Accessing raw text of TextFlow and performing a search

this works fine for me:

var s:String=TextFilter.export(_flow,TextFilter.PLAIN_TEXT_FORMAT,ConversionType.STRING_TYPE) as String;

I think what you are looking for is the "SpellCheckAdornment" example in the Flash examples download. See:

http://labs.adobe.com/downloads/textlayout.html

At the bottom of the page is a link for Flash examples.

You could do an export to plain text, and search that to find the offset in the TextFlow. This should work if the text you are using does not have discretionary hyphens (since they would be filtered out by the plain text filter). However, I would expect you would getter better performance by searching paragraph by paragraph. Iterate through all the paragraphs in the TextFlow, and get the text of each paragraph from the textBlock: paragraph.textBlock.content.rawText.

I think so - here's how I'd reccomend approaching the problem.

Given a TextFlow I'd do a getFirstLeaf. That returns a FlowLeafElement. I'd then look at the FlowLeafElement.text property to access the text for matching. To advance to the next leaf element use FlowLeafElement.getNextLeaf.

You'll have to do some book keeping when you find a match to figure out the absolute position of the beginning of the match. Next step would be to use textFlow.flowComposer.findControllerIndexAtPosition to figure out which container has the matched text.

Adding niceties you could search a paragraph at a time by taking advantage of the limitElement parameter to getNextLeaf. After calling getFirstLeaf call getParagraph to find the paragrah and use that as the limit element. To advance to the next paragraph use null as the limit element and call getParagraph again for the new limit element.

Hope that helps.
Richard