PlaceHolders and UserControls (i.e. Template-Lite UserControls)
Most of you will want to skip this technical addition to my otherwise prolific and interesting personal chronicles. (Ok ok I admit I’ve been way too consumed with professional goals over the last two years to blog. I’m motivated, almost desperately so, to strike while the iron’s hot and thus have had to reallocate my time. Short-terms goals are roughly: finish my programming reading list, crank out a couple learning programming projects, revolutionize digital information, then get back to the business of documenting the incoherent musings of my scrambled brains.)
Anyway, back to my technical article. I’ve been experimenting lately with increasingly complex browser look-and-feel, and encapsulating that into generic UserControls (I know. I’m still using WebForms. Ugh.) I’ve always believed my only recourse was to develop a custom control for this. I prefer to derive from WebControl, and went down the path of building full-on, by-the-books templated datasource controls (exposing ITemplates qualified with PersistenceMode.InnerProperty, and insantiating them in CreateChildControls, and rejiggering DataBind to fire <template>Created and <template>DataBound events so the consumer can FindControls in his content).
Worked fine, but for the consuming code to have to rely on Created and DataBound events to get visibility into his content was totally cumbersome and counterproductive to the original goal of creating something like a Panel that didn’t parse child controls (i.e. was qualified with ParseChildren(false)). Furthermore, I wasn’t doing anything data-related in the controls themselves, since they’re generic. In fact, my TemplateContainers were glorified objects.
So I realized the data templating is beyond what I need, and I refactored back to exposing mere setter PlaceHolder properties, with PersistenceMode.InnerProperty. The setter would merely add the value (which was also a PlaceHolder) to the Controls collection, and keep a handle around for optimized rendering in an overridden RenderContents. This was much nicer because it could be used like a plain old Panel. The drawback, however, was that that complex look and feel that the control encapsulated was still baked into a custom control, without the declarative manageability (and dynamic recompilation) of UserControls. As I pondered this the other night, I wondered if an actual UserControl could expose a setter PlaceHolder property qualified with PersistenceMode.InnerProperty. Such a setter would add the value not to the Controls collection, but to a PlaceHolder persisted in the declarative portion of the control.
I fired up VisualStudio this morning and put finger to keyboard. And the technique worked!
Example UserControl:
1 <%@ Control Language="C#" AutoEventWireup="true" ClassName="WebUserControl" %>
2 <script runat="server">
3 [PersistenceMode(PersistenceMode.InnerProperty)]
4 public PlaceHolder Bah
5 {
6 set { bah.Controls.Clear(); bah.Controls.Add(value); }
7 }
8 </script>
9 <asp:Label runat="server" ID="before">fancy complicated preceeding content</asp:Label>
10 <asp:PlaceHolder runat="server" ID="bah"></asp:PlaceHolder>
11 fancy complicated succeeding content
Example consumer:
1 <%@ Page Language="C#" AutoEventWireup="true" ClassName="Default" %>
2 <%@ Register Src="~/WebUserControl.ascx" TagName="WebUserControl" TagPrefix="test" %>
3 <script type="text/C#" runat="SErver">
4 protected void Page_Load (object sender, EventArgs e)
5 {
6 dynamic.Text = DateTime.Now.Ticks.ToString();
7 }
8 </script><!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
9
10 <html xmlns="http://www.w3.org/1999/xhtml">
11 <head runat="server">
12 <title>Persist as InnerProperty on UserControl test</title>
13 </head>
14 <body>
15 <form id="form1" runat="server">
16 <test:WebUserControl runat="Server">
17 <bah>test <asp:Label runat="server" ID="dynamic"></asp:Label></bah>
18 </test:WebUserControl>
19 </form>
20 </body>
21 </html>
Example output:
1 <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
2
3 <html xmlns="http://www.w3.org/1999/xhtml">
4 <head><title>
5 Persist as InnerProperty on UserControl test
6 </title></head>
7 <body>
8 <form name="form1" method="post" action="default.aspx" id="form1">
9 <div>
10 <input type="hidden" name="__VIEWSTATE" id="__VIEWSTATE" value="/wEPDwULLTE4NzM3NDU2NjcPZBYCAgMPZBYCAgEPZBYCAgIPZBYCZg9kFgICAQ8PFgIeBFRleHQFEjYzMzczMTUyMTk3MDA3NTk0MmRkZIkdLkHRWXYFy657lSmO9z5WqoSr" />
11 </div>
12
13 <span id="ctl02_before">before</span>
14 test <span id="ctl02_dynamic">633731521970075942</span>
15 after
16 </form>
17 </body>
18 </html>
Perfect ID namespacing, content visibility (no need to FindControls) from the consumer, *and* the manageability of UserControls. This is a solution to a problem I’ve faced since I started with WebForms, and I am absolutely stupified that I have not come across this anywhere before. (Maybe in a few days I’ll be back to describe some fatal flaw I found with it.) But so far, it seems like the holy grail of complex UserControls.
Follow me on Twitter for the latest updates, and make sure to check out my community opinion social networking project, Blocvox.
1
5 Comments
Commenting options at bottom.So why male models?
Was that really some kind of program you were talking about? It was only about 1/4 way through your article that I was like oh yeah its Mike Mcg talking so this is all tecnology computer talk from imaginary fictionalry land. . . so let me just give you some constructive feedaback on your experiment. . .
after reviewing your dataface, i inputed the functions by specific retainer order in numeric and symbolic webform browsers, and took the chance on having it broadcast my childhood block executions only to find that its templated style foundation held limits in variance with orinating databots.
Though the instances of counteracting reactions was minimal in regards to datafinding supplement agents, the accessibility and diversity of the Persistance.mode program was definitely affected especially with regards to back up numeric text codes and the broadcast of the hyper drived control trigger itself, and thus content of the databots.
Keep in mind that though a quick catalyst for familiar text finding solutions, the procreation of databots must not be altered by such counteracting agents. Databots are sensitive and reliant upon the assurance of a slow and regulated growing environment. . .
my mind is working too regular and sober right now to continue this topic. a new esteemed colleauge of mine has more to add later. . .
I almost forgot my graph results:
Example User Controls
1- (5%35) gt “dragon” descriptive population.thlmb
2- html.
3- {span, value(cow9!:bunga!} code5.
4- span******$12$
5-
6-tete.text.++++++ circuit”bond”
7-html.
8-(shake.stir) control:3-5
9-
Example What to Watch:
1- $forgale…..@
2- cspan.cspan.
3- latv.org user (foa.22, foa25, foa99, foa101) spot.check
Content of results with regard to creational goal:
:) :( :?/ :`~ … window . . .fly . . . grim
Vince what an ingenious breakthrough you’ve made. I will make sure you are credited as a coauthor when this gets published in next quarter’s issue of Smart Things Journal.
Going to Coachella??
thank you, but i prefer my name be left out and label me as “anonymous genius.”
Sadly those Coachella jerks made their event the weekend before my birthday, and i’m flying home the weekend of my birthday. Sad is this. . . You’re going all days right? I’m going to see Crystal Castles this Tuesday they’re playing over here. I just really want to see the Cure, that’s who i’m missing.
What say you?
Agh, oh well. Ya we saw The Cure last year, and they were great. Looking forward to Crystal Castles as well and many others, but not really making a detailed plan this year. Gonna soak up the rays, and just laze about. I think it’s more that no single band has me super excited, so I’m not really motivated to overthink the schedule.