Struct SourceColumn 

pub struct SourceColumn {
    pub name: String,
    pub native_type: String,
    pub nullable: bool,
    pub precision: Option<u8>,
    pub scale: Option<i8>,
    pub length: Option<u64>,
    pub datetime_precision: Option<u8>,
}

Metadata captured from the source database for a single column, before it is mapped to a RivetType.

Names follow the roadmap struct (§6 SourceColumn) one-for-one so the document and the code stay in lockstep. Optional fields are populated only when the source driver actually provides them — Rivet must never invent a precision/scale to “fill in” a missing one (that is exactly the silent-degradation pattern the roadmap forbids).

Fields

name: String

Column name as reported by the source database.

native_type: String

Native database type identifier — vendor-specific string used for reports and policy decisions (e.g. "numeric", "timestamptz", "jsonb"). Always present.

nullable: bool

True when the source schema declares the column nullable.

Current limitation (see ADR-0016): every driver passes true here unconditionally — Rivet does not query information_schema.columns.is_nullable (MySQL) / pg_attribute.attnotnull (PostgreSQL), so source NOT NULL constraints are not propagated to the output Parquet schema. Downstream catalog tools therefore see every column as nullable regardless of source declaration. Conservative for write paths (any value passes), lossy for read paths (cannot distinguish “schema-mandated NOT NULL” from “allows NULL but happened to have no NULLs in this run”). Tracked for v0.8 Phase A type-report extension (per ADR-0014).

precision: Option<u8>

Decimal precision — number of total significant digits. Only present for fixed-precision numeric types where the source actually declared a precision (e.g. numeric(18,2) → Some(18); numeric → None).

scale: Option<i8>

Decimal scale — number of digits to the right of the decimal point. Signed because PostgreSQL allows negative scale for numeric (e.g. numeric(5,-2) rounds to hundreds).

length: Option<u64>

Length for variable-length character/binary types when the source declares one (e.g. varchar(255) → Some(255)).

datetime_precision: Option<u8>

Fractional-second precision for temporal types when declared (e.g. timestamp(6) → Some(6)).

Implementations

impl SourceColumn

pub fn simple( name: impl Into<String>, native_type: impl Into<String>, nullable: bool, ) -> Self

Convenience constructor for the most common case: a column whose native type doesn’t carry length / precision metadata.

pub fn decimal( name: impl Into<String>, native_type: impl Into<String>, nullable: bool, precision: u8, scale: i8, ) -> Self

Convenience constructor for declared decimal columns.

Trait Implementations

impl Clone for SourceColumn

fn clone(&self) -> SourceColumn

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SourceColumn

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

Formats the value using the given formatter.

impl Eq for SourceColumn

impl PartialEq for SourceColumn

fn eq(&self, other: &SourceColumn) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Serialize for SourceColumn

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.

impl StructuralPartialEq for SourceColumn