changelog

0.8.0   lexer has new option to output an invalid token on some kinds of
          parse errors
        switch tests to hspec
        improve parse error messages
	* and x.* changed to only parse in some expression contexts -
  	  select items and function application argument lists
	support sqlite 'without rowid'
	make types in columndefs optional
	allow column constraints and defaults to be in arbitrary order
	partially support parentheses at query expression level (some nested parens don't parse yet)
0.7.1   fix error message source quoting
0.7.0   support autoincrement for sqlite
	support table constraints without separating comma for sqlite
	switch source from literate to regular haskell
	use megaparsec instead of parsec
	use prettyprinter lib instead of pretty
	parsing nested block comments regressed - post a bug if you need this
	fixed fixity parsing of union, except and intersect (matches postgres docs now)
	removed the Errors module - the pretty printer function for errors is in the Parse module
	parses from and pretty prints to strict Text
	use strict Text instead of String everywhere
	tested with latest three main ghc releases (9.8.1, 9.6.4, and 9.4.8) and stack lts 22.5
	the makeSelect helper is now a distinct type, code using it will need some trivial
	  tweaks, this is change so that code using makeSelect doesn't emit warnings
	overhaul website
0.6.1   added odbc handling to sqlsqerver dialect
		added sqlserver dialect case for convert function
0.6.0
        tested with ghc 8.8.1 also
        change the dialect handling - now a dialect is a bunch of flags
          plus a keyword list, and custom dialects are now feasible
          (still incomplete)
        fix parsing for a lot of things which are keywords in the standard
        fix bug with cte pretty printing an extra 'as', which the parser
          also incorrectly accepted
        bug fix: allow keywords that are quoted to be parsed as identifiers

0.5.0
        update to work with ghc 8.6.5, also tested with 8.4.4 and 8.2.1
	rename some of the modules Lexer -> Lex, Parser -> Parse
	add a separate lexer to simplify code and speed up parsing
	replace SqlIndent with new tool, SimpleSqlParserTool (amazing
	  name) which can indent, and parse and lex.
	experiments in new approach to dealing with fixities with separate
	  pass after parsing
	dml :add support for insert, update, delete and truncate
	ddl: add limited support for create schema, plus drop schema
	  create, alter and drop table with defaults and constraints
	  create, alter and drop for domain, view, sequence
	  create and drop for assertion
	access control: simple create and drop for role
	simple grant and revoke
	limited support for transaction management: start transation,
	rollback, commit, savepoint
	fix the precendence of operators which was following the weird
	  postgresql 9.4 and earlier precendences instead of the standard
	refactor the syntax for names, identifiers and strings slightly
	refactor the dialect support, add some support for postgresql
	  syntax
	change parsing of identifiers and strings to not unescape the
	  identifier or string text during parsing
	add some explicit parse failures for probably ambiguous text
	  */ without /* (outside quoted identifier, string) will fail
	  .,e,E following a number without whitespace always fails
	  three symbols together fails explicitly, instead of trying to
	    lex and giving a less good error at parse time (applies to |
	    and : in postgres dialect)
	fix parsing of functions whose name is a keyword (e.g. abs)
	add basic support for parsing odbc syntax ({d 'literals'} {fn
	  app(something)} and {oj t1 left outer join ... }
	rename ValueExpr -> ScalarExpr (I think scalar expression is
	  slightly less incorrect)
        rename CombineQueryExpr to QueryExprSetOp and CombineOp to SetOperatorName
        use explicit data type for sign in interval literals
        add comments to statement syntax (aimed at codegen)
        add support for oracle type size units 'char' and 'byte', example: varchar2(55 byte)
        updated the makefile to use cabal v2 commands
        fix for parsing window functions with keyword names
0.4.4
	tested with ghc 8.2.1 and 8.4.3
0.4.3
	tested with ghc 8.0.2 and 8.2.1
0.4.1 (commit c156c5c34e91e1f7ef449d2c1ea14e282104fd90)
	tested with ghc 7.4.2, 7.6.3, 7.8.4,7.10.0.20150123
	simple demonstration of how dialects could be handled internally
	add ability to add comments to syntax tree to help with generating
	SQL code
0.4.0 (commit 7914898cc8f07bbaf8358d208469392346341964)
	now targets SQL:2011
	update to ghc 7.8.2
	remove dependency on haskell-src-exts
	derive Data and Typeable in all the syntax types
	improve the error messages a great deal
	sql features:
	parse schema qualified table names in from clause (thanks to Sönke
	  Hahn)
	support multiline string literals
	support colon prefix host parameters and introducer
	support unique predicate
	support match predicate
	support array constructors and subscripting
	support character set literals
	support collate
	support escape for string literals as a postfix operator
	parse schema/whatever qualified ids in various places: identifiers
	  (replaces equivalent functionality using '.' operator), function,
	  aggregate, window function names, explicit tables and functions in
	  from clauses, typenames
	support almost all typename syntax for SQL:2011 (just missing refs)
	support most multiset operations (missing some predicates only,
	  likely to be added before next release)
	support two double quotes in a quoted identifier to represent a
	  quote character in the identifier
	support filter and within group for aggregates
	support next value for
	parse special nullary functions
	annoying changes:
	replace Int with Integer in the syntax
	remove support for parsing clauses after the from clause if there
	  is no from clause
	change the syntax representation of quantified comparison
	  predicates
	change the hardcoded collate keyword in substring and trim to use
	  the new collate postfix operator, this also changes the collation
	  name to be an identifier instead of a string
	represent missing setquantifier as a literal default instead of as
	  the actual default value (all in select, distinct in set
	  operators)
	same for sort directions in order by
	implement complete interval literals (fixed the handling of the
	  interval qualifier)
	make most of the standard reserved words actually reserved (still
	  some gaps)
	change the natural in join abstract syntax to match the concrete
	  syntax instead of combining natural, on and using into one field
	remove support for postgresql limit syntax
	bug fixes:
	fix some trailing whitespace issues in the keyword style functions,
	  e.g. extract(day from x), dealing with trailing whitespace on
	  the parens was fixed
	improve some cases of parsing chained prefix or postfix operators
	  (still some issues here)
	fix bug where the 'as' was incorrectly optional in a 'with
	  expression list item'
	fix bug in set operations where 'all' was assumed as the default
	  instead of 'distinct', e.g. 'select * from t union select * from
	  u' was parsed to 'select * from t union all select * from u'
	  instead of 'select * from t union distinct select * from u'.
	fix corresponding bug where 'distinct' was being pretty printed in
	  this case and 'all' was not since the assumed default was the
	  wrong way round
	fix some trailing junk lexing issues with symbols and number
	  literals
	fix number literals to accept upper case E
0.3.1 (commit 5cba9a1cac19d66166aed2876d809aef892ff59f)
	update to work with ghc 7.8.1
0.3.0 (commit 9e75fa93650b4f1a08d94f4225a243bcc50445ae)
	fix the basic operator fixity parsing
	swap the order in select item abstract syntax so it is now
	  (expression, alias) which matches the order in the concrete
	  syntax
	rename ScalarExpr -> ValueExpr
	rename Duplicates to SetQuantifier
	rename qeDuplicates to qeSetQuantifier
	rename OrderField to SortSpec
	rename InThing to InPredValue
	add support for ? for parameterized queries
	add new abstract syntax for special operators whose concrete
	  syntax is a kind of limited named parameters syntax
	add more parsing for these operators: position, convert,
	  translate, overlay, trim, and improve the substring parsing
	add support for multi keyword type names
	   previously:
	     double precision
	     character varying
	   now:
	     double precision,
	     character varying,
	     char varying,
	     character large object,
	     char large object,
	     national character,
	     national char,
	     national character varying,
	     national char varying,
	     national character large object,
	     nchar large object,
	     nchar varying,
	     bit varying
	rename tools/PrettyIt to tools/SQLIdent and add to cabal file as
	  optional executable (disabled by default)
	rename the qeFetch field in Select to qeFetchFirst
	change the pretty printer to use 'fetch first' instead of
	  'fetch next'

0.2.0 (commit 9ea29c1a0ceb2c3f3157fb161d1ea819ea5d64d4)
	'' quotes in string literal
	parse simple interval literal e.g. "interval '1 week'"
	support . in identifiers as a dot operator
	support quoted identifiers
	partial support for explicit window frames
	support multiple test expressions in when branches in case expressions
	rename CastOp to TypedLit
	support typenames with precision and scale in casts
	support nulls first and nulls last in order by
	support grouping expressions: group by (), grouping sets, cube,
	  rollup and parens
	support with recursive
	support values table literal
	support 'table t' syntax
	rename fields qe1 and qe1 in combinequeryexpr to qe0 and qe1
	add support for functions in from clause
	add support for lateral in from clause
	support column aliases in common table expressions
	refactor the tests and add lots more
	parse * in any scalar context instead of trying to restrict it
	support row ctor without 'row' e.g. (a,b) = (c,d)
	add % ^ | & ~ operators
	support ansi standard syntax for offset n rows and fetch first n
	  rows only
	fix keyword parsing to be case insensitive

0.1.0.0 (commit 9bf4012fc40a74ad9a039fcb936e3b9dfc3f90f0)
	initial release