dimitr
c1cab92cab
1) Reworked error reporting
...
2) Allowed positioned updates for explicit cursors
3) Fixed some possible server crashes
2004-11-17 14:50:33 +00:00
skidder
1e26d6c80d
CURRENT_DATABASE implementation. Established identity may also be used for isc_info_db_id datum
2004-11-11 05:37:52 +00:00
robocop
628b088d2c
Treat string literals and string constants with more respect.
2004-11-10 04:26:45 +00:00
robocop
71fd4ed0a7
Const and use symbolic constants instead of harcoded 20.
2004-11-08 03:15:07 +00:00
robocop
a7498d91e6
Treat string literals and string constants with more respect.
2004-11-08 03:14:17 +00:00
robocop
5a33aeef91
Misc: extra semicolons do not help. :-)
2004-11-07 10:38:13 +00:00
arnobrinkman
4d2a4f6925
bugfix: still were some ambiguous queries possible.
...
fe: select rdb$database.rdb$relation_id from rdb$database, rdb$database
2004-11-02 23:07:09 +00:00
dimitr
7506b3568c
Changed non-standard WEEKDAY and YEARDAY extentions to become non-reserved words.
2004-10-27 12:14:54 +00:00
aafemt
ded2e9ee92
Variable shadowing
2004-10-27 10:14:30 +00:00
dimitr
42c929cfb1
Ported UPDATE/DELETE extentions from Yaffil.
2004-10-27 09:33:08 +00:00
robocop
8cb395414e
Misc.
2004-10-26 07:21:47 +00:00
robocop
34543622a4
a) Error reporting functions must not overflow their buffers.
...
b) Use ISC_STATUS type when dealing with status vector.
2004-10-18 07:19:36 +00:00
robocop
9ac850c3aa
Totally misc.
2004-10-18 06:48:33 +00:00
dimitr
a9370eefa8
Replaced "==" with a SQL-99 compliant distinct predicate.
2004-10-17 08:47:15 +00:00
dimitr
cd413c9b8a
Added the equivalence operator. The major index optimizations are done.
...
Notes:
1) Some optimizations will follow after more deep analysis.
2) The syntax "==" is not yet agreed on. I'd prefer it to be changed.
2004-10-14 18:54:54 +00:00
dimitr
a9772bd861
Restored the old code layout, as suggested by Claudio.
2004-10-14 16:35:13 +00:00
dimitr
a1bbad7ccf
The entire RSE handling in the parser is slightly rewritten. New join syntax is introduced.
...
Notes:
1) Some features are disabled (via isc_wish_list) yet due to their incomplete implementation.
2) The parser may become unstable or cause regressions. It will be fixed ASAP.
2004-10-13 18:37:53 +00:00
robocop
a8da598248
Redundant casts.
2004-10-08 11:08:42 +00:00
robocop
2928b139ae
1.- Style.
...
2.- The const_cast is no longer needed.
2004-10-07 08:28:47 +00:00
robocop
ba48b1d1c9
Style
2004-10-07 08:27:45 +00:00
robocop
ed17cfafbf
Style:
...
if<space>(
for<space>(
switch<space>(
while<space>(
2004-10-04 08:15:00 +00:00
dimitr
9efeab8808
Fixed AV in the parser.
2004-10-03 11:39:03 +00:00
robocop
31ba754981
After talking to Ann, decided to disable several items:
...
- obsolete items like WAL and JOURNAL (cleaned in the past)
- crud from attempts never finished like the shared cache manager for Classic
- DSQL nodes that aren't used. This allowed freeing some keywords.
The functionality was disabled by commenting the code. No code has been deleted.
2004-10-03 04:49:04 +00:00
robocop
cf0f2761a4
parse.cpp needed to be rebuilt after Alex' changes
2004-09-27 03:42:58 +00:00
robocop
20b38f1aad
Got rid of spurious cast.
2004-09-26 11:23:32 +00:00
robocop
600a01d64e
Make compiler happy... nobody wants a cursor name with length 2^31
2004-09-26 07:41:39 +00:00
robocop
592d47e77e
Replace custom name_length() by the general one that uses the same logic
2004-09-26 07:26:42 +00:00
robocop
840bf14b36
If we are using namespace fb_utils, it doesn't make sense to prefix functions in it with fb_ again. Mechanical replacement made one by one (no bulk automatic changes).
2004-09-26 01:49:52 +00:00
robocop
7dcb66a250
When you call va_start, call va_end, don't assume it does nothing because some platforms don't have code for it!
...
Changed some sprintf to SNPRINTF. Work in progress WRT wrapper for this function.
2004-09-25 10:28:09 +00:00
robocop
6ae1fd9918
Style
2004-09-24 06:48:24 +00:00
dimitr
5a67d3a091
Fixed yet another issue with make.cpp noticed by Dimitry Sibiryakov.
2004-09-22 06:54:17 +00:00
dimitr
39a718784b
Fixed another my bug, spotted by Arno.
2004-09-13 19:36:30 +00:00
dimitr
72e9a4923b
Correction.
2004-09-12 10:48:55 +00:00
robocop
cc8beb54e5
Get rid of some reinterpret_cast's.
2004-09-09 09:24:42 +00:00
robocop
33c2a7eb17
Totally misc: typo in comment
2004-09-09 08:58:13 +00:00
robocop
8c2bc04f38
Style
2004-09-09 08:56:33 +00:00
alexpeshkoff
5ba5453a30
fixed AV in DECLARE FILTER
2004-09-08 12:01:30 +00:00
dimitr
0256d424e1
Allowed UNION DISTINCT equal to just UNION (see chapter 7.13 of the SQL spec).
2004-09-06 19:17:16 +00:00
alexpeshkoff
9ff8b3b687
fixed problems, noticed by Claudio
2004-09-06 08:14:34 +00:00
dimitr
f32272d076
Fixed a bug I introduced today.
2004-09-05 21:50:54 +00:00
dimitr
e58ba55a55
Solved issues with literal NULLs in the CASE-based functions.
...
To be reviewed by Arno.
2004-09-05 07:23:42 +00:00
dimitr
f45dc04f6a
Allowed EXTRACT(<part> FROM NULL) to return NULL of SMALLINT/INT (depending on <part>) datatype.
2004-09-04 18:43:11 +00:00
dimitr
cf89467735
Let's describe literal NULLs correctly in expressions.
2004-09-04 18:24:12 +00:00
dimitr
6ad51cf6e8
Fixed my mistake in the plan output. Noticed by Arno.
2004-09-02 07:10:34 +00:00
alexpeshkoff
c3db4aaa97
added (and used in sources) template to change and restore default pool
2004-08-30 18:11:08 +00:00
robocop
092dd6297e
BY SCALAR_ARRAY param can't be used as a return param.
...
There's no support for that. It's only input param to the UDF.
2004-08-27 09:24:39 +00:00
robocop
1994913da7
Bring DSQL up to date with UDF capabilities in the engine.
...
Introduce SCALAR_ARRAY token.
2004-08-27 04:52:32 +00:00
arnobrinkman
ce97981b91
Fix unneeded error-message with simple CASE and NULL (NULL as text in SQL)
...
CASE NULL WHEN NULL THEN 1 END
2004-08-22 00:53:04 +00:00
robocop
27d4dbd374
Not an aesthetical achievement, but prevent us from making mistakes in the same place again. Put variables in their minimal scope.
2004-08-21 09:20:06 +00:00
dimitr
d667e17db5
Fixed uninitialized variable usage reported by Valgrind.
2004-08-20 02:08:39 +00:00
