Recent changes to bugs

date-related functions not working

Eric Safern — Thu, 21 Sep 2006 03:30:58 -0000

The following snippet:
&date(DTOPEN) >= &today()

Produces:

Bareword "_TODAY" not allowed while "strict subs" in
use at load_customers.pql.EXEC line 102.

While this:
&d(DTOPEN) > &last_day(DTOPEN)

Produces this:

Global symbol "@_LAST_DAY_LEAP" requires explicit
package name at load_customers.pql.EXEC line 103.
Global symbol "@_LAST_DAY" requires explicit package
name at load_customers.pql.EXEC line 104.

PDF Generation doesn't work on Linux

Eric Safern — Tue, 19 Sep 2006 19:50:22 -0000

$ pequel -pequeldoc pdf load_customers.pql
Processing pequel
script 'load_customers.pql'...............pod2pdf...
Malformed UTF-8 character (unexpected non-
continuation byte 0xe3, immediately af
ter start byte 0xe2)
at /usr/lib/perl5/site_perl/5.8.0/ETL/Pequel/Pod2Pdf.p
m lin
e 1176.
Malformed UTF-8 character (unexpected non-
continuation byte 0xcf, immediately af
ter start byte 0xe3)
at /usr/lib/perl5/site_perl/5.8.0/ETL/Pequel/Pod2Pdf.p
m lin
e 1176.
Malformed UTF-8 character (unexpected non-
continuation byte 0xd3, immediately af
ter start byte 0xcf)
at /usr/lib/perl5/site_perl/5.8.0/ETL/Pequel/Pod2Pdf.p
m lin
e 1176.
Malformed UTF-8 character (unexpected non-
continuation byte 0x5c, immediately af
ter start byte 0xd3)
at /usr/lib/perl5/site_perl/5.8.0/ETL/Pequel/Pod2Pdf.p
m lin
e 1176.
->load_customers.pql.pdf
---

Here's line 1175 and 1176:

prt("%PDF-1.2\n");
prt("%Ã¢Ã£ÃÃ“\n");

As you can see, some wacky characters are hard-
coded...

'create table' option 'persistant' is misspelled

Eric Safern — Fri, 15 Sep 2006 16:27:12 -0000

The pequel syntax is:

load table [ persistant ]

but that word is normally spelled 'persistent'.

Perhaps for backwards compatibility you can keep the
old spelling while adding the new - but marking the
old spelling as deprecated?

'create table' option 'persistant' is misspelled

Eric Safern — Fri, 15 Sep 2006 00:06:47 -0000

The pequel syntax is:

load table [ persistant ]

but that word is normally spelled 'persistent'.

Perhaps for backwards compatibility you can keep the
old spelling while adding the new - but marking the
old spelling as deprecated?

CRLF in pequel.vim file confuses vim on UNIX

Eric Safern — Thu, 14 Sep 2006 23:00:00 -0000

installing pequel.vim and opening a pequel file with
vim produces this:

----
E171: Missing :endif
line 192:
E15: Invalid expression: version < 600^M
line 8:
E492: Not an editor command: ^M
line 5:
Error detected while
processing /usr/share/vim/vim63/syntax/pequel.vim:
----

this is caused by the CRLF codes at the end of each
line.

tests 5 and 6 fail due to single quote char in description

Eric Safern — Thu, 14 Sep 2006 22:45:44 -0000

cpan> test ETL::Pequel
Running test for module ETL::Pequel
Running make for G/GA/GAFFIE/ETL-Pequel-2.4-6b.tar.gz
Is already unwrapped into
directory /root/.cpan/build/ETL-Pequel-2.4-6b
Has already been processed within this session
Running make test
Prepending /root/.cpan/build/File-HomeDir-
0.58/blib/arch /root/.cpan/build/File-HomeDir-
0.58/blib/lib /root/.cpa
n/build/Data-Dumper-
2.121/blib/arch /root/.cpan/build/Data-Dumper-
2.121/blib/lib to PERL5LIB.
PERL_DL_NONLAZY=1 /usr/bin/perl "-
MExtUtils::Command::MM" "-e" "test_harness
(0, 'lib', 'blib/arch')" t/*.t
t/01_aggregates_1..........ok
t/02_array_fields..........ok
t/03_conditional_aggr......ok
t/04_filter_regex..........ok
t/05_group_by_derived......WARNING:examples/group_by_d
erived.pql:22:53: missing terminating ' character
t/05_group_by_derived......FAILED test 1
Failed 1/1 tests, 0.00% okay
t/06_group_by_derived_2....WARNING:examples/group_by_d
erived_2.pql:22:53: missing terminating ' character
t/06_group_by_derived_2....FAILED test 1
Failed 1/1 tests, 0.00% okay

The immediate problem is solved by removing or
doubling the odd quote character - but I imagine
something deeper is going on...

test 25 (macro_pack) causes segfault

Eric Safern — Thu, 14 Sep 2006 22:39:25 -0000

perl ./examples/make_sample_data.pl 20 |
perl ./scripts/pequel ./examples/macro_pack.pql
Processing pequel
script './examples/macro_pack.pql'..............
[./examples/macro_pack.pql Thu Sep 14 17:38:19 2006]
Init
location|product_code|salesman_list|p_salesman_list|up
_salesman_list
[./examples/macro_pack.pql Thu Sep 14 17:38:19 2006]
Start
Segmentation fault (core dumped)

warning messages when using perl -w

Eric Safern — Thu, 14 Sep 2006 22:33:29 -0000

Running pequel with perl -w produces this:

Duplicate specification "inline_MAKE:s" for
option "inline_make"
Duplicate specification "inline_CC:s" for
option "inline_cc"
Duplicate specification "inline_LIBS:s" for
option "inline_libs"
Duplicate specification "inline_CCFLAGS:s" for
option "inline_ccflags"
Duplicate specification "inline_INC:s" for
option "inline_inc"
Duplicate specification "inline_lddlflags:s" for
option "inline_lddlflags"
Duplicate specification "inline_optimize:s" for
option "inline_optimize"

typing just 'pequel' produces a usage message with an error

Eric Safern — Thu, 14 Sep 2006 22:28:17 -0000

Typing simply 'pequel' produces a long usage
message. Buried in there is:
...
--cpp_cmd < args >
Use of uninitialized value in print
at /usr/lib/perl5/site_perl/5.8.0/ETL/Pequel
/Script.pm line 90.
Use of uninitialized value in print
at /usr/lib/perl5/site_perl/5.8.0/ETL/Pequel
/Script.pm line 90.

--input_record_limit < args >, --irl < args >
Input record process limit count -- process only
initial <limit> records.
--show_synonyms
...

The line in question is below, marked with >>>>:

sub usage : method
{
my $self = shift;

print "@{[ $self->PARAM->VERSION ]}
\n";
print "Usage: @{[ $self->PARAM-
>PEQUEL_EXEC_NAME ]} < script-name.pql > < options...
>\n";
foreach my $o (grep($_->cmdType,
$self->PARAM->options->toArray))
{
my $args = (defined($o-
>cmdFormat) && $o->cmdFormat eq ':s') ? ' < args
>' : '' ;
print join(', ', "--@{[ $o-
>name ]}${args}", map("--@{[ $_->name ]}${args}", $o-
>cmdAlia
s->toArray)), "\n";
>>>> print ' ', $o-
>description, "\n";
}
}

'pequel' script contains CRLF at end of each line

Eric Safern — Thu, 14 Sep 2006 22:25:12 -0000

The script 'pequel' contains \r\n (also known as
CRLF - DOS/Windows-style) at the end of each line.

This breaks the script on UNIX, such that
$perl /usr/bin/pequel

works fine, but
$/usr/bin/pequel
returns the message
: bad interpreter: No such file or directory

The fix is simple - remove that pesky carriage return
from the first line. On UNIX, this will work:

$ tr -d '\r' < pequel > /tmp/p
$ chmod +x /tmp/p
$ mv /tmp/p pequel

more info:

http://faq.perl.org/perlfaq7.html#What_does_bad_interp