…ckage, rather than specifying the arguments on the command line.

…facts. See discussion Turfjs#2629

…o building to work. Simplifies things overall.

…lined in dist/esm/index.js files, greatly increasing their size.

…re causing problems in a previous build because skipNodeModulesBundle was true. That's now been removed and @turf/turf does build fine without them.

…ome degree (except @turf/turf). TS projects we statically type check the index.ts itself. JS projects we type check a minimal types.ts file against the index.d.ts definition.

…ipt infers type dependencies of imported types. These must have surfaced due to this PRs new approach to linking package dependencies. Solution was to explicitly define a few return types that previously were being inferred. Should be fixed permanently in typescript 5.5, though nothing really for us to revert. See microsoft/TypeScript#47663 (comment)

Also synced some JSDoc with the function's actual type definition.

…ve. Adding empty marchingsquares d.ts to isolines (same as we've done for isobands) to retire a ts-expect-error. Also believe prepublishOnly better target to use than prepublish.

…packages the same. Only exception now is @turf/turf.

Run tsc static type checking in all packages (except @turf/turf), even the JS ones. It's a no-op for JS, though standardises our package structure somewhat.
Added a bunch of standard devDependencies to all projects including typescript, tslib, bench, tape, npm-run-all. Same rationale as above.
Upgraded to typescript 5.5.4.
Retired types.ts tests as most of them weren't adding much value any more now that we use tsc --noEmit to type check TS code.

…inting of generated code (escheck). We now run the former during top level test, and the latter only during pre-publish after we've run a top level build.

Upgraded the eslint plugin to match new typescript version, fixed a couple of minor "new" lint errors.