Repository navigation
checkJs skips .mjs/.cjs beside a .d.mts/.d.cts, but not .js beside a .d.ts #64312
Description
🔎 Search Terms
checkJs, allowJs, .d.mts, .d.cts, declaration file adjacent to implementation, file silently dropped from program, extension priority
🕗 Version & Regression Information
Reproduces identically on 5.9.3, 6.0.3, and 7.1.0-dev.20260916.1. Long-standing, not a regression.
⏯ Playground Link
Not applicable: needs two sibling files on disk plus moduleResolution: nodenext.
💻 Code
Three directories, identical but for extensions. Same tsconfig.json in each; package.json has "type": "module" only for the .mjs case.
index.js / index.mjs / index.cjs:
export const n = 1;
const bad = null;
bad.a.b.c();index.d.ts / index.d.mts / index.d.cts:
export declare const n: number;{
"compilerOptions": {
"target": "esnext",
"module": "nodenext",
"moduleResolution": "nodenext",
"allowJs": true,
"checkJs": true,
"strict": true,
"noEmit": true
}
}🙁 Actual behavior
index.js + index.d.ts -> error TS18047: 'bad' is possibly 'null'.
index.mjs + index.d.mts -> (no output)
index.cjs + index.d.cts -> (no output)
The same authoring relationship — a hand-written declaration file beside its implementation — produces opposite outcomes based only on the file extension.
In the .mjs/.cjs cases the implementation is not merely unchecked, it is absent from the program: --listFiles reports only index.d.mts. In the .js case --listFiles reports both index.d.ts and index.js.
🙂 Expected behavior
The three rows should agree. Either all of them check the implementation, or none do.
I'd expect the .js row to be the correct one — the declaration file describes the module for consumers, and checkJs still checks the implementation — but consistency matters more than which way it goes.
Additional information
#47796 reported the .mjs/.mts variant and was closed as "Working as Intended", on the rationale that the JS file is a build artifact of the TS file, and that "the compiler only automatically pulls in one copy of a file with the same basename (the most TS-y one)".
That rationale doesn't extend to this case:
- A
.d.mtsis not something a.mjsis emitted from, so the build-artifact framing doesn't apply..d.ts+.jsis the same relationship, and there the compiler does pull in both. - Neither does the one-copy-per-basename rule, for the same reason —
index.jsandindex.d.tsare both in the program.
The failure is silent, which is what makes it expensive: a package shipping .mjs/.cjs with adjacent hand-written declarations gets no checking of its implementation at all, while every sibling module without a declaration file is checked normally. tsc exits 0 and CI is green. I found this only by deliberately breaking a file and noticing the build still passed. Per #47796, VS Code does report these errors, so the editor and CI silently disagree.
#64098 is the same silent-drop shape from wildcard include (hasFileWithHigherPriorityExtension), if the two are related.
So, concretely:
- Make the three rows consistent, or
- if this is intended, document which sibling extensions suppress checking, and emit a diagnostic when an input file is dropped from the program for this reason.
Activity
[types] type-check `index.mjs`
sbrilenko commentedon Sep 24, 2026
I traced this down to the exact code responsible, in both implementations, and it's a narrower bug than "declaration files suppress their implementation" in general — it's a hardcoded carve-out that only lists two extensions.
Root cause
Wildcard (
include) file discovery de-duplicates same-basename files by extension priority inhasFileWithHigherPriorityExtension(src/compiler/commandLineParser.ts, and identically ported tointernal/tsoptions/tsconfigparsing.goin the native compiler). The relevant branch:This
continue(i.e. "don't drop this file, keep both") is the entire reason.jsbeside.d.tsbehaves the way it does — and the condition only checksExtension.Js/Extension.JsxagainstExtension.Dts. It was never extended toExtension.Mjs/Extension.CjsagainstExtension.Dmts/Extension.Dcts, so those pairs fall through toreturn true— "a higher-priority file exists, drop this one" — exactly the silent exclusion reported here. Confirmed empirically: passing the same two files viafiles: [...]instead ofincludebypassesgetFileNamesFromConfigSpecsentirely and both files are checked, in every extension combination. The bug is specifically in wildcard/glob discovery, not increateProgram's general root-file handling.Why this isn't #47796
#47796 (closed Working as Intended) is
.mjsbeside.mts— a real hand-written implementation in two module dialects, where picking the more TS-y file and ignoring the other is defensible (you don't want to compile two independent implementations of the same module). This issue is.mjsbeside.d.mts— a hand-authored declaration only, the direct.mjs/.d.mtsanalog of the case the comment above already exists to support for.js/.d.ts. The maintainers' own comment frames the.js/.d.tscase as legacy/undesirable-but-kept, not as an intentional design distinct from.mjs/.d.mts— nothing in it argues the two should diverge.Proposed fix
Extend the existing condition symmetrically, in both
commandLineParser.tsandtsconfigparsing.go. A table of (declaration extension, implementation extension) pairs reads more clearly than a chain of mixed&&/||, and makes the parallel between the four pairs explicit:This is a minimal, symmetry-restoring change — it doesn't touch the
.mjs/.mts(#47796) case at all, since that pairing was never routed through this carve-out to begin with. Happy to open a PR with this plus atsc --explainFiles-based regression test if that's the direction maintainers want, given the "Needs Proposal" label — wanted to confirm the approach first since this is shared, security-adjacent-in-spirit (silent exclusion from checking) logic.hardikkaurani commentedon Sep 29, 2026
This issue is being addressed by #64527.
test: add regression test for checkJs with ESM/CJS extensions (micros…
Fix checkJs behavior for .mjs and .cjs files next to declarations (mi…