I made a mistake in the ES WebAssembly build by using the flag -sWASM=0. I initially thought this flag would simply prevent the generation of a WebAssembly binary, but I was wrong. Instead, it embedded everything into a large JavaScript file using wasm2js. In this PR, I removed that flag, and now it generates two WebAssembly files:
drwxr-xr-x 7 hyzyla staff 224 Aug 31 13:11 .
drwxr-xr-x 6 hyzyla staff 192 Aug 31 13:11 ..
-rw-r--r-- 1 hyzyla staff 46380 Aug 31 13:11 index.html
-rw-r--r-- 1 hyzyla staff 159408 Aug 31 13:11 pdfium.esm.js
-rwxr-xr-x 1 hyzyla staff 3899985 Aug 31 13:11 pdfium.esm.wasm
-rw-r--r-- 1 hyzyla staff 159451 Aug 31 13:11 pdfium.js
-rwxr-xr-x 1 hyzyla staff 3899985 Aug 31 13:11 pdfium.wasm
Both pdfium.wasm and pdfium.esm.wasm are identical, but the pdfium.esm.js module links to pdfium.esm.wasm, and pdfium.js links to pdfium.wasm. I'm not sure if it's currently possible to configure em++ to point to the same file for both, but I think this solution is acceptable for now.
I made a mistake in the ES WebAssembly build by using the flag -sWASM=0. I initially thought this flag would simply prevent the generation of a WebAssembly binary, but I was wrong. Instead, it embedded everything into a large JavaScript file using wasm2js. In this PR, I removed that flag, and now it generates two WebAssembly files:
Both
pdfium.wasmandpdfium.esm.wasmare identical, but thepdfium.esm.jsmodule links topdfium.esm.wasm, andpdfium.jslinks topdfium.wasm. I'm not sure if it's currently possible to configureem++to point to the same file for both, but I think this solution is acceptable for now.