[System.__Canon, mscorlib, Version=4.0.0.0, Culture=neutral, PublicKeyToken=b77a5c561934e089]].GetMember(Microsoft.PythonTools.Parsing.Ast.Node, Microsoft.PythonTools.Analysis.AnalysisUnit, System.String)
   at Microsoft.PythonTools.Analysis.AnalysisSetExtensions.GetMember(Microsoft.PythonTools.Analysis.IAnalysisSet, Microsoft.PythonTools.Parsing.Ast.Node, Microsoft.PythonTools.Analysis.AnalysisUnit, System.String)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateMember(Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator, Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateWorker(Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateBinary(Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator, Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateWorker(Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateParenthesis(Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator, Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateWorker(Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateAnd(Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator, Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateWorker(Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateOr(Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator, Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateWorker(Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateParenthesis(Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator, Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateWorker(Microsoft.PythonTools.Parsing.Ast.Node)
   at Microsoft.PythonTools.Analysis.Analyzer.DDG.Walk(Microsoft.PythonTools.Parsing.Ast.IfStatement)
   at Microsoft.PythonTools.Parsing.Ast.IfStatement.Walk(Microsoft.PythonTools.Parsing.Ast.PythonWalker)
   at Microsoft.PythonTools.Analysis.Analyzer.DDG.Walk(Microsoft.PythonTools.Parsing.Ast.SuiteStatement)
   at Microsoft.PythonTools.Parsing.Ast.SuiteStatement.Walk(Microsoft.PythonTools.Parsing.Ast.PythonWalker)
   at Microsoft.PythonTools.Analysis.Analyzer.FunctionAnalysisUnit.AnalyzeWorker(Microsoft.PythonTools.Analysis.Analyzer.DDG, System.Threading.CancellationToken)
   at Microsoft.PythonTools.Analysis.Analyzer.DDG.Analyze(Microsoft.PythonTools.Analysis.Deque`1<Microsoft.PythonTools.Analysis.AnalysisUnit>, System.Threading.CancellationToken, System.Action`1<Int32>, Int32)
   at Microsoft.PythonTools.Analysis.PythonAnalyzer.AnalyzeQueuedEntries(System.Threading.CancellationToken)
   at Microsoft.PythonTools.Intellisense.AnalysisQueue+GroupAnalysis.Analyze(System.Threading.CancellationToken)
   at Microsoft.PythonTools.Intellisense.AnalysisQueue.Worker(System.Object)
   at System.Threading.ThreadHelper.ThreadStart_Context(System.Object)
   at System.Threading.ExecutionContext.RunInternal(System.Threading.ExecutionContext, System.Threading.ContextCallback, System.Object, Boolean)
   at System.Threading.ExecutionContext.Run(System.Threading.ExecutionContext, System.Threading.ContextCallback, System.Object, Boolean)
   at System.Threading.ExecutionContext.Run(System.Threading.ExecutionContext, System.Threading.ContextCallback, System.Object)
   at System.Threading.ThreadHelper.ThreadStart(System.Object)

New project from existing code enhancements

1. Create a blank visual studio solution
2. Right click the solution node, choose Add New Project -> Python -> From Existing Python Code
3. Go through the steps.

At the end of the wizard, your project will have been added to a new solution, not the original one.

Debugger ignores some try ... except ... statements

Using "Start without debugging" the following code executes without problems, but when running with debugging the program breaks with a "No module named invalidImport" error. It presents the options "Break" or "Continue" as for all non-recoverable errors. If I choose "Continue" the program continues without any issue.

The debugger somehow seems to ignore the try ... except ... statement in the context manager while python correctly catches the error.

Clearly PTVS debugging thus should not break here.

from contextlib import contextmanager

@contextmanager
def ignoring(*exceptions):
    try:
        yield
    except exceptions:
        pass

with ignoring(ImportError):
    import invalidImport

print "Test"

Allow separate search paths for x86 and x64

I develop an extension module that needs to work with both x86 and x64 interpreters. These are output by my build system to different build trees, and I need my Python project which uses the extension module to be able to find the correct extension module. I have added both Debug x64 and Debug x86 environments, and I can select the correct environment, but there is no way to have the search path be updated accordingly so that when the x64 environment is selected, it uses the search path to locate the x64 extension module, and similarly for x86.

It seems like this should ultimately play nicely with the Visual Studio solution / project configuration system. This way, both the python environment as well as the search paths could be tied to a particular Project configuration, and the solution configuration could select various project configurations accordingly. It seems like interoperability with the Visual Studio project / solution configuration system is unimplemented.

Support function names in Breakpoints window in mixed mode

"You need to implement IDkmLanguageInstructionDecoder for the Python language and return a reasonable string in GetMethodName(). The debugger will call GetMethodName() for each Python function that has a breakpoint in it. The result of GetMethodName() will get displayed to the user in the function column (it may be hidden by default) of the breakpoints window."

Dev build 2015-04-06 Hangs VS 2013 Desktop Express

I have a mixed solution with three C# projects and one IronPython project in it. When I try to open this solution with the 6th April build of PTVS installed, VS 2013 Desktop Express freezes with the notification "VS Express 2013 Desktop Express is busy ... waiting for an internal operation to complete ...". Solution Explorer shows that the Solution is "(loading...)" but there is little to no CPU activity. It looks like something is deadlocked.

The previous Dev Build 2015-03-26 seemed to work fine except for the occasional freeze that required a restart which all previous builds of PTVS seem to have suffered from when an IronPython project is loaded in VS but with the IronPython source not necessarily being loaded in the editor.

completions: hitting enter when fully typed in interpreter.

there is an option under tools, options, text editor, python, advanced called "add new line on enter at end of fully typed word". can we have this apply to interpreters too?

Intellisense extremely slow in mixed mode debugging.

When placing the mouse cursor over a variable in a python code window, Visual studio will "lock up" for 10 seconds to couple minutes.

This is only occurring in mixed mode debugging where I have an application that has loaded python33.dll. It works fine when not debugging. It also works fine for pure python coding using the .python exe.

I originally saw this problem with VS2012 where it would appear to hang (is it related to this ? https://pytools.codeplex.com/workitem/3015) Eventually, I figured out it was when I moved the cursor over text so suspected Intellisense.

I installed VS2013 and tired again. This time, the intellisense window would pop up but only after a very long delay. All other functions work fine. I can step though code, see variables in the watch windows, etc.

I don't have access to the core application importing python33.dll but can say it is python 3.3.2 64 bit. If there is another application out there that uses python33.dll, I could try that to determine if it is my setup our somehow related to my application (suggestions?)

I'm guessing this must have been tested so that there is some configuration out there that can make this work.

Deadlock(?) loading IronPython project in VS DesktopExpress 2013

I have a mixed VS solution with three C# projects and one IronPython project in it. When I try to open this solution with the 13th April build of PTVS installed VS 2013 Desktop Express freezes with the notification "VS Express 2013 Desktop Express is busy ... waiting for an internal operation to complete ...". Solution Explorer shows that the Solution is "(loading...)" but there is little to no CPU activity. It looks like something is deadlocked.

I originally thought that this was due to some combination of editor windows being open but this seems to be not so. If I rename the solution folder then the project opens fine. Doing nothing further if I then immediately close it and reopen it then the hang occurs. Renaming the solution folder again repeats this behaviour, it opens once the first time and not thereafter.

The behaviour first occurred in the 6th April Dev Build, before then Dev Build 2015-03-26 seemed to work fine except for the occasional freeze that required a restart which all previous builds of PTVS seem to have suffered from when an IronPython project is loaded in VS but with the IronPython source not necessarily being loaded in the editor.

Python Environment window is not switching to horizontal layout

When dragging the window from a vertical layout to a horizontal layout it doesn't switch to the horizontal layout.

OutOfMemoryException when rebuilding Intellisense DB for large packages

When rebuilding the Intellisense DB I get an OutOfMemoryException:

2015-04-23T12:43:12: Starting analysis of 270 modules
2015-04-23T12:46:51: [ERROR] Analysis failed
System.OutOfMemoryException: Exception of type 'System.OutOfMemoryException' was thrown.
    at System.Collections.Generic.HashSet`1.Initialize(Int32 capacity)
    at System.Collections.Generic.HashSet`1.AddIfNotPresent(T value)
    at System.Collections.Generic.HashSet`1.UnionWith(IEnumerable`1 other)
    at Microsoft.PythonTools.Analysis.VariableDef.EstimateTypeCount(IAnalysisSet extraTypes)
    at Microsoft.PythonTools.Analysis.Values.IterableInfo.AddTypes(AnalysisUnit unit, IAnalysisSet[] types)
    at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.MakeSequence(ExpressionEvaluator ee, Node node)
    at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateSequence(ExpressionEvaluator ee, Node node)
    at Microsoft.PythonTools.Analysis.Analyzer.ExpressionEvaluator.EvaluateWorker(Node node)
    at Microsoft.PythonTools.Analysis.Analyzer.DDG.Walk(AssignmentStatement node)
    at Microsoft.PythonTools.Parsing.Ast.AssignmentStatement.Walk(PythonWalker walker)
    at Microsoft.PythonTools.Analysis.Analyzer.DDG.Walk(SuiteStatement node)
    at Microsoft.PythonTools.Parsing.Ast.SuiteStatement.Walk(PythonWalker walker)
    at Microsoft.PythonTools.Analysis.Analyzer.FunctionAnalysisUnit.AnalyzeWorker(DDG ddg, CancellationToken cancel)
    at Microsoft.PythonTools.Analysis.Analyzer.DDG.Analyze(Deque`1 queue, CancellationToken cancel, Action`1 reportQueueSize, Int32 reportQueueInterval)
    at Microsoft.PythonTools.Analysis.PythonAnalyzer.AnalyzeQueuedEntries(CancellationToken cancel)
    at Microsoft.PythonTools.Analysis.PyLibAnalyzer.Analyze()
    at Microsoft.PythonTools.Analysis.PyLibAnalyzer.<Run>d__30.MoveNext()

My development machine has a total of 16 GB memory and there is still ~50% left when the process crashes with a usage of about ~1.7 GB.

Is there anything I can do so that PythonTools uses more memory?

Migrate Doc Page From CodePlex: Profiling

https://github.com/Microsoft/PTVS/wiki/Profiling

Should not auto-list after `as`

Currently with open(filename) as f will auto-complete f to filename. As we're defining a new name here, we should not suggest anything.

Migrate Doc Page From CodePlex: Using-IPython-with-PTVS

https://github.com/Microsoft/PTVS/wiki/Using-IPython-with-PTVS

Migrate Doc Page From CodePlex: Unit-Tests

https://github.com/Microsoft/PTVS/wiki/Unit-Tests

Migrate Doc Page From CodePlex: HPC-and-Cloud-Features

https://github.com/Microsoft/PTVS/wiki/HPC-and-Cloud-Features

Azure remote debugging - breaking doesn't work

When remotely attached to an Azure website, hitting a breakpoint causes debugger to enter the break mode, but there's nothing in call stack or locals, and the editor says "Code not running".

cannot delete auto detected environments

custom added envs have a Remove option under configure. auto detected ones dont. this seems inconsistent.

XAML designer doesn't work in VS 2015 RC

VS 2015 22823.1 D14REL Enterprise (xlab image)
PTVS 30423.00

Create IronPython WPF app
Double click on .xaml file to open in editor

I get the split view of designer/xaml code but the designer part at the top has a progress bar that never ends.

I don't know if it's that same design host dll issue we've had in the past.

[Mixed Mode] Step over os.path.join steps in instead of over from inside a function

Pre-requisites:

• Mixed Mode Debugging enabled
• Debug Python 2.7 Interpreter

Repro script

import os

foo = os.path.join("foo", "bar")   # (1) Step over this line

def some_func():
    foo = os.path.join("foo", "bar")   # (2) Step over this line

some_func()

Stepping over (1) works, but stepping over (2) ends up inside the middle of os.path.join() inside of ntpath.py.

Beta 2.2 slows down Text editor

Hi

I tryed the 2.2 Beta and it slows the IDE sown so every key stroke needs 1-2sec to display on the screen...

Changed back to 2.1 for now.

Test failure in django project

Create Django Project (bootstrap starter template)
Run one of the 3 tests
Result: test failure

I'm using Django 1.8

Test Name:  test_about
Test Outcome:   Failed
Result StandardError:   
======================================================================
ERROR: tearDownClass (app.tests.ViewTest)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "C:\Users\huvalo\AppData\Local\Continuum\Anaconda3\lib\site-packages\django\test\testcases.py", line 962, in tearDownClass
    cls._rollback_atomics(cls.cls_atomics)
AttributeError: type object 'ViewTest' has no attribute 'cls_atomics'
----------------------------------------------------------------------
Ran 1 test in 0.516s
FAILED (errors=1)

Referenced .pyd files are not added to search paths when running tests

If you have a reference to a .pyd file from a Python project, its parent directory is included in the search path when you run/debug, but not when running tests.

(related to, but not the same as, https://pytools.codeplex.com/workitem/2934)

IntelliSense for .pyd requires matching Python architecture

I don't think the following issue is documented anywhere, but for some it may not be obvious.

In order to make Visual Studio PyTools IntelliSense work for a .pyd extension, it is required to select, in Project > Properties > General > Interpreter, Python interpreter matching target architecture of the .pyd binary.

Especially, if is relevant if both 32-bit and 64-bit Python installations are hosted side-by-side.

For instance, given side-by-side installation of the two Python interpreters:
C:\Python32 which is 64-bit version
C:\Python32x86-32 which is 32-bit version
if a .pyd is installed in C:\Python32 only, then one may need to choose "Python 64-bit 3.2" as the project interpreter.

Otherwise,

1. IntelliSense won't be available.
2. An attempt to manually add 64-bit .pyd to References of a project that is configured to target Python 32-bit will fail as well.

PTVS installs to global packages instead of virtualenv

I've tried all approaches I can think of - creating a new virtualenv from with in PTVS, using an existing virtualenn, recreated projects etc. No matter how I try the virtualenv seems to be ignored. Visual studio is showing my virtualenv by name active and all, but is actually listing global packages. And installing packages through PTVS also installs to global packages instead of the virtualenv.

Attaching screenshot to clarify.

I'm using Visual Studio 2013 community edition. PVTS 2.1 (also tried 2.2 beta). Python 2.7.9.

InvalidOperationException when creating project

PTVS Dev 2015-04-06 VS 2013

(this doesn't happen consistently)

Create new bottle project (starter project with bootstrap)
Click OK to create virtual env and dependencies

Result:

System.InvalidOperationException was unhandled
Message: An unhandled exception of type 'System.InvalidOperationException' occurred in mscorlib.dll
Additional information: Collection was modified; enumeration operation may not execute.

With call stack:

    mscorlib.dll!System.ThrowHelper.ThrowInvalidOperationException(System.ExceptionResource resource)   Unknown
    mscorlib.dll!System.Collections.Generic.Dictionary<System.__Canon,System.__Canon>.ValueCollection.Enumerator.MoveNext() Unknown
>   Microsoft.PythonTools.Analysis.dll!Microsoft.PythonTools.Analysis.ProjectEntry.Parse(bool enqueueOnly, System.Threading.CancellationToken cancel)   Unknown
    Microsoft.PythonTools.Analysis.dll!Microsoft.PythonTools.Analysis.ProjectEntry.Analyze(System.Threading.CancellationToken cancel, bool enqueueOnly) Unknown
    Microsoft.PythonTools.dll!Microsoft.PythonTools.Intellisense.AnalysisQueue.Worker(object threadStarted) Unknown
    mscorlib.dll!System.Threading.ThreadHelper.ThreadStart_Context(object state)    Unknown
    mscorlib.dll!System.Threading.ExecutionContext.RunInternal(System.Threading.ExecutionContext executionContext, System.Threading.ContextCallback callback, object state, bool preserveSyncCtx)   Unknown
    mscorlib.dll!System.Threading.ExecutionContext.Run(System.Threading.ExecutionContext executionContext, System.Threading.ContextCallback callback, object state, bool preserveSyncCtx)   Unknown
    mscorlib.dll!System.Threading.ExecutionContext.Run(System.Threading.ExecutionContext executionContext, System.Threading.ContextCallback callback, object state) Unknown
    mscorlib.dll!System.Threading.ThreadHelper.ThreadStart(object obj)  Unknown

I have a dump file, I can upload it to our share when I'm back at the office, if that helps.

Migrate Doc Page From CodePlex: PTVS-Installation

https://github.com/Microsoft/PTVS/wiki/PTVS-Installation

Move/rename the corresponding *.pyc file at the same time when moving/renaming the *.py file

Thank you for PTVS.

I would like to request that PTVS move/rename the corresponding *.pyc file at the same time when moving/renaming the *.py file.

For example, given the following project structure:

package1
    __init__.py
    A.py
package2
    __init__.py
testPackage
    __init__.py
    testA.py

testA.py:

import unittest
from package1.A import A

class TestRoutes(unittest.TestCase):

    def testA(self):
        A = A()
        print A.Name

if __name__ == '__main__':
    unittest.main()

And then if I move A.py to package2, but don't update the from/import statement, and run the test, I expect the test run to fail with an error like "ImportError: No module named A".

However, it succeeds without error (unexpectedly)!

Even cleaning the solution does not resolve the issue because the *.pyc doesn't get deleted during a clean.

Thank you.

Tests with decorators do not appear in the test explorer view

This is the same as issue # 1969 on the CodePlex issue tracker. https://pytools.codeplex.com/workitem/1969

It has been marked as resolved there, but I still see the problem with the latest stable release 2.1 and the latest dev build April 13th.

If a Python unittest has a decorator, it does not appear in the test explorer. If the decorator is removed, it then appears.

    @patch.object(os.path, 'isfile')
    def test_A(self):
        ''' Doesn't appear in the test explorer '''
        self.fail("Not implemented")

    def test_B(self):
        ''' This does appear in the test explorer '''
        self.fail("Not implemented")

I'm using the mock library, so have hundreds of tests with patch decorators.

Types in annotations are not highlighted

In

def f(t : str = str):
    pass

The second str is highlighted as a type but the first str is not.

Undo does not undo committed completions

Type "a = b" and hit enter, and it will complete to an item. Now hit Ctrl+Z to undo and it doesn't.

Undoing should undo the completed part and leave what the user typed.

Code snippets should respect tabs vs spaces

Snippets are hardcoded to do 4 spaces of indention. That may not match the user settings. Instead we should use a tab and let editor settings replace those tabs with proper number of spaces.

Debugger fails to run with Stackless

Using the latest dev version of PTVS, PyPy3 2.4 (32 bit), VS 2015 CTP, Windows 10 TP.

This is the traceback (http://i.imgur.com/X3qNp3F.png):

Traceback (most recent call last):
File "C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\IDE\Extensions\Microsoft\Python Tools for Visual Studio\2.2\visualstudio_py_launcher.py", line 25, in
import visualstudio_py_debugger
File "C:\Program Files (x86)\Microsoft Visual Studio 14.0\Common7\IDE\Extensions\Microsoft\Python Tools for Visual Studio\2.2\visualstudio_py_debugger.py", line 65, in
import stackless
File "C:\PyPy3\lib_pypy\stackless.py", line 542, in
_init()
File "C:\PyPy3\lib_pypy\stackless.py", line 538, in _init
tasklet._init.func(_main_tasklet, label='main')
AttributeError: 'function' object has no attribute 'func'

Internal error detected. Please copy the above traceback and report at
http://go.microsoft.com/fwlink/?LinkId=293415

Press Enter to close. . .

Set next statement is unimplemented

import os

path = os.path.join("foo", "bar")
path = os.path.join("bar", "baz")

if path:  # (1) Set breakpoint here.
    path = os.path.join("baz", "boo")   # While stopped at (1), right click this line and choose Set Next Statement
pass

When you choose Set Next Statement, a dialog pops up saying this is unimplemented.