⚡ Sangsakala Proxy 🏠 Home
Skip to content

Align CSV dialect binding and validation with CPython 3.15 #9684

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Draft
youknowdot wants to merge 3 commits into
base: main
Choose a base branch
from
Draft

Align CSV dialect binding and validation with CPython 3.15 #9684

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
4e047b0
Align CSV dialect binding and validation with CPython 3.15
youknowdot Nov 4, 2026
a37d741
Remove unused CSV binding trait import
youknowdot Nov 4, 2026
d9cf86c
Merge main while preserving CSV dialect controls
youknowdot Nov 4, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
193 changes: 83 additions & 110 deletions crates/stdlib/src/csv.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,7 +7,7 @@ mod _csv {
AsObject, Py, PyObject, PyObjectRef, PyPayload, PyRef, PyResult, TryFromObject,
VirtualMachine,
builtins::{PyBaseExceptionRef, PyInt, PyStr, PyStrRef, PyType, PyTypeRef},
function::{ArgIterable, ArgumentError, FromArgs, FuncArgs, OptionalArg, Param},
function::{ArgIterable, FuncArgs, KwArgs, OptionalArg},
protocol::{PyIter, PyIterReturn, PyNumber},
types::{Constructor, IterNext, Iterable, SelfIter},
vm::thread::PyOperation,
Expand Down Expand Up @@ -195,11 +195,7 @@ mod _csv {
}

#[pyfunction]
fn register_dialect(
name: PyObjectRef,
opts: FormatOptions,
vm: &VirtualMachine,
) -> PyResult<()> {
fn register_dialect(name: PyObjectRef, opts: DialectArg, vm: &VirtualMachine) -> PyResult<()> {
let name = name
.downcast::<PyStr>()
.map_err(|_| vm.new_type_error("argument 0 must be a string"))?;
Expand Down Expand Up @@ -281,7 +277,7 @@ mod _csv {
}

#[pyfunction]
fn reader(iterable: PyIter, options: FormatOptions, vm: &VirtualMachine) -> PyResult<Reader> {
fn reader(iterable: PyIter, options: DialectArg, vm: &VirtualMachine) -> PyResult<Reader> {
let dialect = options.result(vm)?;
Ok(Reader {
iter: iterable,
Expand All @@ -295,11 +291,7 @@ mod _csv {
}

#[pyfunction]
fn writer(
fileobj: PyObjectRef,
options: FormatOptions,
vm: &VirtualMachine,
) -> PyResult<Writer> {
fn writer(fileobj: PyObjectRef, options: DialectArg, vm: &VirtualMachine) -> PyResult<Writer> {
let write = match vm.get_attribute_opt(&fileobj, "write")? {
Some(write_meth) => write_meth,
None if fileobj.is_callable() => fileobj,
Expand Down Expand Up @@ -365,109 +357,89 @@ mod _csv {
}
}

#[derive(Default)]
struct FormatOptions {
dialect: Option<PyObjectRef>,
delimiter: Option<PyObjectRef>,
doublequote: Option<PyObjectRef>,
escapechar: Option<PyObjectRef>,
lineterminator: Option<PyObjectRef>,
quotechar: Option<PyObjectRef>,
quoting: Option<PyObjectRef>,
skipinitialspace: Option<PyObjectRef>,
strict: Option<PyObjectRef>,
#[derive(FromArgs)]
struct DialectArg {
#[pyarg(positional, optional)]
dialect: OptionalArg<PyObjectRef>,
#[pyarg(flatten)]
fmtparams: KwArgs,
}

impl FromArgs for FormatOptions {
const PARAMS: Option<&'static [Param]> = Some(&[
Param {
name: "dialect",
kind: rustpython_vm::function::ParamKind::PositionalOrKeyword,
default: Some(rustpython_vm::function::DefaultRepr::Str("excel")),
},
Param::var_keyword("fmtparams"),
]);

fn arity() -> core::ops::RangeInclusive<usize> {
0..=1
impl DialectArg {
fn result(self, vm: &VirtualMachine) -> PyResult<PyDialect> {
// Reader, writer and registration accept only one positional dialect.
// Its format parameters are bound by Dialect after their own inputs.
FuncArgs::new(
self.dialect.into_option().into_iter().collect::<Vec<_>>(),
self.fmtparams,
)
.bind::<FormatOptions>(vm)?
.result(vm)
}
}

fn from_args(vm: &VirtualMachine, args: &mut FuncArgs) -> Result<Self, ArgumentError> {
let res = Self {
dialect: args.take_positional_keyword("dialect"),
delimiter: args.kwargs.swap_remove("delimiter"),
doublequote: args.kwargs.swap_remove("doublequote"),
escapechar: args.kwargs.swap_remove("escapechar"),
lineterminator: args.kwargs.swap_remove("lineterminator"),
quotechar: args.kwargs.swap_remove("quotechar"),
quoting: args.kwargs.swap_remove("quoting"),
skipinitialspace: args.kwargs.swap_remove("skipinitialspace"),
strict: args.kwargs.swap_remove("strict"),
};

if let Some(last_arg) = args.kwargs.pop() {
// The dialect is parsed by a parser of its own, which has no
// name to give the message.
return Err(ArgumentError::Exception(
vm.new_unexpected_keyword_type_error(None, &last_arg.0.to_string()),
));
}
Ok(res)
}
#[derive(FromArgs)]
struct FormatOptions {
#[pyarg(any, optional)]
dialect: OptionalArg<PyObjectRef>,
#[pyarg(any, optional)]
delimiter: OptionalArg<PyObjectRef>,
#[pyarg(any, optional)]
doublequote: OptionalArg<PyObjectRef>,
#[pyarg(any, optional)]
escapechar: OptionalArg<PyObjectRef>,
#[pyarg(any, optional)]
lineterminator: OptionalArg<PyObjectRef>,
#[pyarg(any, optional)]
quotechar: OptionalArg<PyObjectRef>,
#[pyarg(any, optional)]
quoting: OptionalArg<PyObjectRef>,
#[pyarg(any, optional)]
skipinitialspace: OptionalArg<PyObjectRef>,
#[pyarg(any, optional)]
strict: OptionalArg<PyObjectRef>,
}

fn validate_dialect(vm: &VirtualMachine, dialect: &PyDialect) -> PyResult<()> {
let special = |name: &str, value: char| {
if matches!(value, '\r' | '\n') {
Err(vm.new_value_error(format!(
"{name} must be a single character, not a line break"
)))
} else {
Ok(())
let check_char = |name: &str, value: Option<char>, allow_space: bool| {
let Some(value) = value else {
return Ok(());
};
if matches!(value, '\r' | '\n') || (value == ' ' && !allow_space) {
return Err(vm.new_value_error(format!("bad {name} value")));
}
if dialect.lineterminator.contains(value) {
return Err(vm.new_value_error(format!("bad {name} or lineterminator value")));
}
Ok(())
};

special("delimiter", dialect.delimiter)?;
if let Some(quotechar) = dialect.quotechar {
special("quotechar", quotechar)?;
}
if let Some(escapechar) = dialect.escapechar {
special("escapechar", escapechar)?;
}

if dialect.skipinitialspace
&& (matches!(dialect.escapechar, Some(' ')) || matches!(dialect.quotechar, Some(' ')))
{
return Err(vm.new_value_error(
"escapechar or quotechar cannot be a space when skipinitialspace is enabled",
));
}
check_char("delimiter", Some(dialect.delimiter), true)?;
check_char("escapechar", dialect.escapechar, !dialect.skipinitialspace)?;
check_char("quotechar", dialect.quotechar, !dialect.skipinitialspace)?;

let values: [(&str, Option<char>); 3] = [
let values = [
("delimiter", Some(dialect.delimiter)),
("quotechar", dialect.quotechar),
("escapechar", dialect.escapechar),
("quotechar", dialect.quotechar),
];
for (index, (left_name, left)) in values.iter().enumerate() {
for (right_name, right) in values.iter().skip(index + 1) {
if left.is_some() && left == right {
return Err(vm.new_value_error(format!(
"{left_name} and {right_name} cannot be the same"
)));
return Err(
vm.new_value_error(format!("bad {left_name} or {right_name} value"))
);
}
}
if left.is_some_and(|value| dialect.lineterminator.contains(value)) {
return Err(vm.new_value_error(format!(
"{left_name} and lineterminator cannot be the same"
)));
}
}
Ok(())
}

impl FormatOptions {
fn result(mut self, vm: &VirtualMachine) -> PyResult<PyDialect> {
if let Some(obj) = self.dialect.take() {
if let OptionalArg::Present(obj) =
core::mem::replace(&mut self.dialect, OptionalArg::Missing)
{
let obj = if obj.downcast_ref::<PyStr>().is_some() {
get_dialect(DialectName { name: obj }, vm)?
.into_ref(&vm.ctx)
Expand All @@ -478,8 +450,9 @@ mod _csv {
// Read all non-overridden attributes before converting their values.
macro_rules! fill_from_dialect {
($($name:ident),* $(,)?) => {
$(if self.$name.is_none() {
self.$name = vm.get_attribute_opt(&obj, stringify!($name))?;
$(if self.$name.is_missing() {
self.$name = vm.get_attribute_opt(&obj, stringify!($name))?
.map_or(OptionalArg::Missing, OptionalArg::Present);
})*
};
}
Expand All @@ -493,23 +466,23 @@ mod _csv {
skipinitialspace,
strict,
);
self.dialect = Some(obj);
self.dialect = OptionalArg::Present(obj);
}

let delimiter = match &self.delimiter {
Some(obj) => parse_char(vm, obj, "delimiter", false)?.unwrap(),
None => ',',
OptionalArg::Present(obj) => parse_char(vm, obj, "delimiter", false)?.unwrap(),
OptionalArg::Missing => ',',
};
let doublequote = match &self.doublequote {
Some(obj) => obj.try_to_bool(vm)?,
None => true,
OptionalArg::Present(obj) => obj.try_to_bool(vm)?,
OptionalArg::Missing => true,
};
let escapechar = match &self.escapechar {
Some(obj) => parse_char(vm, obj, "escapechar", true)?,
None => None,
OptionalArg::Present(obj) => parse_char(vm, obj, "escapechar", true)?,
OptionalArg::Missing => None,
};
let lineterminator = match &self.lineterminator {
Some(obj) => {
OptionalArg::Present(obj) => {
let value = obj.downcast_ref::<PyStr>().ok_or_else(|| {
vm.new_type_error(format!(
r#""lineterminator" must be a string, not {}"#,
Expand All @@ -522,14 +495,14 @@ mod _csv {
// sentinel terminator (see `writerow`).
parse_lineterminator(vm, value)?.to_owned()
}
None => "\r\n".to_owned(),
OptionalArg::Missing => "\r\n".to_owned(),
};
let quotechar = match &self.quotechar {
Some(obj) => parse_char(vm, obj, "quotechar", true)?,
None => Some('"'),
OptionalArg::Present(obj) => parse_char(vm, obj, "quotechar", true)?,
OptionalArg::Missing => Some('"'),
};
let quoting = match &self.quoting {
Some(obj) => {
OptionalArg::Present(obj) => {
if !obj.class().is(vm.ctx.types.int_type) {
return Err(vm.new_type_error(format!(
r#""quoting" must be an integer, not {}"#,
Expand All @@ -540,21 +513,21 @@ mod _csv {
.unwrap()
.try_to_primitive::<i32>(vm)?
}
None => QuoteStyle::Minimal as i32,
OptionalArg::Missing => QuoteStyle::Minimal as i32,
};
let skipinitialspace = match &self.skipinitialspace {
Some(obj) => obj.try_to_bool(vm)?,
None => false,
OptionalArg::Present(obj) => obj.try_to_bool(vm)?,
OptionalArg::Missing => false,
};
let strict = match &self.strict {
Some(obj) => obj.try_to_bool(vm)?,
None => false,
OptionalArg::Present(obj) => obj.try_to_bool(vm)?,
OptionalArg::Missing => false,
};

// Validate the quoting value only after all field conversions.
let quoting = QuoteStyle::try_from(quoting as isize)
.map_err(|_| vm.new_type_error(r#"bad "quoting" value"#))?;
let quoting = if quotechar.is_none() && self.quoting.is_none() {
let quoting = if quotechar.is_none() && self.quoting.is_missing() {
QuoteStyle::None
} else {
quoting
Expand Down