My colleagues in QA have brought a perplexing problem to my attention.
I am using RH7 and produce WebHelp. Our end-users are primarily on XP with IE.
The issue is that when a space character falls inside a <span> tag it displays as run-on text in the browser (e.g. "<span>this_</span>is_some_text" would display as "thisis some text". The could equally be a space before: this<span>_is_some_text</span>). Testing has shown that the problem only occurs in IE on XP machines and doesn't occur in Firefox on XP or Windows 7, or IE on Windows 7.
My method for inserting the span is, using the WYSIWYG view in RH, I enter the whole of my text, and then go back and highlight the word(s), followed by a CTRL+B (I am quite keyboard-centric due to RSI). In the RH source it displays correctly, which means I can't catch it when proofing.
So, I use FAR to find instances of the space inside a tag, but this means correcting the Help virtually on a line-by-line basis across multiple sub-projects searching for a number of different spans in use, e.g. bold, italic, bold italic, etc... and then replacing the space so that it falls outside the span. This is hugely time consuming, and is proving largely unsuccessful - which in turn is making my productivity and quality stats stink.
So, the question is - is there a way to prevent this from happening? Or, is there a better way to identify and correct the problem?
Again - I have no control over the end-users OS or browser, and the majority of users use the combination that produces the error.
You've already identified your problem and your solution.
This is the way I've always worked, even in the old Word days.
BTW, the RH Find/Replace won't always catch the existing instances that you might need to fix, simply because it can't ignore spaces and line breaks. Go with FAR or BK ReplaceEm.