Compute a schema for a MongoDB collection — mongo

A DRDL schema is required to match SQL queries with MongoDB aggregation language because MongoDB do not store structured data. However, your data may have a intrinsic structure, which is computed. mtcars_schema() provides precomputed schema for the mtcars dataset, see also ?mtcars for example and testing purpose in the absence of the "mongodrdl" binaries (correct providing the mtcars data are inserted unmodified in an empty collection).

Usage

mongo_schema(
  mongo,
  max_scan = 100L,
  recalc = FALSE,
  path = getOption("mongotranslate.path")
)

# S3 method for mongo_schema
print(x, ...)

mtcars_schema(db = "test", collection = "mtcars")

Arguments

mongo: A mongo object as obtained from mongo() or attr(tbl_mongo(), "mongo").
max_scan: The maximum of documents to scan to elaborate the structure (100 by default).
recalc: Is the schema recalculated, in case it is present in the mongo object (FALSE by default)?
path: The path to the "mongodrdl" binaries. If it is accessible on the search path, or indicated in the "mongotranslate.path" option, no need to specify it.
x: A mongo_schema object.
...: Further arguments passed to print() (not used currently).
db: The name of the database.
collection: The name of the collection.

Value

A mongo_schema object with the schema as character string and a sample attribute that contains a data.frame with the three first documents in the collection, as an example of what is in that collection.

Details

The MongoDB BI Connector's "mongodrdl" external program is used to compute the schema, unless a "schema" attribute is found in the mongo object and recalc = FALSE.

Examples

if (FALSE) {
# We use the same little MongoDB server with mtcars set up for {mongolite}
library(mongoplyr)
database <- "test"
collection <- "mtcars"
mongodb_url <- "mongodb+srv://readwrite:test@cluster0-84vdt.mongodb.net"

# Connect and make sure the collection contains the mtcars dataset
mcon <- mongolite::mongo(collection, database, mongodb_url)
mcon$drop()
mcon$insert(mtcars)

# mtcars_schema() returns a precomputed schema for the mtcars dataset
schema <- mtcars_schema(db = database, collection = collection)
schema

# This schema can be added as an attribute to the connection to avoid
# recalculating it every time
attr(mcon, 'schema') <- schema

# Calling mongo_schema() on a mongo objet that already has a schema attached
# just retrieves it (no recalculation)
schema2 <- mongo_schema(mcon)
identical(schema, schema2)

# ... unless there is nothing attached, or you use recalc = TRUE
# In this case, "mongodrdl" must be installed and accessible
# see vignette("mongoplyr")
schema3 <- mongo_schema(mcon, recalc = TRUE)
schema3
# There is no particular order in the documents collection. So, do not expect
# to get always the same three documents as example
# However, the schema in itself should be fairly consistent (extracted with
#`as.character()` here)
identical(as.character(schema), as.character(schema3))

mcon$disconnect()
}