SQL::Translator::Utils

(source, CPAN)

version 1.66

NAME
SYNOPSIS
DESCSIPTION
EXPORTED FUNCTIONS AND CONSTANTS
AUTHORS

#NAME

SQL::Translator::Utils - SQL::Translator Utility functions

#SYNOPSIS

use SQL::Translator::Utils qw(debug);
debug("PKG: Bad things happened");

#DESCSIPTION

SQL::Translator::Utils contains utility functions designed to be used from the other modules within the SQL::Translator modules.

Nothing is exported by default.

#EXPORTED FUNCTIONS AND CONSTANTS

#debug

debug takes 0 or more messages, which will be sent to STDERR using warn. Occurances of the strings PKG, SUB, and LINE will be replaced by the calling package, subroutine, and line number, respectively, as reported by caller(1).

For example, from within foo in SQL/Translator.pm, at line 666:

debug("PKG: Error reading file at SUB/LINE");

Will warn

[SQL::Translator: Error reading file at foo/666]

The entire message is enclosed within [ and ] for visual clarity when STDERR is intermixed with STDOUT.

#normalize_name

normalize_name takes a string and ensures that it is suitable for use as an identifier. This means: ensure that it starts with a letter or underscore, and that the rest of the string consists of only letters, numbers, and underscores. A string that begins with something other than [a-zA-Z] will be prefixer with an underscore, and all other characters in the string will be replaced with underscores. Finally, a trailing underscore will be removed, because that's ugly.

normalize_name("Hello, world");

Produces:

Hello_world

A more useful example, from the SQL::Translator::Parser::Excel test suite:

normalize_name("silly field (with random characters)");

returns:

silly_field_with_random_characters

#header_comment

Create the header comment. Takes 1 mandatory argument (the producer classname), an optional comment character (defaults to $DEFAULT_COMMENT), and 0 or more additional comments, which will be appended to the header, prefixed with the comment character. If additional comments are provided, then a comment string must be provided ($DEFAULT_COMMENT is exported for this use). For example, this:

package My::Producer;

use SQL::Translator::Utils qw(header_comment $DEFAULT_COMMENT);

print header_comment(__PACKAGE__,
                     $DEFAULT_COMMENT,
                     "Hi mom!");

produces:

--
-- Created by My::Prodcuer
-- Created on Fri Apr 25 06:56:02 2003
--
-- Hi mom!
--

Note the gratuitous spacing.

#parse_list_arg

Takes a string, list or arrayref (all of which could contain comma-separated values) and returns an array reference of the values. All of the following will return equivalent values:

parse_list_arg('id');
parse_list_arg('id', 'name');
parse_list_arg( 'id, name' );
parse_list_arg( [ 'id', 'name' ] );
parse_list_arg( qw[ id name ] );

#truncate_id_uniquely

Takes a string ($desired_name) and int ($max_symbol_length). Truncates $desired_name to $max_symbol_length by including part of the hash of the full name at the end of the truncated name, giving a high probability that the symbol will be unique. For example,

truncate_id_uniquely( 'a' x 100, 64 )
truncate_id_uniquely( 'a' x 99 . 'b', 64 );
truncate_id_uniquely( 'a' x 99,  64 )

Will give three different results; specifically:

aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa_7f900025
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa_6191e39a
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa_8cd96af2

#$DEFAULT_COMMENT

This is the default comment string, '--' by default. Useful for header_comment.

#parse_mysql_version

Used by both Parser::MySQL and Producer::MySQL in order to provide a consistent format for both parser_args->{mysql_parser_version} and producer_args->{mysql_version} respectively. Takes any of the following version specifications:

5.0.3
4.1
3.23.2
5
5.001005  (perl style)
30201     (mysql style)

#parse_dbms_version

Takes a version string (X.Y.Z) or perl style (XX.YYYZZZ) and a target ('perl' or 'native') transforms the string to the given target style. to

#throw

Throws the provided string as an object that will stringify back to the original string. This stops it from being mangled by Moo's isa code.

#ex2err

Wraps an attribute accessor to catch any exception raised using "throw" and store them in $self->error(), finally returning undef. A reference to this function can be passed directly to "around" in Moo.

around foo => \&ex2err;

around bar => sub {
    my ($orig, $self) = (shift, shift);
    return ex2err($orig, $self, @_) if @_;
    ...
};

#carp_ro

Takes a field name and returns a reference to a function can be used around a read-only accessor to make it carp instead of die when passed an argument.

#batch_alter_table_statements

Takes diff and argument hashes as passed to batch_alter_table and an optional list of producer functions to call on the calling package. Returns the list of statements returned by the producer functions.

If no producer functions are specified, the following functions in the calling package are called:

#1. rename_table
#2. alter_drop_constraint
#3. alter_drop_index
#4. drop_field
#5. add_field
#5. alter_field
#6. rename_field
#7. alter_create_index
#8. alter_create_constraint
#9. alter_table

If the corresponding array in the hash has any elements, but the caller doesn't implement that function, an exception is thrown.

#AUTHORS

Darren Chamberlain <darren@cpan.org>, Ken Y. Clark <kclark@cpan.org>.

CONTENTS