@cvsystem/core

Pure TypeScript package — the heart of cvsystem. No browser APIs, no React, no framework coupling.

Install

pnpm add @cvsystem/core

Contents

| Module | What it provides | |---|---| | Domain model | DomainCV, Skill, Role, Training, Education, Commitment, RenderSpec | | Zod schemas | Runtime validation at I/O boundaries (cvSchema, cvFileSchema) | | File formats | Import/export — Cinode JSON → DomainCV, DomainCV → CVFile | | Storage interface | StorageAdapter, CVStorageAdapter, MemoryAdapter | | Brand config | BrandConfig, ThemeLayoutConfig, DEFAULT_BRAND_CONFIG | | Pagination engine | paginateCv, buildSectionsFromCv | | RenderSpec | applyRenderSpec, resolveRoleSkills | | Ontology | MIND tech ontology lookup | | Utilities | getBilingualText, formatDate, cvReducer, competenceService |

Domain model

import type { DomainCV, Skill, Role, RenderSpec, Locale } from '@cvsystem/core';

interface DomainCV {
  id: string;
  locales: Locale[];          // ['sv', 'en']
  name: { first: string; last: string };
  title: BilingualText;       // { sv: '...', en: '...' }
  summary: BilingualText;
  skills: Skill[];
  roles: Role[];
  trainings: Training[];
  educations: Education[];
  commitments: Commitment[];
  renderSpecs?: RenderSpec[];
  activeRenderSpecId?: string | null;
}

Creating a CV

import { createMinimalCvFromSeed } from '@cvsystem/core';

const cv = createMinimalCvFromSeed({ firstName: 'Jane', lastName: 'Doe' });

Validating at I/O boundaries

import { cvFileSchema } from '@cvsystem/core';

const result = cvFileSchema.safeParse(rawJson);
if (!result.success) console.error(result.error.issues);
const file = result.data; // CVFile — validated

Importing from Cinode JSON

import { importFromJson, detectFormat } from '@cvsystem/core';

const format = detectFormat(rawJson);      // 'cinode-sv' | 'cinode-en' | 'cv-system'
const result = await importFromJson(rawJson, { extractors: [...] });
// result.cv — DomainCV

File format

import { createCVFile, serializeToJson } from '@cvsystem/core';

const file = createCVFile(cv, { metadata: { source: 'manual' } });
// { format: 'cv-system', version: '2.0', cv, metadata }

const json = serializeToJson(file); // stable JSON string

Storage adapter interface

import type { StorageAdapter } from '@cvsystem/core';

class MyAdapter implements StorageAdapter {
  readonly id = 'my-adapter';
  async save(cv: DomainCV, options: SaveOptions): Promise<void> { ... }
  async load(): Promise<CVFile | null> { ... }
}

See docs/storage-adapters.md for a full implementation guide.

RenderSpec (audience views)

import { applyRenderSpec } from '@cvsystem/core';

// Filter and transform a CV for a specific audience
const rendered = applyRenderSpec(cv, spec, 'en');
// rendered.roles — hidden roles excluded, skills grouped by groupAs

Brand config

import type { BrandConfig } from '@cvsystem/core';
import { DEFAULT_BRAND_CONFIG } from '@cvsystem/core';

const myBrand: BrandConfig = {
  ...DEFAULT_BRAND_CONFIG,
  name: 'Acme',
  defaultPrintThemeId: 'print-acme',
  logoPath: '/brands/acme/logo.svg',
  footerLogoPath: null,
  themeLayoutConfigs: {
    'print-acme': {
      footers: [{ id: 'footer', heightPx: 80, templateId: 'page-footer', repeatOn: 'allPages' }],
      marginTopPx: 76,
      marginBottomPx: 80,
      blockStartsNewPage: ['education'],
      showHeaderOnCover: false,
      showFooterOnCover: false,
    },
  },
};

See docs/brands.md for a full brand package guide.

Schema versioning

CVFile.version follows MAJOR.MINOR:

| Version | Change | |---|---| | 1.0 | Initial geisli-cv format | | 2.0 | Renamed format to cv-system; added renderSpecs, ontologyRef, groupAs |

migrateFile(raw) upgrades old files to the current schema automatically on import. Consumers must not assume a specific version on load — always run through migrateFile or use cvFileSchema.parse.