Question

I’ve got an internationalised web application in which all the text that eventually gets sent to the browser is generated by code-behind, either by usercontrols that override Render, or by code that sets the Text of Label etc controls.

For example, rather than an aspx containing

<h2>Subtitle</h2>
<p>Some paragraph</p>

this application would have instead

<h2><asp:Label runat="server" ID="SubTitle" /></h2>
<p><mycontrols:Literal runat="server" ID="Paragraph" /></p>

with the actual text only getting supplied at runtime.

I would like to ensure that I haven’t inadvertently left any static text in the markup anywhere. However, the ways I can think of all have flaws:

• I can’t use a tool that tries to parse the markup as xml, because *.as?x markup definitely isn’t valid xml
• I can’t even use an html parser (such as HtmlAgilityPack), because *.as?x markup (remembering this is before it’s rendered by ASP.NET) also isn’t necessarily valid html (consider how repeater templates can interleave tags)
• Just using Visual Studio Find won’t help me find just static text and not, say, CSS class names

Is there any way other than manually checking through every page?

Answer 1

Ok, so if it’s a a site supporting internationalisation – point one of the many free web site downloaders apps at it – and pull the output onto disk. Run it in French (or language of your choice) then German. Any harcoded text strings will not change. You’ll then have to find a WinMerge tool (I think WinMerge can do folders) or write something that can file compare the output. Then PowerGrep to locate the results back in your source.