Mercurial > vim
view runtime/doc/usr_45.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_45.txt* For Vim version 9.0. Last change: 2022 May 13 VIM USER MANUAL - by Bram Moolenaar Select your language (locale) The messages in Vim can be given in several languages. This chapter explains how to change which one is used. Also, the different ways to work with files in various languages is explained. |45.1| Language for Messages |45.2| Language for Menus |45.3| Using another encoding |45.4| Editing files with a different encoding |45.5| Entering language text Next chapter: |usr_50.txt| Advanced Vim script writing Previous chapter: |usr_44.txt| Your own syntax highlighted Table of contents: |usr_toc.txt| ============================================================================== *45.1* Language for Messages When you start Vim, it checks the environment to find out what language you are using. Mostly this should work fine, and you get the messages in your language (if they are available). To see what the current language is, use this command: > :language If it replies with "C", this means the default is being used, which is English. Note: Using different languages only works when Vim was compiled to handle it. To find out if it works, use the ":version" command and check the output for "+gettext" and "+multi_lang". If they are there, you are OK. If you see "-gettext" or "-multi_lang" you will have to find another Vim. What if you would like your messages in a different language? There are several ways. Which one you should use depends on the capabilities of your system. The first way is to set the environment to the desired language before starting Vim. Example for Unix: > env LANG=de_DE.ISO_8859-1 vim This only works if the language is available on your system. The advantage is that all the GUI messages and things in libraries will use the right language as well. A disadvantage is that you must do this before starting Vim. If you want to change language while Vim is running, you can use the second method: > :language fr_FR.ISO_8859-1 This way you can try out several names for your language. You will get an error message when it's not supported on your system. You don't get an error when translated messages are not available. Vim will silently fall back to using English. To find out which languages are supported on your system, find the directory where they are listed. On my system it is "/usr/share/locale". On some systems it's in "/usr/lib/locale". The manual page for "setlocale" should give you a hint where it is found on your system. Be careful to type the name exactly as it should be. Upper and lowercase matter, and the '-' and '_' characters are easily confused. You can also set the language separately for messages, edited text and the time format. See |:language|. DO-IT-YOURSELF MESSAGE TRANSLATION If translated messages are not available for your language, you could write them yourself. To do this, get the source code for Vim and the GNU gettext package. After unpacking the sources, instructions can be found in the directory src/po/README.txt. It's not too difficult to do the translation. You don't need to be a programmer. You must know both English and the language you are translating to, of course. When you are satisfied with the translation, consider making it available to others. Upload it at vim-online (http://vim.sf.net) or e-mail it to the Vim maintainer <maintainer@vim.org>. Or both. ============================================================================== *45.2* Language for Menus The default menus are in English. To be able to use your local language, they must be translated. Normally this is automatically done for you if the environment is set for your language, just like with messages. You don't need to do anything extra for this. But it only works if translations for the language are available. Suppose you are in Germany, with the language set to German, but prefer to use "File" instead of "Datei". You can switch back to using the English menus this way: > :set langmenu=none It is also possible to specify a language: > :set langmenu=nl_NL.ISO_8859-1 Like above, differences between "-" and "_" matter. However, upper/lowercase differences are ignored here. The 'langmenu' option must be set before the menus are loaded. Once the menus have been defined changing 'langmenu' has no direct effect. Therefore, put the command to set 'langmenu' in your vimrc file. If you really want to switch menu language while running Vim, you can do it this way: > :source $VIMRUNTIME/delmenu.vim :set langmenu=de_DE.ISO_8859-1 :source $VIMRUNTIME/menu.vim There is one drawback: All menus that you defined yourself will be gone. You will need to redefine them as well. DO-IT-YOURSELF MENU TRANSLATION To see which menu translations are available, look in this directory: $VIMRUNTIME/lang ~ The files are called menu_{language}.vim. If you don't see the language you want to use, you can do your own translations. The simplest way to do this is by copying one of the existing language files, and change it. First find out the name of your language with the ":language" command. Use this name, but with all letters made lowercase. Then copy the file to your own runtime directory, as found early in 'runtimepath'. For example, for Unix you would do: > :!cp $VIMRUNTIME/lang/menu_ko_kr.euckr.vim ~/.vim/lang/menu_nl_be.iso_8859-1.vim You will find hints for the translation in "$VIMRUNTIME/lang/README.txt". ============================================================================== *45.3* Using another encoding Vim guesses that the files you are going to edit are encoded for your language. For many European languages this is "latin1". Then each byte is one character. That means there are 256 different characters possible. For Asian languages this is not sufficient. These mostly use a double-byte encoding, providing for over ten thousand possible characters. This still isn't enough when a text is to contain several different languages. This is where Unicode comes in. It was designed to include all characters used in commonly used languages. This is the "Super encoding that replaces all others". But it isn't used that much yet. Fortunately, Vim supports these three kinds of encodings. And, with some restrictions, you can use them even when your environment uses another language than the text. Nevertheless, when you only edit files that are in the encoding of your language, the default should work fine and you don't need to do anything. The following is only relevant when you want to edit different languages. USING UNICODE IN THE GUI The nice thing about Unicode is that other encodings can be converted to it and back without losing information. When you make Vim use Unicode internally, you will be able to edit files in any encoding. Unfortunately, the number of systems supporting Unicode is still limited. Thus it's unlikely that your language uses it. You need to tell Vim you want to use Unicode, and how to handle interfacing with the rest of the system. Let's start with the GUI version of Vim, which is able to display Unicode characters. This should work: > :set encoding=utf-8 :set guifont=-misc-fixed-medium-r-normal--18-120-100-100-c-90-iso10646-1 The 'encoding' option tells Vim the encoding of the characters that you use. This applies to the text in buffers (files you are editing), registers, Vim script files, etc. You can regard 'encoding' as the setting for the internals of Vim. This example assumes you have this font on your system. The name in the example is for the X Window System. This font is in a package that is used to enhance xterm with Unicode support. If you don't have this font, you might find it here: http://www.cl.cam.ac.uk/~mgk25/download/ucs-fonts.tar.gz ~ For MS-Windows, some fonts have a limited number of Unicode characters. Try using the "Courier New" font. You can use the Edit/Select Font... menu to select and try out the fonts available. Only fixed-width fonts can be used though. Example: > :set guifont=courier_new:h12 If it doesn't work well, try getting a fontpack. If Microsoft didn't move it, you can find it here: http://www.microsoft.com/typography/fonts/default.aspx ~ Now you have told Vim to use Unicode internally and display text with a Unicode font. Typed characters still arrive in the encoding of your original language. This requires converting them to Unicode. Tell Vim the language from which to convert with the 'termencoding' option. You can do it like this: > :let &termencoding = &encoding :set encoding=utf-8 This assigns the old value of 'encoding' to 'termencoding' before setting 'encoding' to utf-8. You will have to try out if this really works for your setup. It should work especially well when using an input method for an Asian language, and you want to edit Unicode text. USING UNICODE IN A UNICODE TERMINAL There are terminals that support Unicode directly. The standard xterm that comes with XFree86 is one of them. Let's use that as an example. First of all, the xterm must have been compiled with Unicode support. See |UTF8-xterm| how to check that and how to compile it when needed. Start the xterm with the "-u8" argument. You might also need so specify a font. Example: > xterm -u8 -fn -misc-fixed-medium-r-normal--18-120-100-100-c-90-iso10646-1 Now you can run Vim inside this terminal. Set 'encoding' to "utf-8" as before. That's all. USING UNICODE IN AN ORDINARY TERMINAL Suppose you want to work with Unicode files, but don't have a terminal with Unicode support. You can do this with Vim, although characters that are not supported by the terminal will not be displayed. The layout of the text will be preserved. > :let &termencoding = &encoding :set encoding=utf-8 This is the same as what was used for the GUI. But it works differently: Vim will convert the displayed text before sending it to the terminal. That avoids that the display is messed up with strange characters. For this to work the conversion between 'termencoding' and 'encoding' must be possible. Vim will convert from latin1 to Unicode, thus that always works. For other conversions the |+iconv| feature is required. Try editing a file with Unicode characters in it. You will notice that Vim will put a question mark (or underscore or some other character) in places where a character should be that the terminal can't display. Move the cursor to a question mark and use this command: > ga Vim will display a line with the code of the character. This gives you a hint about what character it is. You can look it up in a Unicode table. You could actually view a file that way, if you have lots of time at hand. Note: Since 'encoding' is used for all text inside Vim, changing it makes all non-ASCII text invalid. You will notice this when using registers and the 'viminfo' file (e.g., a remembered search pattern). It's recommended to set 'encoding' in your vimrc file, and leave it alone. ============================================================================== *45.4* Editing files with a different encoding Suppose you have setup Vim to use Unicode, and you want to edit a file that is in 16-bit Unicode. Sounds simple, right? Well, Vim actually uses utf-8 encoding internally, thus the 16-bit encoding must be converted, since there is a difference between the character set (Unicode) and the encoding (utf-8 or 16-bit). Vim will try to detect what kind of file you are editing. It uses the encoding names in the 'fileencodings' option. When using Unicode, the default value is: "ucs-bom,utf-8,latin1". This means that Vim checks the file to see if it's one of these encodings: ucs-bom File must start with a Byte Order Mark (BOM). This allows detection of 16-bit, 32-bit and utf-8 Unicode encodings. utf-8 utf-8 Unicode. This is rejected when a sequence of bytes is illegal in utf-8. latin1 The good old 8-bit encoding. Always works. When you start editing that 16-bit Unicode file, and it has a BOM, Vim will detect this and convert the file to utf-8 when reading it. The 'fileencoding' option (without s at the end) is set to the detected value. In this case it is "utf-16le". That means it's Unicode, 16-bit and little-endian. This file format is common on MS-Windows (e.g., for registry files). When writing the file, Vim will compare 'fileencoding' with 'encoding'. If they are different, the text will be converted. An empty value for 'fileencoding' means that no conversion is to be done. Thus the text is assumed to be encoded with 'encoding'. If the default 'fileencodings' value is not good for you, set it to the encodings you want Vim to try. Only when a value is found to be invalid will the next one be used. Putting "latin1" first doesn't work, because it is never illegal. An example, to fall back to Japanese when the file doesn't have a BOM and isn't utf-8: > :set fileencodings=ucs-bom,utf-8,sjis See |encoding-values| for suggested values. Other values may work as well. This depends on the conversion available. FORCING AN ENCODING If the automatic detection doesn't work you must tell Vim what encoding the file is. Example: > :edit ++enc=koi8-r russian.txt The "++enc" part specifies the name of the encoding to be used for this file only. Vim will convert the file from the specified encoding, Russian in this example, to 'encoding'. 'fileencoding' will also be set to the specified encoding, so that the reverse conversion can be done when writing the file. The same argument can be used when writing the file. This way you can actually use Vim to convert a file. Example: > :write ++enc=utf-8 russian.txt < Note: Conversion may result in lost characters. Conversion from an encoding to Unicode and back is mostly free of this problem, unless there are illegal characters. Conversion from Unicode to other encodings often loses information when there was more than one language in the file. ============================================================================== *45.5* Entering language text Computer keyboards don't have much more than a hundred keys. Some languages have thousands of characters, Unicode has over hundred thousand. So how do you type these characters? First of all, when you don't use too many of the special characters, you can use digraphs. This was already explained in |24.9|. When you use a language that uses many more characters than keys on your keyboard, you will want to use an Input Method (IM). This requires learning the translation from typed keys to resulting character. When you need an IM you probably already have one on your system. It should work with Vim like with other programs. For details see |mbyte-XIM| for the X Window system and |mbyte-IME| for MS-Windows. KEYMAPS For some languages the character set is different from latin, but uses a similar number of characters. It's possible to map keys to characters. Vim uses keymaps for this. Suppose you want to type Hebrew. You can load the keymap like this: > :set keymap=hebrew Vim will try to find a keymap file for you. This depends on the value of 'encoding'. If no matching file was found, you will get an error message. Now you can type Hebrew in Insert mode. In Normal mode, and when typing a ":" command, Vim automatically switches to English. You can use this command to switch between Hebrew and English: > CTRL-^ This only works in Insert mode and Command-line mode. In Normal mode it does something completely different (jumps to alternate file). The usage of the keymap is indicated in the mode message, if you have the 'showmode' option set. In the GUI Vim will indicate the usage of keymaps with a different cursor color. You can also change the usage of the keymap with the 'iminsert' and 'imsearch' options. To see the list of mappings, use this command: > :lmap To find out which keymap files are available, in the GUI you can use the Edit/Keymap menu. Otherwise you can use this command: > :echo globpath(&rtp, "keymap/*.vim") DO-IT-YOURSELF KEYMAPS You can create your own keymap file. It's not very difficult. Start with a keymap file that is similar to the language you want to use. Copy it to the "keymap" directory in your runtime directory. For example, for Unix, you would use the directory "~/.vim/keymap". The name of the keymap file must look like this: keymap/{name}.vim ~ or keymap/{name}_{encoding}.vim ~ {name} is the name of the keymap. Chose a name that is obvious, but different from existing keymaps (unless you want to replace an existing keymap file). {name} cannot contain an underscore. Optionally, add the encoding used after an underscore. Examples: keymap/hebrew.vim ~ keymap/hebrew_utf-8.vim ~ The contents of the file should be self-explanatory. Look at a few of the keymaps that are distributed with Vim. For the details, see |mbyte-keymap|. LAST RESORT If all other methods fail, you can enter any character with CTRL-V: encoding type range ~ 8-bit CTRL-V 123 decimal 0-255 8-bit CTRL-V x a1 hexadecimal 00-ff 16-bit CTRL-V u 013b hexadecimal 0000-ffff 31-bit CTRL-V U 001303a4 hexadecimal 00000000-7fffffff Don't type the spaces. See |i_CTRL-V_digit| for the details. ============================================================================== Next chapter: |usr_50.txt| Advanced Vim script writing Copyright: see |manual-copyright| vim:tw=78:ts=8:noet:ft=help:norl: