Struct Row 

pub struct Row<'stmt> { /* private fields */ }

A single result row of a query.

Implementations

impl Row<'_>

pub fn get_unwrap<I, T>(&self, idx: I) -> T

where I: RowIndex, T: FromSql,

Get the value of a particular column of the result row.

Panics

Panics if calling row.get(idx) would return an error, including:

• If the underlying SQLite column type is not a valid type as a source for T
• If the underlying SQLite integral value is outside the range representable by T
• If idx is outside the range of columns in the returned query

pub fn get<I, T>(&self, idx: I) -> Result<T, Error>

where I: RowIndex, T: FromSql,

Get the value of a particular column of the result row.

Failure

Returns an Error::InvalidColumnType if the underlying SQLite column type is not a valid type as a source for T.

Returns an Error::InvalidColumnIndex if idx is outside the valid column range for this row.

Returns an Error::InvalidColumnName if idx is not a valid column name for this row.

If the result type is i128 (which requires the i128_blob feature to be enabled), and the underlying SQLite column is a blob whose size is not 16 bytes, Error::InvalidColumnType will also be returned.

Examples found in repository

examples/stress_mixed.rs (line 131)

fn execute_workload(
    connection: &mut Connection,
    config: &WorkloadConfig,
) -> Result<WorkloadStats, ExampleError> {
    let started = Instant::now();
    let transaction = connection.transaction()?;
    let mut read_operations = 0;
    let mut write_operations = 0;
    let mut checksum = 0_i64;
    {
        let mut point_lookup = transaction.prepare("SELECT value FROM kv WHERE id = ?1")?;
        let mut point_update =
            transaction.prepare("UPDATE kv SET value = value + ?2 WHERE id = ?1")?;
        for operation_index in 0..config.operation_count {
            let key_index = operation_index % config.key_count;
            let key = i64::try_from(key_index + 1)?;
            let cycle_index = operation_index % OPERATIONS_PER_CYCLE;
            if cycle_index < READ_OPERATIONS_PER_CYCLE {
                let value = point_lookup.query_row(params![key], |row| row.get::<_, i64>(0))?;
                checksum = checksum.wrapping_add(value);
                read_operations += 1;
            } else {
                let changed_rows = point_update.execute(params![key, WRITE_INCREMENT])?;
                if changed_rows != 1 {
                    return Err(io::Error::other(format!(
                        "point update changed {changed_rows} rows for key {key}"
                    ))
                    .into());
                }
                write_operations += 1;
            }
        }
    }
    transaction.commit()?;
    let elapsed = started.elapsed();
    let stats = WorkloadStats {
        read_operations,
        write_operations,
        checksum,
        elapsed,
    };
    if stats.operation_count() != config.operation_count {
        return Err(io::Error::other(format!(
            "completed {} of {} workload operations",
            stats.operation_count(),
            config.operation_count
        ))
        .into());
    }
    Ok(stats)
}

pub fn get_ref<I>(&self, idx: I) -> Result<ValueRef<'_>, Error>

where I: RowIndex,

Get the value of a particular column of the result row as a ValueRef, allowing data to be read out of a row without copying.

This ValueRef is valid only as long as this Row, which is enforced by its lifetime. This means that while this method is completely safe, it can be somewhat difficult to use, and most callers will be better served by get or get_unwrap.

Failure

Returns an Error::InvalidColumnIndex if idx is outside the valid column range for this row.

Returns an Error::InvalidColumnName if idx is not a valid column name for this row.

pub fn get_ref_unwrap<I>(&self, idx: I) -> ValueRef<'_>

where I: RowIndex,

Get the value of a particular column of the result row as a ValueRef, allowing data to be read out of a row without copying.

This ValueRef is valid only as long as this Row, which is enforced by its lifetime. This means that while this method is completely safe, it can be difficult to use, and most callers will be better served by get or get_unwrap.

Panics

Panics if calling row.get_ref(idx) would return an error, including:

• If idx is outside the range of columns in the returned query.
• If idx is not a valid column name for this row.

Trait Implementations

impl<'stmt> AsRef<Statement<'stmt>> for Row<'stmt>

fn as_ref(&self) -> &Statement<'stmt>

Converts this type into a shared reference of the (usually inferred) input type.

impl Debug for Row<'_>

Debug Row like an ordered Map<Result<&str>, Result<(Type, ValueRef)>> with column name as key except that for Type::Blob only its size is printed (not its content).

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.