- developers-list - monetdb.org

Re: [Monetdb-developers] php
by Arjan Scherpenisse 08 Sep '04

08 Sep '04

De extension_dir .ini variabele kan maar 1 pad bevatten, helaas niet meerdere paden gescheiden door ':'s :(... En de waarde van extension_dir alleen op MONETDB_PREFIX/share/MonetDB/php zetten lijkt me ook geen optie omdat dan andere PHP applicaties gaan breken... Arjen P. de Vries wrote: > hoi Arjan, > > De variable extension_dir kan ook gezet worden in php.ini of httpd.conf > (volgens http://aspn.activestate.com/ASPN/docs/PHP/function.ini-set.html) > > misschien dat het dan geen probleem is, als het monet-path wordt toegevoegd > aan php.ini (even ervanuitgaand dat dat ook per-user kan)? > > Groetjes, > > Arjen > > | De monetdb.so file moet echt in PHP_EXTENSION_DIR staan, anders kan de > | module niet worden gebruikt binnen PHP. Hoe zullen we dit fixen? Goed > | documenteren, of mischien een phpinstall.sh scriptje toevoegen wat het > | voor je regelt (wat als root moet worden uitgevoerd)? > | -- > | Arjan Scherpenisse > | > | Centrum voor Wiskunde en Informatica, > | Amsterdam, the Netherlands > | > | > | ------------------------------------------------------- > | This SF.Net email is sponsored by BEA Weblogic Workshop > | FREE Java Enterprise J2EE developer tools! > | Get your free copy of BEA WebLogic Workshop 8.1 today. > | http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click > | _______________________________________________ > | Monetdb-developers mailing list > | Monetdb-developers(a)lists.sourceforge.net > | https://lists.sourceforge.net/lists/listinfo/monetdb-developers > | -- Arjan Scherpenisse Centrum voor Wiskunde en Informatica, Amsterdam, the Netherlands

2 2

[Monetdb-developers] php
by Arjan Scherpenisse 08 Sep '04

08 Sep '04

De monetdb.so file moet echt in PHP_EXTENSION_DIR staan, anders kan de module niet worden gebruikt binnen PHP. Hoe zullen we dit fixen? Goed documenteren, of mischien een phpinstall.sh scriptje toevoegen wat het voor je regelt (wat als root moet worden uitgevoerd)? -- Arjan Scherpenisse Centrum voor Wiskunde en Informatica, Amsterdam, the Netherlands

1 0

[Monetdb-developers] mapi: how to check mil query on errors?
by Arjan Scherpenisse 19 Aug '04

19 Aug '04

For SQL code, I check now with: hdl = mapi_query(mid, "very wrong query."); while( mapi_next_result(hdl) ) { if (mapi_result_error(hdl)) { // error handling here how to do this equivalent for mil queries? Or dont we need mil queries anymore? My personal opinion is to keep doing it the old way for mil-type queries (if (mapi_error()) etc), but that does not seem to work... -- Arjan Scherpenisse Centrum voor Wiskunde en Informatica, Amsterdam, the Netherlands

2 1

[Monetdb-developers] Weird server/client behaviour?
by Fabian 18 Aug '04

18 Aug '04

Because I don't really know if it's a bug on whose side, I decided to at least put in the issue via this list. When I try to load the VOC dataset using the JdbcClient with the new -Xbatching flag, the client hangs when it wants to send its last batch. I'm batching in chunks of 8000 statements. Research pointed out that the Java streams are trying to write data to monet, but it simply can't write it, and locks. I pinned down the problem to be in the very low level code regions which are outside of the Java reachable scope and go into the level of the operating system. It just cannot write to the stream. Monet on the other hand seems to be waiting for data. When I use a smaller dataset the batching goes fine. Any clues on what to look for? GDB couldn't help me out much.

1 0

[Monetdb-developers] Re: DBI
by Arjan Scherpenisse 10 Aug '04

10 Aug '04

Martin Kersten wrote: > That's a real pity. Would it require many changes? > Otherwise, we would have to invalidate the perl API > in the upcoming release. I dont know if it requires many changes. The reply-parsing code in monet.pm looks quite 'unstable' to me; I have not found the problem yet. In my opinion it would be better to base programming language APIs on the existing libMapi C code. That way protocol changes etc. only need to be implemented in a single place... Anyway I do not think I have much time to look further into this, I have a thesis to finish ;-) Arjan > > Arjan Scherpenisse wrote: > >> Martin Kersten wrote: >> >>> Does it run against our current DBI? >> >> >> >> No, it does not run. Connecting works, and executing queries too, but >> fetching results from queries does not. >> >> It is likely that this is the case because it is a pure-perl >> implementation and it is not up-to-date with some of the changes in >> the protocol lately. >> >> Even the simple test script which comes with the current >> implementation does not run anymore :( >> >> Arjan >> > -- Arjan Scherpenisse Centrum voor Wiskunde en Informatica, Amsterdam, the Netherlands

2 1

[Monetdb-developers] How to create Tests
by Stefan.Manegold＠cwi.nl 15 Jul '04

15 Jul '04

Hi, I haven't had any time, yet, to prepare a real HowTo on this topic, but for those of you who are interested, here's a trace of what I did to make Arjan's sql/src/test/bugs/bug-sf-990096.sql script into a "real" test. Stefan ======== 14:51:06 manegold@draco:~/Monet/4.3/sql $ echo 'bug-sf-990096' >> src/test/bugs/Tests/All 14:51:13 manegold@draco:~/Monet/4.3/sql $ echo '$RELSRCDIR/../bug-sf-990096.sql' > src/test/bugs/Tests/bug-sf-990096.sql.src 14:51:21 manegold@draco:~/Monet/4.3/sql $ Mtest_sql src/test/bugs bug-sf-990096 IGNORE = ^# TIMEOUT = 60 RECURSIVE = 0 CONTEXT = -C1 ACCURACY = 1 MONET_SOURCE = /ufs/manegold/Monet/4.3/MonetDB MONET_BUILD = /var/tmp/_BUILD_MONET_GNU_32_d__ MONET_PREFIX = /var/tmp/_PREFIX__GNU_32_d__ TSTSRCBASE = /ufs/manegold/Monet/4.3/sql TSTBLDBASE = /var/tmp/_BUILD_SQL_GNU_32_d__ TSTTRGBASE = /var/tmp/_PREFIX__GNU_32_d__ SQL_PREFIX = /var/tmp/_PREFIX__GNU_32_d__ SQLCLIENT = MapiClient --language=sql MAPICLIENT = MapiClient X100CLIENT = x100run_mapi MAPIPORT = 40891 SQLPORT = 52395 GDK_DEBUG = 10 MONET_MOD_PATH = /var/tmp/_PREFIX__GNU_32_d__/lib/MonetDB SQL_LOGDIR = /var/tmp/_PREFIX__GNU_32_d__/var/MonetDB/log GDK_DBFARM = /var/tmp/_PREFIX__GNU_32_d__/var/MonetDB/dbfarm setMONET_MOD_PATH = --set "monet_mod_path=/var/tmp/_PREFIX__GNU_32_d__/lib/MonetDB" MAPI_MIL = /var/tmp/_BUILD_MONET_GNU_32_d__/src/mapi/mapi.mil SQL_SERVER = sql_server SQL_SERVER_MIL = /var/tmp/_BUILD_SQL_GNU_32_d__/src/backends/monet4/sql_server.mil MDIFF = /var/tmp/_PREFIX__GNU_32_d__/bin/Mdiff : Mdiff MSERVER = /var/tmp/_PREFIX__GNU_32_d__/bin/Mserver : Mserver "--config=/var/tmp/_PREFIX__GNU_32_d__/etc/monet.conf" --debug=10 --set "monet_mod_path=/var/tmp/_PREFIX__GNU_32_d__/lib/MonetDB" --set "gdk_dbfarm=/var/tmp/_PREFIX__GNU_32_d__/var/MonetDB/dbfarm" --set "sql_logdir=/var/tmp/_PREFIX__GNU_32_d__/var/MonetDB/log" --set mapi_port=40891 --set sql_port=52395 --set monet_prompt= --trace MTIMEOUT = /var/tmp/_PREFIX__GNU_32_d__/bin/Mtimeout : Mtimeout -timeout 60 MSHUTDOWN = /var/tmp/_PREFIX__GNU_32_d__/bin/Mshutdown.py : Mshutdown.py "--config=/var/tmp/_PREFIX__GNU_32_d__/etc/monet.conf" --host=draco --port=40891 PYTHON = /usr/bin/python : python SQL_CLIENT = /var/tmp/_PREFIX__GNU_32_d__/bin/MapiClient : MapiClient --language=sql -u monetdb -P monetdb --host=draco --port=52395 X100_CLIENT = : x100run_mapi -u monetdb -P monetdb "--config=/var/tmp/_PREFIX__GNU_32_d__/etc/monet.conf" --host=draco --port=40891 MAPI_CLIENT = /var/tmp/_PREFIX__GNU_32_d__/bin/MapiClient : MapiClient -u monetdb -P monetdb "--config=/var/tmp/_PREFIX__GNU_32_d__/etc/monet.conf" --host=draco --port=40891 Bits: 32bit Modules: ('main', 'kernel', '_kernel', 'arith', '_arith', 'str', '_str', 'uchr', '_uchr', 'bat', '_bat', 'algebra', '_algebra', 'sys', '_sys', 'trans', '_trans', 'streams', 'xtables', 'mkey', 'profiler', 'aggr', 'aggrX3', 'alarm', 'ascii_io', 'bat_arith', 'blob', 'decimal', 'enum', 'lock', 'mmath', 'monettime', 'radix', 'stats', 'tcpip', 'unix', 'url', 'counters', 'pcl', 'bitset', 'bitvector', 'cluster', 'iterator', 'malalgebra', 'mel', 'mprof', 'oo7', 'qt', 'salgebra', 'tpcd', 'vector', 'wisc', 'ddbench', 'bat_mmath', 'bat_cast', 'array', 'color', 'image', 'txtsim', 'calib', 'mapi', 'sql_server', 'pf_support') Running tests ['bug-sf-990096'] in directory /ufs/manegold/Monet/4.3/sql/src/test/bugs. 14:51:29> src/test/bugs/bug-sf-990096.sql (<=240s) ... 1.167s bug-sf-990096.stable.out.FILTERED and bug-sf-990096.test.out.FILTERED differ SIGNIFICANTLY! bug-sf-990096.stable.err.FILTERED and bug-sf-990096.test.err.FILTERED differ SIGNIFICANTLY! !ERROR: Testing FAILED SIGNIFICANTLY ! 1 out of 1 tests produced SIGNIFICANTLY different output First, check the testing results in /var/tmp/_PREFIX__GNU_32_d__/mTests/index.html ! Then, fix the problems by: - fixing sources and test scripts - fixing stable output by hand - approving test output by Mapprove.py (cf. Mapprove.py -?) After that, re-run Mtest. 14:51:32 manegold@draco:~/Monet/4.3/sql $ Mapprove_sql src/test/bugs bug-sf-990096 FORCE = 0 RECURSIVE = 0 EXTENSION = ['out', 'err'] SYSTEM = MONET_SOURCE = /ufs/manegold/Monet/4.3/MonetDB MONET_BUILD = /var/tmp/_BUILD_MONET_GNU_32_d__ MONET_PREFIX = /var/tmp/_PREFIX__GNU_32_d__ TSTSRCBASE = /ufs/manegold/Monet/4.3/sql TSTBLDBASE = /var/tmp/_BUILD_SQL_GNU_32_d__ TSTTRGBASE = /var/tmp/_PREFIX__GNU_32_d__ SQL_PREFIX = /var/tmp/_PREFIX__GNU_32_d__ Running tests ['bug-sf-990096'] in directory /ufs/manegold/Monet/4.3/sql/src/test/bugs. Approving src/test/bugs/bug-sf-990096.test.out -> stable.out (creating new file) patching file /ufs/manegold/Monet/4.3/sql/src/test/bugs/Tests/bug-sf-990096.stable.out Approving src/test/bugs/bug-sf-990096.test.err -> stable.err (creating new file) Mapprove.py Warning: Skipping new (error) message: "+ERROR = !Column: a.i unknown" patching file /ufs/manegold/Monet/4.3/sql/src/test/bugs/Tests/bug-sf-990096.stable.err First, run 'cvs -q diff to check what you have changed. Then, re-run Mtest.py. In case (some of) the skipped error messages are correct/expected, re-run Mapprove.py with -f to force their approval. 14:51:51 manegold@draco:~/Monet/4.3/sql $ joe src/test/bugs/Tests/bug-sf-990096.stable.??? Processing '/ufs/manegold/.joerc'...done Processing '/ufs/manegold/.joerc'...done File src/test/bugs/Tests/bug-sf-990096.stable.err not changed so no update needed. 14:52:45 manegold@draco:~/Monet/4.3/sql $ diff -u src/test/bugs/Tests/bug-sf-990096.stable.out{~,} --- src/test/bugs/Tests/bug-sf-990096.stable.out~ 2004-07-13 14:52:43.000000000 +0200 +++ src/test/bugs/Tests/bug-sf-990096.stable.out 2004-07-13 14:52:43.000000000 +0200 @@ -27,6 +27,12 @@ # 0, # id [ 1, 1 ] +# i, i # name +# mediumint, mediumint # type +# 1, # tuplecount +# 0, # id +[ 1, 1 ] + # 14:51:31 > # 14:51:31 > Done. # 14:51:31 > 14:52:54 manegold@draco:~/Monet/4.3/sql $ diff -u src/test/bugs/Tests/bug-sf-990096.stable.err{~,} --- src/test/bugs/Tests/bug-sf-990096.stable.err~ 2004-07-13 14:52:36.000000000 +0200 +++ src/test/bugs/Tests/bug-sf-990096.stable.err 2004-07-13 14:52:36.000000000 +0200 @@ -56,10 +56,6 @@ # 14:51:30 > Mtimeout -timeout 60 MapiClient --language=sql -u monetdb -P monetdb --host=draco --port=52395 < bug-sf-990096.sql # 14:51:30 > -MAPI = monetdb@draco:52395 -ACTION= read_into_cache -QUERY = SELECT a.i FROM - # 14:51:31 > # 14:51:31 > Done. 14:52:58 manegold@draco:~/Monet/4.3/sql $ Mtest_sql src/test/bugs bug-sf-990096 IGNORE = ^# TIMEOUT = 60 RECURSIVE = 0 CONTEXT = -C1 ACCURACY = 1 MONET_SOURCE = /ufs/manegold/Monet/4.3/MonetDB MONET_BUILD = /var/tmp/_BUILD_MONET_GNU_32_d__ MONET_PREFIX = /var/tmp/_PREFIX__GNU_32_d__ TSTSRCBASE = /ufs/manegold/Monet/4.3/sql TSTBLDBASE = /var/tmp/_BUILD_SQL_GNU_32_d__ TSTTRGBASE = /var/tmp/_PREFIX__GNU_32_d__ SQL_PREFIX = /var/tmp/_PREFIX__GNU_32_d__ SQLCLIENT = MapiClient --language=sql MAPICLIENT = MapiClient X100CLIENT = x100run_mapi MAPIPORT = 46225 SQLPORT = 53058 GDK_DEBUG = 10 MONET_MOD_PATH = /var/tmp/_PREFIX__GNU_32_d__/lib/MonetDB SQL_LOGDIR = /var/tmp/_PREFIX__GNU_32_d__/var/MonetDB/log GDK_DBFARM = /var/tmp/_PREFIX__GNU_32_d__/var/MonetDB/dbfarm setMONET_MOD_PATH = --set "monet_mod_path=/var/tmp/_PREFIX__GNU_32_d__/lib/MonetDB" MAPI_MIL = /var/tmp/_BUILD_MONET_GNU_32_d__/src/mapi/mapi.mil SQL_SERVER = sql_server SQL_SERVER_MIL = /var/tmp/_BUILD_SQL_GNU_32_d__/src/backends/monet4/sql_server.mil MDIFF = /var/tmp/_PREFIX__GNU_32_d__/bin/Mdiff : Mdiff MSERVER = /var/tmp/_PREFIX__GNU_32_d__/bin/Mserver : Mserver "--config=/var/tmp/_PREFIX__GNU_32_d__/etc/monet.conf" --debug=10 --set "monet_mod_path=/var/tmp/_PREFIX__GNU_32_d__/lib/MonetDB" --set "gdk_dbfarm=/var/tmp/_PREFIX__GNU_32_d__/var/MonetDB/dbfarm" --set "sql_logdir=/var/tmp/_PREFIX__GNU_32_d__/var/MonetDB/log" --set mapi_port=46225 --set sql_port=53058 --set monet_prompt= --trace MTIMEOUT = /var/tmp/_PREFIX__GNU_32_d__/bin/Mtimeout : Mtimeout -timeout 60 MSHUTDOWN = /var/tmp/_PREFIX__GNU_32_d__/bin/Mshutdown.py : Mshutdown.py "--config=/var/tmp/_PREFIX__GNU_32_d__/etc/monet.conf" --host=draco --port=46225 PYTHON = /usr/bin/python : python SQL_CLIENT = /var/tmp/_PREFIX__GNU_32_d__/bin/MapiClient : MapiClient --language=sql -u monetdb -P monetdb --host=draco --port=53058 X100_CLIENT = : x100run_mapi -u monetdb -P monetdb "--config=/var/tmp/_PREFIX__GNU_32_d__/etc/monet.conf" --host=draco --port=46225 MAPI_CLIENT = /var/tmp/_PREFIX__GNU_32_d__/bin/MapiClient : MapiClient -u monetdb -P monetdb "--config=/var/tmp/_PREFIX__GNU_32_d__/etc/monet.conf" --host=draco --port=46225 Bits: 32bit Modules: ('main', 'kernel', '_kernel', 'arith', '_arith', 'str', '_str', 'uchr', '_uchr', 'bat', '_bat', 'algebra', '_algebra', 'sys', '_sys', 'trans', '_trans', 'streams', 'xtables', 'mkey', 'profiler', 'aggr', 'aggrX3', 'alarm', 'ascii_io', 'bat_arith', 'blob', 'decimal', 'enum', 'lock', 'mmath', 'monettime', 'radix', 'stats', 'tcpip', 'unix', 'url', 'counters', 'pcl', 'bitset', 'bitvector', 'cluster', 'iterator', 'malalgebra', 'mel', 'mprof', 'oo7', 'qt', 'salgebra', 'tpcd', 'vector', 'wisc', 'ddbench', 'bat_mmath', 'bat_cast', 'array', 'color', 'image', 'txtsim', 'calib', 'mapi', 'sql_server', 'pf_support') Running tests ['bug-sf-990096'] in directory /ufs/manegold/Monet/4.3/sql/src/test/bugs. 14:53:17> src/test/bugs/bug-sf-990096.sql (<=240s) ... 1.065s bug-sf-990096.stable.out.FILTERED and bug-sf-990096.test.out.FILTERED differ SIGNIFICANTLY! bug-sf-990096.stable.err.FILTERED and bug-sf-990096.test.err.FILTERED differ SIGNIFICANTLY! !ERROR: Testing FAILED SIGNIFICANTLY ! 1 out of 1 tests produced SIGNIFICANTLY different output First, check the testing results in /var/tmp/_PREFIX__GNU_32_d__/mTests/index.html ! Then, fix the problems by: - fixing sources and test scripts - fixing stable output by hand - approving test output by Mapprove.py (cf. Mapprove.py -?) After that, re-run Mtest. 14:53:19 manegold@draco:~/Monet/4.3/sql $ rm src/test/bugs/Tests/bug-sf-990096.stable.???~ -rw-r----- 1 manegold ins 2807 Jul 13 14:52 src/test/bugs/Tests/bug-sf-990096.stable.err~ -rw-r----- 1 manegold ins 1140 Jul 13 14:52 src/test/bugs/Tests/bug-sf-990096.stable.out~ Delete these 2 files/directories? (y/N) \c y 2 files deleted. 14:53:56 manegold@draco:~/Monet/4.3/sql $ cvs add src/test/bugs/Tests/bug-sf-990096.* cvs add: warning: unrecognized response '' from cvs server cvs add: use 'cvs commit' to add these files permanently 14:54:16 manegold@draco:~/Monet/4.3/sql $ cvs up src/test/bugs/Tests M src/test/bugs/Tests/All M src/test/bugs/Tests/bug-00001.stable.out A src/test/bugs/Tests/bug-sf-990096.sql.src A src/test/bugs/Tests/bug-sf-990096.stable.err A src/test/bugs/Tests/bug-sf-990096.stable.out 14:54:34 manegold@draco:~/Monet/4.3/sql $ cvs diff src/test/bugs/Tests Index: src/test/bugs/Tests/All =================================================================== RCS file: /cvsroot/monetdb/sql/src/test/bugs/Tests/All,v retrieving revision 1.17 diff -u -r1.17 All --- src/test/bugs/Tests/All 24 Jun 2004 15:26:59 -0000 1.17 +++ src/test/bugs/Tests/All 13 Jul 2004 12:54:42 -0000 @@ -25,3 +25,4 @@ bug-sf-975308 bug-sf-978043 bug-sf-978045 +bug-sf-990096 Index: src/test/bugs/Tests/bug-00001.stable.out =================================================================== RCS file: /cvsroot/monetdb/sql/src/test/bugs/Tests/bug-00001.stable.out,v retrieving revision 1.3 diff -u -r1.3 bug-00001.stable.out --- src/test/bugs/Tests/bug-00001.stable.out 20 Jun 2004 21:46:22 -0000 1.3 +++ src/test/bugs/Tests/bug-00001.stable.out 13 Jul 2004 12:54:42 -0000 @@ -60,8 +60,8 @@ [ 1005711, "privileges", 1005548, "", 1 ] [ 1005740, "history", 1005548, "", 1 ] [ 1005778, "env", 1005548, "", 1 ] -[ 1007047, "foo", 1005548, "", 0 ] -[ 1007212, "ff", 1005548, "", 0 ] +[ 1007035, "foo", 1005548, "", 0 ] +[ 1007200, "ff", 1005548, "", 0 ] # i # name # int # type # 2 # tuplecount @@ -91,9 +91,9 @@ [ 1005711, "privileges", 1005548, "", 1 ] [ 1005740, "history", 1005548, "", 1 ] [ 1005778, "env", 1005548, "", 1 ] -[ 1007047, "foo", 1005548, "", 0 ] -[ 1007212, "ff", 1005548, "", 0 ] -[ 1007468, "s4", 1005548, "", 0 ] +[ 1007035, "foo", 1005548, "", 0 ] +[ 1007200, "ff", 1005548, "", 0 ] +[ 1007456, "s4", 1005548, "", 0 ] # i # name # int # type # 1 # tuplecount cvs diff: src/test/bugs/Tests/bug-sf-990096.sql.src is a new entry, no comparison available cvs diff: src/test/bugs/Tests/bug-sf-990096.stable.err is a new entry, no comparison available cvs diff: src/test/bugs/Tests/bug-sf-990096.stable.out is a new entry, no comparison available 14:54:43 manegold@draco:~/Monet/4.3/sql $ ======== -- | Dr. Stefan Manegold | mailto:Stefan.Manegold@cwi.nl | | CWI, P.O.Box 94079 | http://www.cwi.nl/~manegold/ | | 1090 GB Amsterdam | Tel.: +31 (20) 592-4212 | | The Netherlands | Fax : +31 (20) 592-4312 |

1 0

[Monetdb-developers] Auto-commit
by Fabian 04 Jul '04

04 Jul '04

Consider the following multi query statement: (1) update x set somecol = 'newval'; select * form x; (mind the explicit typo in the second query (from)) Now consider this whole statement being sent at once with auto_commit disabled. The second query will fail due to a syntax error and the whole connection will have to be rollbacked. If the error would be reported, and the --- theoretical --- update count would be ignored, there would be no problem at all, I think. However, suppose auto_commit is switched on. Now the scenario brings some problems: 1. the first update is executed, and committed; thus an update count should be returned. 2. the second query fails and generates an error, however this error can't be displayed if the update count is displayed, the statement executes or fails, not partial. As it looks now, using JDBC, a statement has a mutually exclusive result; either success, or failure. The current implementation of auto_commit, commits after each query, I think it should commit after each statement, so the statement as in (1) gets rollbacked as a whole and results in an error, instead of in a result and an error. Is this a reasonable thought of mine, or does it clash with other things? How does ODBC handle this situation?

2 3

[Monetdb-developers] Bus Error?
by Arjan Scherpenisse 01 Jul '04

01 Jul '04

While executing a big query, sometimes I get a Bus error. Now, I've looked up that "This signal is raised when your program accesses memory in an invalid way"; but what i am wondering, is this signal a) A bug in Monet b) A way of Monet telling me that it has run out of memory? -- Arjan Scherpenisse Centrum voor Wiskunde en Informatica, Amsterdam, the Netherlands

1 0

[Monetdb-developers] php module compiling
by Arjan Scherpenisse 14 Jun '04

14 Jun '04

The following code compiles the php module: gcc -fpic -DCOMPILE_DL=1 `php-config --includes` `monet-config --cflags` -c -o monetdb.o php_monetdb.c gcc -shared `php-config --libs` `monet-config --libs` -lMapi -rdynamic -o monetdb.so monetdb.o And installing: cp monetdb.so `php-config --extension-dir` Maybe someone (stefan?) with more knowledge of the build process can integrate this stuff. The only thing is that monet-config --cflags does not list $MONET_PREFIX/include/MonetDB/C as include dir, which is the directory Mapi.h resides in. Can this be fixed, or is this done on purpose? -- Arjan Scherpenisse Centrum voor Wiskunde en Informatica, Amsterdam, the Netherlands

2 1

[Monetdb-developers] readline things
by Arjan Scherpenisse 02 Jun '04

02 Jun '04

Hi, I, as a frequent MapiClient user, have added some convenience functions to the readline interface over time: - commandline-selectable "save/load history" option (default=off). - basic 'tab' command completion (for SQL only, as of now). Completes on table names and SQL commands. Could be improved, but this already saves a lot of typing. I have added the .mx code and patch for MapiClient.mx and Makefile.ag If someone could look over this code; it would be nice to have it in cvs. -- Arjan Scherpenisse Centrum voor Wiskunde en Informatica, Amsterdam, the Netherlands Index: MapiClient.mx =================================================================== RCS file: /cvsroot/monetdb/MonetDB/src/mapi/clients/C/MapiClient.mx,v retrieving revision 1.57 diff -u -r1.57 MapiClient.mx --- MapiClient.mx 11 May 2004 11:03:04 -0000 1.57 +++ MapiClient.mx 2 Jun 2004 14:56:16 -0000 @@ -56,6 +56,7 @@ -t & --trace & trace Monet interaction \\ -T & --time & time commands \\ -u user & --user=user & user id \\ + -H & --history & load/save cmdline history (default off) \\ -? & --help & show this usage message \\ \end{tabular} @@ -83,6 +84,7 @@ #ifdef HAVE_LIBREADLINE #include <readline/readline.h> #include <readline/history.h> +#include "ReadlineTools.h" #endif #ifndef S_ISCHR @@ -805,6 +807,7 @@ fprintf(stderr, " -t | --trace /* trace Monet interaction */\n"); fprintf(stderr, " -T | --time /* time commands */\n"); fprintf(stderr, " -u user | --user=user /* user id */\n" ); + fprintf(stderr, " -H | --history /* load/save cmdline history (default off) */\n" ); fprintf(stderr, " -? | --help /* show this usage message */\n"); exit(-1); } @@ -832,8 +835,9 @@ int guest = 0; int linemode = 1; int c; - int quiet = 0; Mapi mid; + int quiet = 0; + int save_history = 0; static struct option long_options[] = { {"blocked", 1, 0, 'b'}, {"config", 1, 0, 'c'}, @@ -848,6 +852,7 @@ {"time", 0, 0, 'T'}, {"trace", 0, 0, 't'}, {"user", 1, 0, 'u'}, + {"history", 0, 0, 'H'}, {"quiet", 0, 0, 'q'}, {"help", 0, 0, '?'}, {0, 0, 0, 0} @@ -860,7 +865,7 @@ if ((setlen = mo_builtin_settings(&set)) == 0) usage(argv[0]); - while ((c = getopt_long(argc, argv, "b:c:l:u:p:P:qh:s:DtTd::?", long_options, NULL)) != -1) { + while ((c = getopt_long(argc, argv, "b:c:l:u:p:P:qHh:s:DtTd::?", long_options, NULL)) != -1) { switch (c) { case 'b': if (*optarg == 't' || *optarg == 'T' || *optarg == '1') @@ -924,6 +929,9 @@ *eq = '='; break; } + case 'H': + save_history = 1; + break; case '?': usage(argv[0]); default: @@ -998,6 +1006,10 @@ optind++; } +#ifdef HAVE_LIBREADLINE + init_readline(mid, language, save_history); +#endif + if (command) { c = doRequest(mid, command); } else { @@ -1014,6 +1026,11 @@ } c = doFileByLines(mid, stdin, prompt, linemode); } + +#ifdef HAVE_LIBREADLINE + deinit_readline(); +#endif + mapi_disconnect(mid); return c; } Index: Makefile.ag =================================================================== RCS file: /cvsroot/monetdb/MonetDB/src/mapi/clients/C/Makefile.ag,v retrieving revision 1.3 diff -u -r1.3 Makefile.ag --- Makefile.ag 25 May 2004 21:02:46 -0000 1.3 +++ Makefile.ag 2 Jun 2004 14:56:16 -0000 @@ -27,7 +27,7 @@ MAPI_LIBS = $(SOCKET_LIBS) lib_Mapi = { - SOURCES = Mapi.mx + SOURCES = Mapi.mx LIBS = $(MAPI_LIBS) ../../../common/libmutils ../../../common/libstream HEADERS = h } @@ -37,9 +37,9 @@ SOURCES = Mapi.mx } -BINS = { - SOURCES = MapiClient.mx - MapiClient_LIBS = libMapi $(MAPI_LIBS) $(READLINE_LIBS) \ +bin_MapiClient = { + SOURCES = MapiClient.mx ReadlineTools.mx + LIBS = libMapi $(MAPI_LIBS) $(READLINE_LIBS) \ ../../../common/libmutils ../../../common/libstream \ $(ICONV_LIBS) } @h #ifndef READLINETOOLS_H_INCLUDED #define READLINETOOLS_H_INCLUDED #include <monet_options.h> #include <readline/readline.h> #include <readline/history.h> #include "Mapi.h" void init_readline(Mapi mid, const char *language, int save_history); void deinit_readline(); char **sql_completion (const char *text, int start, int end); char *sql_tablename_generator (const char *text, int state); char *sql_command_generator (const char *text, int state); #endif @c /* * Readline specific stuff */ #include "ReadlineTools.h" #define PATHLENGTH 256 /* maximum file pathname length. */ const char *SQL_COMMANDS[] = { "SELECT", "INSERT", "UPDATE", "SET", "DELETE", "COMMIT", "ROLLBACK", "DROP TABLE", "CREATE", "ALTER", "RELEASE SAVEPOINT", "START TRANSACTION", 0}; Mapi _mid; char _history_file[PATHLENGTH]; int _save_history = 0; void init_readline(Mapi mid, const char *language, int save_history) { _mid = mid; /* Allow conditional parsing of the ~/.inputrc file. */ rl_readline_name = "MapiClient"; /* Tell the completer that we want to try our own completion before std completion (filename) kicks in. */ if (strcmp(language, "sql")==0) { rl_attempted_completion_function = sql_completion; } else { // FIXME: mil completion } if (save_history) { #ifndef NATIVE_WIN32 if (getenv("HOME")!=NULL) { snprintf(_history_file, PATHLENGTH, "%s/.mapiclient_history_%s", getenv("HOME"), language); _save_history = 1; } #else snprintf(_history_file, PATHLENGTH, "%s%c_mapiclient_history_%s", mo_find_option(&set, setlen, "prefix"), DIR_SEP, language); _save_history = 1; #endif read_history(_history_file); } } void deinit_readline() { if (_save_history) { write_history(_history_file); } } char ** sql_completion (const char *text, int start, int end) { char **matches; matches = (char **)NULL; if (start==0 || end == 0) { } if (start == 0) { matches = rl_completion_matches (text, sql_command_generator); } else { matches = rl_completion_matches (text, sql_tablename_generator); } return (matches); } char *sql_tablename_generator (const char *text, int state) { static int seekpos, len, rowcount; static MapiHdl table_hdl; char *name; if (!state) { seekpos = 0; len = strlen (text); if ((table_hdl = mapi_query(_mid, "SELECT name FROM tables"))==NULL || mapi_error(_mid)) { if (table_hdl) { mapi_explain_query(table_hdl, stderr); mapi_close_handle(table_hdl); } else mapi_explain(_mid, stderr); return NULL; } rowcount = mapi_get_row_count(table_hdl); } while (seekpos < rowcount) { mapi_seek_row(table_hdl, seekpos++, MAPI_SEEK_SET); mapi_fetch_row(table_hdl); name = mapi_fetch_field(table_hdl, 0); if (strncmp(name, text, len) == 0) return strdup(name); } return NULL; } // SQL commands (at start of line) char *sql_command_generator (const char *text, int state) { static int index, len; const char *name; if (!state) { index = 0; len = strlen(text); } while ((name = SQL_COMMANDS[index++])) { #ifdef HAVE_STRNCASECMP if (strncasecmp(name, text, len)==0) #else if (strncmp(name, text, len)==0) #endif return strdup(name); } return NULL; }

1 0