Add VECTOR field type support to Quick SQL

[pallasinfotech]

## Title

Add VECTOR field type support to Quick SQL

## Summary

Quick SQL should support generation of Oracle Database `VECTOR` columns so developers can create AI-ready table definitions directly from Quick SQL shorthand.

Oracle Database supports the `VECTOR` data type for AI Vector Search and embedding-based applications. APEX developers building semantic search, RAG, recommendation engines, document search, and similarity-matching apps increasingly need tables with vector embedding columns.

Currently, developers must generate the table DDL with Quick SQL and then manually edit the SQL to add `VECTOR` columns. Native Quick SQL support would remove that friction.

## Requested Enhancement

Add support for `VECTOR` as a valid Quick SQL field type.

Example Quick SQL input:

```sql
documents
  title vc200
  body clob
  embedding vector
  embedding_1536 vector(1536, float32)
  embedding_int8 vector(768, int8)
  sparse_embedding vector(*, float32, sparse)

Expected generated SQL:

create table documents (
    id                number generated by default on null as identity,
    title             varchar2(200),
    body              clob,
    embedding         vector,
    embedding_1536    vector(1536, float32),
    embedding_int8    vector(768, int8),
    sparse_embedding  vector(*, float32, sparse),
    constraint documents_pk primary key (id)
);

Suggested Syntax

Quick SQL should support the following forms:

vector
vector(n)
vector(n, float32)
vector(n, float64)
vector(n, int8)
vector(n, binary)
vector(*, float32)
vector(n, float32, dense)
vector(n, float32, sparse)

Optional shorthand aliases could also be considered:

vec
vec1536
vec1536f32

However, supporting the full vector(...) syntax would be clearer, more explicit, and more future-proof.

Optional Validation

Quick SQL could optionally provide warnings when:

• A BINARY vector dimension is not a multiple of 8.
• A vector column is used as a primary key.
• A vector column is used as a foreign key.
• A vector column is used in a unique constraint.
• A default value is specified for a vector column.
• A vector index is requested on a column without fixed dimensions.

Optional Future Enhancement: Vector Index Generation

A future enhancement could add syntax for generating vector indexes.

Example Quick SQL input:

documents
  title vc200
  body clob
  embedding vector(1536, float32) /vector_index hnsw cosine

Possible generated SQL:

create vector index documents_embedding_hnsw_idx
on documents (embedding)
organization inmemory neighbor graph
distance cosine;

Use Cases

This enhancement would help developers quickly prototype and build:

• RAG document repositories
• Semantic search applications
• AI-powered knowledge bases
• Product recommendation engines
• Similarity-matching applications
• Image/text embedding storage
• Hybrid relational + vector search applications

Business Value

Adding VECTOR support would make Quick SQL more useful for modern AI application development in Oracle APEX.

It would allow developers to create AI-ready schemas directly from shorthand without manually editing generated DDL, helping align Quick SQL with Oracle Database AI Vector Search capabilities.