Unsure if this is possible or not, Either way, the documentation should be more clear as to if this is or is not possible. Regardless, I would like to be able to parse a CSV contained in a ReadableStream (as returned by the browser implementation of fetch). Currently, attempting to use PipeThrough to send a ReadableStream through a parser fails due to the parser's readable property being Boolean instead of a ReadableStream
Motivation
Currently, I'm trying to parse a CSV returned by a fetch. As the CSVs returned here can be significantly large, the line-by-line parsing in the pipe recipe would be extremely useful and would be very useful in avoiding blowing up memory. However, I don't see a way to use ReadableStream.PipeThrough here. This may be a gap in my own knowledge, but the documentation on using this parser in the browser after being run through webpack is fairly sparse. Reading data from a stream seems like a fairly common use case, so this would be nice to support.
Disregard, my issues wound up being caused webpack > 5 not polyfilling dependencies of the node standard library that I had configured it to polyfill. Sorry about the extra issue!
Summary
Unsure if this is possible or not, Either way, the documentation should be more clear as to if this is or is not possible. Regardless, I would like to be able to parse a CSV contained in a ReadableStream (as returned by the browser implementation of fetch). Currently, attempting to use PipeThrough to send a ReadableStream through a parser fails due to the parser's
readable
property being Boolean instead of a ReadableStreamMotivation
Currently, I'm trying to parse a CSV returned by a fetch. As the CSVs returned here can be significantly large, the line-by-line parsing in the
pipe
recipe would be extremely useful and would be very useful in avoiding blowing up memory. However, I don't see a way to useReadableStream.PipeThrough
here. This may be a gap in my own knowledge, but the documentation on using this parser in the browser after being run through webpack is fairly sparse. Reading data from a stream seems like a fairly common use case, so this would be nice to support.