Parsers set body even when they should not parse

[bricka]

Hello! I just ran into a bug in another piece of software based on an interesting behavior in body-parser:
apigee-127/swagger-tools#533

The problem here is that even if a body parser should not parse (e.g. because the content type does not match), the parser still sets req.body = {}. This led another tool to think that parsing had already occurred, because req.body was defined.

I was wondering if it is possible to not modify the request at all if parsing should not be performed. That is the behavior I would expect.

You can see an example of this behavior in the raw parser here:

body-parser/lib/types/raw.js

Lines 62 to 78 in 090a92b

	req.body = req.body || {}
	// skip requests without bodies
	if (!typeis.hasBody(req)) {
	debug('skip empty body')
	next()
	return
	}
	debug('content-type %j', req.headers['content-type'])
	// determine if request should be parsed
	if (!shouldParse(req)) {
	debug('skip parsing')
	next()
	return
	}

Thank you!

[dougwilson]

This is how 2.0 will work; you can test it out at #66

[dougwilson]

The current behavior is documented at the top of the readme as well:

All middlewares will populate the req.body property with the parsed body when the Content-Type request header matches the type option, or an empty object ({}) if there was no body to parse, the Content-Type was not matched, or an error occurred.

[dougwilson]

Closing since I haven't heard anything further.