Previously the version banner only printed via PersistentPreRun, which
Cobra skips for --help and bare invocations. It now prints from main()
before Cobra parses anything, so it's the first line for every command.
Suppressible with --no-version; skipped for the version subcommand to
avoid duplicating its own output.
* Implemented `Concurrent` field in index model
* Updated index creation template to support `CREATE INDEX CONCURRENTLY`
* Added tests for concurrent index creation in migration writer
SQLite can't ALTER TABLE ADD CONSTRAINT, so foreign keys are now written
as inline FOREIGN KEY clauses in CREATE TABLE instead of commented-out
ALTER statements. The default schema (public/main) now produces bare
table names instead of a "public_" prefix; other schemas are still
prefixed to avoid collisions. Also adds direct-to-file execution: the
sqlite writer can now apply generated DDL straight to a .db file via
Metadata["connection_string"], wired into `relspec merge --output-conn`.
Merge conflict detection compared bigserial (DBML) against bigint (live
PostgreSQL read of an existing serial column) as incompatible types, since
serial is sugar over an integer column plus a sequence default and
PostgreSQL always reports back the underlying integer type. Add
SerialUnderlyingType and use it when comparing column types for conflicts.
QuoteDefaultValue also wrapped bare keyword expressions like CURRENT_DATE
in string quotes because they contain no parentheses, unlike function-call
defaults such as now(). Recognize known bare keyword defaults and leave
them unquoted across CREATE TABLE, ALTER TABLE ADD COLUMN, and
ALTER COLUMN SET DEFAULT generation.
Backtick-wrapped defaults (e.g. from GORM tags like `now()`) were only
stripped in the CREATE TABLE column-definition path, leaving raw
backticks in the ALTER COLUMN ... SET DEFAULT migration statement and
in the migration-generated CREATE TABLE template, producing invalid
SQL. Default-drift comparisons also compared raw values, so a
backtick-wrapped model default never matched the live DB default and
kept re-emitting redundant ALTER statements.
Existing databases that already ran an old migration kept stale NOT
NULL constraints and mismatched column types/defaults, because the
schema writer only emitted idempotent ADD COLUMN IF NOT EXISTS guards
and never altered columns that already existed.
- Emit guarded ALTER COLUMN ... SET/DROP NOT NULL when a column's
nullability differs from the model.
- Emit guarded ALTER COLUMN ... TYPE, falling back to renaming the old
column and adding a fresh one when the in-place conversion fails.
- Emit guarded ALTER COLUMN ... SET/DROP DEFAULT for default drift.
- Collapse the previously duplicated plain/guarded templates so
WriteSchema (full-schema, live-state-checking) and WriteMigration
(diff-based) share the same guarded SQL templates and Go helpers
instead of maintaining the logic twice.
Table.Columns/Constraints/Indexes/Relationships are Go maps, and every
writer, reader, diff, inspector, and merge code path that iterated them
directly was subject to Go's randomized map order, so identical input
could produce different output (or a different in-report violation/diff
order) on every run. Most visibly this showed up as bun/gorm `unique:`
struct tags changing order across consecutive `make models` runs with no
source change.
Fixed by sorting map iteration (by Sequence then Name, or alphabetically
for string-keyed maps) everywhere the order affects generated output or
first-match tie-break logic, across the bun, gorm, sqlite, dbml, drawdb,
pgsql, prisma, graphql, typeorm, drizzle, and dctx writers; the dctx,
prisma, and typeorm readers; the shared models.GetPrimaryKey/
GetForeignKeys helpers; pkg/diff, pkg/inspector, and pkg/merge; and the
TUI column/relationship pickers in pkg/ui.
Bun's pgdialect scans/appends native slices directly, so array columns
(text[], integer[], uuid[], ...) always generate as plain []string,
[]int32, etc. with an explicit "array" bun tag, regardless of --types
(sqltypes/stdlib/baselib). The SqlXxxArray wrapper types are no longer
used for Bun array columns (gorm is unaffected and keeps using them).
Adds --array-nullable pointer_slice to represent nullable array columns
as *[]T instead of []T, so callers can distinguish SQL NULL (nil) from
'{}' (pointer to an empty slice). Verified end-to-end against a live
PostgreSQL instance for NULL/{}/populated arrays in every --types mode.
Closes#13
Adds `relspec report bug|feature <title>` which files an issue directly
against the RelSpec tracker, tagging the title with the system's OS
machine-id (falling back to a persisted UUID) and rate-limiting
submissions to one per minute via a local state file.
Closes#14
bun's pgdialect overrides Field.Scan/Append with its own slice-only array
handling whenever the tag's type: value ends in "[]", clobbering the
sql.Scanner/driver.Valuer implemented on SqlXxxArray wrapper types and
causing "bun: Scan(unsupported sqltypes.SqlStringArray)" at query time.
Emit the underscore-prefixed internal type name (e.g. _text) instead,
which is DDL-valid but doesn't end in "[]" so bun leaves our scanner alone.
Implements a new `relspec assets` command (list/execute subcommands) that
loads local binary and text asset files into PostgreSQL by binding file bytes
as native pgx query parameters — never as SQL text literals — so binary data
stays byte-exact with no escaping overhead.
Key design points:
- YAML manifest (assets.yaml) colocated with files describes each entry:
file path, SQL call with :bytes/:filename/:param named placeholders, and
optional static params map.
- Placeholder substitution converts :name to positional $N params; PostgreSQL
::cast syntax is protected before substitution to avoid false matches.
- Directory scan follows the existing {priority}_{sequence}_{name} naming
convention, enabling asset-loading steps to be correctly interleaved with
relspec scripts execute in a migrate-apply pipeline.
- Symlink components and path traversal (../) are silently skipped to prevent
directory escape attacks.
- 14 unit tests cover manifest loading, directory scanning, ordering, symlink
skipping, path traversal rejection, placeholder substitution edge cases
(repeated, cast protection, binary byte-exact, unknown).
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* Implement tests for inserting, updating, and retrieving UUIDs in a SQLite database.
* Verify that Value() returns a string representation of the UUID.
* Ensure compatibility with custom types implementing fmt.Stringer.
* Update type mapper imports to reflect new package path.
* change default nullable type from resolvespec to baselib
* update type mappings in tests to reflect new nullable types
* adjust comments for clarity on nullable type options
* Implement SQLiteToCanonicalTypes for type mapping
* Add ConvertSQLiteToCanonical and ConvertCanonicalToSQLite functions
* Update mapDataType to utilize new conversion logic
Hein
warkanum
SG Command
sgcommand
Hermes Agent