datum-merge is a modern typescript library that simplifies merge and diff operations for deeply nested objects. Supports object configuration to customize the behavior for primitive, array and nested attributes including specific keys or key patterns.
datum-merge is a modern typescript library that simplifies merge and diff operations for deeply nested objects.
Source Code : https://github.com/therohk/datum-merge
NPM library : https://www.npmjs.com/package/datum-merge
Sample Usage
Merge with default config:
import { merge, detailMerge, UpdateCode } from "datum-merge";
let changed = merge(target, source, UpdateCode.I, UpdateCode.XM, UpdateCode.B);
//same as
changed = customMerge(target, source, {
scalar: UpdateCode.I,
vector: UpdateCode.XM,
nested: UpdateCode.B
});
Exact nestable config that ignores all other fields:
changed = detailMerge(target, source, {
mykey: UpdateCode.I,
myarr: UpdateCode.XM,
anobj: UpdateCode.B,
myobj: { myid: UpdateCode.I },
});
Deep merge with generic config patterns:
import { customMerge, MergeConfig, UpdateCode } from "datum-merge";
const conf: MergeConfig = {
"*_id": UpdateCode.I,
scalar: UpdateCode.B,
field1: UpdateCode.D,
"arr*": UpdateCode.XM,
nested: UpdatedCode.N,
obj1: {
scalar: UpdateCode.B,
vector: UpdateCode.XM,
},
};
let diff = customMerge(target, source, conf);
Upcoming Features
-
inline the unmaintained deep-diff library which contains serious bugs. (available)
-
support merging for top level arrays or primitives.
-
option to throw error if field datatype changes during merge.
-
project should compile in strict mode.
-
formalize config schema for deeply nested objects (for v1).
Code contributions are welcome via issues and pull requests.
Merge Strategy
This string code describes how modifications to an attribute for a PUT/UPDATE operation should be handled. It decides whether a change to the value of the field is allowed during a merge between two entities.
Strategy Codes
The same field within a source and target object is represented by s
and t
respectively.
Whether the strategy requires data to be present for the field, is shown by { 0=no, 1=yes, X=irrelavant }.
The value is migrated from the source field to the target field only if the predicate passes.
Code Value | Predicate | Meaning |
---|---|---|
C | n/a | always create new instance |
T | n/a | touch datum ; empty merge |
N | 0 | reject any change ; skip merge |
Y | sX & tX | accept any change ; bypass merge |
B | s1 & tX | insert or update, no delete |
H | s1 & t1 | update only if exists |
U | sX & t1 | update or delete only, no insert |
I | sX & t0 | insert only, no update or delete |
D | s0 & tX | delete only, no update or insert |
XR | sX & tX | full vector replacement |
XM | s ∪ t | set union, vector merge |
XD | s - t | set difference, delete given values |
XI | s ∩ t | set intersection, delete missing values |
XS | t + s | preserve order insert (allows dupes) |
XF | s + t | insert from start (allows dupes) |
Diff Codes
Applying the merge results in one of these transitions per primitive value in the target object.
Code Value | Meaning | Transitions |
---|---|---|
Z | noop / ignore | null <-- null or non-null == non-null |
N | new / insert | null <-- non-null |
E | edit / update | non-null <-- non-null |
D | unset / delete | non-null <-- null |
AN | array size increase (tbd) | non-null > non-null |
AD | array size decrease (tbd) | non-null < non-null |
No comments yet.
Sign in to be the first to comment.