Underscore might not be the cause of it. Does it matter if the "db" folder
is empty initially? If I run that multiple times, first time will hit the
error (when db & db_tmp are empty) but the second time will succeed.

Thanks,
Greg

On Mon, Jul 30, 2012 at 2:05 AM, Greg Barker <[email protected]> wrote:

> Sorry meant to include the snippet of code that causes the issue:
>
> std::auto_ptr<ibis::tablex> tmpTblx(ibis::tablex::create());
> tmpTblx->addColumn("col_1", CATEGORY);
> tmpTblx->addColumn("col_2", DOUBLE);
> int ierr = tmpTblx->readCSV("data.csv", 0, "db_tmp, ",");
> ierr = tmpTblx->write("db_tmp", "working", NULL, NULL, NULL);
> tmpTblx->clearData();
> ibis::part working_part("db");
> working_part.append("db_tmp");
>
> data.csv is simply:
> "value_1",12.34
>
> The cause appears to be the underscore in the data file. If I change
> "value_1" to "value1" I no longer hit the error.
>
> On Sun, Jul 29, 2012 at 11:13 PM, Greg Barker <[email protected]>wrote:
>
>> Hi John,
>>
>> Thank you for the quick response and the updated code. The issue with the
>> long where in clause appears to be fixed now, but I ran into a different
>> issue with the program I use to load data:
>>
>> Error -- direkte::construct0 failed with error code -1
>> Segmentation fault (core dumped)
>>
>> #0  0x0074a228 in ?? () from /lib/tls/i686/cmov/libc.so.6
>> #1  0x01de781d in ibis::category::fillIndex (this=0x98adb20,
>> dir=0x98a8e68 "/home/greg/fastbit_test/db") at category.cpp:363
>> #2  0x01dec5ef in ibis::category::append (this=0x98adb20, dt=0x98a8e68
>> "/home/greg/fastbit_test/db", df=0x98a62fc
>> "/home/greg/fastbit_test/db_tmp", nold=0, nnew=22458, nbuf=16777216,
>>     buf=0xa1ee9008 "value0") at category.cpp:1253
>> #3  0x01f659d2 in ibis::part::appendToBackup (this=0xbfa245c4,
>> dir=0x98a62fc "/home/greg/fastbit_test/db_tmp") at parti.cpp:1143
>> #4  0x01f687ad in ibis::part::append1 (this=0xbfa245c4, dir=0x98a62fc
>> "/home/greg/fastbit_test/db_tmp") at parti.cpp:633
>> #5  0x01f69dae in ibis::part::append (this=0xbfa245c4, dir=0x98a62fc
>> "/home/greg/fastbit_test/db_tmp") at parti.cpp:584
>>
>> I'm not sure if the problem lies within my code though. We have been
>> using fastbit 1.2.8 before this. Do you know what the problem might be?
>>
>> Thanks again,
>> Greg
>>
>> On Sat, Jul 28, 2012 at 12:17 AM, K. John Wu <[email protected]> wrote:
>>
>>> Hi, Greg,
>>>
>>> Thanks for your patience.  Your issue has been getting a little
>>> tangled with a couple of other changes.  There has been a large update
>>> to SVN.  Please give revision 528 a try when you get a chance and let
>>> us know how it works for you.
>>>
>>> John
>>>
>>>
>>> On 7/27/12 4:54 PM, Greg Barker wrote:
>>> > Hi John,
>>> >
>>> > Sorry to bug you about this, but do you know when the fix for this
>>> > might be ready?
>>> >
>>> > Thanks,
>>> > Greg
>>> >
>>> > On Wed, Jul 25, 2012 at 11:24 AM, K. John Wu <[email protected]
>>> > <mailto:[email protected]>> wrote:
>>> >
>>> >     Hi, Greg,
>>> >
>>> >     I just realized the problem myself.  Please bear with us while as
>>> we
>>> >     work through this.
>>> >
>>> >     John
>>> >
>>> >
>>> >     On 7/25/12 11:21 AM, Greg Barker wrote:
>>> >     > John,
>>> >     >
>>> >     > Thanks again for the quick response. Unfortunately I am running
>>> >     into a
>>> >     > different issue now with 527.
>>> >     >
>>> >     > When I try to run a query (using ibis) like my original example
>>> >     I hit
>>> >     > the following:
>>> >     >
>>> >     > Warning -- query[RWYNnK-FCtB----0]::doEvaluate -- unable to
>>> >     evaluate a
>>> >     > query term of unknown type, copy the mask as the solution
>>> >     > doQuery -- failed to evaluate "col_a in ('value1', 'value2')",
>>> error
>>> >     > code = -21
>>> >     >
>>> >     > Thanks,
>>> >     > Greg
>>> >     >
>>> >     > On Tue, Jul 24, 2012 at 9:39 PM, K. John Wu <[email protected]
>>> >     <mailto:[email protected]>
>>> >     > <mailto:[email protected] <mailto:[email protected]>>> wrote:
>>> >     >
>>> >     >     Hi, Greg,
>>> >     >
>>> >     >     Please give SVN 527 a try.  It should have fixed the
>>> >     problems you've
>>> >     >     reported.  Feel free to let us know if you encounter any
>>> other
>>> >     >     problems.
>>> >     >
>>> >     >     Thanks.
>>> >     >
>>> >     >     John
>>> >     >
>>> >     >
>>> >     >     On 7/24/12 7:27 PM, K. John Wu wrote:
>>> >     >     > Hi, Greg,
>>> >     >     >
>>> >     >     > Thanks for the note.  I am working on another change and
>>> >     this is one
>>> >     >     > of the problem I need to fix.  I will let you when I am
>>> >     done with my
>>> >     >     > testing.
>>> >     >     >
>>> >     >     > John
>>> >     >     >
>>> >     >     >
>>> >     >     >
>>> >     >     >
>>> >     >     > On 7/24/12 4:35 PM, Greg Barker wrote:
>>> >     >     >> Thank you for the quick reply. Saw you checked in some
>>> >     changes so I
>>> >     >     >> checked out revision 526 and rebuilt. Now when I try to
>>> >     compile my
>>> >     >     >> program I am hitting the following errors:
>>> >     >     >>
>>> >     >     >> ./lib/libfastbit.so: undefined reference to
>>> >     `ibis::slice::clear()'
>>> >     >     >> ./lib/libfastbit.so: undefined reference to
>>> >     >     >> `ibis::slice::getSerialSize() const'
>>> >     >     >> collect2: ld returned 1 exit status
>>> >     >     >>
>>> >     >     >> Do you know what the cause of this would be?
>>> >     >     >>
>>> >     >     >> Thank you,
>>> >     >     >> Greg
>>> >     >     >>
>>> >     >     >> On Mon, Jul 23, 2012 at 8:03 PM, K. John Wu <[email protected]
>>> >     <mailto:[email protected]>
>>> >     >     <mailto:[email protected] <mailto:[email protected]>>
>>> >     >     >> <mailto:[email protected] <mailto:[email protected]>
>>> >     <mailto:[email protected] <mailto:[email protected]>>>> wrote:
>>> >     >     >>
>>> >     >     >>     Hi, Greg,
>>> >     >     >>
>>> >     >     >>     Thanks for report.  The problem appears to be due to a
>>> >     >     silly attempt
>>> >     >     >>     to optimize that converts the expression you
>>> >     mentioned into
>>> >     >     a list of
>>> >     >     >>     equality comparisons.  This will be removed shortly.
>>> >     >     >>
>>> >     >     >>     John
>>> >     >     >>
>>> >     >     >>
>>> >     >     >>     On 7/23/12 2:54 PM, Greg Barker wrote:
>>> >     >     >>     > Hello -
>>> >     >     >>     >
>>> >     >     >>     > I have a query that has an "in" clause which
>>> contains a
>>> >     >     large number
>>> >     >     >>     > of strings:
>>> >     >     >>     >
>>> >     >     >>     > select col_a, col_b where col_a in ('value1',
>>> 'value2',
>>> >     >     'value3',
>>> >     >     >>     > 'value4', ...., 'valueN')
>>> >     >     >>     >
>>> >     >     >>     > When N is very large (thousands of items) I am
>>> >     hitting stack
>>> >     >     >>     overflow.
>>> >     >     >>     > This appears to be caused by the recursive nature of
>>> >     >     verifyExpr in
>>> >     >     >>     > whereClause.cpp:552. Was wondering if you had any
>>> >     advice
>>> >     >     on how
>>> >     >     >>     to fix
>>> >     >     >>     > this.
>>> >     >     >>     >
>>> >     >     >>     > Thanks!
>>> >     >     >>     > Greg
>>> >     >     >>     >
>>> >     >     >>     >
>>> >     >     >>     > _______________________________________________
>>> >     >     >>     > FastBit-users mailing list
>>> >     >     >>     > [email protected]
>>> >     <mailto:[email protected]>
>>> >     >     <mailto:[email protected]
>>> >     <mailto:[email protected]>>
>>> >     >     <mailto:[email protected]
>>> >     <mailto:[email protected]>
>>> >     >     <mailto:[email protected]
>>> >     <mailto:[email protected]>>>
>>> >     >     >>     >
>>> >     https://hpcrdm.lbl.gov/cgi-bin/mailman/listinfo/fastbit-users
>>> >     >     >>     >
>>> >     >     >>
>>> >     >     >>
>>> >     >
>>> >     >
>>> >
>>> >
>>>
>>
>>
>
_______________________________________________
FastBit-users mailing list
[email protected]
https://hpcrdm.lbl.gov/cgi-bin/mailman/listinfo/fastbit-users

Reply via email to