Mercurial > vim
view runtime/doc/usr_23.txt @ 32936:c517845bd10e v9.0.1776
patch 9.0.1776: No support for stable Python 3 ABI
Commit: https://github.com/vim/vim/commit/c13b3d1350b60b94fe87f0761ea31c0e7fb6ebf3
Author: Yee Cheng Chin <ychin.git@gmail.com>
Date: Sun Aug 20 21:18:38 2023 +0200
patch 9.0.1776: No support for stable Python 3 ABI
Problem: No support for stable Python 3 ABI
Solution: Support Python 3 stable ABI
Commits:
1) Support Python 3 stable ABI to allow mixed version interoperatbility
Vim currently supports embedding Python for use with plugins, and the
"dynamic" linking option allows the user to specify a locally installed
version of Python by setting `pythonthreedll`. However, one caveat is
that the Python 3 libs are not binary compatible across minor versions,
and mixing versions can potentially be dangerous (e.g. let's say Vim was
linked against the Python 3.10 SDK, but the user sets `pythonthreedll`
to a 3.11 lib). Usually, nothing bad happens, but in theory this could
lead to crashes, memory corruption, and other unpredictable behaviors.
It's also difficult for the user to tell something is wrong because Vim
has no way of reporting what Python 3 version Vim was linked with.
For Vim installed via a package manager, this usually isn't an issue
because all the dependencies would already be figured out. For prebuilt
Vim binaries like MacVim (my motivation for working on this), AppImage,
and Win32 installer this could potentially be an issue as usually a
single binary is distributed. This is more tricky when a new Python
version is released, as there's a chicken-and-egg issue with deciding
what Python version to build against and hard to keep in sync when a new
Python version just drops and we have a mix of users of different Python
versions, and a user just blindly upgrading to a new Python could lead to
bad interactions with Vim.
Python 3 does have a solution for this problem: stable ABI / limited API
(see https://docs.python.org/3/c-api/stable.html). The C SDK limits the
API to a set of functions that are promised to be stable across
versions. This pull request adds an ifdef config that allows us to turn
it on when building Vim. Vim binaries built with this option should be
safe to freely link with any Python 3 libraies without having the
constraint of having to use the same minor version.
Note: Python 2 has no such concept and this doesn't change how Python 2
integration works (not that there is going to be a new version of Python
2 that would cause compatibility issues in the future anyway).
---
Technical details:
======
The stable ABI can be accessed when we compile with the Python 3 limited
API (by defining `Py_LIMITED_API`). The Python 3 code (in `if_python3.c`
and `if_py_both.h`) would now handle this and switch to limited API
mode. Without it set, Vim will still use the full API as before so this
is an opt-in change.
The main difference is that `PyType_Object` is now an opaque struct that
we can't directly create "static types" out of, and we have to create
type objects as "heap types" instead. This is because the struct is not
stable and changes from version to version (e.g. 3.8 added a
`tp_vectorcall` field to it). I had to change all the types to be
allocated on the heap instead with just a pointer to them.
Other functions are also simply missing in limited API, or they are
introduced too late (e.g. `PyUnicode_AsUTF8AndSize` in 3.10) to it that
we need some other ways to do the same thing, so I had to abstract a few
things into macros, and sometimes re-implement functions like
`PyObject_NEW`.
One caveat is that in limited API, `OutputType` (used for replacing
`sys.stdout`) no longer inherits from `PyStdPrinter_Type` which I don't
think has any real issue other than minor differences in how they
convert to a string and missing a couple functions like `mode()` and
`fileno()`.
Also fixed an existing bug where `tp_basicsize` was set incorrectly for
`BufferObject`, `TabListObject, `WinListObject`.
Technically, there could be a small performance drop, there is a little
more indirection with accessing type objects, and some APIs like
`PyUnicode_AsUTF8AndSize` are missing, but in practice I didn't see any
difference, and any well-written Python plugin should try to avoid
excessing callbacks to the `vim` module in Python anyway.
I only tested limited API mode down to Python 3.7, which seemes to
compile and work fine. I haven't tried earlier Python versions.
2) Fix PyIter_Check on older Python vers / type##Ptr unused warning
For PyIter_Check, older versions exposed them as either macros (used in
full API), or a function (for use in limited API). A previous change
exposed PyIter_Check to the dynamic build because Python just moved it
to function-only in 3.10 anyway. Because of that, just make sure we
always grab the function in dynamic builds in earlier versions since
that's what Python eventually did anyway.
3) Move Py_LIMITED_API define to configure script
Can now use --with-python-stable-abi flag to customize what stable ABI
version to target. Can also use an env var to do so as well.
4) Show +python/dyn-stable in :version, and allow has() feature query
Not sure if the "/dyn-stable" suffix would break things, or whether we
should do it another way. Or just don't show it in version and rely on
has() feature checking.
5) Documentation first draft. Still need to implement v:python3_version
6) Fix PyIter_Check build breaks when compiling against Python 3.8
7) Add CI coverage stable ABI on Linux/Windows / make configurable on Windows
This adds configurable options for Windows make files (both MinGW and
MSVC). CI will also now exercise both traditional full API and stable
ABI for Linux and Windows in the matrix for coverage.
Also added a "dynamic" option to Linux matrix as a drive-by change to
make other scripting languages like Ruby / Perl testable under both
static and dynamic builds.
8) Fix inaccuracy in Windows docs
Python's own docs are confusing but you don't actually want to use
`python3.dll` for the dynamic linkage.
9) Add generated autoconf file
10) Add v:python3_version support
This variable indicates the version of Python3 that Vim was built
against (PY_VERSION_HEX), and will be useful to check whether the Python
library you are loading in dynamically actually fits it. When built with
stable ABI, it will be the limited ABI version instead
(`Py_LIMITED_API`), which indicates the minimum version of Python 3 the
user should have, rather than the exact match. When stable ABI is used,
we won't be exposing PY_VERSION_HEX in this var because it just doesn't
seem necessary to do so (the whole point of stable ABI is the promise
that it will work across versions), and I don't want to confuse the user
with too many variables.
Also, cleaned up some documentation, and added help tags.
11) Fix Python 3.7 compat issues
Fix a couple issues when using limited API < 3.8
- Crash on exit: In Python 3.7, if a heap-allocated type is destroyed
before all instances are, it would cause a crash later. This happens
when we destroyed `OptionsType` before calling `Py_Finalize` when
using the limited API. To make it worse, later versions changed the
semantics and now each instance has a strong reference to its own type
and the recommendation has changed to have each instance de-ref its
own type and have its type in GC traversal. To avoid dealing with
these cross-version variations, we just don't free the heap type. They
are static types in non-limited-API anyway and are designed to last
through the entirety of the app, and we also don't restart the Python
runtime and therefore do not need it to have absolutely 0 leaks.
See:
- https://docs.python.org/3/whatsnew/3.8.html#changes-in-the-c-api
- https://docs.python.org/3/whatsnew/3.9.html#changes-in-the-c-api
- PyIter_Check: This function is not provided in limited APIs older than
3.8. Previously I was trying to mock it out using manual
PyType_GetSlot() but it was brittle and also does not actually work
properly for static types (it will generate a Python error). Just
return false. It does mean using limited API < 3.8 is not recommended
as you lose the functionality to handle iterators, but from playing
with plugins I couldn't find it to be an issue.
- Fix loading of PyIter_Check so it will be done when limited API < 3.8.
Otherwise loading a 3.7 Python lib will fail even if limited API was
specified to use it.
12) Make sure to only load `PyUnicode_AsUTF8AndSize` in needed in limited API
We don't use this function unless limited API >= 3.10, but we were
loading it regardless. Usually it's ok in Unix-like systems where Python
just has a single lib that we load from, but in Windows where there is a
separate python3.dll this would not work as the symbol would not have
been exposed in this more limited DLL file. This makes it much clearer
under what condition is this function needed.
closes: #12032
Signed-off-by: Christian Brabandt <cb@256bit.org>
Co-authored-by: Yee Cheng Chin <ychin.git@gmail.com>
author | Christian Brabandt <cb@256bit.org> |
---|---|
date | Sun, 20 Aug 2023 21:30:04 +0200 |
parents | f8116058ca76 |
children | 4635e43f2c6f |
line wrap: on
line source
*usr_23.txt* For Vim version 9.0. Last change: 2020 Dec 19 VIM USER MANUAL - by Bram Moolenaar Editing other files This chapter is about editing files that are not ordinary files. With Vim you can edit files that are compressed or encrypted. Some files need to be accessed over the internet. With some restrictions, binary files can be edited as well. |23.1| DOS, Mac and Unix files |23.2| Files on the internet |23.3| Encryption |23.4| Binary files |23.5| Compressed files Next chapter: |usr_24.txt| Inserting quickly Previous chapter: |usr_22.txt| Finding the file to edit Table of contents: |usr_toc.txt| ============================================================================== *23.1* DOS, Mac and Unix files Back in the early days, the old Teletype machines used two characters to start a new line. One to move the carriage back to the first position (carriage return, <CR>), another to move the paper up (line feed, <LF>). When computers came out, storage was expensive. Some people decided that they did not need two characters for end-of-line. The UNIX people decided they could use <New Line> or <NL> only for end-of-line. The Apple people standardized on <CR>. The Microsoft Windows folks decided to keep the old <CR><NL> (we use <NL> for line feed in the help text). This means that if you try to move a file from one system to another, you have line-break problems. The Vim editor automatically recognizes the different file formats and handles things properly behind your back. The option 'fileformats' contains the various formats that will be tried when a new file is edited. The following command, for example, tells Vim to try UNIX format first and MS-DOS format second: > :set fileformats=unix,dos You will notice the format in the message you get when editing a file. You don't see anything if you edit a native file format. Thus editing a Unix file on Unix won't result in a remark. But when you edit a dos file, Vim will notify you of this: "/tmp/test" [dos] 3L, 71C ~ For a Mac file you would see "[mac]". The detected file format is stored in the 'fileformat' option. To see which format you have, execute the following command: > :set fileformat? The three names that Vim uses are: unix <NL> dos <CR><NL> mac <CR> USING THE MAC FORMAT On Unix, <NL> is used to break a line. It's not unusual to have a <CR> character halfway a line. Incidentally, this happens quite often in Vi (and Vim) scripts. On the Macintosh, where <CR> is the line break character, it's possible to have a <NL> character halfway a line. The result is that it's not possible to be 100% sure whether a file containing both <CR> and <NL> characters is a Mac or a Unix file. Therefore, Vim assumes that on Unix you probably won't edit a Mac file, and doesn't check for this type of file. To check for this format anyway, add "mac" to 'fileformats': > :set fileformats+=mac Then Vim will take a guess at the file format. Watch out for situations where Vim guesses wrong. OVERRULING THE FORMAT If you use the good old Vi and try to edit an MS-DOS format file, you will find that each line ends with a ^M character. (^M is <CR>). The automatic detection avoids this. Suppose you do want to edit the file that way? Then you need to overrule the format: > :edit ++ff=unix file.txt The "++" string is an item that tells Vim that an option name follows, which overrules the default for this single command. "++ff" is used for 'fileformat'. You could also use "++ff=mac" or "++ff=dos". This doesn't work for any option, only "++ff" and "++enc" are currently implemented. The full names "++fileformat" and "++encoding" also work. CONVERSION You can use the 'fileformat' option to convert from one file format to another. Suppose, for example, that you have an MS-DOS file named README.TXT that you want to convert to UNIX format. Start by editing the MS-DOS format file: > vim README.TXT Vim will recognize this as a dos format file. Now change the file format to UNIX: > :set fileformat=unix :write The file is written in Unix format. ============================================================================== *23.2* Files on the internet Someone sends you an e-mail message, which refers to a file by its URL. For example: You can find the information here: ~ ftp://ftp.vim.org/pub/vim/README ~ You could start a program to download the file, save it on your local disk and then start Vim to edit it. There is a much simpler way. Move the cursor to any character of the URL. Then use this command: > gf With a bit of luck, Vim will figure out which program to use for downloading the file, download it and edit the copy. To open the file in a new window use CTRL-W f. If something goes wrong you will get an error message. It's possible that the URL is wrong, you don't have permission to read it, the network connection is down, etc. Unfortunately, it's hard to tell the cause of the error. You might want to try the manual way of downloading the file. Accessing files over the internet works with the netrw plugin. Currently URLs with these formats are recognized: ftp:// uses ftp rcp:// uses rcp scp:// uses scp http:// uses wget (reading only) Vim doesn't do the communication itself, it relies on the mentioned programs to be available on your computer. On most Unix systems "ftp" and "rcp" will be present. "scp" and "wget" might need to be installed. Vim detects these URLs for each command that starts editing a new file, also with ":edit" and ":split", for example. Write commands also work, except for http://. For more information, also about passwords, see |netrw|. ============================================================================== *23.3* Encryption Some information you prefer to keep to yourself. For example, when writing a test on a computer that students also use. You don't want clever students to figure out a way to read the questions before the exam starts. Vim can encrypt the file for you, which gives you some protection. To start editing a new file with encryption, use the "-x" argument to start Vim. Example: > vim -x exam.txt Vim prompts you for a key used for encrypting and decrypting the file: Enter encryption key: ~ Carefully type the secret key now. You cannot see the characters you type, they will be replaced by stars. To avoid the situation that a typing mistake will cause trouble, Vim asks you to enter the key again: Enter same key again: ~ You can now edit this file normally and put in all your secrets. When you finish editing the file and tell Vim to exit, the file is encrypted and written. When you edit the file with Vim, it will ask you to enter the same key again. You don't need to use the "-x" argument. You can also use the normal ":edit" command. Vim adds a magic string to the file by which it recognizes that the file was encrypted. If you try to view this file using another program, all you get is garbage. Also, if you edit the file with Vim and enter the wrong key, you get garbage. Vim does not have a mechanism to check if the key is the right one (this makes it much harder to break the key). SWITCHING ENCRYPTION ON AND OFF To disable the encryption of a file, set the 'key' option to an empty string: > :set key= The next time you write the file this will be done without encryption. Setting the 'key' option to enable encryption is not a good idea, because the password appears in the clear. Anyone shoulder-surfing can read your password. To avoid this problem, the ":X" command was created. It asks you for an encryption key, just like the "-x" argument did: > :X Enter encryption key: ****** Enter same key again: ****** LIMITS ON ENCRYPTION The encryption algorithm used by Vim is not very strong. It is good enough to keep out the casual prowler, but not good enough to keep out a cryptology expert with lots of time on his hands. The text in the swap file and the undo file is also encrypted. However, this is done block-by-block and may reduce the time needed to crack a password. You can disable the swap file, but then a crash will cause you to lose your work, since Vim keeps all the text in memory only. The undo file can be disabled with the only disadvantage that you can't undo after unloading the buffer. To avoid using a swap file, supply the -n argument on the command line. For example, to edit the encrypted file "file.txt" without a swap file use the following command: > vim -x -n file.txt When already editing a file, the swapfile can be disabled with: > :setlocal noswapfile Since there is no swapfile, recovery will be impossible. Save the file a bit more often to avoid the risk of losing your changes. While the file is in memory, it is in plain text. Anyone with privilege can look in the editor's memory and discover the contents of the file. If you use a viminfo file, be aware that the contents of text registers are written out in the clear as well. If you really want to secure the contents of a file, edit it only on a portable computer not connected to a network, use good encryption tools, and keep the computer locked up in a big safe when not in use. ============================================================================== *23.4* Binary files You can edit binary files with Vim. Vim wasn't really made for this, thus there are a few restrictions. But you can read a file, change a character and write it back, with the result that only that one character was changed and the file is identical otherwise. To make sure that Vim does not use its clever tricks in the wrong way, add the "-b" argument when starting Vim: > vim -b datafile This sets the 'binary' option. The effect of this is that unexpected side effects are turned off. For example, 'textwidth' is set to zero, to avoid automatic formatting of lines. And files are always read in Unix file format. Binary mode can be used to change a message in a program. Be careful not to insert or delete any characters, it would stop the program from working. Use "R" to enter replace mode. Many characters in the file will be unprintable. To see them in Hex format: > :set display=uhex Otherwise, the "ga" command can be used to see the value of the character under the cursor. The output, when the cursor is on an <Esc>, looks like this: <^[> 27, Hex 1b, Octal 033 ~ There might not be many line breaks in the file. To get some overview switch the 'wrap' option off: > :set nowrap BYTE POSITION To see on which byte you are in the file use this command: > g CTRL-G The output is verbose: Col 9-16 of 9-16; Line 277 of 330; Word 1806 of 2058; Byte 10580 of 12206 ~ The last two numbers are the byte position in the file and the total number of bytes. This takes into account how 'fileformat' changes the number of bytes that a line break uses. To move to a specific byte in the file, use the "go" command. For example, to move to byte 2345: > 2345go USING XXD A real binary editor shows the text in two ways: as it is and in hex format. You can do this in Vim by first converting the file with the "xxd" program. This comes with Vim. First edit the file in binary mode: > vim -b datafile Now convert the file to a hex dump with xxd: > :%!xxd The text will look like this: 0000000: 1f8b 0808 39d7 173b 0203 7474 002b 4e49 ....9..;..tt.+NI ~ 0000010: 4b2c 8660 eb9c ecac c462 eb94 345e 2e30 K,.`.....b..4^.0 ~ 0000020: 373b 2731 0b22 0ca6 c1a2 d669 1035 39d9 7;'1.".....i.59. ~ You can now view and edit the text as you like. Vim treats the information as ordinary text. Changing the hex does not cause the printable character to be changed, or the other way around. Finally convert it back with: > :%!xxd -r Only changes in the hex part are used. Changes in the printable text part on the right are ignored. See the manual page of xxd for more information. ============================================================================== *23.5* Compressed files This is easy: You can edit a compressed file just like any other file. The "gzip" plugin takes care of decompressing the file when you edit it. And compressing it again when you write it. These compression methods are currently supported: .Z compress .gz gzip .bz2 bzip2 Vim uses the mentioned programs to do the actual compression and decompression. You might need to install the programs first. ============================================================================== Next chapter: |usr_24.txt| Inserting quickly Copyright: see |manual-copyright| vim:tw=78:ts=8:noet:ft=help:norl: