- developers-list - monetdb.org

Re: [Monetdb-developers] [Monetdb-checkins] clients/src/mapiclient MapiClient.mx, Clients_1-20, 1.88.2.4, 1.88.2.5
by Peter Boncz 21 Dec '07

21 Dec '07

Hi Sjoerd, Thanks for fixing some things I was unaware of (empty lines in other languages?) > The timer is again started when the first line of a query is typed. > This is the case for *all* languages. I cannot image cases where it is useful for mclient to keeps stays on someone's typing speed, while I can image cases where one would like to have information about query performance. Do others agree? (I maybe unjustly assumed that timing query performance was the intended behavior) Peter > -----Original Message----- > From: monetdb-checkins-bounces(a)lists.sourceforge.net > [mailto:monetdb-checkins-bounces@lists.sourceforge.net] On > Behalf Of Sjoerd Mullender > Sent: Wednesday, December 19, 2007 3:04 PM > To: monetdb-checkins(a)lists.sourceforge.net > Subject: [Monetdb-checkins] clients/src/mapiclient > MapiClient.mx,Clients_1-20, 1.88.2.4, 1.88.2.5 > > > Update of /cvsroot/monetdb/clients/src/mapiclient > In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv6918 > > Modified Files: > Tag: Clients_1-20 > MapiClient.mx > Log Message: > Did Peter's checkin and Niels' fix correctly: In XQuery mode, when > typing <> when there is no query yet will result in just asking for > more data, and empty lines in other modes will not end the session. > Also fixed \f processing in that the new value is now remembered for a > subsequent bare \f. > The timer is again started when the first line of a query is typed. > This is the case for *all* languages. > > > Index: MapiClient.mx > =================================================================== > RCS file: /cvsroot/monetdb/clients/src/mapiclient/MapiClient.mx,v > retrieving revision 1.88.2.4 > retrieving revision 1.88.2.5 > diff -u -d -r1.88.2.4 -r1.88.2.5 > --- MapiClient.mx 19 Dec 2007 08:35:21 -0000 1.88.2.4 > +++ MapiClient.mx 19 Dec 2007 14:04:09 -0000 1.88.2.5 > @@ -152,6 +152,7 @@ > XMLformatter > }; > static enum formatters formatter = NOformatter; > +char *output = NULL; /* output format as string */ > > #define DEFWIDTH 80 > > @@ -247,8 +248,6 @@ > } > } > > -char *output = NULL; /* output format is a global variable */ > - > static void > SQLsetSpecial(const char *command) > { > @@ -796,9 +795,14 @@ > } > > static void > -setFormatter(char *s) > +setFormatter(Mapi mid, char *s) > { > - if (strcmp(s, "sql") == 0) > + if (mode == XQUERY) { > + mapi_output(mid, s); > + if (output != NULL) > + free(output); > + output = strdup(s); > + } else if (strcmp(s, "sql") == 0) > formatter = TABLEformatter; > else if (strcmp(s, "csv") == 0) > formatter = CSVformatter; > @@ -1123,13 +1127,13 @@ > fprintf(toConsole, "\\a - disable auto commit\n"); > } > if (mode == XQUERY) { > - fprintf(toConsole, "\\f - result format: > dm or xml[-noheader][-typed|-noroot|-root-FOOBAR] \n"); > - } else { > + fprintf(toConsole, "\\f - result format: > dm or xml[-noheader][-typed|-noroot|-root-FOOBAR]\n"); > + } else { > fprintf(toConsole, "\\e - echo the query > in sql formatting mode\n"); > - fprintf(toConsole, "\\f - format using a > built-in renderer {csv,tab,raw,sql,xml} \n"); > + fprintf(toConsole, "\\f - format using a > built-in renderer {csv,tab,raw,sql,xml}\n"); > fprintf(toConsole, "\\w# - set maximal page > width (-1=raw,0=no limit, >0 max char)\n"); > - fprintf(toConsole, "\\r# - set maximum rows > per page (-1=raw)\n "); > - } > + fprintf(toConsole, "\\r# - set maximum rows > per page (-1=raw)\n"); > + } > fprintf(toConsole, "\\L file - save client/server > interaction\n"); > fprintf(toConsole, "\\X - trace mclient code\n"); > fprintf(toConsole, "\\q - terminate session\n"); > @@ -1140,9 +1144,10 @@ > { > char *line = NULL; > char *buf = NULL; > - ssize_t length; > + size_t length; > MapiHdl hdl = mapi_get_active(mid); > MapiMsg rc = MOK; > + int sent = 0; /* whether we sent any data to > the server */ > > #ifdef HAVE_LIBREADLINE > if (prompt == NULL) > @@ -1185,13 +1190,17 @@ > line = fgets(buf, BUFSIZ, fp); > } > if (line == NULL || (mode == XQUERY && line[0] > == '<' && line[1] == '>')) { > - length = 0; > - if (line) > - length = -1; > - else if (hdl == NULL) > - return 0; /* EOF: nothing more to do */ > + /* end of file */ > + if (hdl == NULL) { > + if (line != NULL) > + continue; > + /* nothing more to do */ > + return 0; > + } > + > /* hdl != NULL, we should finish the > current query */ > line = NULL; > + length = 0; > } else > length = strlen(line); > if (hdl == NULL && length > 0 && line[length - > 1] == '\n') { > @@ -1217,7 +1226,7 @@ > if (mark2) > free(mark2); > mark2 = strdup(line + 2); > - if (mode == XQUERY) > + if (mode == XQUERY) > > mapi_profile(mid, mark != NULL); > continue; > case 'X': > @@ -1378,31 +1387,31 @@ > if (*line == 0) { > > fprintf(toConsole, "Current formatter: "); > if (mode == XQUERY) > - > fprintf(toConsole, "%s\n", output); > - else switch > (formatter) { > - case RAWformatter: > - > fprintf(toConsole, "raw\n"); > - break; > - case TABLEformatter: > - > fprintf(toConsole, "sql\n"); > - break; > - case CSVformatter: > - > fprintf(toConsole, "csv\n"); > - break; > - case TABformatter: > - > fprintf(toConsole, "tab\n"); > - break; > - case XMLformatter: > - > fprintf(toConsole, "xml\n"); > - break; > - default: > - > fprintf(toConsole, "none\n"); > - break; > + > fprintf(toConsole, "%s\n", output == NULL ? "dm" : output); > + else { > + switch > (formatter) { > + case > RAWformatter: > + > fprintf(toConsole, "raw\n"); > + break; > + case > TABLEformatter: > + > fprintf(toConsole, "sql\n"); > + break; > + case > CSVformatter: > + > fprintf(toConsole, "csv\n"); > + break; > + case > TABformatter: > + > fprintf(toConsole, "tab\n"); > + break; > + case > XMLformatter: > + > fprintf(toConsole, "xml\n"); > + break; > + default: > + > fprintf(toConsole, "none\n"); > + break; > + } > } > - } else if (mode == XQUERY) > - mapi_output(mid, line); > - else > - setFormatter(line); > + } else > + setFormatter(mid, line); > continue; > default: > showCommands(); > @@ -1412,34 +1421,37 @@ > } > > if (hdl == NULL) { > + timerStart(); > hdl = mapi_query_prep(mid); > CHECK_RESULT(mid, hdl, buf, continue); > } > assert(hdl != NULL); > > if (length > 0) { > + sent = 1; > SQLsetSpecial(line); > mapi_query_part(hdl, line, length); > CHECK_RESULT(mid, hdl, buf, continue); > } > > - if (mode == XQUERY && line) > - continue; /* XQuery always wants more data */ > - > /* If the server wants more but we're at the > end of file (line == NULL), notify the > server that we don't have anything more. > If the server still wants more (shouldn't > happen according to the protocol) we break > out of the loop (via the continue). The > - assertion at the end will then go off. > - */ > - timerStart(); > - if (mapi_query_done(hdl) == MMORE) { > - assert(mode != XQUERY); /* XQuery never > sends MMORE */ > - if (line != NULL) continue; /* get more data */ > - timerStart(); > - if (mapi_query_done(hdl) == MMORE) { > + assertion at the end will then go off. > + > + Note that XQuery is weird: we continue > + sending more until we reach end-of-file, > + and *then* we send the mapi_query_done. To > + exit, you need to send an end-of-file > + again. */ > + if (mode == XQUERY || mapi_query_done(hdl) == MMORE) { > + if (line != NULL) { > + continue; /* get more data */ > + } else if (mapi_query_done(hdl) == MMORE) { > + assert(mode != XQUERY); /* > XQuery never sends MMORE */ > hdl = NULL; > continue; /* done */ > } > @@ -1459,7 +1471,7 @@ > > /* for XQuery, only exit when end-of-file and we > didn't send any data */ > - } while (hdl != NULL); > + } while (line != NULL || (mode == XQUERY && sent)); > /* reached on end of file */ > assert(hdl == NULL); > return 0; > @@ -1603,7 +1615,6 @@ > strcmp(optarg, "x") == 0) { > language = "xquery"; > mode = XQUERY; > - if (!output) output = "dm"; > } else { > fprintf(stderr, "language > option needs to be one of sql, mil, mal, or xquery\n"); > exit(-1); > @@ -1618,7 +1629,7 @@ > passwd = optarg; /* can be NULL */ > break; > case 'f': > - output = optarg; /* output format */ > + output = strdup(optarg); /* output format */ > break; > case 'I': > input = optarg; > @@ -1726,13 +1737,8 @@ > > mapi_profile(mid, mark != NULL); > mapi_trace(mid, trace); > - > - if (output) { > - if (mode == XQUERY) > - mapi_output(mid, output); > - else > - setFormatter(output); > - } > + if (output) > + setFormatter(mid, output); > > c = 0; > > > > -------------------------------------------------------------- > ----------- > SF.Net email is sponsored by: > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services > for just about anything Open Source. > http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.n et/marketplace _______________________________________________ Monetdb-checkins mailing list Monetdb-checkins(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/monetdb-checkins

5 6

Re: [Monetdb-developers] [Monetdb-checkins] MonetDB/src/gdk gdk_posix.mx, , 1.150, 1.151
by Sjoerd Mullender 20 Dec '07

20 Dec '07

Peter Boncz wrote: > Update of /cvsroot/monetdb/MonetDB/src/gdk > In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv19538 > > Modified Files: > gdk_posix.mx > Log Message: > - try making the file writable if CreateFile fails I doubt that this will work since Windows reports: "The process cannot access the file because it is being used by another process." > > > Index: gdk_posix.mx > =================================================================== > RCS file: /cvsroot/monetdb/MonetDB/src/gdk/gdk_posix.mx,v > retrieving revision 1.150 > retrieving revision 1.151 > diff -u -d -r1.150 -r1.151 > --- gdk_posix.mx 19 Dec 2007 18:58:33 -0000 1.150 > +++ gdk_posix.mx 20 Dec 2007 08:45:23 -0000 1.151 > @@ -1290,8 +1290,12 @@ > > h1 = CreateFile(path, mode0, mode1, &sa, OPEN_ALWAYS, mode2, NULL); > if (h1 == INVALID_HANDLE_VALUE) { > - GDKsyserror("MT_mmap: CreateFile('%s', %lu, %lu, &sa, %lu, %lu, NULL) failed\n", path, mode0, mode1, (DWORD) OPEN_ALWAYS, mode2); > - return (void *) -1; > + (void) SetFileAttributes(path, FILE_ATTRIBUTE_NORMAL); > + h1 = CreateFile(path, mode0, mode1, &sa, OPEN_ALWAYS, mode2, NULL); > + if (h1 == INVALID_HANDLE_VALUE) { > + GDKsyserror("MT_mmap: CreateFile('%s', %lu, %lu, &sa, %lu, %lu, NULL) failed\n", path, mode0, mode1, (DWORD) OPEN_ALWAYS, mode2); > + return (void *) -1; > + } > } > > h2 = CreateFileMapping(h1, &sa, mode3, (DWORD) ((((gdk_int64) off + (gdk_int64) len) >> 32) & LL_CONSTANT(0xFFFFFFFF)), (DWORD) ((off + len) & LL_CONSTANT(0xFFFFFFFF)), NULL); > > > ------------------------------------------------------------------------- > SF.Net email is sponsored by: > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services > for just about anything Open Source. > http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace > _______________________________________________ > Monetdb-checkins mailing list > Monetdb-checkins(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/monetdb-checkins -- Sjoerd Mullender

1 0

[Monetdb-developers] Replication services
by Aaron Craelius 19 Dec '07

19 Dec '07

I see a mention of replication services being planned for the Winter of 2007 for SQL in the roadmap. Does anyone have any idea as to when they will actually be available and what the development status/plan is? Thanks, Aaron

1 0

[Monetdb-developers] Inconsistent code in pftijah.mx?
by Stefan Manegold 17 Dec '07

17 Dec '07

Hoi, I have no idea what the below code snippets from pathfinder/modules/pftijah/pftijah.mx are supposed to do, but to me, they look (at least) "inconsistent". Maybe, an expert might want to look at them ... Stefan ======== PROC tj_delete_collection(str ftiName, bit doAdmLock) : void { [...] bat("tj_" + ftiName + "_fragments")@batloop() { ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ var fpfx := str(int($h)); ^^ bat("tj_" + ftiName + "_tid"+fpfx).persists(false); bat("tj_" + ftiName + "_size"+fpfx).persists(false); } [...] } ======== PROC _tj_collection_str(str ftiName) : BAT[void,bat] { [...] bat("tj_" + ftiName + "_fragments")@batloop() ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ { tjCollBat.append("tj_" + ftiName + "_tid" + str(int($t))); ^^ tjCollBat.append("tj_" + ftiName + "_size"+ str(int($t))); ^^ } [...] ======== -- | 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

Re: [Monetdb-developers] [Monetdb-pf-checkins] pathfinder/compiler/algebra/opt opt_const.c, , 1.24, 1.25
by Stefan Manegold 14 Dec '07

14 Dec '07

Intel's icc (IMHO correctly) says: ======== icc -no-gcc -DHAVE_CONFIG_H -I. -I/scratch/monet.Int.64.64.d.29718/pathfinder/compiler/algebra/opt -I../../.. -I/scratch/monet.Int.64.64.d.29718/pathfinder/compiler/algebra/opt -I../../include -I/scratch/monet.Int.64.64.d.29718/pathfinder/compiler/algebra/opt/../../include -I/scratch/monet.Int.64.64.d.29718/MonetDB/.Int.64.64.d-Fedora6/include/MonetDB -I/scratch/monet.Int.64.64.d.29718/MonetDB/.Int.64.64.d-Fedora6/include/MonetDB/common -I/scratch/monet.Int.64.64.d.29718/MonetDB/.Int.64.64.d-Fedora6/include/MonetDB/gdk -DLIBOPT -we140 -wd1418 -Wall -w2 -c99 -mp1 -O3 -restrict -unroll -tpp7 -axWP -we266 -Werror -wd1418,1419,279,981,810,193,111,1357,1572,1599 -c /scratch/monet.Int.64.64.d.29718/pathfinder/compiler/algebra/opt/opt_const.c -KPIC -DPIC -o .libs/libopt_la-opt_const.o /scratch/monet.Int.64.64.d.29718/pathfinder/compiler/algebra/opt/opt_const.c(125): error #589: transfer of control bypasses initialization of: variable "items" (declared at line 168) variable "count" (declared at line 169) switch (p->kind) { ^ compilation aborted for /scratch/monet.Int.64.64.d.29718/pathfinder/compiler/algebra/opt/opt_const.c (code 2) make[8]: *** [libopt_la-opt_const.lo] Error 1 ======== cf., http://monetdb.cwi.nl/testing/projects/monetdb/Current/pathfinder/.Int.64.6… Stefan On Thu, Dec 13, 2007 at 10:08:57PM +0000, Tom Schreiber wrote: > Update of /cvsroot/monetdb/pathfinder/compiler/algebra/opt > In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv11953/compiler/algebra/opt > > Modified Files: > opt_const.c > Log Message: > -- Introduced a rewrite for the la_serialize_rel operator. > -- For plans that use the la_serialize_rel operator, we ensure (analogous to the existing rewrite for la_serialize_seq) that attributes with constant values are introduced at the latest possible point in the plan. > > Index: opt_const.c > =================================================================== > RCS file: /cvsroot/monetdb/pathfinder/compiler/algebra/opt/opt_const.c,v > retrieving revision 1.24 > retrieving revision 1.25 > diff -u -d -r1.24 -r1.25 > --- opt_const.c 10 Dec 2007 15:10:00 -0000 1.24 > +++ opt_const.c 13 Dec 2007 22:08:55 -0000 1.25 > @@ -139,6 +139,48 @@ > } > break; > > + > + case la_serialize_rel: > + > + /* Introduce (superfluous) attach-ops for constant attributes. > + This rewrite ensures, that constants are introduced at > + the latest possible point in the plan. > + */ > + > + /* Rewrite for the iter-attribute. */ > + if (PFprop_const_left (p->prop, p->sem.ser_rel.iter)) { > + L(p) = add_attach (L(p), p->sem.ser_rel.iter, > + PFprop_const_val_left ( > + p->prop, > + p->sem.ser_rel.iter)); > + > + } > + > + /* Rewrite for the pos-attribute. */ > + if (PFprop_const_left (p->prop, p->sem.ser_rel.pos)) { > + L(p) = add_attach (L(p), p->sem.ser_rel.pos, > + PFprop_const_val_left ( > + p->prop, > + p->sem.ser_rel.pos)); > + } > + > + /* Rewrite for the item-attributes. */ > + PFalg_attlist_t items = p->sem.ser_rel.items; > + unsigned int count = items.count; > + for (unsigned int i = 0; i < count; i++) > + { > + PFalg_att_t item = items.atts[i]; > + if (PFprop_const_left (p->prop, item)) { > + L(p) = add_attach (L(p), item, > + PFprop_const_val_left ( > + p->prop, > + item)); > + } > + } > + > + break; > + > + > case la_eqjoin: > { /** > * If both join columns are constant > > > ------------------------------------------------------------------------- > SF.Net email is sponsored by: > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services > for just about anything Open Source. > http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace > _______________________________________________ > Monetdb-pf-checkins mailing list > Monetdb-pf-checkins(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/monetdb-pf-checkins -- | 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

Re: [Monetdb-developers] [Monetdb-checkins] clients/src/mapiclient MapiClient.mx, , 1.96, 1.97
by Sjoerd Mullender 13 Dec '07

13 Dec '07

On 2007-12-13 19:35, Niels Nes wrote: > Update of /cvsroot/monetdb/clients/src/mapiclient > In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv17600/src/mapiclient > > Modified Files: > MapiClient.mx > Log Message: > don't leak memory > > > Index: MapiClient.mx > =================================================================== > RCS file: /cvsroot/monetdb/clients/src/mapiclient/MapiClient.mx,v > retrieving revision 1.96 > retrieving revision 1.97 > diff -u -d -r1.96 -r1.97 > --- MapiClient.mx 13 Dec 2007 15:44:24 -0000 1.96 > +++ MapiClient.mx 13 Dec 2007 18:35:21 -0000 1.97 > @@ -1224,6 +1224,8 @@ > fflush(stdout); > } > #endif > + if (buf != oldbuf) > + free(buf); > buf = oldbuf; > line = fgets(buf, BUFSIZ, fp); > } > @@ -1237,7 +1239,7 @@ > line = to; > iconv(cd_in, &from, &fromlen, &to, &tolen); > *to = 0; > - if (oldbuf) > + if (!oldbuf) I prefer oldbuf == NULL. oldbuf is not a Boolean. > free(buf); > buf = line; > } > > > ------------------------------------------------------------------------- > SF.Net email is sponsored by: > Check out the new SourceForge.net Marketplace. > It's the best place to buy or sell services > for just about anything Open Source. > http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace > _______________________________________________ > Monetdb-checkins mailing list > Monetdb-checkins(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/monetdb-checkins -- Sjoerd Mullender

1 0

[Monetdb-developers] mclient query output....
by Romulo Goncalves 13 Dec '07

13 Dec '07

I have a file with several queries: if I run mclient like this: mclient -lsql file.sql I get output only for the first 5 queries.. If run mclient -lsql < file.sql I get the complete output. Should this be reported as a bug or is there a logical explanation? Regards, Romulo

2 1

Re: [Monetdb-developers] [Monetdb-checkins] buildtools/autogen README, MonetDB_1-20, 1.5, 1.5.8.1
by Fabian Groffen 13 Dec '07

13 Dec '07

On 13-12-2007 15:22:43 +0000, Stefan Manegold wrote: > Update of /cvsroot/monetdb/buildtools/autogen > In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv12301/autogen > > Modified Files: > Tag: MonetDB_1-20 > README > Log Message: > > documenting a Mac-peculiarity we recently encountered Keep in mind that Apple builds their Python, Perl, etc. etc. in very strange non-standard ways which break in many ways for other applications. (And it indeed changes inbetween major releases of the OS.) I agree with the noting of this difference, but would like to note that I think we should encourage people to use non-Apple built versions of software when developing on MonetDB. The amount of patches that Apple applies (and rejected upstream) is not good for our bugfixabilities, though easily solvable by using the respective non-changed versions via any of the widely known providers of "Unix" software for Mac OS X. > Index: README > =================================================================== > RCS file: /cvsroot/monetdb/buildtools/autogen/README,v > retrieving revision 1.5 > retrieving revision 1.5.8.1 > diff -u -d -r1.5 -r1.5.8.1 > --- README 17 Oct 2006 08:59:45 -0000 1.5 > +++ README 13 Dec 2007 15:22:41 -0000 1.5.8.1 > @@ -14,6 +14,16 @@ > <python-version> is the version of Python used (e.g. 2.4). > You can automatically get the respective directory via > "<install-dir-prefix>/`python -c 'import distutils.sysconfig; print distutils.sysconfig.get_python_lib(0,0,"")'`". > +( > +In case your are using Apple-built Python 2.5.1 on Mac OS X 10.5.1 (Darwin 9.1.0), > +it can be that the above command yields > +"<install-dir-prefix>/Library/Python/2.5/site-packages", > +while Python's distutils.core.setup() actually installs the python modules into > +"<install-dir-prefix>/lib/python2.5/site-packages". > +In that case, replacing the above command by > +"<install-dir-prefix>/`python -c 'import distutils.sysconfig; print distutils.sysconfig.get_python_lib(0,1,"")'`/site-packages" > +might help to overcome this inconsistency. > +) > > On Windows, to follow these instructions, Python must of course be > found in your PATH. The default location where autogen.py is

1 0

Re: [Monetdb-developers] [Monetdb-checkins] MonetDB4/src/modules/plain arith.mx, MonetDB_4-20, 1.3, 1.3.6.1
by Stefan Manegold 13 Dec '07

13 Dec '07

On Thu, Dec 13, 2007 at 11:40:49AM +0000, Niels Nes wrote: > Update of /cvsroot/monetdb/MonetDB4/src/modules/plain > In directory sc8-pr-cvs16.sourceforge.net:/tmp/cvs-serv24024/src/modules/plain > > Modified Files: > Tag: MonetDB_4-20 > arith.mx > Log Message: > added function sign (sql allready expected this to exist) > > returns 1 for positive numbers > 0 for 0 > -1 for negative > > > Index: arith.mx > =================================================================== > RCS file: /cvsroot/monetdb/MonetDB4/src/modules/plain/arith.mx,v > retrieving revision 1.3 > retrieving revision 1.3.6.1 > diff -u -d -r1.3 -r1.3.6.1 > --- arith.mx 20 Feb 2007 11:32:12 -0000 1.3 > +++ arith.mx 13 Dec 2007 11:40:47 -0000 1.3.6.1 > @@ -251,6 +251,7 @@ > @= mel_unary_ops > .COMMAND abs(@1) : @1 = @1_unary_ABS; "absolute value" > .OPERATOR "-" (@1) : @1 = @1_unary_NEG; "negative value" > + .COMMAND sign(@1) : @1 = @1_unary_SIGN; "Returns +1, 0, -1 based on the sign of the given expression" > @= unary_ops > @:@1_unary_ops(chr)@ > @:@1_unary_ops(bte)@ > @@ -411,9 +412,11 @@ > > #define arith_abs(s) ((s)<0)?-(s):(s) > #define arith_neg(s) (-(s)) > +#define arith_sign(s) ((s)<0)?-1:(s==0)?0:1 ^^^ ^ should(n't) the second "s" be in parentheses, too? Stefan > @= c_unary_ops > @:arith_unop(_ABS,arith_abs,@1)@ > @:arith_unop(_NEG,arith_neg,@1)@ > +@:arith_unop(_SIGN,arith_sign,@1)@ > @c > @:unary_ops(c)@ -- | 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 |

2 1

[Monetdb-developers] Nightly Testing: ("Stable", "Current") -> ("Current", Stable")
by Stefan Manegold 13 Dec '07

13 Dec '07

Dear fellow MonetDB developers, for your information: To improve overlap between the actual testing (on the "clients") and the "aftermath" (results analysis/assessment & TestWeb generation) on the "server", I'll switch the order of nightly MonetDB testing (as usually starting just after midnight CET) from "first Stable, then Current" to "first Current, then Stable" as of tonight/tomorrow morning. I plan to keep the "first Current, then Stable" order (at least) until we start preparing a new release (i.e., create new release branches), or until there is any good reason/need/request to return to the "first Stable, then Current" order. Kind regards, Stefan -- | 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