API reference

Lexshift exposes two main functions:

• identify(record, options?) returns the lexicon revision a record matches.
• shift(record, targetRevision, options?) migrates a record to a different revision.

Both functions accept a historyProvider so you can supply older lexicons when the current revision does not match.

identify

import { identify } from "lexshift";

const identified = await identify(record, {
  historyProvider: async ({ nsid, current }) => {
    return loadPreviousLexicons(nsid, current.revision).map((candidate) => ({
      revision: candidate.revision,
      lexicon: candidate.lexicon,
      cid: candidate.cid,
    }));
  },
});

identify returns the current revision when the record matches the latest lexicon, or the first matching historical revision when older lexicons are provided.

shift

import { shift } from "lexshift";

const migrated = await shift(record, 5, {
  historyProvider: async ({ nsid, current }) => {
    return loadPreviousLexicons(nsid, current.revision).map((candidate) => ({
      revision: candidate.revision,
      lexicon: candidate.lexicon,
    }));
  },
});

shift first identifies the current revision, then applies the lexicon differences in order until the target revision is reached.

Notes

• record.$type must be a valid NSID.
• If no matching revision is found, Lexshift throws an error.
• Records are preserved when fields stay compatible; otherwise Lexshift converts or fills values where possible.