From ed7130bba8a24c914534769e5b15d8b29ccaa679 Mon Sep 17 00:00:00 2001
From: Hein <hein.puth@gmail.com>
Date: Sun, 26 Apr 2026 17:35:15 +0200
Subject: [PATCH] refactor(pkg): canonicalize base types and adjust length
 handling * Update base types to keep explicit modifier forms * Modify length
 handling for vector types in tests

---
 pkg/pgsql/types_registry.go       | 12 ++++++++----
 pkg/pgsql/types_registry_test.go  |  2 +-
 pkg/readers/bun/reader.go         |  1 +
 pkg/readers/bun/reader_test.go    |  6 +++---
 pkg/readers/drawdb/reader_test.go |  2 +-
 pkg/readers/gorm/reader.go        |  1 +
 pkg/readers/gorm/reader_test.go   |  6 +++---
 7 files changed, 18 insertions(+), 12 deletions(-)

diff --git a/pkg/pgsql/types_registry.go b/pkg/pgsql/types_registry.go
index 8b7f0f8..d0d507d 100644
--- a/pkg/pgsql/types_registry.go
+++ b/pkg/pgsql/types_registry.go
@@ -1,6 +1,9 @@
 package pgsql
 
-import "strings"
+import (
+	"sort"
+	"strings"
+)
 
 // TypeSpec describes PostgreSQL type capabilities used by parsers/writers.
 type TypeSpec struct {
@@ -106,9 +109,9 @@ var postgresBaseTypes = map[string]TypeSpec{
 	"ltree":     {},
 	"lquery":    {},
 	"ltxtquery": {},
-	"vector":    {SupportsLength: true}, // pgvector: vector(dim)
-	"halfvec":   {SupportsLength: true}, // pgvector: halfvec(dim)
-	"sparsevec": {SupportsLength: true}, // pgvector: sparsevec(dim)
+	"vector":    {}, // pgvector: keep explicit modifier form (vector(dim))
+	"halfvec":   {}, // pgvector: keep explicit modifier form (halfvec(dim))
+	"sparsevec": {}, // pgvector: keep explicit modifier form (sparsevec(dim))
 }
 
 var postgresTypeAliases = map[string]string{
@@ -148,6 +151,7 @@ func GetPostgresBaseTypes() []string {
 	for t := range postgresBaseTypes {
 		result = append(result, t)
 	}
+	sort.Strings(result)
 	return result
 }
 
diff --git a/pkg/pgsql/types_registry_test.go b/pkg/pgsql/types_registry_test.go
index 5ea920b..3a6c104 100644
--- a/pkg/pgsql/types_registry_test.go
+++ b/pkg/pgsql/types_registry_test.go
@@ -53,7 +53,7 @@ func TestPostgresTypeRegistry_TypeParsingAndCapabilities(t *testing.T) {
 			wantBase:          "vector",
 			wantCanonicalBase: "vector",
 			wantKnown:         true,
-			wantLength:        true,
+			wantLength:        false,
 		},
 		{
 			input:             "numeric(10,2)",
diff --git a/pkg/readers/bun/reader.go b/pkg/readers/bun/reader.go
index 3dfebaf..efcb263 100644
--- a/pkg/readers/bun/reader.go
+++ b/pkg/readers/bun/reader.go
@@ -711,6 +711,7 @@ func (r *Reader) parseTypeWithLength(typeStr string) (baseType string, length in
 		rawBaseType := strings.TrimSpace(matches[1])
 		if pgsql.SupportsLength(rawBaseType) {
 			if _, err := fmt.Sscanf(matches[2], "%d", &length); err == nil {
+				baseType = pgsql.CanonicalizeBaseType(rawBaseType)
 				return
 			}
 		}
diff --git a/pkg/readers/bun/reader_test.go b/pkg/readers/bun/reader_test.go
index a5f2c33..62cf501 100644
--- a/pkg/readers/bun/reader_test.go
+++ b/pkg/readers/bun/reader_test.go
@@ -367,9 +367,9 @@ func TestParseTypeWithLength_PreservesExplicitTypeModifiers(t *testing.T) {
 		wantType   string
 		wantLength int
 	}{
-		{"varchar(255)", "varchar(255)", 255},
-		{"character varying(120)", "character varying(120)", 120},
-		{"vector(1536)", "vector(1536)", 1536},
+		{"varchar(255)", "varchar", 255},
+		{"character varying(120)", "character varying", 120},
+		{"vector(1536)", "vector(1536)", 0},
 		{"numeric(10,2)", "numeric(10,2)", 0},
 	}
 
diff --git a/pkg/readers/drawdb/reader_test.go b/pkg/readers/drawdb/reader_test.go
index 3ac2911..d5a736a 100644
--- a/pkg/readers/drawdb/reader_test.go
+++ b/pkg/readers/drawdb/reader_test.go
@@ -317,7 +317,7 @@ func TestConvertToColumn_PreservesExplicitTypeModifiers(t *testing.T) {
 			name:       "custom vector modifier",
 			fieldType:  "vector(1536)",
 			wantType:   "vector(1536)",
-			wantLength: 1536,
+			wantLength: 0,
 		},
 	}
 
diff --git a/pkg/readers/gorm/reader.go b/pkg/readers/gorm/reader.go
index b2bc698..0550153 100644
--- a/pkg/readers/gorm/reader.go
+++ b/pkg/readers/gorm/reader.go
@@ -804,6 +804,7 @@ func (r *Reader) parseTypeWithLength(typeStr string) (baseType string, length in
 		// This avoids converting custom modifiers like vector(1536) into Length.
 		if pgsql.SupportsLength(rawBaseType) && !strings.Contains(parens, ",") {
 			if _, err := fmt.Sscanf(parens, "%d", &length); err == nil {
+				baseType = pgsql.CanonicalizeBaseType(rawBaseType)
 				return
 			}
 		}
diff --git a/pkg/readers/gorm/reader_test.go b/pkg/readers/gorm/reader_test.go
index b3bd9bc..9a5de45 100644
--- a/pkg/readers/gorm/reader_test.go
+++ b/pkg/readers/gorm/reader_test.go
@@ -374,9 +374,9 @@ func TestParseTypeWithLength_PreservesExplicitTypeModifiers(t *testing.T) {
 		wantType   string
 		wantLength int
 	}{
-		{"varchar(255)", "varchar(255)", 255},
-		{"character varying(120)", "character varying(120)", 120},
-		{"vector(1536)", "vector(1536)", 1536},
+		{"varchar(255)", "varchar", 255},
+		{"character varying(120)", "character varying", 120},
+		{"vector(1536)", "vector(1536)", 0},
 		{"numeric(10,2)", "numeric(10,2)", 0},
 	}