Python 3 deals with this reasonable sensibly, too, I think. They use UTF-8 by default, but allow you to specify other encodings.
It uses Latin-1 for ASCII strings, UCS-2 for strings that contain code points in the BMP and UCS-4 only for strings that contain code points outside the BMP.
It would be pretty silly for them to explode all strings to 4-byte characters.
They need at most 21 bits. The bits may only be available in multiples of 8, but the implementation also doesn't byte-pack them into 24-bit units, so that's moot.
My understanding of the current "always and only utf-8/unicode" zeitgeist is that is comes mostly from encoding issues among which the complexity of detecting encoding.
I think that the current status quo is better than what came before, but I also think it could be improved.
The languages that i really dont get are those that force valid utf-8 everywhere but dont enforce NFC. Which is most of them but seems like the worst of both worlds.
Non normalized unicode is just as problematic as non validated unicode imo.
But if you do want a sequence of bytes for whatever reason, you can trivially obtain that in any version of Python.
I'll probably just use rust for that script if python2 ever gets dropped by my distro. Reminds me of https://gregoryszorc.com/blog/2020/01/13/mercurial%27s-journ...
Show me.
This is a script created by someone on #nethack a long time ago. It works great with other things as well like old BBS games. It was intended to transparently rewrite single byte encodings to multibyte with an optional conversion array.
It almost works as-is in my testing. (By the way, there's a typo in the usage message.) Here is my test process:
#!/usr/bin/env python
import random, sys, time
def out(b):
# ASCII 0..7 for the second digit of the color code in the escape sequence
color = random.randint(48, 55)
sys.stdout.buffer.write(bytes([27, 91, 51, color, 109, b]))
sys.stdout.flush()
for i in range(32, 256):
out(i)
time.sleep(random.random()/5)
while True:
out(random.randint(32, 255))
time.sleep(0.1)
(My initial version of this didn't flush the output, which mistakenly lead me to try a bunch of unnecessary things in the main script.)
After fixing the `print` calls, the only thing I was forced to change (although I would do the code differently overall) is the output step:
# sys.stdout.write(out.encode("UTF-8"))
sys.stdout.buffer.write(out.encode("UTF-8"))
sys.stdout.flush()
There's a warning that `bufsize=1` no longer actually means a byte buffer of size 1 for reading (instead it's magically interpreted as a request for line buffering), but this didn't cause a failure when I tried it. (And setting the size to e.g. `2` didn't break things, either.)
I also tried having my test process read from standard input; the handling of ctrl-C and ctrl-D seems to be a bit different (and in general, setting up a Python process to read unbuffered bytes from stdin isn't the most fun thing), but I generally couldn't find any issues here, either. Which is to say, the problems there are in the test process, not in `ibmfilter`. The input is still forwarded to, and readable from, the test process via the `Popen` object. And any problems of this sort are definitely still fixable, as demonstrated by the fact that `curses` is still in the standard library.
Of course, keys in the `special` mapping need to be defined as bytes literals now. Although that could trivially be adapted if you insist.
As for typo, yep. But then, I've left this script essentially untouched for a couple of decades since I was given it.
Here's a diff:
diff --git a/ibmfilter b/ibmfilter
index 245d32c..2633335 100755
--- a/ibmfilter
+++ b/ibmfilter
@@ -1,6 +1,5 @@
-#!/usr/bin/python2 -tt
-# vim:set fileencoding=utf-8
-
+#!/usr/bin/python3
+
from subprocess import *
import sys
import os, select
@@ -10,8 +9,8 @@ special = {
}
if len(sys.argv) < 2:
- print "usage: ibmfilter [command]"
- print "Runs command in a subshell and translates its output from ibm473 codepage to UTF-8."
+ print("usage: ibmfilter [command]")
+ print("Runs command in a subshell and translates its output from ibm473 codepage to UTF-8.")
sys.exit(0)
handle = Popen(sys.argv[1:], stdout=PIPE, bufsize=1)
@@ -26,8 +25,10 @@ while buf != '':
os.kill(handle.pid)
os.system('reset')
raise Exception("Timed out while waiting for stdout to be writeable...")
- sys.stdout.write(out.encode("UTF-8"))
-
+ sys.stdout.buffer.write(out.encode("UTF-8"))
+ sys.stdout.flush()
+
buf = handle.stdout.read(1)
handle.wait()
If you want to add overrides, you must use bytes literals for the keys. That looks like:
b'\xff': 'X'
Some Python programs have mutually incompatible dependencies, and you can't really have two versions of the same dependency loaded in the same runtime. This has always been a problem; you're just looking at the current iteration of pip trying to cooperate with Linux distros to help you not break your system as a result.
"Using the same set" is not actually desirable for development.
And with that out of the way. This one seems to mostly work!
So python3 did not significantly change handling this sort of byte stream and while Mercurial folks might well have had their own woes, I have no idea what the issues were in all those prior attempts with this file.
... that said, it does do one odd thing (following is output on launching):
/usr/lib/python3.12/subprocess.py:1016: RuntimeWarning: line buffering (buffering=1) isn't supported in binary mode, the default buffer size will be used
self.stdout = io.open(c2pread, 'rb', bufsize)
At least for now, I'm going to tentatively say it seems fine. Hm. You know what. Let me try with some more obvious things that might fail if the buffer size is wrong.
So. Now I'm wondering why, given how relatively minor this change is (aside from the odd error message, and the typical python3 changes just one slightly modified line and one inserted line), why did so many pythonistas have so much difficulty over the many years I asked about this? I mean, I only formed my opinion that maybe there was a problem with python3 byte/string handling to just how many attempts there were... Were they trying to do things in a more idiomatic python3 fashion? Did the python3 APIs change? Does the error hint at something more concerning? Well, whatever. Clearly it's (mostly) fine now. And my carefully tweaked nethack profile is safe if python2 is removed without needing to make my own stream filter. Yay! Thanks!
... further updates.. ok there are a few issues.
1) the warning
2) there's an odd ghost flicker that jumps around the nethack level as if a cursor is appearing - does not happen in the python2 one.
3) on quitting it no longer it no longer exits gracefully and I have to ctrl-c the script.
4) It is much slower to render. the python2 one draws a screen almost instantly for most uses (although still a bit slower than not filtered, at least on this computer, for things that change a lot, like video). This one ripples down - that might explain the ghost flickering in ② and might be related to the buffer warning. This becomes much more noticeable with BBSes although it is usually fine in nethack. You can see the difference on a simpler testcase without setting up a BBS account by streaming a bit more data at once say by running: ibmfilter curl ascii.live/nyan
So, clearly not perfect but.. eh. functional? Still far better than prior attempts, and at least it mostly works with nethack.
Yes, that would be exactly why. You can use e.g. `sed` to remove leading whitespace from each line (I used it to add the leading whitespace for posting).
> ... that said, it does do one odd thing (following is output on launching):
Yes, that's the warning I mentioned. The original code requests to use a buffer size of 1, which is no longer supported (it now means to use line buffering).
> It is much slower to render.
Avoiding line buffering (by requiring a buffer size of 2 or more) might fix that. Actually, it might be a good idea to use a significantly larger buffer, so that e.g. an entire ANSI colour code can be read all at once.
The other issues are, I'm pretty sure, because of other things that changed in how `subprocess` works. Fixing things at this level would indeed require quite a bit more hacking around with the low-level terminal APIs.
> I mean, I only formed my opinion that maybe there was a problem with python3 byte/string handling to just how many attempts there were... Were they trying to do things in a more idiomatic python3 fashion? Did the python3 APIs change? Does the error hint at something more concerning?
Most likely, other attempts either a) didn't understand what the original code was doing in precise enough detail, or b) didn't know how to send binary data to standard output properly (Python 3 defaults to opening standard output as a text stream).
All of that said: I think that nowadays you should just be able to get a build of NetHack that just outputs UTF-8 characters directly; failing that, you can use the `locale` command to tell your terminal to expect cp437 data.
The unfortunate thing is the "lag" is a bit annoying with some apps, so I'll probably still use the python2 one for now.