Today is the publication of the third edition of Autism Adulthood: Insights and Creative Strategies for a Fulfilling Life. It’s my wife Susan’s book collecting stories and experiences from people all along the autism spectrum, from the self-diagnosed to the profound.

The book includes dozens of interviews with autistic adults, their parents, caregivers, researchers, and professionals. Everyone’s experience of autism is different. Reading others’ stories and perspectives can give us a glimpse into other possibilities for ourselves and our loved ones.

If you have someone in your life on the spectrum, or are on it yourself, I guarantee you will find new ways to understand the breadth of what autism means and what it can be.

Susan has also written two other non-fiction autism books, including a memoir of our early days with our son Nat. Of course I highly recommend all of them.

In Why your mock doesn’t work I explained this rule of mocking:

Mock where the object is used, not where it’s defined.

That blog post explained why that rule was important: often a mock doesn’t work at all if you do it wrong. But in some cases, the mock will work even if you don’t follow this rule, and then it can break much later. Why?

Let’s say you have code like this:

# user.py

def get_user_settings():
    with open(Path("~/settings.json").expanduser()) as f:
        return json.load(f)

def add_two_settings():
    settings = get_user_settings()
    return settings["opt1"] + settings["opt2"]

You write a simple test:

def test_add_two_settings():
    # NOTE: need to create ~/settings.json for this to work:
    #   {"opt1": 10, "opt2": 7}
    assert add_two_settings() == 17

As the comment in the test points out, the test will only pass if you create the correct settings.json file in your home directory. This is bad: you don’t want to require finicky environments for your tests to pass.

The thing we want to avoid is opening a real file, so it’s a natural impulse to mock out open():

# test_user.py

from io import StringIO
from unittest.mock import patch

@patch("builtins.open")
def test_add_two_settings(mock_open):
    mock_open.return_value = StringIO('{"opt1": 10, "opt2": 7}')
    assert add_two_settings() == 17

Nice, the test works without needing to create a file in our home directory!

Much later...

One day your test suite fails with an error like:

...
  File ".../site-packages/coverage/python.py", line 55, in get_python_source
    source_bytes = read_python_source(try_filename)
  File ".../site-packages/coverage/python.py", line 39, in read_python_source
    return source.replace(b"\r\n", b"\n").replace(b"\r", b"\n")
           ~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^
TypeError: replace() argument 1 must be str, not bytes

What happened!? Coverage.py code runs during your tests, invoked by the Python interpreter. The mock in the test changed the builtin open, so any use of it anywhere during the test is affected. In some cases, coverage.py needs to read your source code to record the execution properly. When that happens, coverage.py unknowingly uses the mocked open, and bad things happen.

When you use a mock, patch it where it’s used, not where it’s defined. In this case, the patch would be:

@patch("myproduct.user.open")
def test_add_two_settings(mock_open):
    ... etc ...

With a mock like this, the coverage.py code would be unaffected.

Keep in mind: it’s not just coverage.py that could trip over this mock. There could be other libraries used by your code, or you might use open yourself in another part of your product. Mocking the definition means anything using the object will be affected. Your intent is to only mock in one place, so target that place.

Postscript

I decided to add some code to coverage.py to defend against this kind of over-mocking. There is a lot of over-mocking out there, and this problem only shows up in coverage.py with Python 3.14. It’s not happening to many people yet, but it will happen more and more as people start testing with 3.14. I didn’t want to have to answer this question many times, and I didn’t want to force people to fix their mocks.

From a certain perspective, I shouldn’t have to do this. They are in the wrong, not me. But this will reduce the overall friction in the universe. And the fix was really simple:

open = open

This is a top-level statement in my module, so it runs when the module is imported, long before any tests are run. The assignment to open will create a global in my module, using the current value of open, the one found in the builtins. This saves the original open for use in my module later, isolated from how builtins might be changed later.

This is an ad-hoc fix: it only defends one builtin. Mocking other builtins could still break coverage.py. But open is a common one, and this will keep things working smoothly for those cases. And there’s precedent: I’ve already been using a more involved technique to defend against mocking of the os module for ten years.

Even better!

No blog post about mocking is complete without encouraging a number of other best practices, some of which could get you out of the mocking mess:

• Use autospec=True to make your mocks strictly behave like the original object: see Why your mock still doesn’t work.
• Make assertions about how your mock was called to be sure everything is connected up properly.
• Use verified fakes instead of auto-generated mocks: Fast tests for slow services: why you should use verified fakes.
• Separate your code so that computing functions like our add_two_settings don’t also do I/O. This makes the functions easier to test in the first place. Take a look at Function Core, Imperative Shell.
• Dependency injection lets you explicitly pass test-specific objects where they are needed instead of relying on implicit access to a mock.

It’s been a busy, bumpy week with coverage.py. Some things did not go smoothly, and I didn’t handle everything as well as I could have.

It started with trying to fix issue 2064 about conflicts between the “sysmon” measurement core and a concurrency setting.

To measure your code, coverage.py needs to know what code got executed. To know that, it collects execution events from the Python interpreter. CPython now has two mechanisms for this: trace functions and sys.monitoring. Coverage.py has two implementations of a trace function (in C and in Python), and an implementation of a sys.monitoring listener. These three components are the measurement cores, known as “ctrace”, “pytrace”, and “sysmon”.

The fastest is sysmon, but there are coverage.py features it doesn’t yet support. With Python 3.14, sysmon is the default core. Issue 2064 complained that when the defaulted core conflicted with an explicit concurrency choice, the conflict resulted in an error. I agreed with the issue: since the core was defaulted, it shouldn’t be an error, we should choose a different core.

But I figured if you explicitly asked for the sysmon core and also a conflicting setting, that should be an error because you’ve got two settings that can’t be used together.

Implementing all that got a little involved because of “metacov”: coverage.py coverage-measuring itself. The sys.monitoring facility in Python was added in 3.12, but wasn’t fully fleshed out enough to do branch coverage until 3.14. When we measure ourselves, we use branch coverage, so 3.12 and 3.13 needed some special handling to avoid causing the error that sysmon plus branch coverage would cause.

I got it all done, and released 7.11.1 on Friday.

Soon, issue 2077 arrived. Another fix in 7.11.1 involved some missing branches when using the sysmon core. That fix required parsing the source code during execution. But sometimes the “code” can’t be parsed: Jinja templates compile html files to Python and use the html file as the file name for the code. When coverage.py tries to parse the html file as Python, of course it fails. My fix didn’t account for this. I fixed that on Saturday and released 7.11.2.

In the meantime, issue 2076 and issue 2078 both pointed out that now some settings combinations that used to produce warnings now produced errors. This is a breaking change, they said, and should not have been released as a patch version.

To be honest, my first reaction was that it wasn’t that big a deal, the settings were in conflict. Fix the settings and all will be well. It’s hard to remember all of the possibilities when making changes like this, it’s easy to make mistakes, and semantic versioning is bound to have judgement calls anyway. I had already spent a while getting 7.11.1 done, and .2 followed just a day later. I was annoyed and didn’t want to have to re-think everything.

But the more I thought about it, I decided they were right: it does break pipelines that used to work. And falling back to a different core is fine: the cores differ in speed and compatibility but (for the most part) produce the same results. Changing the requested core with a warning is a fine way to deal with the settings conflict without stopping test suites from running.

So I just released 7.11.3 to go back to the older behavior. Maybe I won’t have to do another release tomorrow!

While all this was going on, I also moved the code from my personal GitHub account to a new coveragepy GitHub organization!

Coverage.py is basically a one-man show. Maybe the GitHub organization will make others feel more comfortable chiming in, but I doubt it. I’d like to have more people to talk through changes with. Maybe I wouldn’t have had to make three releases in three days if someone else had been around as a sounding board.

I’m in the #coverage-py channel if you want to talk about any aspect of coverage.py, or I can be reached in lots of other ways. I’d love to talk to you.

Last night was a Boston Python project night where I had a good conversation with a few people that was mostly guided by questions from a nice guy named Mark.

How to write nice code in research

Mark works in research and made the classic observation that research code is often messy, and asked about how to make it nicer.

I pointed out that for software engineers, the code is the product. For research, the results are the product, so there’s a reason the code can be and often is messier. It’s important to keep the goal in mind. I mentioned it might not be worth it to add type annotations, detailed docstrings, or whatever else would make the code “nice”.

But the more you can make “nice” a habit, the less work it will be to do it as a matter of course. Even in a result-driven research environment, you’ll be able to write code the way you want, or at least push back a little bit. Code usually lives longer than people expect, so the nicer you can make it, the better it will be.

Side projects

Side projects are a good opportunity to work differently. If work means messy code, your side project could be pristine. If work is very strict, your side project can be thrown together just for fun. You get to set the goals.

And different side projects can be different. I develop coverage.py very differently than fun math art projects. Coverage.py has an extensive test suite run on many versions of Python (including nightly builds of the tip of main). The math art projects usually have no tests at all.

Side projects are a great place to decide how you want to code and to practice that style. Later you can bring those skills and learnings back to a work environment.

Forgive yourself

Mark said one of his difficulties with side projects is perfectionism. He’ll come back to a project and find he wants to rewrite the whole thing.

My advice is: forgive yourself. It’s OK to rewrite the whole thing. It’s OK to not rewrite the whole thing. It’s OK to ignore it for months at a time. It’s OK to stop in the middle of a project and never come back to it. It’s OK to obsess about “irrelevant” details.

The great thing about a side project is that you are the only person who decides what and how it should be.

How to stay motivated

But how to stay motivated on side projects? For me, it’s very motivating that many people use and get value from coverage.py. It’s a service to the community that I find rewarding. Other side projects will have other motivations: a chance to learn new things, flex different muscles, stretch myself in new ways.

Find a reason that motivates you, and structure your side projects to lean into that reason. Don’t forget to forgive yourself if it doesn’t work out the way you planned or if you change your mind.

How to write something people will use

Sure, it’s great to have a project that many people use, but how do you find a project that will end up like that? The best way is to write something that you find useful. Then talk about it with people. You never know what will catch on.

I mentioned my cog project, which I first wrote in 2004 for one reason, but which is now being used by other people (including me) for different purposes. It took years to catch on.

Of course there’s no guarantee something like that will happen: it most likely won’t. But I don’t know of a better way to make something people will use than to start by making something that you will use.

Other topics

The discussion wasn’t as linear as this. We touched on other things along the way: unit tests vs system tests, obligations to support old versions of software, how to navigate huge code bases. There were probably other tangents that I’ve forgotten.

Project nights are almost never just about projects: they are about connecting with people in lots of different ways. This discussion felt like a good connection. I hope the ideas of choosing your own paths and forgiving yourself hit home.

This post continues where Hobby Hilbert Simplex left off. If you haven’t read it yet, start there. It explains the basics of Hobby curves, Hilbert sorting and Simplex noise that I’m using.

Animation

To animate one of our drawings, instead of considering 40 lines, we’ll think about 140 lines. The first frame of the animation will draw lines 1 through 40, the second draws lines 2 through 41, and so on until the 100th frame is lines 100 through 140:

I’ve used a single Hilbert sorter for all of the frames to remove some jumping, but the Hobby curves still hop around. Also the animation doesn’t loop smoothly, so there’s a giant jump from frame 100 back to frame 1.

Natural cubics

Hobby curves look nice, but have this unfortunate discontinuity where a small change in a point can lead to a radical change in the curve. There’s another way to compute curves through points automatically, called natural cubic curves. These curves don’t jump around the way Hobby curves can.

Jake Low’s page about Hobby curves has interactive examples of natural cubic curves which you should try. Natural cubics don’t look as nice to our eyes as Hobby curves. Below is a comparison. Each row has the same points, with Hobby curves on the left and natural cubic curves on the right:

The “natural” cubics actually have a quite unnatural appearance. But in an animation, those quirks could be a good trade-off for smooth transitions. Here’s an animation with the same points as our first one, but with natural cubic curves:

Now the motion is smooth except for the jump from frame 100 back to frame 1. Let’s do something about that.

Circular Simplex

So far, we’ve been choosing points by sampling the simplex noise in small steps along a horizontal line: use a fixed u value, then take tiny steps along the v axis. That gave us our x coordinates, and a similar line with a different u value gave us the y coordinates. The ending point will be completely unrelated to the starting point. To make a seamlessly looping animation, we need our x,y values to cycle seamlessly, returning to where they started.

We can make our x,y coordinates loop by choosing u,v values in a circle. Because the u,v values return to their starting point in the continuous simplex noise, the x,y coordinates will return as well. We use two circles: one for the x coordinates and another for the y. The circles are far from each other to keep x and y independent of each other. The size of the circle is determined by the distance we want for each step and how many steps we want in the loop.

Here are three point paths created two ways, with linear sampling on the right and circular sampling on the left. Because simplex provides values between -1 and 1, the points wander within a square:

It can get a bit confusing at this point: these traces are not the curves we are drawing. They are the paths of the control points for successive curves. We draw curves through corresponding sets of points to get our animation. The first curve connects the first red/green/blue points, the second curve connects the second set, and so on.

Using circular sampling of the simplex noise, we can make animations that loop perfectly:

Colophon

If you are interested, the code is available on GitHub at nedbat/fluidity.

I saw a generative art piece I liked and wanted to learn how it was made. Starting with the artist’s Kotlin code, I dug into three new algorithms, hacked together some Python code, experimented with alternatives, and learned a lot. Now I can explain it to you.

It all started with this post by aBe on Mastodon:

I love how these lines separate and reunite. And the fact that I can express this idea in 3 or 4 lines of code.

For me they’re lives represented by closed paths that end where they started, spending part of the journey together, separating while we go in different directions and maybe reconnecting again in the future.

#CreativeCoding #algorithmicart #proceduralArt #OPENRNDR #Kotlin

The drawing is made by choosing 10 random points, drawing a curve through those points, then slightly scooching the points and drawing another curve. There are 40 curves, each slightly different than the last. Occasionally the next curve makes a jump, which is why they separate and reunite.

Eventually I made something similar:

Along the way I had to learn about three techniques I got from the Kotlin code: Hobby curves, Hilbert sorting, and simplex noise.

Each of these algorithms tries to do something “natural” automatically, so that we can generate art that looks nice without any manual steps.

Hobby curves

To draw swoopy curves through our random points, we use an algorithm developed by John Hobby as part of Donald Knuth’s Metafont type design system. Jake Low has a great interactive page for playing with Hobby curves, you should try it.

Here are three examples of Hobby curves through ten random points:

The curves are nice, but kind of a scribble, because we’re joining points together in the order we generated them (shown by the green lines). If you asked a person to connect random points, they wouldn’t jump back and forth across the canvas like this. They would find a nearby point to use next, producing a more natural tour of the set.

We’re generating everything automatically, so we can’t manually intervene to choose a natural order for the points. Instead we use Hilbert sorting.

Hilbert sorting

The Hilbert space-filling fractal visits every square in a 2D grid. Hilbert sorting uses a Hilbert fractal traversing the canvas, and sorts the points by when their square is visited by the fractal. This gives a tour of the points that corresponds more closely to what people expect. Points that are close together in space are likely (but not guaranteed) to be close in the ordering.

If we sort the points using Hilbert sorting, we get much nicer curves. Here are the same points as last time:

Here are pairs of the same points, unsorted and sorted side-by-side:

If you compare closely, the points in each pair are the same, but the sorted points are connected in a better order, producing nicer curves.

Simplex noise

Choosing random points would be easy to do with a random number generator, but we want the points to move in interesting graceful ways. To do that, we use simplex noise. This is a 2D function (let’s call the inputs u and v) that produces a value from -1 to 1. The important thing is the function is continuous: if you sample it at two (u,v) coordinates that are close together, the results will be close together. But it’s also random: the continuous curves you get are wavy in unpredictable ways. Think of the simplex noise function as a smooth hilly landscape.

To get an (x,y) point for our drawing, we choose a (u,v) coordinate to produce an x value and a completely different (u,v) coordinate for the y. To get the next (x,y) point, we keep the u values the same and change the v values by just a tiny bit. That makes the (x,y) points move smoothly but interestingly.

Here are the trails of four points taking 50 steps using this scheme:

If we use seven points taking five steps, and draw curves through the seven points at each step, we get examples like this:

I’ve left the points visible, and given them large steps so the lines are very widely spaced to show the motion. Taking out the points and drawing more lines with smaller steps gives us this:

With 40 lines drawn wider with some transparency, we start to see the smoky fluidity:

Jumps

In his Mastodon post, aBe commented on the separating of the lines as one of the things he liked about this. But why do they do that? If we are moving the points in small increments, why do the curves sometimes make large jumps?

The first reason is because of Hobby curves. They do a great job drawing a curve through a set of points as a person might. But a downside of the algorithm is sometimes changing a point a small amount makes the entire curve take a different route. If you play around with the interactive examples on Jake Low’s page you will see the curve can unexpectedly take a different shape.

As we inch our points along, sometimes the Hobby curve jumps.

The second reason is due to Hilbert sorting. Each of our lines is sorted independently of how the previous line was sorted. If a point’s small motion moves it into a different grid square, it can change the sorting order, which changes the Hobby curve even more.

If we sort the first line, and then keep that order of points for all the lines, the result has fewer jumps, but the Hobby curves still act unpredictably:

Colophon

This was all done with Python, using other people’s implementations of the hard parts: hobby.py, hilbertcurve, and super-simplex. My code is on GitHub (nedbat/fluidity), but it’s a mess. Think of it as a woodworking studio with half-finished pieces and wood chips strewn everywhere.

A lot of the learning and experimentation was in my Jupyter notebook. Part of the process for work like this is playing around with different values of tweakable parameters and seeds for the random numbers to get the effect you want, either artistic or pedagogical. The notebook shows some of the thumbnail galleries I used to pick the examples to show.

I went on to play with animations, which led to other learnings, but those will have to wait for another blog post. Update: I animated these in Natural cubics, circular Simplex.