Initial _ctypes implementation (#5519)

* initial _ctypes implementation with _CData, get_errno, and set_errno

Signed-off-by: Ashwin Naren <arihant2math@gmail.com>

Cargo.lock

@@ -2258,6 +2258,7 @@ dependencies = [
  "cfg-if",
  "chrono",
  "crossbeam-utils",
+ "errno",
  "exitcode",
  "flame",
  "flamer",

extra_tests/snippets/builtins_ctypes.py

@@ -0,0 +1,105 @@
+from _ctypes import sizeof
+from _ctypes import _SimpleCData
+from struct import calcsize as _calcsize
+
+def _check_size(typ, typecode=None):
+    # Check if sizeof(ctypes_type) against struct.calcsize.  This
+    # should protect somewhat against a misconfigured libffi.
+    from struct import calcsize
+    if typecode is None:
+        # Most _type_ codes are the same as used in struct
+        typecode = typ._type_
+    actual, required = sizeof(typ), calcsize(typecode)
+    if actual != required:
+        raise SystemError("sizeof(%s) wrong: %d instead of %d" % \
+                          (typ, actual, required))
+
+class c_short(_SimpleCData):
+    _type_ = "h"
+_check_size(c_short)
+
+class c_ushort(_SimpleCData):
+    _type_ = "H"
+_check_size(c_ushort)
+
+class c_long(_SimpleCData):
+    _type_ = "l"
+_check_size(c_long)
+
+class c_ulong(_SimpleCData):
+    _type_ = "L"
+_check_size(c_ulong)
+
+if _calcsize("i") == _calcsize("l"):
+    # if int and long have the same size, make c_int an alias for c_long
+    c_int = c_long
+    c_uint = c_ulong
+else:
+    class c_int(_SimpleCData):
+        _type_ = "i"
+    _check_size(c_int)
+
+    class c_uint(_SimpleCData):
+        _type_ = "I"
+    _check_size(c_uint)
+
+class c_float(_SimpleCData):
+    _type_ = "f"
+_check_size(c_float)
+
+class c_double(_SimpleCData):
+    _type_ = "d"
+_check_size(c_double)
+
+class c_longdouble(_SimpleCData):
+    _type_ = "g"
+if sizeof(c_longdouble) == sizeof(c_double):
+    c_longdouble = c_double
+
+if _calcsize("l") == _calcsize("q"):
+    # if long and long long have the same size, make c_longlong an alias for c_long
+    c_longlong = c_long
+    c_ulonglong = c_ulong
+else:
+    class c_longlong(_SimpleCData):
+        _type_ = "q"
+    _check_size(c_longlong)
+
+    class c_ulonglong(_SimpleCData):
+        _type_ = "Q"
+    ##    def from_param(cls, val):
+    ##        return ('d', float(val), val)
+    ##    from_param = classmethod(from_param)
+    _check_size(c_ulonglong)
+
+class c_ubyte(_SimpleCData):
+    _type_ = "B"
+c_ubyte.__ctype_le__ = c_ubyte.__ctype_be__ = c_ubyte
+# backward compatibility:
+##c_uchar = c_ubyte
+_check_size(c_ubyte)
+
+class c_byte(_SimpleCData):
+    _type_ = "b"
+c_byte.__ctype_le__ = c_byte.__ctype_be__ = c_byte
+_check_size(c_byte)
+
+class c_char(_SimpleCData):
+    _type_ = "c"
+c_char.__ctype_le__ = c_char.__ctype_be__ = c_char
+_check_size(c_char)
+
+class c_char_p(_SimpleCData):
+    _type_ = "z"
+    def __repr__(self):
+        return "%s(%s)" % (self.__class__.__name__, c_void_p.from_buffer(self).value)
+_check_size(c_char_p, "P")
+
+class c_void_p(_SimpleCData):
+    _type_ = "P"
+c_voidp = c_void_p # backwards compatibility (to a bug)
+_check_size(c_void_p)
+
+class c_bool(_SimpleCData):
+    _type_ = "?"
+_check_size(c_bool)

vm/Cargo.toml

@@ -99,6 +99,8 @@ uname = "0.1.1"
 [target.'cfg(not(target_arch = "wasm32"))'.dependencies]
 rustyline = { workspace = true }
 which = "6"
+errno = "0.3"
+widestring = { workspace = true }
 
 [target.'cfg(any(not(target_arch = "wasm32"), target_os = "wasi"))'.dependencies]
 num_cpus = "1.13.1"
@@ -106,7 +108,6 @@ num_cpus = "1.13.1"
 [target.'cfg(windows)'.dependencies]
 junction = { workspace = true }
 schannel = { workspace = true }
-widestring = { workspace = true }
 winreg = "0.52"
 
 [target.'cfg(windows)'.dependencies.windows]

vm/src/stdlib/ctypes.rs

@@ -0,0 +1,107 @@
+pub(crate) mod base;
+
+use crate::builtins::PyModule;
+use crate::{PyRef, VirtualMachine};
+
+pub(crate) fn make_module(vm: &VirtualMachine) -> PyRef<PyModule> {
+    let module = _ctypes::make_module(vm);
+    base::extend_module_nodes(vm, &module);
+    module
+}
+
+#[pymodule]
+pub(crate) mod _ctypes {
+    use super::base::PyCSimple;
+    use crate::builtins::PyTypeRef;
+    use crate::class::StaticType;
+    use crate::function::Either;
+    use crate::{AsObject, PyObjectRef, PyResult, TryFromObject, VirtualMachine};
+    use crossbeam_utils::atomic::AtomicCell;
+    use std::ffi::{
+        c_double, c_float, c_int, c_long, c_longlong, c_schar, c_short, c_uchar, c_uint, c_ulong,
+        c_ulonglong,
+    };
+    use std::mem;
+    use widestring::WideChar;
+
+    pub fn get_size(ty: &str) -> usize {
+        match ty {
+            "u" => mem::size_of::<WideChar>(),
+            "c" | "b" => mem::size_of::<c_schar>(),
+            "h" => mem::size_of::<c_short>(),
+            "H" => mem::size_of::<c_short>(),
+            "i" => mem::size_of::<c_int>(),
+            "I" => mem::size_of::<c_uint>(),
+            "l" => mem::size_of::<c_long>(),
+            "q" => mem::size_of::<c_longlong>(),
+            "L" => mem::size_of::<c_ulong>(),
+            "Q" => mem::size_of::<c_ulonglong>(),
+            "f" => mem::size_of::<c_float>(),
+            "d" | "g" => mem::size_of::<c_double>(),
+            "?" | "B" => mem::size_of::<c_uchar>(),
+            "P" | "z" | "Z" => mem::size_of::<usize>(),
+            _ => unreachable!(),
+        }
+    }
+
+    const SIMPLE_TYPE_CHARS: &str = "cbBhHiIlLdfguzZPqQ?";
+
+    pub fn new_simple_type(
+        cls: Either<&PyObjectRef, &PyTypeRef>,
+        vm: &VirtualMachine,
+    ) -> PyResult<PyCSimple> {
+        let cls = match cls {
+            Either::A(obj) => obj,
+            Either::B(typ) => typ.as_object(),
+        };
+
+        if let Ok(_type_) = cls.get_attr("_type_", vm) {
+            if _type_.is_instance((&vm.ctx.types.str_type).as_ref(), vm)? {
+                let tp_str = _type_.str(vm)?.to_string();
+
+                if tp_str.len() != 1 {
+                    Err(vm.new_value_error(
+                        format!("class must define a '_type_' attribute which must be a string of length 1, str: {tp_str}"),
+                    ))
+                } else if !SIMPLE_TYPE_CHARS.contains(tp_str.as_str()) {
+                    Err(vm.new_attribute_error(format!("class must define a '_type_' attribute which must be\n a single character string containing one of {SIMPLE_TYPE_CHARS}, currently it is {tp_str}.")))
+                } else {
+                    Ok(PyCSimple {
+                        _type_: tp_str,
+                        _value: AtomicCell::new(vm.ctx.none()),
+                    })
+                }
+            } else {
+                Err(vm.new_type_error("class must define a '_type_' string attribute".to_string()))
+            }
+        } else {
+            Err(vm.new_attribute_error("class must define a '_type_' attribute".to_string()))
+        }
+    }
+
+    #[pyfunction(name = "sizeof")]
+    pub fn size_of(tp: Either<PyTypeRef, PyObjectRef>, vm: &VirtualMachine) -> PyResult<usize> {
+        match tp {
+            Either::A(type_) if type_.fast_issubclass(PyCSimple::static_type()) => {
+                let zelf = new_simple_type(Either::B(&type_), vm)?;
+                Ok(get_size(zelf._type_.as_str()))
+            }
+            Either::B(obj) if obj.has_attr("size_of_instances", vm)? => {
+                let size_of_method = obj.get_attr("size_of_instances", vm)?;
+                let size_of_return = size_of_method.call(vec![], vm)?;
+                Ok(usize::try_from_object(vm, size_of_return)?)
+            }
+            _ => Err(vm.new_type_error("this type has no size".to_string())),
+        }
+    }
+
+    #[pyfunction]
+    fn get_errno() -> i32 {
+        errno::errno().0
+    }
+
+    #[pyfunction]
+    fn set_errno(value: i32) {
+        errno::set_errno(errno::Errno(value));
+    }
+}

vm/src/stdlib/ctypes/base.rs

@@ -0,0 +1,161 @@
+use crate::builtins::{PyBytes, PyFloat, PyInt, PyModule, PyNone, PyStr};
+use crate::class::PyClassImpl;
+use crate::{Py, PyObjectRef, PyResult, TryFromObject, VirtualMachine};
+use crossbeam_utils::atomic::AtomicCell;
+use num_traits::ToPrimitive;
+use rustpython_common::lock::PyRwLock;
+use std::fmt::Debug;
+
+#[allow(dead_code)]
+fn set_primitive(_type_: &str, value: &PyObjectRef, vm: &VirtualMachine) -> PyResult {
+    match _type_ {
+        "c" => {
+            if value
+                .clone()
+                .downcast_exact::<PyBytes>(vm)
+                .is_ok_and(|v| v.len() == 1)
+                || value
+                    .clone()
+                    .downcast_exact::<PyBytes>(vm)
+                    .is_ok_and(|v| v.len() == 1)
+                || value
+                    .clone()
+                    .downcast_exact::<PyInt>(vm)
+                    .map_or(Ok(false), |v| {
+                        let n = v.as_bigint().to_i64();
+                        if let Some(n) = n {
+                            Ok((0..=255).contains(&n))
+                        } else {
+                            Ok(false)
+                        }
+                    })?
+            {
+                Ok(value.clone())
+            } else {
+                Err(vm.new_type_error(
+                    "one character bytes, bytearray or integer expected".to_string(),
+                ))
+            }
+        }
+        "u" => {
+            if let Ok(b) = value.str(vm).map(|v| v.to_string().chars().count() == 1) {
+                if b {
+                    Ok(value.clone())
+                } else {
+                    Err(vm.new_type_error("one character unicode string expected".to_string()))
+                }
+            } else {
+                Err(vm.new_type_error(format!(
+                    "unicode string expected instead of {} instance",
+                    value.class().name()
+                )))
+            }
+        }
+        "b" | "h" | "H" | "i" | "I" | "l" | "q" | "L" | "Q" => {
+            if value.clone().downcast_exact::<PyInt>(vm).is_ok() {
+                Ok(value.clone())
+            } else {
+                Err(vm.new_type_error(format!(
+                    "an integer is required (got type {})",
+                    value.class().name()
+                )))
+            }
+        }
+        "f" | "d" | "g" => {
+            if value.clone().downcast_exact::<PyFloat>(vm).is_ok() {
+                Ok(value.clone())
+            } else {
+                Err(vm.new_type_error(format!("must be real number, not {}", value.class().name())))
+            }
+        }
+        "?" => Ok(PyObjectRef::from(
+            vm.ctx.new_bool(value.clone().try_to_bool(vm)?),
+        )),
+        "B" => {
+            if value.clone().downcast_exact::<PyInt>(vm).is_ok() {
+                Ok(vm.new_pyobj(u8::try_from_object(vm, value.clone())?))
+            } else {
+                Err(vm.new_type_error(format!("int expected instead of {}", value.class().name())))
+            }
+        }
+        "z" => {
+            if value.clone().downcast_exact::<PyInt>(vm).is_ok()
+                || value.clone().downcast_exact::<PyBytes>(vm).is_ok()
+            {
+                Ok(value.clone())
+            } else {
+                Err(vm.new_type_error(format!(
+                    "bytes or integer address expected instead of {} instance",
+                    value.class().name()
+                )))
+            }
+        }
+        "Z" => {
+            if value.clone().downcast_exact::<PyStr>(vm).is_ok() {
+                Ok(value.clone())
+            } else {
+                Err(vm.new_type_error(format!(
+                    "unicode string or integer address expected instead of {} instance",
+                    value.class().name()
+                )))
+            }
+        }
+        _ => {
+            // "P"
+            if value.clone().downcast_exact::<PyInt>(vm).is_ok()
+                || value.clone().downcast_exact::<PyNone>(vm).is_ok()
+            {
+                Ok(value.clone())
+            } else {
+                Err(vm.new_type_error("cannot be converted to pointer".to_string()))
+            }
+        }
+    }
+}
+
+pub struct RawBuffer {
+    #[allow(dead_code)]
+    pub inner: Box<[u8]>,
+    #[allow(dead_code)]
+    pub size: usize,
+}
+
+#[pyclass(name = "_CData", module = "_ctypes")]
+pub struct PyCData {
+    _objects: AtomicCell<Vec<PyObjectRef>>,
+    _buffer: PyRwLock<RawBuffer>,
+}
+
+#[pyclass]
+impl PyCData {}
+
+#[pyclass(
+    name = "_SimpleCData",
+    base = "PyCData",
+    module = "_ctypes"
+    // TODO: metaclass
+)]
+#[derive(PyPayload)]
+pub struct PyCSimple {
+    pub _type_: String,
+    pub _value: AtomicCell<PyObjectRef>,
+}
+
+impl Debug for PyCSimple {
+    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
+        f.debug_struct("PyCSimple")
+            .field("_type_", &self._type_)
+            .finish()
+    }
+}
+
+#[pyclass(flags(BASETYPE))]
+impl PyCSimple {}
+
+pub fn extend_module_nodes(vm: &VirtualMachine, module: &Py<PyModule>) {
+    let ctx = &vm.ctx;
+    extend_module!(vm, module, {
+        "_CData" => PyCData::make_class(ctx),
+        "_SimpleCData" => PyCSimple::make_class(ctx),
+    })
+}

vm/src/stdlib/mod.rs

@@ -37,6 +37,8 @@ pub mod posix;
 #[path = "posix_compat.rs"]
 pub mod posix;
 
+#[cfg(any(target_family = "unix", target_family = "windows"))]
+mod ctypes;
 #[cfg(windows)]
 pub(crate) mod msvcrt;
 #[cfg(all(unix, not(any(target_os = "android", target_os = "redox"))))]
@@ -124,5 +126,9 @@ pub fn get_module_inits() -> StdlibMap {
             "_winapi" => winapi::make_module,
             "winreg" => winreg::make_module,
         }
+        #[cfg(any(target_family = "unix", target_family = "windows"))]
+        {
+            "_ctypes" => ctypes::make_module,
+        }
     }
 }