How to enumerate all S4 methods implemented by a package? - r

Question

I'm looking for a way to query all S4 methods implemented by a particular package (given through its namespace environment). I think I could enumerate all objects that start with .__T__, but I'd rather prefer using a documented and/or less hackish way.
> ls(asNamespace("RSQLite"), all.names = TRUE, pattern = "^[.]__T__")
[1] ".__T__dbBegin:DBI" ".__T__dbBeginTransaction:RSQLite"
[3] ".__T__dbBind:DBI" ".__T__dbClearResult:DBI"
[5] ".__T__dbColumnInfo:DBI" ".__T__dbCommit:DBI"
[7] ".__T__dbConnect:DBI" ".__T__dbDataType:DBI"
[9] ".__T__dbDisconnect:DBI" ".__T__dbExistsTable:DBI"
[11] ".__T__dbFetch:DBI" ".__T__dbGetException:DBI"
[13] ".__T__dbGetInfo:DBI" ".__T__dbGetPreparedQuery:RSQLite"
[15] ".__T__dbGetQuery:DBI" ".__T__dbGetRowCount:DBI"
[17] ".__T__dbGetRowsAffected:DBI" ".__T__dbGetStatement:DBI"
[19] ".__T__dbHasCompleted:DBI" ".__T__dbIsValid:DBI"
[21] ".__T__dbListFields:DBI" ".__T__dbListResults:DBI"
[23] ".__T__dbListTables:DBI" ".__T__dbReadTable:DBI"
[25] ".__T__dbRemoveTable:DBI" ".__T__dbRollback:DBI"
[27] ".__T__dbSendPreparedQuery:RSQLite" ".__T__dbSendQuery:DBI"
[29] ".__T__dbUnloadDriver:DBI" ".__T__dbWriteTable:DBI"
[31] ".__T__fetch:DBI" ".__T__isSQLKeyword:DBI"
[33] ".__T__make.db.names:DBI" ".__T__show:methods"
[35] ".__T__sqlData:DBI" ".__T__SQLKeywords:DBI"

Answer 1

I think showMethods is the only thing available in methods, but it does not actually return the functions as an object, just prints them to the screen.
The following will return a list of the methods defined in an environment. Adapted from covr::replacements_S4(), which is used to modify all methods in a package to track coverage.
S4_methods <- function(env) {
generics <- methods::getGenerics(env)
res <- Map(generics#.Data, generics#package, USE.NAMES = FALSE,
f = function(name, package) {
what <- methods::methodsPackageMetaName("T", paste(name, package, sep = ":"))
table <- get(what, envir = env)
mget(ls(table, all.names = TRUE), envir = table)
})
res[lengths(res) > 0]
}
m <- S4_methods(asNamespace("DBI"))
length(m)
#> [1] 21
m[1:3]
#> [[1]]
#> [[1]]$DBIObject
#> function(dbObj, obj, ...) {
#> dbiDataType(obj)
#> }
#> <environment: namespace:DBI>
#> attr(,"target")
#> An object of class "signature"
#> dbObj
#> "DBIObject"
#> attr(,"defined")
#> An object of class "signature"
#> dbObj
#> "DBIObject"
#> attr(,"generic")
#> [1] "dbDataType"
#> attr(,"generic")attr(,"package")
#> [1] "DBI"
#> attr(,"class")
#> [1] "MethodDefinition"
#> attr(,"class")attr(,"package")
#> [1] "methods"
#>
#>
#> [[2]]
#> [[2]]$character
#> function(drvName, ...) {
#> findDriver(drvName)(...)
#> }
#> <environment: namespace:DBI>
#> attr(,"target")
#> An object of class "signature"
#> drvName
#> "character"
#> attr(,"defined")
#> An object of class "signature"
#> drvName
#> "character"
#> attr(,"generic")
#> [1] "dbDriver"
#> attr(,"generic")attr(,"package")
#> [1] "DBI"
#> attr(,"class")
#> [1] "MethodDefinition"
#> attr(,"class")attr(,"package")
#> [1] "methods"
#>
#>
#> [[3]]
#> [[3]]$`DBIConnection#character`
#> function(conn, statement, ...) {
#> rs <- dbSendStatement(conn, statement, ...)
#> on.exit(dbClearResult(rs))
#> dbGetRowsAffected(rs)
#> }
#> <environment: namespace:DBI>
#> attr(,"target")
#> An object of class "signature"
#> conn statement
#> "DBIConnection" "character"
#> attr(,"defined")
#> An object of class "signature"
#> conn statement
#> "DBIConnection" "character"
#> attr(,"generic")
#> [1] "dbExecute"
#> attr(,"generic")attr(,"package")
#> [1] "DBI"
#> attr(,"class")
#> [1] "MethodDefinition"
#> attr(,"class")attr(,"package")
#> [1] "methods"

Answer 2

I think you want the showMethods function, as in:
showMethods(where=asNamespace("RSQLite"))
The output is:
Function: dbBegin (package DBI)
conn="SQLiteConnection"
Function: dbBeginTransaction (package RSQLite)
conn="ANY"
Function: dbClearResult (package DBI)
res="SQLiteConnection"
res="SQLiteResult"
Function: dbColumnInfo (package DBI)
res="SQLiteResult"
and this goes on for many more rows. ?showMethods will has some additional arguments for tailoring the results.