Enum CatalogQuery 

pub enum CatalogQuery {
    PgVersion,
    Schemas,
    Tables,
    Columns,
    Constraints,
    Indexes,
    Sequences,
    Comments,
    Dependencies,
    ViewsAndMvs,
    ViewColumns,
    UserTypes,
    EnumValues,
    DomainDetails,
    DomainChecks,
    CompositeAttributes,
    Functions,
    Aggregates,
    Extensions,
    Triggers,
    PartitionedTables,
    Partitions,
    ClusterRoles,
    ClusterMembers,
    ClusterTablespaces,
    DefaultPrivileges,
    Policies,
    Publications,
    PublicationRel,
    PublicationNamespace,
    PublicationAttributes,
    EventTriggers,
    Subscriptions,
    Statistics,
    StatisticAttributes,
    StatisticExpressions,
    Collations,
    Casts,
}

Identifier for each catalog query the reader runs. Adapters dispatch on this enum to pick the per-version SQL string.

Variants

PgVersion

SHOW server_version_num.

Schemas

pg_namespace rows for managed schemas.

Tables

pg_class (relkind=‘r’) for managed tables.

Columns

pg_attribute joined with pg_attrdef/pg_type for managed tables.

Constraints

pg_constraint for managed tables (PK/UNIQUE/FK/CHECK).

Indexes

pg_index for managed tables (excluding constraint-backing indexes).

Sequences

pg_class (relkind=‘S’) joined with pg_sequence.

Comments

pg_description (currently inlined into the per-object queries).

Dependencies

pg_depend rows linking sequences to their owning columns.

ViewsAndMvs

pg_class (relkind IN (‘v’,‘m’)) joined with pg_get_viewdef.

ViewColumns

pg_attribute for view and materialized view columns.

UserTypes

pg_type filtered to typtype IN ('e','d','c') for user-defined types.

EnumValues

pg_enum labels for enum types.

DomainDetails

Base-type and nullability details for domain types.

DomainChecks

Named CHECK constraints attached to domain types.

CompositeAttributes

Attributes (fields) of composite types.

Functions

pg_proc rows for functions and procedures (prokind IN ‘f’,‘p’).

Aggregates

pg_aggregate rows joined to their wrapper pg_proc entry.

Schema-filtered via $1::text[] (same param group as Self::Functions). The assembler skips ordered-set / hypothetical-set aggregates and any aggregate whose state/final function is in an unmanaged language, recording those in DriftReport::unmanaged_aggregates.

Extensions

pg_extension rows for installed extensions.

Triggers

pg_trigger rows for user triggers (excluding internal + extension-owned).

PartitionedTables

pg_class (relkind=‘p’) rows for partitioned-table parents.

Partitions

pg_class (relispartition=true) rows for child partitions.

ClusterRoles

pg_authid rows for cluster roles (with pg_shdescription for comments).

Uses $1::text[] as the bootstrap-role filter (names to exclude), not a managed-schema list. takes_text_array_param returns true so the adapter passes the parameter; the cluster reader supplies bootstrap role names.

ClusterMembers

pg_auth_members edges joined to pg_authid for role names.

Same $1::text[] bootstrap-role filter as Self::ClusterRoles.

ClusterTablespaces

pg_tablespace rows (with pg_shdescription for comments) for managed cluster tablespaces.

Built-in pg_default / pg_global are always excluded. Uses $1::text[] as the bootstrap filter (names to exclude), in the same param-group as Self::ClusterRoles; takes_text_array_param returns true.

DefaultPrivileges

pg_default_acl rows joined to pg_authid and pg_namespace.

Returns one row per (target_role, schema, object_type) tuple. Rows for predefined pg_* roles are filtered out. Takes no $1::text[] parameter; takes_text_array_param returns false for this variant.

Policies

pg_policies rows for managed schemas.

Returns one row per policy, scoped to schemaname = ANY($1::text[]). Decoded into crate::ir::policy::Policy and attached to their owning Table by the assembler. Policies on unmanaged tables are silently dropped.

Publications

pg_publication rows for all publications in the database.

Publications are database-global (not schema-scoped); takes no $1::text[] parameter (takes_text_array_param returns false).

PublicationRel

pg_publication_rel rows — one per (publication, table) membership.

PG 15+ includes prqual (row filter) and prattrs (column list); PG 14 variant substitutes NULL for both. Takes no parameter.

PublicationNamespace

pg_publication_namespace rows — one per (publication, schema) membership (PG 15+ only). PG 14 variant returns zero rows. Takes no parameter.

PublicationAttributes

pg_attribute rows for every column of every table referenced by any publication. Used to resolve column attnums to names. Takes no parameter.

EventTriggers

pg_event_trigger rows for all event triggers in the database.

Event triggers are database-global (not schema-scoped); takes no $1::text[] parameter (takes_text_array_param returns false). Extension-owned event triggers (pg_depend.deptype = 'e') are excluded at the SQL layer.

Subscriptions

pg_subscription rows for all subscriptions in the database.

Subscriptions are database-global (not schema-scoped). Takes no $1::text[] parameter (takes_text_array_param returns false).

pg_subscription is superuser-readable only. Non-super connections will receive an empty result or a permission error; the assembler catches the error and sets DriftReport::unreadable_subscriptions.

Statistics

pg_statistic_ext rows for managed schemas. Takes $1::text[] (managed schema names).

StatisticAttributes

Column-attnum resolver for statistics target tables. Bulk-fetched once; grouped by target_oid in the assembler. Takes $1::text[].

StatisticExpressions

Bulk expression decode via pg_get_statisticsobjdef_expressions for all statistics in managed schemas. Returns one row per expression entry with columns (stat_oid, expr_index, expr_sql). Takes $1::text[] (managed schema names).

Collations

pg_collation rows for managed schemas — user-defined collations only (built-in and extension-owned collations are filtered out at the SQL layer). Takes $1::text[] (managed schema names).

Casts

pg_cast rows for all user-defined casts in the database.

Casts are database-global (not schema-scoped); takes no $1::text[] parameter (takes_text_array_param returns false). System / built-in casts (oid < 16384) and extension-owned casts (pg_depend.deptype = 'e') are excluded at the SQL layer. The assembler skips casts whose conversion function is in an unmanaged language (anything other than sql / plpgsql), recording those in DriftReport::unmanaged_casts.

Implementations

impl CatalogQuery

pub const fn takes_text_array_param(self) -> bool

Whether this query accepts a $1::text[] argument.

The semantic meaning of the array varies by variant: managed-schema names for per-DB queries, bootstrap-role names for cluster queries. The adapter is responsible for passing the right slice to the right variant.

A few variants (PgVersion, Extensions) take no parameters at all; this method returns false for those.

Trait Implementations

impl Clone for CatalogQuery

fn clone(&self) -> CatalogQuery

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for CatalogQuery

impl Debug for CatalogQuery

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

Formats the value using the given formatter.

impl Eq for CatalogQuery

impl Hash for CatalogQuery

fn hash<__H: Hasher>(&self, state: &mut __H)

Feeds this value into the given Hasher.

fn hash_slice<H>(data: &[Self], state: &mut H)

where H: Hasher, Self: Sized,

Feeds a slice of this type into the given Hasher.

impl PartialEq for CatalogQuery

fn eq(&self, other: &CatalogQuery) -> 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 StructuralPartialEq for CatalogQuery