sofa

  _ _ _ _ _ _ _ _ _ _ _
 /|                   |\
/ |_ _ _ _ _ _ _ _ _ _| \
\ /                    \/
 \ ___________________ /

An easy interface to CouchDB from R

Note: Check out R4couchdb, another R package to interact with CouchDB.

CouchDB versions

sofa is built targeting CouchDB v2 or greater.

CouchDB Info

Docs: http://docs.couchdb.org/en/latest/index.html
Installation: http://docs.couchdb.org/en/latest/install/index.html

Connect to CouchDB

This may be starting it on your terminal/shell

couchdb

Or opening the CouchDB app on your machine, or running it in Docker. Whatever it is, start it up.

You can interact with your CouchDB databases as well in your browser. Navigate to http://localhost:5984/_utils

Install sofa

From CRAN

install.packages("sofa")

Development version from GitHub

devtools::install_github("ropensci/sofa")

library('sofa')

Cushions

Cushions? What? Since it’s couch we gotta use cushions somehow. cushions are a connection class containing all connection info to a CouchDB instance. See ?Cushion for help.

As an example, connecting to a Cloudant couch:

z <- Cushion$new(
  host = "stuff.cloudant.com",
  transport = 'https',
  port = NULL,
  user = 'foobar',
  pwd = 'things'
)

Break down of parameters:

host: the base url, without the transport (http/https)
path: context path that is appended to the end of the url
transport: http or https
port: The port to connect to. Default: 5984. For Cloudant, have to set to NULL
user: User name for the service.
pwd: Password for the service, if any.
headers: headers to pass in all requests

If you call Cushion$new() with no arguments you get a cushion set up for local use on your machine, with all defaults used.

x <- Cushion$new()

Ping the server

x$ping()
#> $couchdb
#> [1] "Welcome"
#> 
#> $version
#> [1] "2.1.1"
#> 
#> $features
#> $features[[1]]
#> [1] "scheduler"
#> 
#> 
#> $vendor
#> $vendor$name
#> [1] "The Apache Software Foundation"

Nice, it’s working.

Create a new database, and list available databases

#> $ok
#> [1] TRUE

db_create(x, dbname = 'sofadb')
#> $ok
#> [1] TRUE

see if its there now

db_list(x)
#> [1] "cats"       "flights"    "sofadb"     "testing123"

Create documents

Write a document WITH a name (uses PUT)

doc1 <- '{"name":"sofa","beer":"IPA"}'
doc_create(x, doc1, dbname = "sofadb", docid = "a_beer")
#> $ok
#> [1] TRUE
#> 
#> $id
#> [1] "a_beer"
#> 
#> $rev
#> [1] "1-a48c98c945bcc05d482bc6f938c89882"

Write a json document WITHOUT a name (uses POST)

doc2 <- '{"name":"sofa","icecream":"rocky road"}'
doc_create(x, doc2, dbname = "sofadb")
#> $ok
#> [1] TRUE
#> 
#> $id
#> [1] "901e4bf214fb50db456d3ef8ec0516c9"
#> 
#> $rev
#> [1] "1-fd0da7fcb8d3afbfc5757d065c92362c"

More docs

See the vignettes for more documentation.