open-brainy/src/db/whereMatcher.ts

/**
 * @module db/whereMatcher
 * @description In-memory evaluation of `find()` metadata filters against a
 * single resolved entity — the overlay half of historical and speculative
 * reads on a `Db` value.
 *
 * A `Db` pinned at a past generation answers metadata-level `find()` by
 * combining the live index's results (for entities untouched since the pin)
 * with per-entity evaluation of the same filter against generation records
 * (for entities that DID change). The same evaluator backs `db.with()`
 * speculative overlays. The semantics here deliberately mirror the index
 * evaluator in `src/utils/metadataIndex.ts` (`getIdsForFilter`) — operator
 * aliases, array-membership equality, `exists`/`missing`, and the
 * `allOf`/`anyOf`/`not` logical forms — so an entity matches in-memory if and
 * only if it would have matched through the index.
 *
 * **Honesty contract:** an operator this module does not recognize throws
 * {@link UnsupportedWhereOperatorError} instead of guessing — a historical
 * read must never silently return wrong results.
 */

import type { Entity, FindParams } from '../types/brainy.types.js'

/**
 * @description Thrown when a `where` clause uses an operator this in-memory
 * evaluator does not implement. Callers (historical `find()` on a `Db`)
 * convert this into the documented historical-query error rather than
 * returning silently-wrong results.
 */
export class UnsupportedWhereOperatorError extends Error {
  /** The unrecognized operator name. */
  public readonly operator: string

  /**
   * @param operator - The operator name that could not be evaluated.
   */
  constructor(operator: string) {
    super(
      `The where-operator '${operator}' is not supported for historical/speculative ` +
        `in-memory evaluation. Supported: eq/equals/is, ne/notEquals/isNot, in/oneOf, ` +
        `gt/greaterThan, gte/greaterThanOrEqual/greaterEqual, lt/lessThan, ` +
        `lte/lessThanOrEqual/lessEqual, between, contains, exists, missing, ` +
        `plus allOf/anyOf/not.`
    )
    this.name = 'UnsupportedWhereOperatorError'
    this.operator = operator
  }
}

/**
 * @description Resolve a filter field name to its value on an entity,
 * mirroring how `extractIndexableFields` lays entities out for the metadata
 * index: standard fields live at the top level (with the `noun` ↔ `type`
 * alias), everything else is a custom field inside the `metadata` bag.
 * Dotted paths (`metadata.priority`, `address.city`) traverse nested objects.
 *
 * @param entity - The resolved entity to read from.
 * @param field - The filter field name.
 * @returns The field's value, or `undefined` when absent.
 */
export function resolveEntityField(entity: Entity, field: string): unknown {
  switch (field) {
    case 'noun':
    case 'type':
      return entity.type
    case 'subtype':
      return entity.subtype
    case 'id':
      return entity.id
    case 'createdAt':
      return entity.createdAt
    case 'updatedAt':
      return entity.updatedAt
    case 'service':
      return entity.service
    case 'createdBy':
      return entity.createdBy
    case 'confidence':
      return entity.confidence
    case 'weight':
      return entity.weight
    case '_rev':
      return entity._rev
    case 'data':
      return entity.data
  }

  if (field.includes('.')) {
    // Dotted path: resolve against the whole entity first (`metadata.x`),
    // then against the metadata bag (`address.city` on nested metadata).
    const fromEntity = resolvePath(entity as unknown as Record<string, unknown>, field)
    if (fromEntity !== undefined) return fromEntity
    return resolvePath((entity.metadata ?? {}) as Record<string, unknown>, field)
  }

  return ((entity.metadata ?? {}) as Record<string, unknown>)[field]
}

/** Walk a dotted path through nested plain objects. */
function resolvePath(obj: Record<string, unknown>, path: string): unknown {
  let current: unknown = obj
  for (const segment of path.split('.')) {
    if (current === null || typeof current !== 'object') return undefined
    current = (current as Record<string, unknown>)[segment]
  }
  return current
}

/**
 * @description Index-equality semantics: scalar strict equality, with
 * array-valued fields matching by membership (the index stores one posting
 * per array element, so `eq` on an array field means "contains").
 */
function eqMatches(value: unknown, operand: unknown): boolean {
  if (Array.isArray(value)) {
    return value.some((element) => element === operand)
  }
  return value === operand
}

/** Ordered comparison for range operators (numbers, or both-strings lexicographic). */
function compare(value: unknown, operand: unknown): number | null {
  if (typeof value === 'number' && typeof operand === 'number') {
    return value - operand
  }
  if (typeof value === 'string' && typeof operand === 'string') {
    return value < operand ? -1 : value > operand ? 1 : 0
  }
  return null
}

/**
 * @description Evaluate one field condition (shorthand value or operator
 * object) against a resolved field value, mirroring the operator table in
 * `metadataIndex.getIdsForFilter`.
 *
 * @param value - The entity's field value (possibly `undefined`).
 * @param condition - The filter condition for this field.
 * @returns Whether the value satisfies the condition.
 * @throws UnsupportedWhereOperatorError for unrecognized operators.
 */
function fieldConditionMatches(value: unknown, condition: unknown): boolean {
  if (condition === null || typeof condition !== 'object' || Array.isArray(condition)) {
    // Shorthand for 'eq'.
    return eqMatches(value, condition)
  }

  for (const [op, operand] of Object.entries(condition as Record<string, unknown>)) {
    let matches: boolean
    switch (op) {
      case 'is':
      case 'equals':
      case 'eq':
        matches = eqMatches(value, operand)
        break
      case 'isNot':
      case 'notEquals':
      case 'ne':
        matches = !eqMatches(value, operand)
        break
      case 'oneOf':
      case 'in':
        matches = Array.isArray(operand) && operand.some((candidate) => eqMatches(value, candidate))
        break
      case 'greaterThan':
      case 'gt': {
        const cmp = compare(value, operand)
        matches = cmp !== null && cmp > 0
        break
      }
      case 'greaterEqual':
      case 'greaterThanOrEqual':
      case 'gte': {
        const cmp = compare(value, operand)
        matches = cmp !== null && cmp >= 0
        break
      }
      case 'lessThan':
      case 'lt': {
        const cmp = compare(value, operand)
        matches = cmp !== null && cmp < 0
        break
      }
      case 'lessEqual':
      case 'lessThanOrEqual':
      case 'lte': {
        const cmp = compare(value, operand)
        matches = cmp !== null && cmp <= 0
        break
      }
      case 'between': {
        if (!Array.isArray(operand) || operand.length !== 2) {
          matches = false
          break
        }
        const lower = compare(value, operand[0])
        const upper = compare(value, operand[1])
        matches = lower !== null && upper !== null && lower >= 0 && upper <= 0
        break
      }
      case 'contains':
        // Index semantics: array fields post one entry per element, so
        // 'contains' is the same lookup as 'eq' (membership on arrays).
        matches = eqMatches(value, operand)
        break
      case 'exists':
        matches = operand ? value !== undefined : value === undefined
        break
      case 'missing':
        matches = operand ? value === undefined : value !== undefined
        break
      default:
        throw new UnsupportedWhereOperatorError(op)
    }
    if (!matches) return false // Multiple operators on one field AND together.
  }
  return true
}

/**
 * @description Evaluate a full `where` filter (field conditions plus
 * `allOf`/`anyOf`/`not` logical composition) against one entity.
 *
 * @param entity - The resolved entity (historical record or speculative overlay).
 * @param where - The `where` clause from `FindParams`.
 * @returns Whether the entity satisfies every clause.
 * @throws UnsupportedWhereOperatorError for unrecognized operators.
 */
export function whereMatches(entity: Entity, where: Record<string, unknown>): boolean {
  for (const [field, condition] of Object.entries(where)) {
    if (field === 'allOf') {
      if (!Array.isArray(condition)) return false
      if (!condition.every((sub) => whereMatches(entity, sub as Record<string, unknown>))) return false
      continue
    }
    if (field === 'anyOf') {
      if (!Array.isArray(condition)) return false
      if (!condition.some((sub) => whereMatches(entity, sub as Record<string, unknown>))) return false
      continue
    }
    if (field === 'not') {
      if (whereMatches(entity, condition as Record<string, unknown>)) return false
      continue
    }
    if (!fieldConditionMatches(resolveEntityField(entity, field), condition)) return false
  }
  return true
}

/**
 * @description Evaluate the metadata-level portions of a `find()` query
 * (`type`, `subtype`, `where`, `service`, `excludeVFS`) against one resolved
 * entity. Used by historical and speculative `find()` to decide whether a
 * changed/overlaid entity belongs in the result set. The caller guarantees
 * the query carries no index-only dimensions (semantic `query`/`vector`,
 * `connected` traversal, …) — those are rejected with the documented
 * historical-query error before evaluation starts.
 *
 * @param entity - The resolved entity.
 * @param params - The metadata-level find parameters.
 * @returns Whether the entity matches every requested filter.
 * @throws UnsupportedWhereOperatorError for unrecognized `where` operators.
 */
export function entityMatchesFind(entity: Entity, params: FindParams): boolean {
  if (params.type !== undefined) {
    const types = Array.isArray(params.type) ? params.type : [params.type]
    if (!types.includes(entity.type)) return false
  }

  if (params.subtype !== undefined) {
    const subtypes = Array.isArray(params.subtype) ? params.subtype : [params.subtype]
    if (entity.subtype === undefined || !subtypes.includes(entity.subtype)) return false
  }

  if (params.service !== undefined && entity.service !== params.service) {
    return false
  }

  if (params.excludeVFS === true) {
    // Mirror of find()'s exclusion filter (`vfsType: { exists: false }`,
    // `isVFSEntity: { ne: true }`) and Brainy's VFS-marker helper.
    const metadata = (entity.metadata ?? {}) as Record<string, unknown>
    if (metadata.vfsType !== undefined) return false
    if (metadata.isVFSEntity === true || metadata.isVFS === true) return false
  }

  if (params.where !== undefined) {
    if (!whereMatches(entity, params.where as Record<string, unknown>)) return false
  }

  return true
}