Pure Python Vim clone(github.com) |
Pure Python Vim clone(github.com) |
- Core editor (internal representation, etc.)
- Key bindings (so you could easily create an emacs instead of a vi)
- Rendering
- Scripting language (for customized behavior)
Finally, make sure you thoroughly document these building blocks, so others can create really cool stuff with it. Also, think of possible use-cases when defining the modules. A smart architecture could allow for a collaborative editor, for example.
The key bindings are also separate. Getting emacs bindings is not much more than changing this line [0]. Only adding the bindings for the window management and emacs command line is still to be done. (I know that emacs is actually much more than only its key bindings, but you know what I mean.)
The rendering is also independent. There are two backends: vt100 terminals and the windows console. (Honesly, my main focus is vt100, but any render back-end is possible. I think even graphical)
The same for the event loops by the way, it can run on a couple of event loops. For instance asyncio.
Documentation will follow. prompt-toolkit has already quite a lot of examples, and there's a lot of documentation in the code itself. But I agree that we should keep improving.
Cheers!
[0] https://github.com/jonathanslenders/pyvim/blob/master/pyvim/...
There are plenty of examples in the repo that cover most of the features in the library. The awesome thing is all of the examples fit in a single page, a testament to the power and simplicity of the library.
I built pgcli (https://github.com/dbcli/pgcli) almost entirely by reading the examples.
"Python, huh? Seems like typos in uncommon branches of the code would cause it to randomly fail at runtime, losing your work!"
"Now evmar, don't be such a internet nay-sayer, plenty of people write reliable Python code. You just need tests and... yep, there's a tests directory right there in the repository."
"Let's take a look. ...there's only one test!?"
It looks pretty neat other than that, though.
In the process I have been writing unit tests. Previously the program didn't have any, which was really dumb. It was dumb because, in the three days I've spent doing this, I have found (and fixed) so, _so_ many bugs.
And now I have a decent set of tests, I can change something, rerun them, and have a pretty good idea of whether it worked or not. I don't even have to run the program!
Unit tests. They Will Save You Time™.
But to be clear, in my day job I work on an app with 350k lines of Python in it and the reason I know it mostly works is due to our test coverage. As someone else mentioned in this thread, a lack of tests should not give you confidence regardless of the language.
Maybe not - Python itself can provide some crash protection. Run it this way: python -i run_pyvim.py. Then if pyvim crashes, Python will still be running with all your work still there. Maybe you can just type run() at the Python prompt to resume the session in almost the state you left it -- depending on how run() is coded, how much re-initialization it does.
Q: Why Python?
A: The only alternative would be Haskell, but I still have to learn that.
However I managed to build a _very_ basic proof-of-concept editor (no dependencies) in just a few hundreds lines of code which I could explain, but until now I was too shy to share it as it did not involve magic abstract Haskell-foo ... ;)
Please don't be, simple understandable Haskell code is very nice :)
Plus if there is a better way of doing it you get to find that out too.
"Gets the job done in just a few hundred lines of easily-explained code" is a terrific standard to meet.
"Magic abstract Haskell-fu" cannot improve such a program very much.
It's an automatic translation of the C version.
Like tomp said 4 hours ago: Like vim, just slower.™
Why did I create Pyvim?
There are several reasons.
The main reason is maybe because it was a small step after I created the Python prompt-toolkit library. That is a library which is actually only a simply pure Python readline replacement, but with some nice additions like syntax highlighting and multiline editing. It was never intended to be a toolkit for full-screen terminal applications, but at some point I realised that everything we need for an editor was in there and I liked to challenge its design. So, I started an editor and the first proof of concept was literally just a few hundred lines of code, but it was already a working editor.
The creation of pyvim will make sure that we have a solid architecture for prompt-toolkit, but it also aims to demonstrate the flexibility of the library. When it makes sense, features of pyvim will move back to prompt-toolkit, which in turn also results in a better Python REPL. (see ptpython, an alternative REPL.)
Above all, it is really fun to create an editor.
Every clone I see try to reproduce vim usage, but none really try to run vimfiles. I think running vimfiles is a must have for a clone to get real users.
Does it work with unicode Indic characters?
"There is no roadmap. I mostly implement the stuff which I need or interests me, or which gives me the opportunity to learn."
Pretty much says it all. Seems it was primarily written to satisfy personal needs, not be the end-all, be-all of professionally written software. It is, incidentally, potentially interesting to others who might want to learn from it, or use it, knowing the background.
So, you might cut the guy a break. The requirements for personal projects aren't the same.
I don't remember all of the issues, but there are a ton of small things that make the editor unusable to me. I used it for a couple of weeks, and I spent some time working on these issues, but never had PR-worthy code. Here's what I can remember off the top of my head:
- Startup time is very slow because of the way configuration works. In my local copy, I made a version without runtime configuration, and that solved this problem. This conflicts pretty badly with the whole architecture, so I didn't make a PR.
- :n :N don't work. Opening multiple files from the command line doesn't work.
- :cq doesn't work. I fixed this, but my fix was a hack, so I didn't make a PR.
- Operating on regions with '{' and '}' is off by one line in some directions.
- You can't replace regions with shell commands. For example, using '!}sort' to sort a paragraph.
I had a really hilarious bug today where under some circumstances all the text in the display would be replaced by numbers. Small integers, each placed where the word should be.
What had happened is that I'd added a layer of indirection; where previously, after line wrapping, the data structure for a rendered paragraph was an array of pointers to the word objects, now it was an array of indices into the paragraph's word array. And I'd forgotten about one particular exotic code path. Lua was seeing the array of indices, automatically casting the ints to strings, and then using those strings instead of the word data itself...
Static types would have made this bug impossible.
...way back when, there was a really nice and thoroughly obscure language called Strongtalk; it was a Smalltalk 80 clone with optional strict types. You could annotate your classes and methods with type information. If it was there, it would be checked; if it wasn't, you got the traditional behaviour. The JIT knew about the type information and could use it to produce really fast code. It combined the ultimate dynamic language with an expressive static type system (complete with parametric polymorphism).
It was open sourced in 2006 and sank without trace. Sigh.
int[] bodies = {1, 2, 3};
for (int body : bodies) {
String formatted_body = "<p>" + body + "</p>";
callMethodWithStrArg(formatted_body);
}
bodies = [1, 2, 3]
for body in bodies:
formatted_body = '<p>' + body + '</p>'
Dynamically typed languages still have types.
I had totally forgotten that Java does it too, despite having done `""+i` lots of times as a cheap and easy and evil way to convert numbers to strings.
...I am currently rewriting a big chunk of the primary data storage to use immutable data structures, because it makes implementing Undo easier. I am having to fight the urge to redo it all in Haskell.
Ignoring trivial cases and dependent types, types tell you one of two things: "this might be correct" or "this is incorrect". Again ignoring trivial cases where exhaustive checking is possible, that's the same thing that tests tell you - and it's a hugely useful thing to be told!
Types are one mechanism for static analysis. Better contracts (nullability, valid ranges, etc) goes much further.
Compiles, runs, fails if i is out of bounds. Which means that you need to test the code that you write, and that even if you have 100% line coverage (or branch coverage or MC/DC coverage or...), it doesn't mean i won't get the wrong value.
People who claim that "once my C++/Haskell/Agda/whatever program compiles, I know it probably has no bugs" thus tempt others to mention "a false sense of security." (Agda might be going further than anyone else with proving that i cannot be out of bounds, AFAIK... though a generally undecidable problem will remain generally undecidable.)
It is a substitute for those unit tests that are essentially checking type invariants. It is not a substitute for all unit tests, but I don't think anyone made (much less intended to make) that claim anywhere in this thread.
"They only tell you that you are matching function signatures correctly."
All of computation can be expressed as application of functions, so for sufficiently expressive function signatures that's not much of a limitation. Of course, if you want to guarantee that your compiles terminate, you need to apply some limits to the expressiveness of your function signatures... but there are powerful guarantees you can get out of even so simplistic a type system as C's, if you work with it rather than against it.
If however you say, "hey, I seriously don't want to cover all the lines including say error messages", then in Python,
if error: print obj.name
...might be a problem because obj doesn't have a name; in C++,
if(error) cout << obj.name;
...might only blow up because obj is a null reference (which is what you get when you dereference the null pointer, in practice, even though null references aren't supposed to exist); and in a language with non-nullable pointers, the equivalent of the above can only blow up if printing the name (which is surely valid if obj is non-nullable and type-checks as having a name) somehow blows up, which for a string in a memory-safe language is very very unlikely.
So if you leave uncovered lines in your code which we all do then yes, the stricter the type system, the better your chances are, statistically, all else being equal (for instance, the number of lines being the same... which might not be the case.)
Overall the silly tests people sometimes write in dynamic languages are needless IMO and result from over-applying "TDD" or "unit testing" or some other buzzword and/or paranoia of someone coming from a statically typed language background. I personally think I have pretty much the same amount of tests regardless of the type system.
Quite often I write some Rust code and know that it will work if it compiles (though I still write tests)
I have seen claims and have a hunch that more than just very silly tests can be eliminated by types, but I'm struggling to come up with or remember any examples.
I'm hoping someone who has one will reply with one of these claims or any examples.
print(arr[i])
print (arr `V.unsafeIndex` i)