enhance: Change Schema.normalize interface

[ntucker]

Motivation

Schema methods should have limitations in args sent

• Input data (value, parent, key, args)
• Functions to operate on and interface with larger normalize/denormalize process
  • NOT direct access to entire data tree like entities

Recursive calls should not require re-passing functions

• Functions can all be closed around those operations, so it makes it easier for schema implementers, makes the code easier to read, and possibly improves performance

Ordering of args should be as consistent as possible to help with readability

• schema always first if it’s an arg
• next is inputs, if they exist (value, parent, key, args)
• next is any recursive functions (like visit, unvisit)
• next is remaining context acting functions (like addEntity, getEntity, checkLoop)

Solution

class MemoCache {
  denormalize(schema, input, entities, args): { data, paths };
  query(schema, args, entities, indexes): data;
  buildQueryKey(schema, args, entities, indexes): normalized;
}
normalize(schema, input, { args, date, expiresAt, fetchedAt }, { entities, indexes, entityMeta });
denormalize(schema, input, entities, args);

interface SchemaSimple {
  normalize(
    input: any,
    parent: any,
    key: any,
    args: any[],
    visit: (schema: any, value: any, parent: any, key: any, args: readonly any[]) => any,
    addEntity: (...args: any) => any,
    getEntity: (...args: any) => any,
    checkLoop: (...args: any) => any,
  ): any;
  denormalize(
    input: {},
    args: readonly any[],
    unvisit: (schema: any, input: any) => any,
  ): T;
}

const { result, entities, indexes, entityMeta } = normalize(
  action.endpoint.schema,
  payload,
  action.meta,
  state,
);

Open questions

Args order: There are legit cases where it's nice to not specify args because there are none. This means we should place args last to give it a default. However, this makes ordering inconsistent with other functions like query which always has an args.

[changeset-bot]

🦋 Changeset detected

Latest commit: 4096ece

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 14 packages

Name	Type
@data-client/normalizr	Minor
@data-client/endpoint	Minor
@data-client/graphql	Minor
@data-client/rest	Minor
@data-client/react	Minor
@data-client/core	Minor
@data-client/test	Minor
@data-client/img	Minor
@data-client/ssr	Patch
example-benchmark	Patch
normalizr-github-example	Patch
normalizr-redux-example	Patch
normalizr-relationships	Patch
coinbase-lite	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.79%. Comparing base (3c95531) to head (4096ece).

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #3134   +/-   ##
=======================================
  Coverage   98.79%   98.79%           
=======================================
  Files         127      126    -1     
  Lines        2243     2249    +6     
  Branches      455      460    +5     
=======================================
+ Hits         2216     2222    +6     
  Misses         16       16           
  Partials       11       11           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[github-actions]

Benchmark

Benchmark suite	Current: 4096ece	Previous: 7bd322d	Ratio
normalizeLong	503 ops/sec (±0.65%)	450 ops/sec (±1.81%)	0.89
infer All	9696 ops/sec (±0.47%)	9576 ops/sec (±1.57%)	0.99
denormalizeLong	342 ops/sec (±1.29%)	328 ops/sec (±2.51%)	0.96
denormalizeLong donotcache	889 ops/sec (±0.52%)	888 ops/sec (±1.56%)	1.00
denormalizeShort donotcache 500x	1355 ops/sec (±0.23%)	1351 ops/sec (±0.14%)	1.00
denormalizeShort 500x	973 ops/sec (±0.22%)	986 ops/sec (±0.26%)	1.01
denormalizeShort 500x withCache	4821 ops/sec (±0.18%)	4966 ops/sec (±0.33%)	1.03
denormalizeLong with mixin Entity	301 ops/sec (±0.27%)	302 ops/sec (±0.30%)	1.00
denormalizeLong withCache	7129 ops/sec (±0.17%)	7339 ops/sec (±0.24%)	1.03
denormalizeLong All withCache	6690 ops/sec (±0.30%)	6888 ops/sec (±0.17%)	1.03
denormalizeLong Query-sorted withCache	6612 ops/sec (±0.59%)	6540 ops/sec (±0.97%)	0.99
denormalizeLongAndShort withEntityCacheOnly	1488 ops/sec (±0.53%)	1571 ops/sec (±0.42%)	1.06
getResponse	5946 ops/sec (±1.09%)	6046 ops/sec (±1.32%)	1.02
getResponse (null)	5346945 ops/sec (±0.84%)	6178925 ops/sec (±0.96%)	1.16
getResponse (clear cache)	292 ops/sec (±0.26%)	297 ops/sec (±0.34%)	1.02
getSmallResponse	2593 ops/sec (±0.16%)	2683 ops/sec (±0.53%)	1.03
getSmallInferredResponse	2069 ops/sec (±0.24%)	2027 ops/sec (±0.61%)	0.98
getResponse Query-sorted	6553 ops/sec (±0.66%)	6579 ops/sec (±0.76%)	1.00
getResponse Collection	6308 ops/sec (±1.13%)	6281 ops/sec (±1.16%)	1.00
get Collection	5517 ops/sec (±1.01%)	5455 ops/sec (±1.11%)	0.99
setLong	507 ops/sec (±0.24%)	454 ops/sec (±1.98%)	0.90
setLongWithMerge	194 ops/sec (±0.47%)	188 ops/sec (±1.48%)	0.97
setLongWithSimpleMerge	205 ops/sec (±0.54%)	195 ops/sec (±1.35%)	0.95
setSmallResponse 500x	871 ops/sec (±0.51%)	831 ops/sec (±1.14%)	0.95

This comment was automatically generated by workflow using github-action-benchmark.