maria-developers team mailing list archive

[Oleksandr Byelkin <sanja@xxxxxxxxxxxxxxxx>]

Hi!

On 21.01.2016 17:16, Sergei Golubchik wrote:
> Hi, Sanja!
>
> On Dec 25, OleksandrByelkin wrote:
> > revision-id: 684bd68c33284f6106551188d9567f35b18593ed (mariadb-10.1.9-21-g684bd68)
> > parent(s): 59fcd7ff2315d007045eb987da5f21abbea6f6f1
> > committer: Oleksandr Byelkin
> > timestamp: 2015-12-25 18:39:43 +0100
> > message:
> >
> > MDEV-8615: Assertion `m_cpp_buf <= begin_ptr && begin_ptr <= m_cpp_buf + m_buf_length' failed in Lex_input_stream::body_utf8_start
> >
> > Nothing should be done before any keyword recognised.
> >
> > diff --git a/mysql-test/r/mysqltest.result b/mysql-test/r/mysqltest.result
> > index e258b1d..a9980a7 100644
> > --- a/mysql-test/r/mysqltest.result
> > +++ b/mysql-test/r/mysqltest.result
> > @@ -29,40 +29,40 @@ select 0 as "after_successful_stmt_errno" ;
> >   after_successful_stmt_errno
> >   0
> >   garbage ;
> > -ERROR 42000: You have an error in your SQL syntax; <...> near 'garbage' at line 1
> > +ERROR 42000: You have an error in your SQL syntax; <...> near '' at line 1
> Why is that?
> I mean, how does the parser parse "garbage" now?
New rules. That is what I didi not liked. I think it take it as possible 
label, but I have no idea how to rewrite rules to avoid this problem.
> > diff --git a/mysql-test/suite/funcs_1/r/storedproc.result b/mysql-test/suite/funcs_1/r/storedproc.result
> > index f0e1ee7..485082a 100644
> > --- a/mysql-test/suite/funcs_1/r/storedproc.result
> > +++ b/mysql-test/suite/funcs_1/r/storedproc.result
> > @@ -15982,8 +15982,7 @@ insert into res_t3_itisalongname_1381742_itsaverylongname_1381742 values( 'xyz'
> >   until count1 < 3
> >   END repeat label1;
> >   END//
> > -ERROR 42000: You have an error in your SQL syntax; check the manual that corresponds to your MariaDB server version for the right syntax to use near 'label1;
> > -END' at line 8
> > +ERROR 42000: End-label label1 without match
> New error message is correct, great!
But in other places it parses more then we want :(
> > diff --git a/sql/sql_yacc.yy b/sql/sql_yacc.yy
> > index a5a62ae..74d308f 100644
> > --- a/sql/sql_yacc.yy
> > +++ b/sql/sql_yacc.yy
> > @@ -248,6 +248,36 @@ static bool maybe_start_compound_statement(THD *thd)
> >     return 0;
> This looks quite good!
> But it was a bit difficult to see how the new sql_yacc.yy will look
> like. Did you push this patch? In what branch?

No, I have not pushed was waiting for review.

[skip]