PH Industry Intel

Framework-agnostic TypeScript library for Philippine industry lookup using the official PSIC hierarchy.

@bzync/ph-industry-intel helps you work with the Philippine Standard Industrial Classification in apps, forms, CRMs, onboarding flows, and back-office systems.

Features

• Official PSIC hierarchy bundled locally
• Section → Division → Group → Class → Subclass traversal
• Code lookups for all PSIC levels
• Search by code or title
• Full path resolution from any valid code
• Zero runtime dependencies
• TypeScript-first API
• ESM + CJS support

Installation

npm install @bzync/ph-industry-intel

yarn add @bzync/ph-industry-intel
pnpm add @bzync/ph-industry-intel
bun add @bzync/ph-industry-intel

Quick Start

import {
  getSections,
  getDivisions,
  getGroups,
  getClasses,
  getSubclasses,
  getPath,
  getNode,
  search,
  isValidCode,
} from '@bzync/ph-industry-intel'

const sections = getSections()
const divisions = getDivisions('J')
const groups = getGroups('62')
const classes = getClasses('620')
const subclasses = getSubclasses('6201')

const node = getNode('62011')
const path = getPath('62011')
const hits = search('software', { levels: ['class', 'subclass'], includePath: true })
const valid = isValidCode('62011')

API

Hierarchy getters

getSections(): PsicSection[]
getDivisions(sectionCode?: string): PsicDivision[]
getGroups(divisionCode?: string): PsicGroup[]
getClasses(groupCode?: string): PsicClass[]
getSubclasses(classCode?: string): PsicSubclass[]

Single-item lookups

getSection(code: string): PsicSection | undefined
getDivision(code: string): PsicDivision | undefined
getGroup(code: string): PsicGroup | undefined
getClass(code: string): PsicClass | undefined
getSubclass(code: string): PsicSubclass | undefined
getNode(code: string): PsicNode | null

Path resolution

getPath(code: string): PsicPath | null

Search

search(query: string, options?: SearchOptions): SearchResult[]

Supported search options:

• levels: filter by section, division, group, class, subclass
• limit: maximum number of results
• exact: exact code/title match
• startsWith: prefix matching
• includePath: attach the full PSIC path to each result

Validation

isValidCode(code: string): boolean

Raw dataset access

getDataset(): PsicDataset
getFlatEntries(): PsicFlatEntry[]
getCounts(): PsicCounts
getSource(): PsicSource

Example output

getPath('62011')

{
  "section": { "code": "J", "title": "INFORMATION AND COMMUNICATION" },
  "division": { "code": "62", "title": "COMPUTER PROGRAMMING, CONSULTANCY AND RELATED ACTIVITIES" },
  "group": { "code": "620", "title": "Computer programming, consultancy and related activities" },
  "class": { "code": "6201", "title": "Computer programming activities" },
  "subclass": { "code": "62011", "title": "Game design and development" }
}

Intended use cases

• SaaS tenant onboarding
• Employer setup forms
• Business registration support
• Analytics segmentation
• Government-aligned master data
• Searchable industry pickers

Data source

Bundled from the official PSIC hierarchy prepared from the PSA-issued 2019 updates to the 2009 PSIC.

Security

Threat model

@bzync/ph-industry-intel is a static, offline library with zero runtime dependencies and no network access. The attack surface at runtime is limited to:

• Public API input handling
• The integrity of the bundled JSON dataset

The library does not:

• Make network requests
• Execute dynamic code
• Access the filesystem at runtime
• Accept executable input

Input safety

All public search and lookup functions validate inputs and enforce safety limits:

| Constant | Value | Purpose | |---|---|---| | PSIC_MAX_QUERY_LENGTH | 200 | Queries longer than this are rejected | | PSIC_MAX_RESULT_LIMIT | 200 | Result counts are clamped to this ceiling |

For inputs from untrusted or untyped sources (e.g., user-submitted form data in plain JavaScript), prefer the hardened wrapper:

import { safeSearch, validatePsicCode } from '@bzync/ph-industry-intel'

// Accepts unknown type — safe to call with unvalidated input
const results = safeSearch(req.query.industry, { limit: 10 })

// Detailed validation with error message
const { valid, level, message } = validatePsicCode(req.body.psicCode)
if (!valid) throw new Error(message)

search() and isValidCode() also guard against non-string inputs, oversized strings, and uncapped limits.

Data provenance

The bundled PSIC dataset is generated from the official PSA-issued Excel file (2019 updates to the 2009 PSIC). The transformation script (scripts/prepare-psic-data.ts) is committed to the repository. Dataset changes between versions are visible via git diff on src/data/.

Release integrity

Published releases include npm provenance attestations linking each package version to the exact commit and CI workflow run that produced it. Verify with:

npm audit signatures

Reporting vulnerabilities

Please report security issues privately. See SECURITY.md for the full policy, scope, and response SLA.

License

MIT