oracle.pm

视频监控网络部分的协议ddns,的模块的实现代码,请大家大胆指正.

=item ORA_OCI

Oracle doesn't provide a formal API for determining the exact version
number of the OCI client library used, so DBD::Oracle has to go digging
(and sometimes has to more or less guess).  The ORA_OCI constant
holds the result of that process.

In string context ORA_OCI returns the full "A.B.C.D" version string.

In numeric context ORA_OCI returns the major.minor version number
(8.1, 9.2, 10.0 etc).  But note that version numbers are not actually
floating point and so if Oracle ever makes a release that has a two
digit minor version, such as C<9.10> it will have a lower numeric
value than the preceding C<9.9> release. So use with care.

The contents and format of ORA_OCI are subject to change (it may,
for example, become a I<version object> in later releases).
I recommend that you avoid checking for exact values.


=head2 Connect Attributes

=over 4

=item ora_session_mode

The ora_session_mode attribute can be used to connect with SYSDBA
authorization and SYSOPER authorization.
The ORA_SYSDBA and ORA_SYSOPER constants can be imported using

  use DBD::Oracle qw(:ora_session_modes);

This is one case where setting ORACLE_SID may be useful since
connecting as SYSDBA or SYSOPER via SQL*Net is frequently disabled
for security reasons.

Example:

  $dsn = "dbi:Oracle:";       # no dbname here
  $ENV{ORACLE_SID} = "orcl";  # set ORACLE_SID as needed
  delete $ENV{TWO_TASK};      # make sure TWO_TASK isn't set

  $dbh = DBI->connect($dsn, "", "", { ora_session_mode => ORA_SYSDBA });

It has been reported that this only works if $dsn does not contain a SID
so that Oracle then uses the value of the ORACLE_SID (not TWO_TASK)
environment variable to connect to a local instance. Also the username
and password should be empty, and the user executing the script needs
to be part of the dba group or osdba group.

=item ora_oratab_orahome

Passing a true value for the ora_oratab_orahome attribute will make
DBD::Oracle change $ENV{ORACLE_HOME} to make the Oracle home directory
specified in the C</etc/oratab> file I<if> the database to connect to
is specified as a SID that exists in the oratab file, and DBD::Oracle was
built to use the Oracle 7 OCI API (not Oracle 8+).

=item ora_module_name

After connecting to the database the value of this attribute is passed
to the SET_MODULE() function in the C<DBMS_APPLICATION_INFO> PL/SQL
package. This can be used to identify the application to the DBA for
monitoring and performance tuning purposes. For example:

  DBI->connect($dsn, $user, $passwd, { ora_module_name => $0 });

=item ora_dbh_share

Needs at least Perl 5.8.0 compiled with ithreads. Allows to share database
connections between threads. The first connect will make the connection, 
all following calls to connect with the same ora_dbh_share attribute
will use the same database connection. The value must be a reference
to a already shared scalar which is initialized to an empty string.

  our $orashr : shared = '' ;

  $dbh = DBI->connect ($dsn, $user, $passwd, {ora_dbh_share => \$orashr}) ;

=item ora_use_proc_connection

This attribute allows to create a DBI handle for an existing SQLLIB
database connection. This can be used to share database connections
between Oracle ProC code and DBI running in an embedded Perl interpreter.
The SQLLIB connection id is appended after the "dbi:Oracle:" initial
argument to DBI::connect.

For example, if in ProC a connection is made like

    EXEC SQL CONNECT 'user/pass@db' AT 'CONID';

the connection may be used from DBI after running something like

    my $dbh = DBI->connect("dbi:Oracle:CONID", "", "",
                           { ora_use_proc_connection => 1 });

To disconnect, first call $dbh->disconnect(), then disconnect in ProC.

This attribute requires DBD::Oracle to be built with the -ProC
option to Makefile.PL.  It is not available with OCI_V7. Not tested
with Perl ithreads or with the ora_dbh_share connect attribute.

=item ora_envhp

The first time a connection is made a new OCI 'environment' is
created by DBD::Oracle and stored in the driver handle.
Subsequent connects reuse (share) that same OCI environment
by default.

The ora_envhp attribute can be used to disable the reuse of the OCI
environment from a previous connect. If the value is C<0> then
a new OCI environment is allocated and used for this connection.

The OCI environment is what holds information about the client side
context, such as the local NLS environment. So by altering %ENV and
setting ora_envhp to 0 you can create connections with different
NLS settings. This is most useful for testing.

=item ora_charset, ora_ncharset

For oracle versions >= 9.2 you can specify the client charset and
ncharset with the ora_charset and ora_ncharset attributes.  You
still need to pass C<ora_envhp = 0> for all but the first connect.

These attributes override the settings from environment variables.

  $dbh = DBI->connect ($dsn, $user, $passwd,
                       {ora_charset => 'AL32UTF8'});

=back

=head2 Database Handle Attributes

=over 4

=item C<ora_ph_type>

The default placeholder data type for the database session.
The C<TYPE> or L</ora_type> attributes to L<DBI/bind_param> and
L<DBI/bind_param_inout> override the data type for individual placeholders.
The most frequent reason for using this attribute is to permit trailing spaces
in values passed by placeholders.

Constants for the values allowed for this attribute can be imported using

  use DBD::Oracle qw(:ora_types);

Only the following values are permitted for this attribute.

=over 4

=item ORA_VARCHAR2

Oracle clients using OCI 8 will strip trailing spaces and allow embedded \0 bytes.
Oracle clients using OCI 9.2 do not strip trailing spaces and allow embedded \0 bytes.
This is the normal default placeholder type.

=item ORA_STRING

Don't strip trailing spaces and end the string at the first \0.

=item ORA_CHAR

Don't strip trailing spaces and allow embedded \0.
Force 'blank-padded comparison semantics'.

For example:

  use DBD::Oracle qw(:ora_types);
  
  $sql="select username from all_users where username = ?";
  #username is a char(8)
   
  $sth=$dbh->prepare($sql)";
   
  $sth->bind_param(1,'bloggs',{ ora_type => ORA_CHAR});
   
Will pad bloggs out to 8 chracters and return the username.  

=back

=item ora_parse_error_offset

If the previous error was from a failed C<prepare> due to a syntax error,
this attribute gives the offset into the C<Statement> attribute where the
error was found.

=back

=over 4

=item ora_array_chunk_size

Because of OCI limitations, DBD::Oracle needs to buffer up rows of
bind values in its C<execute_for_fetch> implementation. This attribute
sets the number of rows to buffer at a time (default value is 1000).

The C<execute_for_fetch> function will collect (at most) this many
rows in an array, send them of to the DB for execution, then go back
to collect the next chunk of rows and so on. This attribute can be
used to limit or extend the number of rows processed at a time.

Note that this attribute also applies to C<execute_array>, since that
method is implemented using C<execute_for_fetch>.

=back

=head2 Prepare Attributes

These attributes may be used in the C<\%attr> parameter of the
L<DBI/prepare> database handle method.

=over 4

=item ora_placeholders

Set to false to disable processing of placeholders. Used mainly for loading a
PL/SQL package that has been I<wrapped> with Oracle's C<wrap> utility.

=item ora_parse_lang

Tells the connected database how to interpret the SQL statement.
If 1 (default), the native SQL version for the database is used.
Other recognized values are 0 (old V6, treated as V7 in OCI8),
2 (old V7), 7 (V7), and 8 (V8).
All other values have the same effect as 1.

=item ora_auto_lob

If true (the default), fetching retrieves the contents of the CLOB or
BLOB column in most circumstances.  If false, fetching retrieves the
Oracle "LOB Locator" of the CLOB or BLOB value.

See L</Handling LOBs> for more details.
See also the LOB tests in 05dbi.t of Oracle::OCI for examples
of how to use LOB Locators.

=item ora_check_sql

If 1 (default), force SELECT statements to be described in prepare().
If 0, allow SELECT statements to defer describe until execute().

See L</Prepare postponed till execute> for more information.

=back

=head2 Placeholder Binding Attributes

These attributes may be used in the C<\%attr> parameter of the
L<DBI/bind_param> or L<DBI/bind_param_inout> statement handle methods.

=over 4

=item ora_type

Specify the placeholder's data type using an Oracle data type.
A fatal error is raised if C<ora_type> and the DBI C<TYPE> attribute
are used for the same placeholder.
Some of these types are not supported by the current version of
DBD::Oracle and will cause a fatal error if used.
Constants for the Oracle datatypes may be imported using

  use DBD::Oracle qw(:ora_types);

Potentially useful values when DBD::Oracle was built using OCI 7 and later:

  ORA_VARCHAR2, ORA_STRING, ORA_LONG, ORA_RAW, ORA_LONGRAW,
  ORA_CHAR, ORA_MLSLABEL, ORA_RSET

Additional values when DBD::Oracle was built using OCI 8 and later:

  ORA_CLOB, ORA_BLOB, ORA_NTY, ORA_VARCHAR2_TABLE, ORA_NUMBER_TABLE

See L</Binding Cursors> for the correct way to use ORA_RSET.

See L</Handling LOBs> for how to use ORA_CLOB and ORA_BLOB.

See L</SYS.DBMS_SQL datatypes> for ORA_VARCHAR2_TABLE, ORA_NUMBER_TABLE.

See L</Other Data Types> for more information.

See also L<DBI/Placeholders and Bind Values>.

=item ora_csform

Specify the OCI_ATTR_CHARSET_FORM for the bind value. Valid values
are SQLCS_IMPLICIT (1) and SQLCS_NCHAR (2). Both those constants can
be imported from the DBD::Oracle module. Rarely needed.

=item ora_csid

Specify the I<integer> OCI_ATTR_CHARSET_ID for the bind value. 
Character set names can't be used currently.

=item ora_maxdata_size

Specify the integer OCI_ATTR_MAXDATA_SIZE for the bind value. 
May be needed if a character set conversion from client to server
causes the data to use more space and so fail with a truncation error.

=item ora_maxarray_numentries

Specify the maximum number of array entries to allocate. Used with
ORA_VARCHAR2_TABLE, ORA_NUMBER_TABLE. Define the maximum number of
array entries Oracle can pass back to you in OUT variable of type
TABLE OF ... .

=item ora_internal_type

Specify internal data representation. Currently is supported only for
ORA_NUMBER_TABLE.

=back

=head2 Trailing Spaces

Please note that only the Oracle OCI 8 strips trailing spaces from VARCHAR placeholder
values and uses Nonpadded Comparison Semantics with the result. 
This causes trouble if the spaces are needed for
comparison with a CHAR value or to prevent the value from
becoming '' which Oracle treats as NULL.
Look for Blank-padded Comparison Semantics and Nonpadded
Comparison Semantics in Oracle's SQL Reference or Server
SQL Reference for more details.

To preserve trailing spaces in placeholder values for Oracle clients that use OCI 8, 
either change the default placeholder type with L</ora_ph_type> or the placeholder
type for a particular call to L<DBI/bind> or L<DBI/bind_param_inout>
with L</ora_type> or C<TYPE>.
Using L<ORA_CHAR> with L<ora_type> or C<SQL_CHAR> with C<TYPE>
allows the placeholder to be used with Padded Comparison Semantics
if the value it is being compared to is a CHAR, NCHAR, or literal.

Please remember that using spaces as a value or at the end of
a value makes visually distinguishing values with different
numbers of spaces difficult and should be avoided.

Oracle Clients that use OCI 9.2 do not strip trailing spaces.

=head2 Padded Char Fields

Oracle Clients after OCI 9.2 will automatically pad CHAR placeholder values to the size of the CHAR.
As the default placeholder type value in DBD::Oracle is ORA_VARCHAR2 to access this behavior you will 
have to change the default placeholder type with L</ora_ph_type> or placeholder 
type for a particular call with L<DBI/bind> or L<DBI/bind_param_inout>
with L</ORA_CHAR> or C<ORA_CHARZ>.

=head1 Metadata

=head2 C<get_info()>

DBD::Oracle supports C<get_info()>, but (currently) only a few info types.

=head2 C<table_info()>

DBD::Oracle supports attributes for C<table_info()>.

In Oracle, the concept of I<user> and I<schema> is (currently) the
same. Because database objects are owned by an user, the owner names
in the data dictionary views correspond to schema names.
Oracle does not support catalogs so TABLE_CAT is ignored as
selection criterion.

Search patterns are supported for TABLE_SCHEM and TABLE_NAME.

TABLE_TYPE may contain a comma-separated list of table types.
The following table types are supported:

  TABLE
  VIEW
  SYNONYM
  SEQUENCE

The result set is ordered by TABLE_TYPE, TABLE_SCHEM, TABLE_NAME.

The special enumerations of catalogs, schemas and table types are
supported. However, TABLE_CAT is always NULL.

An identifier is passed I<as is>, i.e. as the user provides or
Oracle returns it.
C<table_info()> performs a case-sensitive search. So, a selection
criterion should respect upper and lower case.
Normally, an identifier is case-insensitive. Oracle stores and
returns it in upper case. Sometimes, database objects are created
with quoted identifiers (for reserved words, mixed case, special
characters, ...). Such an identifier is case-sensitive (if not all
upper case). Oracle stores and returns it as given.
C<table_info()> has no special quote handling, neither adds nor
removes quotes.

=head2 C<primary_key_info()>

Oracle does not support catalogs so TABLE_CAT is ignored as
selection criterion.