Jest custom transformer - can its performance be improved? - jestjs

Question

I'm testing the performance of a custom transformer with Jest. Currently, the transformer does nothing but return the code that it gets from Jest. The transformer has implemented the getCacheKey function.
Here's the entire code for the transformer:
function process(src, path, config, transformOptions) {
return src;
}
exports.process = process;
function getCacheKey(fileData, filePath, configStr, options) {
return crypto.createHash('md5')
.update(fileData + filePath + configStr, 'utf8')
.digest('hex');
}
exports.getCacheKey = getCacheKey;
Link to the transformer
The jest config, in package.json is as follows:
"jest": {
"transform": {
"^.+\\.tsx?$": "<rootDir>/ts-transformer.js"
},
"testMatch": [
"<rootDir>/test-jest/**/*.ts"
],
"moduleFileExtensions": [
"ts",
"tsx",
"js",
"json"
]
}
Link to package.json
When testing this setup with Jest, it takes the same amount of time with and without --no-cache (around 9 seconds)
When testing this setup with Mocha, the first run takes around 7 seconds and subsequent runs take around 4 seconds.
In both cases (with jest and mocha), the subsequent runs were tested without changing any source or test file.
My questions:
Shouldn't subsequent Jest runs be faster due to caching?
Is there something in the transformer that is preventing an improvement in the testing duration?
Is there a minimum overhead that Jest incurs which is clouding this issue?

Answer 1

It might be faster to update the pieces (fileData, filePath, configStr) separately so there does not have to be a copy of the file contents on the concatenation.
function getCacheKey(fileData, filePath, configStr, options) {
const hash = crypto.createHash('md5');
hash.update(fileData);
hash.update(filePath);
hash.update(configStr);
return hash.digest('hex');
}
Note: If encoding is not provided, and the data is a string, an encoding of 'utf8' is enforced.