Tutorial Extension_Provider_Part3 - getrailo/railo GitHub Wiki

Back to Main Extension Provider Tutorial

Writing an extension - Part 3

After we have created the provider and created the form, we now want to validate the form so that one can check for different circumstances in the form editing. So whether combinations of fields are allowed only or similar other checks.

For this you can use the method called validate of the install.cfc component . In order to validate the forms for mango blog I created the following method:

    <cffunction name="validate" returntype="void" output="no" hint="called to validate the entered">
        <cfargument name="error" type="struct">
        <cfargument name="path" type="string">
        <cfargument name="config" type="struct">
        <cfargument name="step" type="numeric">
        
        <cfset var mixed = config.mixed>
        <cfif arguments.step eq 1>
            <cfif not structKeyExists(config, "datasource_select") or len(trim(mixed.datasource_select)) eq 0>
                <cfset error.fields.datasource_select="Please choose an existing datasource or create a new one">
            </cfif>
            <cfif mixed.datasource_select eq "new">
                <cfif len(trim(mixed.datasource_new)) eq 0>
                    <cfset error.fields.datasource_new = "The datasource name must not be empty">
                </cfif>
                <cfif not structKeyExists(config, "dbtype_new") OR len(trim(mixed.dbtype_new)) eq 0>
                    <cfset error.fields.dbtype_new = "Please select a database type">
                </cfif>
                <cfif len(trim(mixed.server_new)) eq 0>
                    <cfset error.fields.server_new = "The server name must not be empty">
                </cfif>
                <cfif len(trim(mixed.database_new)) eq 0>
                    <cfset error.fields.database_new = "The database name must not be empty">
                </cfif>
                <cfif len(trim(mixed.username_new)) eq 0>
                    <cfset error.fields.username_new = "The username must not be empty">
                </cfif>
                <cfif len(trim(mixed.password_new)) eq 0>
                    <cfset error.fields.password_new = "The password must not be empty">
                </cfif>
            </cfif> 
        <cfelseif arguments.step eq 2>
            <cfif not structKeyExists(config, "isblognew") or len(trim(mixed.isblognew)) eq 0>
                <cfset error.fields.isblognew = "Please choose whether the blog is new or it will be imported from either BlogCFC or Wordpress">
            </cfif>
            <cfif mixed.isblognew eq "new">
                <cfif len(trim(mixed.name)) eq 0>
                    <cfset error.fields.name = "The name of the blog owner">
                </cfif>
                <cfif len(trim(mixed.username)) eq 0>
                    <cfset error.fields.username = "The login username must not be empty">
                </cfif>
                <cfif len(trim(mixed.password)) eq 0>
                    <cfset error.fields.password = "The login password must not be empty">
                </cfif>
                <cfif len(trim(mixed.email)) eq 0>
                    <cfset error.fields.email = "The blog owner email must not be empty">
                </cfif>
                <cfif len(trim(mixed.blog_title)) eq 0>
                    <cfset error.fields.blog_title = "The blog title must not be empty">
                </cfif>
                <cfif len(trim(mixed.blog_address)) eq 0>
                    <cfset error.fields.blog_address = "The blog address must not be empty">
                </cfif>
            </cfif>
        </cfif>
    </cffunction>

So simply by setting keys (named after their corresponding form fields) in the error.fields struct you can validate your form. If the error.fields struct contains no key, the form validation is completed successfully. As you can see above the method gets passed the following attributes: error, path, config and step.

By adding several steps to the config.xml you can create a kind of wizard which you can use for filling the necessary variables for the configuration of your installer. In the validate method you receive an argument that contains the current step to validate. After the last step has been validated, the install method is called which we will have a look at next. The install method simply executes whatever actions you require it to do. So it's mostly up to you to define what is going to happen there. You can unzip the archive, create mappings, create tables and datasources and even other things like add jars to the local context or create CFX tags.

Back to Main Extension Provider Tutorial

⚠️ **GitHub.com Fallback** ⚠️