-
Notifications
You must be signed in to change notification settings - Fork 29
/
changelog
224 lines (221 loc) · 9.97 KB
/
changelog
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
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