Support Functions — postgresqlSupport • RPostgreSQL

These functions are the workhorse behind the RPostgreSQL package, but users need not invoke these directly. For details see PostgreSQL.

Usage

## PostgreSQLDriver-related
   postgresqlInitDriver(max.con=16, fetch.default.rec = 500, force.reload=FALSE)
   postgresqlDriverInfo(obj, what, ...)
   postgresqlDescribeDriver(obj, verbose = FALSE, ...)
   postgresqlCloseDriver(drv, ...)

   ## PostgreSQLConnection-related
   postgresqlNewConnection(drv, user, password, host, dbname, port, 
      tty, options, forceISOdate = TRUE)
   postgresqlCloneConnection(con, ...)
   postgresqlConnectionInfo(obj, what, ...)
   postgresqlDescribeConnection(obj, verbose = FALSE, ...)
   postgresqlCloseConnection(con, ...)

   ## PostgreSQLResult-related
   postgresqlExecStatement(con, statement, params, ...)
   postgresqlFetch(res, n=0, ...)
   postgresqlQuickSQL(con, statement, ...)
   postgresqlResultInfo(obj, what, ...)
   postgresqlDescribeResult(obj, verbose = FALSE, ...)
   postgresqlCloseResult(res, ...)
   postgresqlDescribeFields(res, ...)

   ## data mappings, convenience functions, and extensions
   postgresqlDataType(obj, ...)
   postgresqlReadTable(con, name, row.names = "row.names", check.names = TRUE, ...)
   postgresqlWriteTable(con, name, value, field.types, row.names = TRUE,
      overwrite=FALSE, append=FALSE, ..., allow.keywords = FALSE)
   postgresqlpqExec(con, statement)
   postgresqlCopyIn(con, filename)
   postgresqlgetResult(con)
   postgresqlEscapeStrings(con, preescapedstring)
   postgresqlEscapeBytea(con, raw_data)
   postgresqlUnescapeBytea(escapedbytea)
   postgresqlQuoteId(identifiers)
   postgresqlTableRef(identifiers)
   postgresqlImportFile(con, name, value, field.types, overwrite=FALSE,
      append=FALSE, header, row.names, nrows=50, sep=",", eol="\n", 
      skip = 0, quote="\"", ...)

   ## Transaction Management
   postgresqlTransactionStatement(con, statement)

Arguments

max.con: positive integer specifying maximum number of open connections. The current default of 10 is hardcoded in the C code.
fetch.default.rec: default number of rows to fetch (move to R/S-Plus). This default is used in postgresqlFetch. The default is 500.
force.reload: logical indicating whether to re-initialize the driver. This may be useful if you want to change the defaults (e.g., fetch.default.rec). Note that the driver is a singleton (subsequent inits just returned the previously initialized driver, thus this argument).
obj: any of the PostgreSQL DBI objects (e.g., PostgreSQLConnection, PostgreSQLResult).
what: character vector of metadata to extract, e.g., "version", "statement", "isSelect".
verbose: logical controlling how much information to display. Defaults to FALSE.
drv: an PostgreSQLDriver object as produced by postgresqlInitDriver.
con: an PostgreSQLConnection object as produced by postgresqlNewConnection and postgresqlCloneConnection.
res: an PostgreSQLResult object as produced by by postgresqlExecStatement and postgresqlgetResult.
user: a character string with the PostgreSQL's user name.
password: character string with the PostgreSQL's password.
dbname: character string with the PostgreSQL database name.
host: character string with the name (or IP address) of the machine hosting the database. Default is "", which is interpreted as localhost by the PostgreSQL's API.
port: (optional) positive integer specifying the TCP port number that the PostgreSQL server is listening to. Consult the PostgreSQL documentation for details.
tty: Ignored (formerly, this specified where to send server debug output)
options: Command-line options to be sent to the server
forceISOdate: logical indicating whether "set datestyle to ISO" is issued upon connection. Although this is made as an option, the conversion needs to be ISO style for proper conversion of the date datatype.
force: logical indicating whether to close a connection that has open result sets. The default is FALSE.
statement: character string holding one (and only one) SQL statement.
params: actual values that is written as parameters in the statement.
n: number of rows to fetch from the given result set. A value of -1 indicates to retrieve all the rows. The default of 0 specifies to extract whatever the fetch.default.rec was specified during driver initialization postgresqlInit.
name: character vector of names (table names, fields, keywords).
value: a data.frame.
field.types: a list specifying the mapping from R/S-Plus fields in the data.frame value to SQL data types. The default is sapply(value,SQLDataType), see PostgreSQLSQLType.
header: logical, does the input file have a header line? Default is the same heuristic used by read.table, i.e., TRUE if the first line has one fewer column that the second line.
row.names: a logical specifying whether to prepend the value data.frame row names or not. The default is TRUE.
check.names: a logical specifying whether to convert DBMS field names into legal S names. Default is TRUE.
overwrite: logical indicating whether to replace the table name with the contents of the data.frame value. The defauls is FALSE.
append: logical indicating whether to append value to the existing table name.
nrows: number of lines to rows to import using read.table from the input file to create the proper table definition. Default is 50.
sep: field separator character.
eol: end-of-line separator.
skip: number of lines to skip before reading data in the input file.
quote: the quote character used in the input file (defaults to \".
allow.keywords: logical indicating whether column names that happen to be PostgreSQL keywords be used as column names in the resulting relation (table) being written. Defaults to FALSE, forcing postgresqlWriteTable to modify column names to make them legal PostgreSQL identifiers.
preescapedstring: character string to be escaped
raw_data: RAW data to be escaped
escapedbytea: 'escaped' or hex encoded binary data as character
identifiers: one or more character strings to be used as identfier in SQL statement
filename: character string indicating the file which contains the data to be copied to the PostgreSQL backend
...: placeholder for future use.

Value

postgresqlInitDriver returns an PostgreSQLDriver object.

postgresqlDriverInfo returns a list of name-value metadata pairs.

postgresqlDescribeDriver returns NULL (displays the object's metadata).

postgresqlCloseDriver returns a logical indicating whether the operation succeeded or not.

postgresqlNewConnection returns an PostgreSQLConnection object.

postgresqlCloneConnection returns an PostgreSQLConnection object.

postgresqlConnectionInforeturns a list of name-value metadata pairs.

postgresqlDescribeConnection returns NULL (displays the object's metadata).

postgresqlCloseConnection returns a logical indicating whether the operation succeeded or not.

postgresqlExecStatement returns an PostgreSQLResult object.

postgresqlFetch returns a data.frame.

postgresqlQuickSQL returns either a data.frame if the statement is a select-like or NULL otherwise.

postgresqlDescribeResult returns NULL (displays the object's metadata).

postgresqlCloseResult returns a logical indicating whether the operation succeeded or not.

postgresqlDescribeFields returns a data.frame with one row per field with columns name, Sclass, type, len, precision, scale, and nullOK which fully describe each field in a result set. Except for Sclass (which shows the mapping of the field type into an R/S-Plus class) all the information pertains to PostgreSQL's data storage attributes.

postgresqlReadTable returns a data.frame with the contents of the DBMS table.

postgresqlWriteTable returns a logical indicating whether the operation succeeded or not.

postgresqlpqExec returns NUL (executes the statement but does not try to get result. This is called internally from postgresqlWriteTable before postgresqlCopyInDataframe

postgresqlCopyIn returns NULL (copies the content of the file through the socket connection to postgresql backend. This should be used just after COPY tablename FROM STDIN query. This is not used now.)

postgresqlCopyInDataframe returns NULL (copies the content of the dataframe through the socket connection to postgresql backend. Strings are encoded as UTF-8 for transfer. The client_encoding should be set to UTF-8. This should be used just after COPY tablename FROM STDIN query.)

postgresqlgetResult returns an PostgreSQLResult object. This is called after completion of execution of postgresqlpqExec.

postgresqlEscapeStrings returns a character string which is escaped properly so that it can be surrounded with a single quote and used as literal in SQL. The escape procedure is dependent on the character encoding of the connection.

postgresqlEscapeBytea returns a character string which represents the raw data proper escape so that it can be surrounded with a single quote and used as literal in SQL. Note that on 8.X connection E prefix should exist before the first quote. However this changes the behaviour in 9.0, where the E should not exist.

postgresqlUnescapeBytea returns a raw data specified by the character string. The character string should contain the 'escaped' bytea or hex encoded bytea that was output from the database server.

postgresqlQuoteId returns a character string which is quoted as identifier. Returns vector on vector arguemnt.

postgresqlTableRef returns a character string which is quoted as identifier. Reterns a charcter string concatenated with "." so that "dbname"."schemaname"."tablename" reference is created upon c("dbname", "schemaname", "tablename") arguemnt.

postgresqlDataType retuns a character string with the closest

postgresqlResultInfo returns a list of name-value metadata pairs.

postgresqlTransactionStatement returns a logical indicating whether the operation succeeded or not.

Constants

.PostgreSQLPkgName (currently "RPostgreSQL"), .PostgreSQLPkgVersion (the R package version), .PostgreSQLPkgRCS (the RCS revision), .PostgreSQL.NA.string (character that PostgreSQL uses to denote NULL on input), .PostgreSQLSQLKeywords (a lot!) .conflicts.OK.