@rebasepro/schema-inference 0.4.0 → 0.5.0

package/README.md

@@ -1 +1,76 @@
-# schema_inference
+# @rebasepro/schema-inference
+
+Automatically infer Rebase collection property schemas from sample data — analyzes field types, detects enums, references, validation rules, and nested structures.
+
+## Installation
+
+```bash
+pnpm add @rebasepro/schema-inference
+```
+
+## What This Package Does
+
+`@rebasepro/schema-inference` examines arrays of data objects and produces `Properties` definitions compatible with the Rebase collection schema. It uses statistical analysis to determine the most probable type for each field, detect enum values, identify reference patterns, and build nested map/array structures. Used by the Rebase introspection pipeline and data import tools.
+
+## Key Exports
+
+### Collection Builder
+
+| Export | Type | Description |
+|---|---|---|
+| `buildEntityPropertiesFromData` | `(data: object[], getType: InferenceTypeBuilder) => Promise<Properties>` | Main entry — infer full property schema from data |
+| `buildPropertyFromData` | `(data: unknown[], property: Property, getType: InferenceTypeBuilder) => Property` | Refine an existing property with new sample data |
+| `buildPropertiesOrder` | `(properties: Properties, propertiesOrder?: string[], priorityKeys?: string[]) => string[]` | Sort property keys (title/name first, then images, then alphabetical) |
+| `inferTypeFromValue` | `(value: unknown) => DataType` | Default type inference: string, number, boolean, array, map, vector |
+| `InferenceTypeBuilder` | Type | `(value: unknown) => DataType` — pluggable type detector |
+
+### String Utilities
+
+| Export | Description |
+|---|---|
+| `parseReferenceString` | Parse `"path/entityId"` or `"db:::path/entityId"` format |
+| `looksLikeReference` | Check if a string looks like a document reference |
+| `findCommonInitialStringInPath` | Find shared collection path prefix in sample values |
+| `removeInitialAndTrailingSlashes` | Path cleanup |
+| `removeInitialSlash` / `removeTrailingSlash` | Path cleanup |
+
+### General Utilities (re-exported from `@rebasepro/utils`)
+
+| Export | Description |
+|---|---|
+| `extractEnumFromValues` | Extract enum entries from sample string values |
+| `resolveEnumValues` | Normalize `EnumValues` to `EnumValueConfig[]` |
+| `prettifyIdentifier` | `"snake_case"` → `"Snake Case"` |
+| `unslugify` | `"my-slug"` → `"My Slug"` |
+| `mergeDeep` | Deep merge objects |
+
+## Quick Start
+
+```typescript
+import {
+    buildEntityPropertiesFromData,
+    inferTypeFromValue
+} from "@rebasepro/schema-inference";
+
+const sampleData = [
+    { name: "Alice", age: 30, active: true, role: "admin" },
+    { name: "Bob", age: 25, active: false, role: "editor" },
+    { name: "Carol", age: 28, active: true, role: "admin" },
+];
+
+const properties = await buildEntityPropertiesFromData(
+    sampleData,
+    inferTypeFromValue
+);
+// properties = {
+//   name: { type: "string", name: "Name", ... },
+//   age:  { type: "number", name: "Age", ... },
+//   active: { type: "boolean", name: "Active", ... },
+//   role: { type: "string", name: "Role", enum: [...], ... }
+// }
+```
+
+## Related Packages
+
+- `@rebasepro/types` — Provides `Property`, `DataType`, `Properties`, and related types
+- `@rebasepro/utils` — General utilities re-exported by this package

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@rebasepro/schema-inference",
-  "version": "0.4.0",
+  "version": "0.5.0",
   "type": "module",
   "publishConfig": {
     "access": "public"
@@ -16,8 +16,8 @@
     "ts-jest": "^29.4.10",
     "typescript": "^5.9.3",
     "vite": "^7.3.3",
-    "@rebasepro/types": "0.4.0",
-    "@rebasepro/utils": "0.4.0"
+    "@rebasepro/types": "0.5.0",
+    "@rebasepro/utils": "0.5.0"
   },
   "exports": {
     ".": {

package/dist/test_schemas/test_schema.d.ts

@@ -1 +0,0 @@
-export {};

package/src/test_schemas/test_schema.ts

@@ -1,34 +0,0 @@
-import { buildEntityPropertiesFromData } from "../collection_builder";
-
-// import usage from "./usage.json" assert {
-//         type: "json",
-//         integrity: "sha384-ABC123"
-//         };
-import usage from "./pop_products.json" assert {
-    type: "json",
-    integrity: "sha384-ABC123"
-};
-import * as util from "util";
-import { DataType } from "@rebasepro/types";
-
-buildEntityPropertiesFromData(usage, getType)
-    .then((res) => console.log(util.inspect(res, { showHidden: false,
-depth: null,
-colors: true })));
-
-
-function getType(value: any): DataType {
-    if (typeof value === "number")
-        return "number";
-    else if (typeof value === "string")
-        return "string";
-    else if (typeof value === "boolean")
-        return "boolean";
-    else if (Array.isArray(value))
-        return "array";
-    else if (value && "_seconds" in value && "_nanoseconds" in value)
-        return "date";
-    else if (value && "id" in value && "path" in value)
-        return "reference";
-    return "map";
-}