=head1 NAME CGI::QuickForm - Perl module to provide quick CGI forms. =head1 SYNOPSIS # Minimal example. (Insecure no error checking.) #!/usr/bin/perl -w use strict ; use CGI qw( :standard :html3 ) ; use CGI::QuickForm ; show_form( -ACCEPT => \&on_valid_form, # You must supply this subroutine. -TITLE => 'Test Form', -FIELDS => [ { -LABEL => 'Name', }, # Default field type is textfield. { -LABEL => 'Age', }, # Stored in param( 'Age' ). ], ) ; sub on_valid_form { my $name = param( 'Name' ) ; my $age = param( 'Age' ) ; open PEOPLE, ">>people.tab" ; print PEOPLE "$name\t$age\n" ; close PEOPLE ; print header, start_html( 'Test Form Acceptance' ), h3( 'Test Form Data Accepted' ), p( "Thank you $name for your data." ), end_html ; } # All QuickForm options (aide memoir) #!/usr/bin/perl -w use strict ; use CGI qw( :standard :html3 ) ; use CGI::QuickForm ; show_form( -ACCEPT => \&on_valid_form, -BORDER => 0, -FOOTER => undef, -HEADER => undef, -INTRO => undef, -LANGUAGE => 'en', -USER_REQUIRED => undef, -USER_INVALID => undef, -TITLE => 'Test Form', -REQUIRED_HTML => '+', -INVALID_HTML => '*', -VALIDATE => undef, # Set this to validate the entire record -SIZE => undef, -MAXLENGTH => undef, -ROWS => undef, -COLUMNS => undef, -CHECK => 1, -SPACE => 0, # Output some newlines to assist debugging if 1 -MULTI_COLUMN => 0, -NAME => undef, -ONSUBMIT => undef, -JSCRIPT => '', -STYLE_FIELDNAME => '', -STYLE_FIELDVALUE => '', -STYLE_BUTTONS => '', -STYLE_ROW => '', -STYLE_WHY => '', -TABLE_OPTIONS => '', -FIELDS => [ { -LABEL => 'Personal Details', -HEADLINE => 1, -STYLE_FIELDNAME => '', -COLSPAN => 2, -END_ROW => 1, }, { -LABEL => 'Name', -START_ROW => 1, -END_ROW => 1, -COLSPAN => 1, -REQUIRED => undef, -TYPE => 'textfield', -VALIDATE => undef, # Set this to validate the field -CLEAN => undef, # Set this to clean up valid data -DESC => undef, -STYLE_FIELDNAME => '', # If set over-rides form-level setting -STYLE_FIELDVALUE => '', # If set over-rides form-level setting -STYLE_ROW => '', # If set over-rides form-level setting # Lowercase options are those supplied by CGI.pm -name => undef, # Defaults to -LABEL's value. -default => undef, -size => 30, -maxlength => undef, }, # For all others: same QuickForm options as above # and all CGI.pm options (which vary with -TYPE) available { -LABEL => 'Address', -TYPE => 'textarea', -rows => 3, -columns => 40, }, { -LABEL => 'Password', -TYPE => 'password_field', }, { -LABEL => 'Hair colour', -TYPE => 'scrolling_list', '-values' => [ qw( Red Black Brown Grey White ) ], -size => 1, -multiples => undef, }, { -LABEL => 'Worst Sport', -TYPE => 'radio_group', -values => [ qw( Boxing Cricket Golf ) ], -default => 'Golf', }, # Any other CGI.pm field can be used in the same way. ], -BUTTONS => [ { -name => 'Add' }, { -name => 'Edit' }, { -name => 'List' }, { -name => 'Remove' }, { -name => 'Clear', -DEFAULTS => 1 }, ], ) ; =head1 DESCRIPTION C, provides a quick and simple mechanism for providing on-line CGI forms. When C executes it presents the form with the fields requested. As you can see from the minimal example at the beginning of the synopsis it will default everything it possibly can to get you up and running as quickly as possible. If you have specified any validation it will validate when the user presses the submit button. If there is an error it will re-present the form with the erroneous fields marked and with all the data entered in tact. This is repeated as often as needed. Once the user has corrected all errors and the data is valid then your C<&on_valid_form> subroutine will be called so that you can process the valid data in any way you wish. Note that EXAMPLE #1 and EXAMPLE #2 are in this pod; C, C, etc. are supplied as files. =head2 QuickForm form-level (record-level) options =over =item C<-ACCEPT> Required subroutine reference. This is a reference to the subroutine to execute when the form is successfully completed, i.e. once all the fields and the whole record are valid (either because no validation was requested or because every validation subroutine called returned true). The parameters are accessible via C, so your C<&on_valid_form> may look something like this: sub on_valid_form { my $first_param = param( 'first' ) ; my $second_param = param( 'second' ) ; my $third_param = param( 'third' ) ; # Process, e.g. send an email or write a record to a file or database. # Give the user a thank you. } =item C<-BORDER> Optional integer. This is the border width. Default is zero. You would normally set this to 1 if you are using C<-DESC> to add textual descriptions to your fields. =item C<-BUTTONS> Optional array reference. This is an array of submit buttons. The buttons appear at the bottom of the form, after all the fields. Each button is defined as an anonymous hash, e.g. -BUTTONS => [ { -name => 'New' }, { -name => 'Update' }, ], although any other legitimate C options may also be given, e.g. -BUTTONS => [ { -name => 'New', -value => 'BUTTON_NEW' }, { -name => 'Update', -value => 'BUTTON_UPDATE' }, ], If you want a button which resets the form to its default values then create an entry like this: { -name => 'Clear', -DEFAULTS => 1 }, If no C<-BUTTONS> option array reference is given it will be created with C<{ -name =E 'Submit' }> by default. Note that this option replaces the C<-BUTTONLABEL> option. If C<-BUTTONLABEL> is used it will be converted into the new form automatically so old scripts will I be broken. However use of C<-BUTTONS> is recommended for all new work. To see which button has been pressed you might use code like this in your on_valid_form subroutine: if( param( 'New' ) ) { # New pressed } elsif( param( 'Update' ) ) { # Update pressed } # etc. =item C<-CHECK> Optional boolean, default is true. When C is called it will check (i.e. do validation) providing there are parameters (i.e. the user has filled in the form) I if C<-CHECK> is true. This option would not normally be used. However if you have links which call your form with some parameters (e.g. default values), you will want the form to be displayed with the defaults but I any validation taking place in the first instance. In this situation you would set C<-CHECK> to false. Thus we must cope with the following scenarios: =over =item 1. Form is called with no params - must display blank form and validate when the user presses a button; =item 2. Form is called with params (e.g. by clicking a link we've provided) - must display form with any defaults and I validate until the user presses a button; =item 3. Form is called with params (as the result of the user pressing a button) - validation must take place. =back To achieve the above we need to add an extra field=value pair to the URL we provide and if that is present then skip validation. The field's name must I be one of the form's fields! e.g. # If it is to be called from one of our own URLs with something like # www.mysite.com/cgi-bin/myscript?colour=green&size=large # then we must add in the extra field=value and write the preceeding link # for example as: # www.mysite.com/cgi-bin/myscript?QFCHK=0&colour=green&size=large # We then use query_string() to set -CHECK to 0 and show the form with the # defaults without validating - we'll validate when they press a button. # If its been called as something like www.mysite.com/cgi-bin/myscript # then set -CHECK to 1 which gives us standard behaviour: # i.e. if there are params then show_form will validate; otherwise it will # show the blank form. show_form( -CHECK => ( query_string() =~ /QFCHK=0/o ? 0 : 1 ), # etc ) ; # Or more verbosely: my $Check = 1 ; $Check = 0 if query_string() =~ /QFCHK=0/o ; show_form( -CHECK => $Check, # etc ) ; Note that QuickForm discards any query string if it reinvokes itself because of invalid data. This is useful because it means you can use the query string to distinguish between a 'first time' call and subsequent calls as we do here with -CHECK. However if you want a query string parameter to survive these calls we must extract them and pass them ourselves, e.g. via a hidden field. =item C<-FOOTER> Optional string. This is used to present any text following the form and if used it must include everything up to and including final "", e.g.: my $footer = p( "Thank's for your efforts." ) . h6( "Copyright (c) 1999 Summer plc" ) . end_html ; show_form( -FOOTER => $footer, # etc =item C<-HEADER> Optional string. This is used to present your own title and text before the form proper. If you use this it must include everything from "Content-type: text/html" onwards. For example: my $header = header . start_html( "This is my Title" ) . h2( "My new Form" ) . p( "Please answer the questions!" ) ; show_form( -HEADER => $header, # etc =item C<-INTRO> Optional string. If you specify C<-TITLE> you may want to specify this field too; it puts a paragraph of text before the form. The English default is "Please enter the information.", there is a default for each supported language (see C<-LANGUAGE>). =item C<-INVALID_HTML> Optional HTML string. Default is: * You can over-ride this to set your own marker to indicate an invalid field. You could use an image tag for example: * Note that if you use your own C<-USER_REQUIRED> or C<-USER_INVALID> strings, this string will replace the sequence C<~I~> if it occurs. See C and the companion option C<-REQUIRED_HTML>. =item C<-JSCRIPT> Optional string. If using Javascript you can include the code in this string and it will be put into the html that QuickForm outputs. (See C<-NAME> and C<-ONSUBMIT>.) =item C<-LANGUAGE> Optional string. This option accepts 'en' (english), 'cy' (welsh), 'de' (german), 'es' (spanish) and 'fr' (french) - the French translation was done by Babelfish - see CHANGES for the human translators. ('english' is also supported for backward compatibility.) If people provide me with translations I will add other languages. This is used for the presentation of messages to the user, e.g.: Please enter the information. Fields marked with + are required. Fields marked with * contain errors or are empty. If you want to create your own 'required' or 'invalid' strings using the C<-USER_REQUIRED> and/or C<-USER_INVALID> options you I set C<-LANGUAGE> to 'user'. See C. =item C<-MULTI_COLUMN> Optional boolean (default false). If false QuickForm behaves as it always has producing a two column table, the first column with field names and the second column with field values. If true QuickForm will put all field names and field values in the same row, except where you force a new row to be used by marking some fields with C<-END_ROW => 1,>. See the field-level options C<-START_ROW>, C<-END_ROW> and C<-COLSPAN>. See C and C which have been updated to demonstrate these options. =item C<-NAME> Optional string. If specified this string is given to the start_form() function as its C<-name> option; used for identifying the form for Javascript. (See C<-JSCRIPT> and C<-ONSUBMIT>.) =item C<-ONSUBMIT> Optional string. If specified this string is given to the start_form() function as its C<-onSubmit> option; used with Javascript. (See C<-JSCRIPT> and C<-NAME>.) =item C<-REQUIRED_HTML> Optional HTML string. Default is: + You can over-ride this to set your own marker to indicate a required field. Note that if you use your own C<-USER_REQUIRED> or C<-USER_INVALID> strings, this string will replace the sequence C<~R~> if it occurs. See C and the companion option C<-INVALID_HTML>. =item C<-TITLE> Required string (unless you use C<-HEADER>). This is used as the form's title and as a header on the form's page - unless you use the C<-HEADER> option (see above) in which case this option is ignored. =item C<-USER_INVALID> Optional string. If specified you I set C<-LANGUAGE> to 'user' and if you are not writing English then you'll need to set C<-USER_REQUIRED> too. This string is used as the error text for invalid fields. If it contains the character sequence C<~I~> that sequence will be replaced with the HTML used to signify an invalid field (which you can override by setting C<-INVALID_HTML>). =item C<-USER_REQUIRED> Optional string. If specified you I set C<-LANGUAGE> to 'user' and if you are not writing English then you'll need to set C<-USER_INVALID> too. This string is used as the error text for required fields. If it contains the character sequence C<~R~> that sequence will be replaced with the HTML used to signify a required field (which you can override by setting C<-REQUIRED_HTML>). =item C<-VALIDATE> Optional subroutine reference. This routine is called after each individual field has been validated. It is given the fields in a name=>value hash. It should either return a simple true (valid) or false (invalid) or a two element list, the first element being a true/false value and the second value either an empty string or an (html) string which gives the reason why the record is invalid. Typically it may have this structure: sub valid_record { my %field = @_ ; my $valid = 1 ; # Do some multi-field validation, e.g. if( $field{'colour'} eq 'blue' and $field{'make'} eq 'estate' ) { $valid = 0 ; # No blue estates available. } # etc. $valid ; # Return the valid variable which may now be false. } or now (preferred style): sub valid_record { my %field = @_ ; my $valid = 1 ; my $why = '' ; # Do some multi-field validation, e.g. if( $field{'colour'} eq 'blue' and $field{'make'} eq 'estate' ) { $valid = 0 ; # No blue estates available. $why = 'No blue estates available' ; } # etc. ( $valid, $why ) ; } I If the record is invalid the C<$why> element will be shown near the top of the form just before the fields themselves, otherwise (i.e. if the record is valid) it will be ignored. =item C<-COLUMNS> Optional integer. If set then any C<-TYPE =E textarea> will have a C<-columns> set to this value unless an explicit C<-columns> is given. =item C<-MAXLENGTH> Optional integer. If set then any C<-TYPE =E textfield> will have a C<-maxlength> set to this value unless an explicit C<-maxlength> is given. =item C<-ROWS> Optional integer. If set then any C<-TYPE =E textarea> will have a C<-rows> set to this value unless an explicit C<-rows> is given. =item C<-SIZE> Optional integer. If set then any C<-TYPE =E textfield> will have a C<-size> set to this value unless an explicit C<-size> is given. For example: show_form( -ACCEPT => \&on_valid_form, # You must supply this subroutine. -TITLE => 'Test Form', -SIZE => 50, -FIELDS => [ { -LABEL => 'Name', }, { -LABEL => 'Age', }, ], ) ; # Both the fields will be textfields because that is the default and both # will have a -size of 50. show_form( -ACCEPT => \&on_valid_form, # You must supply this subroutine. -TITLE => 'Test Form', -SIZE => 50, -MAXLENGTH => 70, -FIELDS => [ { -LABEL => 'Name', -CLEAN => \&cleanup, # You must supply this (see later). }, { -LABEL => 'Age', }, { -LABEL => 'Country', # Here we upper case the country. -CLEAN => sub { local $_ = shift ; tr/a-z/A-Z/ ; $_ }, -size => 20, }, ], ) ; # All three fields will be textfields. Name and Age will have a -size of # 50 but Country will have a -size of 20. All three will have a -maxlength # of 70. =item C<-STYLE_*> These options apply globally and are documented under Styles later. =item C<-SPACE> Optional integer. If true then QuickForm will output some newlines to help make the HTML more human readable for debugging; otherwise no additional whitespace is added. Defaults to false. =item C<-FIELDS> Required array reference. This is an array of hashes; there must be at least one. The fields are displayed in the order given. The options available in each field hash are covered in the next section. =back =head2 QuickForm field-level options =over =item C<-CLEAN> Optional subroutine reference. If specified this subroutines will be called for the relevant field if and only if the whole record is valid, i.e. just before calling your C subroutine. It will receive a single parameter (the value of the relevant param), and must return a new value. A typical routine might clean up excess whitespace, e.g.: sub cleanup { local $_ = shift ; # This is the value of param( ) tr/\t \n\r\f/ /s ; # Convert multiple whitespace to one space. s/^\s*//o ; # Remove leading whitespace. s/\s*$//o ; # Remove trailing whitespace. $_ ; } =item C<-COLSPAN> Optional integer. Default is 1; ignored if C<-MULTI_COLUMN> is false. If you choose C<-MULTI_COLUMN> and want some fields to span multiple rows then you can use this option to define how many rows are spanned. (Note that every field is two rows wide, one for the fieldname and one for the fieldvalue.) =item C<-DESC> Optional string. This is a short piece of descriptive text which appears above the field and is used to give the user a little guidance on what they should choose or enter. Normally if you use these then you would set the form-level C<-BORDER> option to 1 to help visually group the field and its descriptive text. =item C<-HEADLINE> Optional boolean. Default is false. If set to true then instead of inserting a field, QuickForm will insert a label. This is used to separate blocks of input fields - see C. If this is true then you will probably want to set C<-STYLE_FIELDNAME>, e.g. to make the text stand out; and also C<-COLSPAN> to the number of columns in the form if using C<-MULTI_COLUMN>. The C<-LABEL> is the text that will be displayed. If using C<-MULTI_COLUMN> any field that preceeds C<-HEADLINE> should normally set C<-END_ROW> to true; this isn't done automatically in case your form has two or more columns and you want to have different headlines above each column. Thus a typical headline field looks like: { -LABEL => 'General Information', -HEADLINE => 1, -COLSPAN => 2, # Probably needs to be more for -MULTI_COLUMN -END_ROW => 1, -STYLE_FIELDNAME => 'style="background-color:black;color:white;font-weight:bold"', }, =item C<-LABEL> Required string. This is the display label for the field. It is also used as the field's name if no C<-name> option is used. =item C<-REQUIRED> Optional boolean. Default is false. If set to true the field must contain something. Should only be used with text fields. It is ignored if C<-VALIDATE> is given since C<-VALIDATE> overrides (see later). =item C<-START_ROW> and C<-END_ROW> Optional booleans, default is true. These options are only relevant if the form-level C<-MULTI_COLUMN> option is set to true in which case these options are used to identify at which fields rows start and end. In practice C<-START_ROW> should never be needed; simply set C<-END_ROW => 1,> in each field which is to be the last field in a given row. Note that because some fields may need to span several columns e.g. a layout where say the first two fields are side-by-side and the following field is so wide it must take the whole width; the wider field's C<-COLSPAN> setting may need to be set to C<-COLSPAN => 3,> or similar. See C and C which have been updated to demonstrate these options. =item C<-STYLE_*> Some of these options may be applied on a per-field basis; they are documented under Styles later. =item C<-TYPE> Optional string. Default is C. May be any field supported by C. =item C<-VALIDATE> Optional subroutine reference. If specified this subroutine will be called when the user presses the submit button; its argument will be the value of the field. It should either return a simple true (valid) or false (invalid) or a two element list, the first element being a true/false value and the second value either an empty string or an (html) string which gives the reason why the field is invalid. Its typical structure may be: sub valid_national_insurance { my $ni = shift ; $ni = uc $ni ; ( $ni =~ /^[A-Z]{2}\d{7}[A-Z]$/o ) ? 1 : 0 ; } or now (preferred style): sub valid_national_insurance { my $ni = shift ; my $why = 'Should be 2 letters followed by 7 ' . 'digits then a letter' ; $ni = uc $ni ; my $valid = ( $ni =~ /^[A-Z]{2}\d{7}[A-Z]$/o ) ? 1 : 0 ; ( $valid, $why ) ; } I If the field is invalid the C<$why> element will be shown immediately to the right of the field it refers to, otherwise (i.e. if the field is valid) it will be ignored. =back =head2 CGI.pm field-level options All the other options passed in the hash should be the lowercase options supported by C for the particular field type. For example for a C<-TYPE> of C the options currently supported are C<-name>, C<-default>, C<-size> and C<-maxlength>; you may use any, all or none of them since C always provides sensible defaults. See "All QuickForm options" in the SYNOPSIS above for examples of the most common field types. =head2 Styles If you wish to use a cascading style sheet with QuickForm then you need to set the -HEADER option to include a tag which includes a reference to your stylesheet. If you wish to use multiple columns see C<-MULTI_COLUMN>, C<-START_ROW>, C<-END_ROW> and C<-COLSPAN> as well as C and C. Whether you use a stylesheet for classes or in-line styles you can set the class or style using the -STYLE_* options, e.g. -STYLE_FIELDNAME => qq{style="font-size:12pt;margin:2em;"}, -STYLE_FIELDVALUE => qq{class="valueclass"}, -STYLE_ROW => qq{class="rowclass"}, -STYLE_DESC => qq{style="color:darkblue"}, The above styles apply globally to all rows, but can be over-ridden, see later. -STYLE_WHY => qq{style="font-style:italic;color:red"}, Because a popular browser cannot cope with this: -STYLE_BUTTONS => qq{style="font-family:Helvetica;text-align:center;"}, which produces: (linux-help) and C (bicycle) for more examples. You can of course also apply your own global styles to the existing tags in the normal way. When C<-STYLE_DESC>, C<-STYLE_FIELDNAME>, C<-STYLE_FIELDVALUE> and C<-STYLE_ROW> are set in C at form-level (i.e. not inside the C<-FIELDS> section) they apply globally to every fieldname cell, fieldvalue cell and row respectively. If you require finer control you can set these styles on a per field basis by including them as field-level options, e.g. show_form( -ACCEPT => \&on_valid_form, -STYLE_FIELDNAME => 'style="background-color:#AAAAAA"', -STYLE_FIELDVALUE => 'style="background-color:#DDDDDD"', -FIELDS => [ { -LABEL => 'Forename', -STYLE_FIELDNAME => 'style="background-color:LIGHTGREEN"', -STYLE_FIELDVALUE => 'style="background-color:YELLOW"', }, # etc. If you have set a style at form-level but do not wish it to apply to a particular row you can over-ride either by setting a new style for the row as in the example above or by coding C<-STYLE_ROW => ' '> for example; we have to use a space because if we used the empty string or undef the global style would be applied. See example 5. =head2 EXAMPLE #1: Using a form to generate email This program is provided as an example of QuickForm's capabilities, it is not a production-quality program: it has no error checking and is I secure. #!/usr/bin/perl -w use strict ; use CGI qw( :standard :html3 ) ; use CGI::QuickForm ; show_form( -TITLE => 'Test Form', -ACCEPT => \&on_valid_form, -FIELDS => [ { -LABEL => 'Forename', -REQUIRED => 1, }, { -LABEL => 'Surname', -REQUIRED => 1, }, { -LABEL => 'Age', }, { -LABEL => 'Sex', -TYPE => 'radio_group', '-values' => [ qw( Female Male ) ], }, ], ) ; # This subroutine will only be called if the name fields contain at # least one character. sub on_valid_form { my $forename = param( 'Forename' ) ; my $surname = param( 'Surname' ) ; my $age = param( 'Age' ) ; open MAIL, "|/usr/lib/sendmail -t" ; print MAIL "From: test\@localhost\n" . "To: user\@localhost\n" . "Subject: Quick Form Email Test\n\n" . "Name: $forename $surname\n" . "Age: $age\n" ; print header, start_html( 'Test Form Data Accepted' ), h3( 'Test Form Data Accepted' ), p( "Thank you $forename for your data." ), end_html ; } =head2 EXAMPLE #2: Appending data to a file This program is provided as an example of QuickForm's capabilities, it is not a production-quality program: it has no error checking and is I secure. #!/usr/bin/perl -w use strict ; use CGI qw( :standard :html3 ) ; use CGI::QuickForm ; show_form( -TITLE => 'Test Form', -ACCEPT => \&on_valid_form, -VALIDATE => \&valid_form, -SIZE => 40, -MAXLENGTH => 60, -FIELDS => [ { -LABEL => 'Forename', -VALIDATE => \&valid_name, -CLEAN => \&cleanup, # (See earlier for definition.) }, { -LABEL => 'Surname', -VALIDATE => \&valid_name, -CLEAN => \&cleanup, # (See earlier for definition.) }, { -LABEL => 'Age', # &mk_valid_number generates a subroutine (a closure) and # returns a reference to that subroutine. -VALIDATE => &mk_valid_number( 3, 130 ), -size => 10, -maxlength => 3, }, ], ) ; # This will only be called if all the validation routines return true. sub on_valid_form { my $forename = param( 'Forename' ) ; my $surname = param( 'Surname' ) ; my $age = param( 'Age' ) ; open FILE, ">>namedata.tab" ; print FILE "$surname\t$forename\t$age\n" ; close FILE ; print header, start_html( 'Test Form Data Accepted' ), h3( 'Test Form Data Accepted' ), p( "Thank you $forename for your data." ), end_html ; } # This is called to validate the entire form (record). # Use a routine like this if there are relationships between fields that # must be tested. sub valid_form { my %rec = @_ ; my $valid = 1 ; # We don't allow (perfectly valid!) names like 'John John'. my $why = 'Not allowed to have identical forename and surname' ; $valid = 0 if lc $rec{'Surname'} eq lc $rec{'Forename'} ; ( $valid, $why ) ; # $why is ignored if valid. } sub valid_name { my $name = shift ; my $valid = 1 ; $valid = 0 if $name !~ /^\w{2,}$/o ; ( $valid, 'Name must have at least 2 letters' ) ; } sub mk_valid_number { my( $min, $max ) = @_ ; sub { my $valid = $_[0] ? ( $min <= $_[0] and $_[0] <= $max ) : 1 ; ( $valid, "Should be between $min and $max inclusive" ) ; } ; } =head2 mod_perl QuickForm appears to run fine in CGI scripts that run under Apache::Registry without requiring any changes. If you want to use QuickForm under pure mod_perl, i.e. outside Apache::Registry they you need to do the following: =over =item 1. Add the following lines at the beginning of your script: require 'CGI/Apache.pm' ; use Apache::Constants qw( :common ) ; =item 2. Ensure that any routines that return to mod_perl return OK. Normally this will be C and C. =item 3. Convert your script into a module by wrapping the C call in a C subroutine etc. =item 4. Copy the module into an Apache subdirectory in your C<@INC> path. =item 5. Edit your Apache httpd.conf (or perl.conf) to add a Location for the module. =item 6. If you are converting a script that uses C you may need to add the name of the script, e.g. C. =back See C for a working example that covers all the points above. C also has notes at the beginning to explain how to set things up. C is a simple conversion of C so you can see the simple changes required. Of course you don't have to change your scripts at all if you run them under Apache::Registry. =head2 USING QUICKFORM WITH OBJECT MODULES If you want to pass QuickForm an C function that is in fact an object's method call then instead of: -ACCEPT => \&on_valid_form, you need to write: -ACCEPT => sub { $object->on_valid_form }, assuming that you have an object reference called C<$object> and you want it to call its C method. =head2 INTRODUCTORY ARTICLE See http://www.perlpress.com/perl/quickform.html =head2 YOUR QUICKFORM URLs If you use QuickForm please let me know the URL and a short one line description so I can add it to the list at http://www.perlpress.com/perl/quickform-users.html Thanks! =head1 BUGS If you get messages like this under pure mod_perl: [error] Undefined subroutine &Apache::xxxxxxx::handler called. the problem is with your configuration I QuickForm, and I won't be able to help you. Please see the copious and high quality documentation at L for help with this problem. If mod_perl prints things like "Content: text/html" at the top of your forms it's a mod_perl issue; check your PerlSendHeader setting (I have it Off). Some browsers get awfully confused about colouring rows and cells in tables, so if you have problems with this please check the HTML that QuickForm produces (via your browsers View Source command) -- in my test cases QuickForm always outputs correct HTML so the problem if any is with the browser -- however if you find that the HTML is invalid please email me your script plus the HTML shown by View Source (in case I can't run your script). =head1 AUTHOR Mark Summerfield. I can be contacted as - please include the word 'quickform' in the subject line. Emails giving the URLs of the pages where you use QuickForm would be appreciated! See CHANGES for acknowledgements. =head1 COPYRIGHT Copyright (c) Mark Summerfield 1999-2000. All Rights Reserved. This module may be used/distributed/modified under the LGPL. =cut