Update feather serializer; Add parquet serializer

[pachadotdev]

Pull Request

Before you submit a pull request, please do the following:

• Add an entry to NEWS concisely describing what you changed. OK

• Add unit tests in the tests/testthat directory. OK

• Run Build->Check Package in the RStudio IDE, or devtools::check(), to make sure your change did not add any messages, warnings, or errors. OK

The relevant tests were added to these files:

══ Testing test-parse-body.R ═══════════════════════════════════════════════════
[ FAIL 0 | WARN 0 | SKIP 0 | PASS 169 ] Done!

══ Testing test-serializer-feather.R ═══════════════════════════════════════════
[ FAIL 0 | WARN 0 | SKIP 0 | PASS 6 ] Done!

══ Testing test-serializer-parquet.R ═══════════════════════════════════════════
[ FAIL 0 | WARN 0 | SKIP 0 | PASS 6 ] Done!

In addition, I separated the PR in different pushes (i.e. 1 for content + 1 for NEWS + 1 for tests)

Minimal reproducible example

I created a mini API to test both from browser and in R with download,file

MIni API

library(plumber)

#* Mtcars JSON
#* @serializer json
#* @get /mtcarsjson
function(){
  mtcars
}

# this is thx to good old PR #520

#* Mtcars JSON
#* @serializer csv
#* @get /mtcarscsv
function(){
  mtcars
}

#* Mtcars JSON
#* @serializer feather
#* @get /mtcarsfeather
function(){
  mtcars
}

#* Mtcars JSON
#* @serializer parquet
#* @get /mtcarsparquet
function(){
  mtcars
}

I run this mini API from RStudio with the command plumber::plumb("plumber.R")$run(port = 1234)

Then I run these reproducible tests in a new RStudio window

download.file(
  "http://127.0.0.1:1234/mtcarscsv",
  "mtcars.csv"
)

download.file(
  "http://127.0.0.1:1234/mtcarsfeather",
  "mtcars.feather"
)

download.file(
  "http://127.0.0.1:1234/mtcarsparquet",
  "mtcars.parquet"
)

readr::read_csv("mtcars.csv")
arrow::read_feather("mtcars.feather")
arrow::read_parquet("mtcars.parquet")

All good!

PR task list:

• Update NEWS
• Add tests
• Update documentation with devtools::document()

In addition, I added myself as ctb, after adding #520, #398 and #360.

[pachadotdev]

this can be interesting in terms of performance for mid-sized tables
the current PR can reduce the time to move data from one the server running plumber to the laptop by ~50%
ropensci/tradestatistics#49

[pachadotdev]

@schloerke all ready now !

[pachadotdev]

@schloerke all ready now !

hi @schloerke !
any news on this?