Trait CElement

pub unsafe trait CElement: CData {
    // Required method
    fn assert_completness(&self);
}

A CData representing a single value rather than an entire buffer of a range of values.

Safety

Considerations for implementers

Callers must be able to rely on all pointers being valid, i.e. the “range” is not empty.

Since the indicator provided by implementation is used to indicate the length of the value in the buffer, care must be taken to prevent out of bounds access in case the implementation also is used as an output parameter, and contains truncated values (i.e. the indicator is longer than the buffer and the value within).

Required Methods

fn assert_completness(&self)

Must panic if the parameter is not complete. I.e. the indicator of a variable length parameter indicates a value larger than what is present in the value buffer.

This is used to prevent using truncacted values as input buffers, which could cause inserting invalid memory with drivers which just copy values for the length of the indicator buffer without checking the length of the target buffer first. The ODBC standard is inconclusive wether the driver has to check for this or not. So we need to check this. We can not manifest this as an invariant expressed by a type for all cases, due to the existence of input/output parameters.

Implementations on Foreign Types

impl CElement for f32

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for f64

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for i8

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for i16

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for i32

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for i64

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for u8

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for u16

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for u32

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for u64

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for Box<dyn InputParameter>

fn assert_completness(&self)

impl CElement for CString

fn assert_completness(&self)

`CString`` is terminated by zero. Indicator can therefore never indicate a truncated value and is always complete.

impl CElement for CStr

fn assert_completness(&self)

CStr is terminated by zero. Indicator can therefore never indicate a truncated value and is always complete.

impl CElement for Date

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for Numeric

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for Time

fn assert_completness(&self)

Fixed sized types are always complete

impl CElement for Timestamp

fn assert_completness(&self)

Fixed sized types are always complete

Implementors

impl CElement for Bit

impl<K: VarKind> CElement for VarCell<&[K::Element], K>

impl<K: VarKind> CElement for VarCell<&mut [K::Element], K>

impl<K: VarKind> CElement for VarCell<Box<[K::Element]>, K>

impl<T> CElement for Nullable<T>

where T: Pod,

impl<T> CElement for WithDataType<T>

where T: CElement,

impl<const LENGTH: usize, K: VarKind> CElement for VarCell<[K::Element; LENGTH], K>