PHP XPath parsing HTML graphics mixed content

Now only Readability can crawl part of the major station content, but there are many Web site content can not be captured accurately, so you want to use XPath rules to crawl the content of specific sites.

However, if the article is mixed with text, the text sequence after the completion of the crawl is disrupted.
Is there a way to record the order, or to use XPath to fully capture the text within a particular area.

Reply content:

Now only Readability can crawl part of the major station content, but there are many Web site content can not be captured accurately, so you want to use XPath rules to crawl the content of specific sites.

However, if the article is mixed with text, the text sequence after the completion of the crawl is disrupted.
Is there a way to record the order, or to use XPath to fully capture the text within a particular area.

Well, I think of a solution.
Pre-use XPath to extract the image URL, and then replace the original image part of node to

image url

So as long as the output to do a URL to determine the line.

