Orafce - Oracle’s compatibility functions and packages

Functions and operators that emulate a subset of functions and packages from the Oracle RDBMS.

There is an associated Google group - https://groups.google.com/forum/?hl=en#!forum/orafce-general

The Orafce is supported in AWS Aurora with PostgreSQL Compatibility and also in Azure Database for PostgreSQL.

Oracle functions and Oracle packages

This module contains some useful functions that can help with porting Oracle application to PostgreSQL or that can be generally useful.

Built-in Oracle date functions have been tested against Oracle 10 for conformance. Date ranges from 1960 to 2070 work correctly. Dates before 1100-03-01 cannot be verified due to a bug in Oracle.

All functions are fully compatibles with Oracle and respect all known format strings. Detailed descriptions can be found on the internet. Use keywords like : oracle round trunc date iyyy.

List of format strings for trunc, round functions

Y,YY,YYY,YYYY,SYYY,SYEAR	year
I,IY,IYY,IYYY			iso year
Q,				quarter
WW				week, day as first day of year
IW				week, beginning Monday
W				week, day as first day of month
DAY,DY,D			first day of week, sunday
MONTH,MON,MM,RM			month
CC,SCC				century
DDD,DD,J			day
HH,HH12,HH24			hour
MI				minute

Functions round up. That is, a date of July 1st will be rounded to the next year. The 16th of July will be rounded to August.

Date Functions

add_months(date, integer) date - Returns date plus n months
```
	add_months(date '2005-05-31',1) -> 2005-06-30
```
last_date(date) date - Returns last day of the month based on a date value
```
	last_day(date '2005-05-24') -> 2005-05-31
```
next_day(date, text) date - Returns the first weekday that is greater than a date value
```
	next_day(date '2005-05-24', 'monday') -> 2005-05-30
```
next_day(date, integer) date - Same as above. The second argument should be 1..7 and interpreted as Sunday..Satday.
```
	next_day(date '2005-05-24', 1) -> 2005-05-30
```
months_between(date, date) numeric - Returns the number of months between date1 and date2. If a fractional month is calculated, the months_between function calculates the fraction based on a 31-day month.
```
	months_between(date '1995-02-02', date '1995-01-01') -> 1.0322580645161
```
trunc(date, text) date - truncate date according to the specified format
```
	trunc(date '2005-07-12', 'iw') -> 2005-07-11
```
round(date, text) date - will round dates according to the specified format
```
	round(date '2005-07-12', 'yyyy') -> 2006-01-01
```
to_date(text) timestamp - will typecast input text to timestamp. The GUC orafce.nls_date_format is used to specify input text format for this function. If the value is left blank or set as DEFAULT then input text format according to PostgreSQL’s datestyle GUC setting.
```
orafce.nls_date_format value to DEFAULT
```
```
	to_date('2014-05-19 17:23:53+5:30') -> 2014-05-19 17:23:53
```
```
orafce.nls_date_format='YYYY-MMDD HH24:MI:SS'
```
```
	to_date('2014-0519 17:23:53+5:30')  -> 2014-05-19 17:23:53
```

oracle.date data type

This module contains implementation of oracle compatible DATE data type "oracle.date" and functions which are using DATE data type like oracle.add_months,oracle.last_day(),oracle.next_day(),oracle.months_between() etc.

Example:

    set search_path TO oracle,"$user", public, pg_catalog;
    create table oracle_date(col1 date);
    insert into oracle_date values('2014-06-24 12:12:11'::date);
    select * from oracle_date;
            col1
    ---------------------
     2014-06-24 12:12:11
    (1 row)

oracle.date functions

oracle.add_months(timestamp with time zone, integer) - Returns date and time plus n months
```
	oracle.add_months(oracle.date'2005-05-31 10:12:12',1) -> 2005-06-30 10:12:12
```
oracle.last_day(timestamp with time zone) - Returns last day of the month based on a date value
```
	oracle.last_day(oracle.date '2005-05-24 11:12:12') -> 2005-05-31 11:12:12
```
oracle.next_day(timestamp with time zone, text) - Returns the first weekday that is greater than a date value
```
	oracle.next_day(oracle.date '2005-05-24 10:12:12', 'monday') -> 2005-05-30 10:12:12
```
oracle.next_day(timestamp with time zone, integer) - Same as above. The second argument should be 1..7 and interpreted as Sunday..Saturday.
```
	oracle.next_day(oracle.date '2005-05-24 11:21:12', 1) -> 2005-05-29 11:21:12
```
oracle.months_between(timestamp with time zone, timestamp with time zone) - Returns the number of months between timestamp1 and timestamp2. If a fractional month is calculated, the months_between function calculates the fraction based on a 31-day month.
```
	oracle.months_between(oracle.date '1995-02-02 10:00:00', oracle.date '1995-01-01 10:21:11') -> 1.03225806451613
```

oracle.to_date(text,text) - Returns timestamp without time zone.

	oracle.to_date('02/16/09 04:12:12', 'MM/DD/YY HH24:MI:SS') -> 2009-02-16 04:12:12

oracle.to_date(text) - Returns oracle.date

	oracle.to_date('02/16/09 04:12:12') -> 2009-02-16 04:12:12

oracle.sysdate() - Returns statement timestamp at server timezone (orafce.timezone)
```
	oracle.sysdate() -> 2015-12-09 17:47:56
```
oracle.dbtimezone - Returns server time zone - emulated via orafce.timezone
```
	oracle.dbtimezone() -> GMT
```
oracle.sessiontimezone() - Returns session timezone - current PostgreSQL timezone
```
	oracle.sessiontimezone() -> Europe/Prague
```

oracle.to_char(timestamp) - Returns timestamp in nls_date_format.

	orafce.nls_date_format='YY-MonDD HH24:MI:SS'

	oracle.to_char(to_date('14-Jan08 11:44:49+05:30')) -> 14-Jan08 11:44:49

	orafce.nls_date_format='YY-MonDD HH24:MI:SS'

	oracle.to_char(oracle.to_date('21052014 12:13:44+05:30','DDMMYYYY HH24:MI:SS')) -> 14-May21 12:13:44

oracle.date Operators

oracle.+(oracle.date,smallint) - Returns oracle.date

	oracle.to_date('2014-07-02 10:08:55','YYYY-MM-DD HH:MI:SS') + 9::smallint -> 2014-07-11 10:08:55

oracle.+(oracle.date,integer) - Returns oracle.date

	oracle.to_date('2014-07-02 10:08:55','YYYY-MM-DD HH:MI:SS') + 9::integer -> 2014-07-11 10:08:55

oracle.+(oracle.date,bigint) - Returns oracle.date

	oracle.to_date('2014-07-02 10:08:55','YYYY-MM-DD HH:MI:SS') + 9::bigint -> 2014-07-11 10:08:55

oracle.+(oracle.date,numeric) - Returns oracle.date

	oracle.to_date('2014-07-02 10:08:55','YYYY-MM-DD HH:MI:SS') + 9::numeric -> 2014-07-11 10:08:55

oracle.-(oracle.date,smallint) - Returns oracle.date

	oracle.to_date('2014-07-02 10:08:55','YYYY-MM-DD HH:MI:SS') - 9::smallint -> 2014-06-23 10:08:55

oracle.-(oracle.date,integer) - Returns oracle.date

	oracle.to_date('2014-07-02 10:08:55','YYYY-MM-DD HH:MI:SS') - 9::integer -> 2014-06-23 10:08:55

oracle.-(oracle.date,bigint) - Returns oracle.date

	oracle.to_date('2014-07-02 10:08:55','YYYY-MM-DD HH:MI:SS') - 9::bigint -> 2014-06-23 10:08:55

oracle.-(oracle.date,numeric) - Returns oracle.date

	oracle.to_date('2014-07-02 10:08:55','YYYY-MM-DD HH:MI:SS') - 9::numeric -> 2014-06-23 10:08:55

oracle.-(oracle.date,oracle.date) - Returns double precision

	oracle.to_date('2014-07-17 11:10:15', 'yyyy-mm-dd hh24:mi:ss') - oracle.to_date('2014-02-01 10:00:00', 'yyyy-mm-dd hh24:mi:ss') -> 166.048785

You need to set search_path TO oracle,"$user", public, pg_catalog because functions like oracle.add_months,oracle.last_day,oracle.next_day,oracle.months_between are installed side-by-side with pg_catalog.add_months,pg_catalog.last_day,pg_catalog.next_day,pg_catalog.months_between.

Table dual

PostgreSQL does not need Oracle’s table 'dual', but since it is intensively used by Oracle users, it has been added in orafce. This table is in schema oracle. Usually you want allow unqualified access - so you should to add this schema to search_path configuration (like search_path = 'oracle, pg_catalog, "$user", public' in postgresql.conf).

Package dbms_output

PostgreSQL sends information to the client via RAISE NOTICE. Oracle uses dbms_output.put_line(). This works differently from RAISE NOTICE. Oracle has a session queue, put_line() adds a line to the queue and the function get_line() reads from queue. If flag 'serveroutput' is set, then client over all sql statements reads queue. You can use:

    select dbms_output.enable();
    select dbms_output.put_line('first_line');
    select dbms_output.put_line('next_line');
    select * from dbms_output.get_lines(0);

    select dbms_output.enable();
    select dbms_output.serveroutput('t');
    select dbms_output.put_line('first_line');

This package contains the following functions: enable(), disable(), serveroutput(), put(), put_line(), new_line(), get_line(), get_lines(). The package queue is implemented in the session’s local memory.

Package utl_file

This package allows PL/pgSQL programs to read from and write to any files that are accessible from server. Every session can open a maximum of ten files and max line size is 32K. This package contains following functions:

utl_file.fclose(file utl_file.file_type) - close file
utl_file.fclose_all() - close all files
utl_file.fcopy(src_location, src_filename, dest_location, dest_filename[, start_line][, end_line]) - copy text file
utl_file.fflush(file utl_file.file_type) - flushes all data from buffers
utl_file.fgetattr(location, filename) - get file attributes
utl_file.fopen(location text, filename text, file_mode text [, maxlinesize int] [, encoding name]) utl_file.file_type - open file
utl_file.fremove(location, filename) - remove file
utl_file.frename(location, filename, dest_dir, dest_file[, overwrite]) - rename file
utl_file.get_line(file utl_file.file_type) text - read one line from file
utl_file.get_nextline(file utl_file.file_type) text - read one line from file or returns NULL
utl_file.is_open(file utl_file.file_type) bool - returns true, if file is opened
utl_file.new_line(file utl_file.file_type [,rows int]) - puts some new line chars to file
utl_file.put(file utl_file.file_type, buffer text) - puts buffer to file
utl_file.put_line(file utl_file.file_type, buffer text) - puts line to file
utl_file.putf(file utl_file.file_type, format buffer [,arg1 text][,arg2 text][..][,arg5 text]) - put formatted text into file
utl_file.tmpdir() - get path of temp directory

Because PostgreSQL doesn’t support call by reference, some functions are slightly different: fclose and get_line.

  declare f utl_file.file_type;
  begin
    f := utl_file.fopen('/tmp', 'sample.txt', 'r');
    <<readl>>
    loop
      begin
        raise notice '%', utl_file.get_line(f);
      exception
        when no_data_found then
          exit readl;
      end;
    end loop;
    f := fclose(f);
  end;

or second (with PostgreSQL specific function get_nextline)

    declare
      f utl_file.file_type;
      line text;
    begin
      f := utl_file.fopen('/tmp', 'sample.txt', 'r');
      loop
        line := utl_file.get_nextline(f);
        exit when line is NULL;
        raise notice '%', line;
    exception
      when others then
        utl_file.fclose_all();
    end;

Before using the package you have to set the utl_file.utl_file_dir table. It contains all allowed directories without ending symbol ('/' or '\'). On WinNT platform, the paths have to end with symbol '\' every time.

Directory entries can be named (second column in table utl_file.utl_file_dir). The location parameter can be either the directory name or the dictionary path. The location is first interpreted and checked as a directory name. If not found (in 2nd column), then the location is interpreted and checked as a path.

Functions from utl_file package (schema on Postgres) requires a access to table utl_file.utl_file_dir. This fact can be used to control what users can use these functions or not. Default setting is READ for PUBLIC. INSERT, UPDATE can do only privileged user (super user). So unprivileged user can use functions from this package, but cannot to change list of safe directories (content of utl_file.utl_file_dir table). The content of this table is visible for PUBLIC (or should be visible for users who uses functions from this package).

package dbms_sql

This is implementation of Oracle’s API of package DBMS_SQL

It doesn’t ensure full compatibility, but should to decrease a work necessary for successful migration.

Attention: PostgreSQL architecture is different than Oracle’s architecture. PL/pgSQL is executed in same context like SQL engine. Then is not any reason to use Oracle’s patterns like bulk collect and iteration over collection in Postgres to get good performance. This code is designed to reduce work related to porting some applications from Oracle to Postgres, and it can work well. But there will not be any performance advantage aganst buildin PL/pgSQL statements. The emulation of Oracle’s API has memory and CPU overhead, that can be significant on bigger data.

Functionality

This extension implements subset of Oracle’s dbms_sql interface. The goal of this extension is not a compatibility with Oracle, it is designed to reduce some work related migration Oracle’s applications to Postgres. Some basic bulk DML functionality is supported:

    do $$
    declare
      c int;
      a int[];
      b varchar[];
      ca numeric[];
    begin
      c := dbms_sql.open_cursor();
      call dbms_sql.parse(c, 'insert into foo values(:a, :b, :c)');
      a := ARRAY[1, 2, 3, 4, 5];
      b := ARRAY['Ahoj', 'Nazdar', 'Bazar'];
      ca := ARRAY[3.14, 2.22, 3.8, 4];

      call dbms_sql.bind_array(c, 'a', a, 2, 3);
      call dbms_sql.bind_array(c, 'b', b, 3, 4);
      call dbms_sql.bind_array(c, 'c', ca);
      raise notice 'inserted rows %d', dbms_sql.execute(c);
    end;
    $$;

    do $$
    declare
      c int;
      a int[];
      b varchar[];
      ca numeric[];
    begin
      c := dbms_sql.open_cursor();
      call dbms_sql.parse(c, 'select i, ''Ahoj'' || i, i + 0.003 from generate_series(1, 35) g(i)');
      call dbms_sql.define_array(c, 1, a, 10, 1);
      call dbms_sql.define_array(c, 2, b, 10, 1);
      call dbms_sql.define_array(c, 3, ca, 10, 1);

      perform dbms_sql.execute(c);
      while dbms_sql.fetch_rows(c) > 0
      loop
        call dbms_sql.column_value(c, 1, a);
        call dbms_sql.column_value(c, 2, b);
        call dbms_sql.column_value(c, 3, ca);
        raise notice 'a = %', a;
        raise notice 'b = %', b;
        raise notice 'c = %', ca;
      end loop;
      call dbms_sql.close_cursor(c);
    end;
    $$;

There is function dbms_sql.describe_columns_f, that is like procedure dbms_sql.describe_columns. Attention, the type ids are related to PostgreSQL type system. The values are not converted to Oracle’s numbers

    do $$
    declare
      c int;
      r record;
      d dbms_sql.desc_rec;
    begin
      c := dbms_sql.open_cursor();
      call dbms_sql.parse(c, 'select * from pg_class');
      r := dbms_sql.describe_columns(c);
      raise notice '%', r.col_cnt;

      foreach d in array r.desc_t
      loop
        raise notice '% %', d.col_name, d.col_type::regtype;
      end loop;

      call dbms_sql.close_cursor(c);
    end;
    $$;

    do $$
    declare
      c int;
      n int;
      d dbms_sql.desc_rec;
      da dbms_sql.desc_rec[];
    begin
      c := dbms_sql.open_cursor();
      call dbms_sql.parse(c, 'select * from pg_class');
      call dbms_sql.describe_columns(c, n, da);
      raise notice '%', n;

      foreach d in array da
      loop
        raise notice '% %', d.col_name, d.col_type::regtype;
      end loop;

      call dbms_sql.close_cursor(c);
    end;
    $$;

Package dbms_pipe

This package is an emulation of dbms_pipe Oracle package. It provides inter-session communication. You can send and read any message with or without waiting; list active pipes; set a pipe as private or public; and, use explicit or implicit pipes.

The maximum number of pipes is 50.

Shared memory is used to send messages.

An example follows:

-- Session A
select dbms_pipe.create_pipe('my_pipe',10,true); -- explicit pipe creating
select dbms_pipe.pack_message('neco je jinak');
select dbms_pipe.pack_message('anything is else');
select dbms_pipe.send_message('my_pipe',20,0); -- change limit and send without waiting
select * from dbms_pipe.db_pipes; -- list of current pipes

-- Session B
select dbms_pipe.receive_message('my_pipe',1); -- wait max 1 sec for message
select dbms_pipe.next_item_type(); -- -> 11, text
select dbms_pipe.unpack_message_text();
select dbms_pipe.next_item_type(); -- -> 11, text
select dbms_pipe.unpack_message_text();
select dbms_pipe.next_item_type(); -- -> 0, no more items
select dbms_pipe.remove_pipe('my_pipe');

There are some differences compared to Oracle, however:

limit for pipes isn’t in bytes but in elements in pipe
you can send message without waiting
you can send empty messages
next_item_type knows about TIMESTAMP (type 13)
PostgreSQL doesn’t know about the RAW type, use bytea instead

Package dbms_alert

Another means of inter-process communication.

-- Session A
select dbms_alert.register('boo');
select * from dbms_alert.waitany(10);

-- Session B
select dbms_alert.register('boo');
select * from dbms_alert.waitany(10);

-- Session C
select dbms_alert.signal('boo','Nice day');

Package PLVdate

This module contains some functions for working with business days from package PLVdate. Detailed documentation can be found in PLVision library. This package is multicultural, but default configurations are only for european countries (see source code).

You should define your own non-business days (max 50 days) and own holidays (max 30 days). A holiday is any non-business day, which is the same every year. For example, Christmas day in Western countries.

Functions

plvdate.add_bizdays(day date, days int) date - Get the date created by adding <n> business days to a date
plvdate.nearest_bizday(day date) date - Get the nearest business date to a given date, user defined
plvdate.next_bizday(day date) date - Get the next business date from a given date, user defined
plvdate.bizdays_between(day1 date, day2 date) int - Get the number of business days between two dates
plvdate.prev_bizday(day date) date - Get the previous business date from a given date
plvdate_isbizday(date) bool - Call this function to determine if a date is a business day
plvdate.set_nonbizday(dow varchar) - Set day of week as non bussines day
plvdate.unset_nonbizday(dow varchar) - Unset day of week as non bussines day
plvdate.set_nonbizday(day date) - Set day as non bussines day
plvdate.unset_nonbizday(day date) - Unset day as non bussines day
plvdate.set_nonbizday(day date, repeat bool) - Set day as non bussines day, if 'repeat' is true, then day is nonbiz every year
plvdate.unset_nonbizday(day date, repeat bool) - Unset day as non bussines day, if 'repeat' is true, then day is nonbiz every year
plvdate.use_easter() - Easter Sunday and easter monday will be holiday
plvdate.unuse_easter();
plvdate.use_easter(useit boolean);
plvdate.using_easter() bool - If we use easter then returns true
plvdate.use_great_friday() - Easter Great Friday will be holiday
plvdate.unuse_easter();
plvdate.use_easter(useit boolean);
plvdate.using_easter() bool - If we use easter Great Friday as holiday then returns true
plvdate.include_start() - Include starting date in bizdays_between calculation
plvdate.noinclude_start();
plvdate.include_start(include boolean);
plvdate.including_start() bool;
plvdate.default_holidays(varchar) - load default configurations. You can use the following configurations: Czech, German, Austria, Poland, Slovakia, Russia, GB and USA at this moment.
configuration contains only common holidays for all regions. You can add your own regional holiday with plvdate.set_nonbizday(nonbizday, true)

Example:

postgres=# select plvdate.default_holidays('czech');
 default_holidays
 -----------------

(1 row)
postgres=# select to_char(current_date, 'day'),
           plvdate.next_bizday(current_date),
	   to_char(plvdate.next_bizday(current_date),'day');
  to_char  | next_bizday |  to_char
 ----------+-------------+-----------
 saturday  | 2006-03-13  | monday
(1 row)

Change for non-European environment:

select plvdate.unset_nonbizday('saturday');
select plvdate.unset_nonbizday('sunday');
select plvdate.set_nonbizday('friday');
select plvdate.set_nonbizday('2006-05-19', true);
select plvdate.unuse_easter();

Package PLVstr and PLVchr

This package contains some useful string and character functions. Each function supports positive and negative offsets — i.e., offset from the end of the string. For example:

plvstr.left('abcdef',2) -> ab
plvstr.left('abcdef',-2) -> abcd
plvstr.substr('abcdef',1,1) -> a
plvstr.substr('abcdef',-1,1) -> f
plvstr.substr('abcde',-2,1) -> d

List of functions:

plvstr.normalize(str text) - Normalize string - Replace white chars by space, replace spaces by space
plvstr.is_prefix(str text, prefix text, cs bool) - Returns true, if prefix is prefix of str
plvstr.is_prefix(str text, prefix text) - Returns true, if prefix is prefix of str
plvstr.is_prefix(str int, prefix int) - Returns true, if prefix is prefix of str
plvstr.is_prefix(str bigint, prefix bigint) - Returns true, if prefix is prefix of str
plvstr.substr(str text, start int, len int) - Returns substring started on start_in to end
plvstr.substr(str text, start int) - Returns substring started on start_in to end
plvstr.instr(str text, patt text, start int, nth int) - Search pattern in string
plvstr.instr(str text, patt text, start int) - Search pattern in string
plvstr.instr(str text, patt text) - Search pattern in string
plvstr.lpart(str text, div text, start int, nth int, all_if_notfound bool) - Call this function to return the left part of a string
plvstr.lpart(str text, div text, start int, nth int) - Call this function to return the left part of a string
plvstr.lpart(str text, div text, start int) - Call this function to return the left part of a string
plvstr.lpart(str text, div text) - Call this function to return the left part of a string
plvstr.rpart(str text, div text, start int, nth int, all_if_notfound bool) - Call this function to return the right part of a string
plvstr.rpart(str text, div text, start int, nth int) - Call this function to return the right part of a string
plvstr.rpart(str text, div text, start int) - Call this function to return the right part of a string
plvstr.rpart(str text, div text) - Call this function to return the right part of a string
plvstr.lstrip(str text, substr text, num int) - Call this function to remove characters from the beginning
plvstr.lstrip(str text, substr text) - Call this function to remove characters from the beginning
plvstr.rstrip(str text, substr text, num int) - Call this function to remove characters from the end
plvstr.rstrip(str text, substr text) - Call this function to remove characters from the end
plvstr.rvrs(str text, start int, _end int) - Reverse string or part of string
plvstr.rvrs(str text, start int) - Reverse string or part of string
plvstr.rvrs(str text) - Reverse string or part of string
plvstr.left(str text, n int) - Returns firs num_in characters. You can use negative num_in
plvstr.right(str text, n int) - Returns last num_in characters. You can use negative num_ni
plvstr.swap(str text, replace text, start int, length int) - Replace a substring in a string with a specified string
plvstr.swap(str text, replace text) - Replace a substring in a string with a specified string
plvstr.betwn(str text, start int, _end int, inclusive bool) - Find the Substring Between Start and End Locations
plvstr.betwn(str text, start text, _end text, startnth int, endnth int, inclusive bool, gotoend bool) - Find the Substring Between Start and End Locations
plvstr.betwn(str text, start text, _end text) - Find the Substring Between Start and End Locations
plvstr.betwn(str text, start text, _end text, startnth int, endnth int) - Find the Substring Between Start and End Locations
plvchr.nth(str text, n int) - Call this function to return the Nth character in a string
plvchr.first(str text) - Call this function to return the first character in a string
plvchr.last(str text) - Call this function to return the last character in a string
plvchr.is_blank(c int) - Is blank
plvchr.is_blank(c text) - Is blank
plvchr.is_digit(c int) - Is digit
plvchr.is_digit(c text) - Is digit
plvchr.is_quote(c int) - Is quote
plvchr.is_quote(c text) - Is quote
plvchr.is_other(c int) - Is other
plvchr.is_other(c text) - Is other
plvchr.is_letter(c int) - Is letter
plvchr.is_letter(c text) - Is letter
plvchr.char_name(c text) - Returns the name of the character to ascii code as a VARCHAR.
plvchr.quoted1(str text) - Quoted text between '''
plvchr.quoted2(str text) - Quoted text between '"'
plvchr.stripped(str text, char_in text) - Strips a string of all instances of the specified characters

Package PLVsubst

The PLVsubst package performs string substitutions based on a substitution keyword.

plvsubst.string(template_in text, vals_in text[]) - Scans a string for all instances of the substitution keyword and replace it with the next value in the substitution values list
plvsubst.string(template_in text, vals_in text[], subst_in text)
plvsubst.string(template_in text, vals_in text, delim_in text)
plvsubst.string(template_in text, vals_in text, delim_in text, subst_in text)
plvsubst.setsubst(str text) - Set substitution keyword to default '%s'
plvsubst.subst() - Retrieve substitution keyword

Examples:

select plvsubst.string('My name is %s %s.', ARRAY['Pavel','Stěhule']);
          string
 --------------------------
 My name is Pavel Stěhule.
(1 row)

select plvsubst.string('My name is %s %s.', 'Pavel,Stěhule');
          string
 --------------------------
 My name is Pavel Stěhule.
(1 row)

select plvsubst.string('My name is $$ $$.', 'Pavel|Stěhule','|','$$');
          string
 --------------------------
 My name is Pavel Stěhule.
(1 row)

Package DBMS_utility

dms_utility.format_call_stack() — return a formatted string with content of call stack

postgres=# select foo2();
               foo2
 ---------------------------------
 -----  Call Stack  -----
   line             object
 number  statement  name
      1  return     function foo
      1  return     function foo1
      1  return     function foo2
(1 row)

Package PLVlex

This package isn’t compatible with original PLVlex.

postgres=# select * from
	plvlex.tokens('select * from a.b.c join d ON x=y', true, true);

 pos | token  | code |  class  | separator | mod
 ----+--------+------+---------+-----------+------
   0 | select |  527 | KEYWORD |           |
   7 | *      |   42 | OTHERS  |           | self
   9 | from   |  377 | KEYWORD |           |
  25 | a.b.c  |      | IDENT   |           |
  20 | join   |  418 | KEYWORD |           |
  25 | d      |      | IDENT   |           |
  27 | on     |  473 | KEYWORD |           |
  30 | x      |      | IDENT   |           |
  31 | =      |   61 | OTHERS  |           | self
  32 | y      |      | IDENT   |           |
(10 rows)

Warning: Keyword’s codes can be changed between PostgreSQL versions! o plvlex.tokens(str text, skip_spaces bool, qualified_names bool) - Returns table of lexical elements in str.

DBMS_ASSERT

This package protects user input against SQL injection.

dbms_assert.enquote_literal(varchar) varchar - Add leading and trailing quotes, verify that all single quotes are paired with adjacent single quotes.
dbms_assert.enquote_name(varchar [, boolean]) varchar - Enclose name in double quotes. Optional second parameter ensure loweralize of name. Attention - On Oracle is second parameter capitalize!
dbms_assert.noop(varchar) varchar - Returns value without any checking.
dbms_assert.qualified_sql_name(varchar) varchar - This function verifies that the input string is qualified SQL name.
dbms_assert.schema_name(varchar) varchar - Function verifies that input string is an existing schema name.
dbms_assert.simple_sql_name(varchar) varchar -This function verifies that the input string is simple SQL name.
dbms_assert.object_name(varchar) varchar - Verifies that input string is qualified SQL identifier of an existing SQL object.

PLUnit

This unit contains some assert functions.

plunit.assert_true(bool [, varchar]) - Asserts that the condition is true.
plunit.assert_false(bool [, varchar]) - Asserts that the condition is false.
plunit.assert_null(anyelement [, varchar]) - Asserts that the actual is null.
plunit.assert_not_null(anyelement [, varchar]) - Asserts that the actual isn’t null.
plunit.assert_equals(anyelement, anyelement [, double precision] [, varchar]) - Asserts that expected and actual are equal.
plunit.assert_not_equals(anyelement, anyelement [, double precision] [, varchar]) - Asserts that expected and actual are equal.
plunit.fail([varchar]) - Fail can be used to cause a test procedure to fail immediately using the supplied message.

Package DBMS_random

dbms_random.initialize(int) - Initialize package with a seed value.
dbms_random.normal() - Returns random numbers in a standard normal distribution.
dbms_random.random() - Returns random number from -2^31 .. 2^31.
dbms_random.seed(int)
dbms_random.seed(text) - Reset seed value.
dbms_random.string(opt text(1), len int) - Create random string
dbms_random.terminate() - Terminate package (do nothing in Pg)
dbms_random.value() - Returns a random number from [0.0 - 1.0)
dbms_random.value(low double precision, high double precision) - Returns a random number from [low - high)

Others functions

This module contains implementation of functions: concat, nvl, nvl2, lnnvl, decode, greatest, least, bitand, nanvl, sinh, cosh, tanh, oracle.substr and oracle.mod.

oracle.substr(str text, start int, len int) - Oracle compatible substring
oracle.substr(str text, start int) - Oracle compatible substring
oracle.substr(str numeric, start numeric) - Oracle compatible substring
oracle.substr(str numeric, start numeric, len numeric) - Oracle compatible substring
oracle.substr(str varchar, start numeric) - Oracle compatible substring
oracle.substr(str varchar, start numeric,len numeric) - Oracle compatible substring
oracle.lpad(string, length [, fill]) - Oracle compatible lpad
oracle.rpad(string, length [, fill]) - Oracle compatible rpad
oracle.ltrim(string text [, characters text]) - Oracle compatible ltrim
oracle.rtrim(string text [, characters text]) - Oracle compatible rtrim
oracle.btrim(string text [, characters text]) - Oracle compatible btrim
oracle.length(string char) - Oracle compatible length
oracle.listagg(str text [, separator text]) - aggregate values to list
oracle.wm_concat(str text) - aggregate values to comma separatated list
oracle.median(float4) - calculate a median
oracle.median(float8) - calculate a median
oracle.to_number(text) - converts a string to a number
oracle.to_number(numeric) - converts a string to a number
oracle.to_number(numeric,numeric) - converts a string to a number
public.to_multi_byte(text) - Convert all single-byte characters to their corresponding multibyte characters
public.to_single_byte(text) - Convert all multi-byte characters to their corresponding single-byte characters
oracle.greatest(anyelement, anyelement[]) - Oracle compatibility greatest, return NULL on NULL input
oracle.least(anyelement, anyelement[]) - Oracle compatibility least, return NULL on NULL input
oracle.mod(int, int) - Oracle compatibility mod, If the second parameter is zero, it returns the first parameter
oracle.remainder(int, int) - returns remainder of number divided by another number
oracle.remainder(numeric, numeric) - returns remainder of number divided by another number
oracle.sys_guid() - returns bytea - 16 bytes of global uniq id

You might need to set search_path to 'oracle, pg_catalog, "$user", public' because oracle.substr, oracle.lpad, oracle.rpad, oracle.ltrim, oracle.rtrim, oracle.btrim, oracle.length are installed side-by-side with pg_catalog.substr, pg_catalog.lpad, pg_catalog.rpad, pg_catalog.ltrim, pg_catalog.rtrim, pg_catalog.btrim, pg_catalog.length respectively.

Functions oracle.decode, oracle.greatest and oracle.least must always be prefixed by the schema name even if the oracle is before pg_catalog in the search_path because these functions are implemented inside PostgreSQL parser and analyzer. Without the schema name the internal functions will always be used.

Note that in case of lpad and rpad, parameters string and fill can be of types CHAR, VARCHAR, TEXT, VARCHAR2 or NVARCHAR2 (note that the last two are orafce-provided types). The default fill character is a half-width space. Similarly for ltrim, rtrim and btrim.

Note that oracle.length has a limitation that it works only in units of characters because PostgreSQL CHAR type only supports character semantics.

The oracle.substr with three arguments can returns different result (null or empty string) in dependency to setting orafce.using_substring_zero_width_in_substr variable (oracle, warning_oracle, orafce, warning_orafce). This different result is returned only when third argument (substring_length) is zero. Default is warning_oracle, thats means raising warning and returning null.

oracle.sys_guid() function

This functions returns global unique id. It calls specified functions from "uuid-ossp" extension, and then this function should be installed before function sys_guid is used. By default this function uses function uuid_generate_v1, but function uuid_generate_v1mc, uuid_generate_v4 can be used too (by setting orafce.sys_guid_source). oracle.sys_guid can use builin gen_random_uuid func too. In this case the extension "uuid-ossp" is not required.

VARCHAR2 and NVARCHAR2 Support

orafce’s VARCHAR2 implements parts of Oracle database specification about VARCHAR2:

Unit of type modifier = 'bytes' (for character semantics, see NVARCHAR2)
Unlike PostgreSQL varchar, implicit cast to VARCHAR2 does not truncate white spaces over declared maximum length

For these types is possible to use null safe || operator, when you enable orafce.varchar2_null_safe_concat TO true . The behaviour is very similar to Oracle.

Attention: - when result is empty string, then result is NULL. This behaviour is
disabled by default.

Attention: - there is possible incompatibility between 3.7 and older Orafce
releases. A operator function is now marked as stable (was immutable before).
It's not possible to create functional indexes over stable or volatile expressions.

-- null safe concat (disabled by default)
SELECT NULL || 'hello'::varchar2 || NULL;

SET orafce.varchar2_null_safe_concat TO true;
SELECT NULL || 'hello'::varchar2 || NULL;

Please note that PostgreSQL does not allow to dynamically specify how we interpret varchar strings. It always interprets them as 'character' strings as determined by database encoding. So, we cannot support both BYTE and CHARACTER semantics for a given varchar type in the same database. We chose to implement the BYTE semantics as that is default in Oracle. For CHARACTER semantics, please see NVARCHAR2 which by default always implements the CHARACTER semantics.

Please be careful when using the above type to store strings consisting of multibyte encoded characters wherein each character may be composed of an arbitrary number of bytes.

NVARCHAR2 implements the following:

Unit of type modifier = 'characters' (using the character set/encoding of the database)

Use this type if character semantics is preferred.

Please note that unlike Oracle, orafce’s VARCHAR2 and NVARCHAR2 do not impose the 4000 bytes limit on the 'declared' size. In fact it is same as that of PostgreSQL varchar, which is about 10MB (although varchar can theoretically store values of size up to 1GB)

Some byte-based string functions to be used with VARCHAR2 strings

substrb(VARCHAR2, int [, int]) - extract a substring of specified length (in bytes) starting at a given byte position (counting from one); if the third argument isnot specified then length to the end of the string is considered
strposb(VARCHAR2, VARCHAR2) - returns the location of specified substring in a given string (counting from one)
lengthb(VARCHAR2) - returns the length (in bytes) of a given string

Triggers

Oracle doesn’t make differences between NULL and empty string (when a value is used as text). For Postgres NULL and empty string are different values. For simplicity is good to ensure (in Postgres database) use only NULLs (and don’t use empty strings) or use only empty strings (and don’t use NULLs) for text type columns. Both variants has some advantages and disadvantages.

This can be enusured with trigger functions:

oracle.replace_empty_strings([raise_warnings boolean])
oracle.replace_null_strings([raise_warnings boolean])

Optional boolean argument is used as indicator so these functions should to raise warning when row was changed inside these functions.

CREATE TABLE test(id serial, name varchar, surname varchar);
CREATE TRIGGER test_trg
  BEFORE INSERT OR UPDATE
  ON test
  FOR EACH ROW
  EXECUTE PROCEDURE oracle.replace_empty_strings();

INSERT INTO test(name, surname) VALUES('', 'Stehule');

-- name will be replaced by NULL

Emulated views

oracle.user_tab_columns
oracle.user_tables
oracle.user_cons_columns
oracle.user_constraints
oracle.product_componenent_version
oracle.user_objects
oracle.dba_segments

TODO

better documentation
better seralization in dbms_pipe (via _send and _recv functions)
alter shared memory structures by temporary tables: only locks are in shmem, (bitmaps), data in tmp tbl

License

This module is released under BSD licence.

Contributors

The project was founded in 2008 by Pavel Stehule <stehule@kix.fsv.cvut.cz>.

Other contributors:

Gabriele Bartolini (gbartolini)
Jeffrey Cohen (jcohen)
Giles Darold (darold)
Pavan Deolasee (pavanvd)
Peter Eisentraut (petere)
Beena Emerson (b-emerson)
Takahiro Itagaki (itagaki)
Zdenek Kotala (hlipa)
Amit Langote (amitlan)
Heikki Linnakangas (hlinnaka)
Fujii Masao
Marco Nenciarini (mnencia)
Vinayak Pokale
Gavin Sherry (swm)
Pavel Stehule (okbob)
Rahila Syed (rahila)

Leo-XM-Zeng/orafce