CVS log for sql/mysql/parser3mysql.C

Up to [parser3project] / sql / mysql

Request diff between arbitrary revisions - Display revisions graphically

Keyword substitution: kv
Default branch: MAIN

support for multiple client libraries specification, comma separated

warning war

msvc does not support dymamic array sizing

Copyright year updated

whitespace

goto removed

mysql_next_result used (fixes bug #1195)

*** empty log message ***

whitespace

optimizaion: no need for dynamic allocation

+ config_file=my.cnf & config_group=parser3 options (implements feature #1194)

multiple mysql hosts support using host1,host2 syntax (implements feature #1168)

old_client option removed, whitespace

no message

port?port:MYSQL_PORT -> port as 0 means default port and MYSQL_PORT is not defined in MariaDB 10.2 (fixes bug #1163)

warning war

Copyright year updated

SET CHARACTER SET -> SET_NAMES (closes bug #985)

~MySQL_Driver removed, related to issue #925

NEWDECIMAL/BIT check added for mysql 4.X compatibility

bugfix: mysql_fetch_fields usage reverted to mysql_fetch_field usage, as it is more compatible with different mysql client libraries versions
(does not depend on MYSQL_FIELD structure size). (fixes #874)

bugfix: lt_dlinit() can fail, but lt_dlerror return 0, and we count 0 for success.

*** empty log message ***

compiler warnings fixed:
-format '%u' expects type 'unsigned int', but argument 4 has type 'long unsigned int'
-strmcat might overflow destination buffer

ident now works under Linux + ident displays filenames (closes issue #818)
Copyright updated

- an option ?local_infile=1|0 was added. this option should be used if mysql command LOAD LOCAL INFILE is exicuted.

- transcode result was optimized ( new feature: #41 )

native quoting impementation (to avoid bug #1)

*** empty log message ***

no message

- now driver can be built with old mysql includes (prior 4.1), however options ?old_client and ?multi_statements definitely will not work
- while building on freebsd4 we don't add descructor because it causes core while executing

- old_client option added. it allow to disable multi result.

- use new API and new version (10.0)
- $.limit(0) fixed
- transcode server error messaged if ClientCharset option was specified
- less calls to transcode method if ClientCharset specified but it's equal to request:charset

- some spaces to tabs

- multi statements now can be switched on with option multi_statements=1

- CLIENT_MULTI_STATEMENTS => CLIENT_MULTI_RESULTS

- call mysql_server_end before unload driver

- charser and ClientCharset options now can be specified together.
  it can be usable for transcoding text from utf8 (in DB) to koi8: &charset=utf-8&ClientCharset=utf8
  in this case sql server return data in utf and parser itself transcode it from utf to koi8

- support storage procedure for MySQL 5 (patch from Sumo: http://www.parser.ru/forum/?id=54595 )

sql connect string, rsplit @

beauty: detailed error message on lt_dlopen failure

new api supported, not implemented yet:
	^void:sql{call paf(:a)}[
		$.bind[hash]
	]

beauty: param renamed to reflect it's nature

merged: bugfixes from 3.1.2

merged bugfixes from 3.1.2

bugfix: tolower/toupper named with _str suffix to work around cases when they are macros [freebsd]

beauty: merged from HEAD [mistakenly commited there]

ctype configured, gcc const wonders worked around

new: mysql option [like that recently added to oracle driver]
            ClientCharset=parser-charset << charset in which parser thinks client works

bugfix: connection struct mem to be controlled by libgc [oracle already fixed before]

beauty: connection handle=0 after close

no message

change: moved to new interface [without extra 'services' passing back/forth]

mysql: transaction support: ?autocommit=0

moved tempate_gc to HEAD

convention changed: all resulting strings are zero-terminated

mistakenly patched HEAD branch. copying changes to template_gc one

merged lt_init + column count bugfix from main trunc

seen some broken client,
which reported "44" for column count of response to "select 2+2"
added check on that

2002->2003

long ago forgotten lt_dlinit() called now

changed exception handling mech in sql handlers #2

changed exception handling mech in sql handlers

name spelling

2002

introducing estimating String::cstr_bufsize, 0

init(not const

2001.11.06 numrows on "HP-UX istok1 B.11.00 A 9000/869 448594332 two-user license"
		3.23.42 & 4.0.0.alfa never worked, both subst & .sl version returned 0

changed connect string format to support options
charset=cp1251_koi8
timeout=3
compress=1&
named_pipe=1

no message

Initial revision