fix: simplify null and undefined checks in get function

[OgabekYuldoshev]

Summary

This PR fixes the get function to properly handle null values by returning the default value when the final resolved value is null. Previously, the function only checked for undefined at the end, allowing null values to be returned even when a default value was provided.

Changes:

• Updated the final return condition to check for both undefined and null values
• Ensures consistent behavior where both null and undefined trigger default value fallback
• Maintains backward compatibility for all existing use cases

This improves the robustness of the utility function and provides more predictable behavior when dealing with nullable data structures.

Related issue, if any:

For any code change,

• Related documentation has been updated, if needed
• Related tests have been added or updated, if needed
• Related benchmarks have been added or updated, if needed
• Release notes in next-minor.md or next-major.md have been added, if needed

Does this PR introduce a breaking change?

No

This change maintains backward compatibility. The only behavioral change is that null values at the end of a path resolution will now return the default value instead of null, which aligns with the expected behavior of most developers using this utility function.

Bundle impact

Status	File	Size 1	Difference
M	src/object/get.ts	167	-23 (-12%)

Footnotes

1. Function size includes the import dependencies of the function. ↩

[radashi-bot]

Benchmark Results

Name	Current	Baseline	Change
get ▶︎ with simple path	2,739,464.13 ops/sec ±0.14%	2,837,296.5 ops/sec ±0.19% 🔗	🐢 -3.45%
get ▶︎ with array index path	1,250,012.87 ops/sec ±0.92%	1,369,164.74 ops/sec ±0.65% 🔗	🐢 -8.7%
get ▶︎ with default value	1,674,100.68 ops/sec ±0.18%	1,788,207.39 ops/sec ±0.16% 🔗	🐢 -6.38%
get ▶︎ with undefined nested path and default value	828,806.32 ops/sec ±0.51%	916,839.99 ops/sec ±0.14% 🔗	🐢 -9.6%

Performance regressions of 30% or more should be investigated, unless they were anticipated. Smaller regressions may be due to normal variability, as we don't use dedicated CI infrastructure.

[aleclarson]

I agree with this change. It makes get behave similarly to a ?? b.

This is a behavior change, so a test needs to be updated or added such that it will fail when using the previous implementation. This behavior should also be documented.

[aleclarson]

On second thought, maybe get shouldn't have a defaultValue parameter at all?

Can't you just do this?

_.get(obj, key) ?? defaultValue

[OgabekYuldoshev]

I think we should keep the defaultValue parameter for consistency with other Radashi utilities, better readability in chained calls, and to avoid extra wrapping with ??. This change focuses on improving nullish handling without removing that convenience.