Using Regular Express or MSHTMl to retrieve HTML Element

Question

Hello experts,

I am wondering if I use Regular Express to get matched HTML Element is faster than using MSHTML by traverse every element to get match. Could you please give me an advice for this?

Thanks.

Answer 1

All depends on multiple factors, the complexity of your HTML file, Complexity of regular expression, ...
I would suggest you measure the 2 solutions.

Answer 2

Could we prove in theory which approach could be faster in which situation that you mention?

Answer 3

I do not think that there are studies available on the topic, did a quick search without success. You could create mutiple HTML files with different complexity and do some testing yourself. It might be interesting for other parties if you publish your results...

Answer 4

check this link this may help you

http://stackoverflow.com/questions/9336989/c-html-parser-to-get-the-content-of-a-given-tag

Answer 5

As I did some searching on Internet, using Regular Expression for parsing HTML element is not recommended. So, is it correct?

Answer 6

I think it depends on what function do you want to achieve.

Best regards,
Jesse

---

Jesse Jiang [MSFT]
MSDN Community Support | Feedback to us

Answer 7

THTML is a Type 2 grammar in Chomsky hierarchy and regex is a type 3. You cannot express all HTML in regex, regex's complexity is simplely not enough. If you want to parse HTML, use an HTML parser that has enough complexity for your input.

---

---

The following is signature, not part of post
Please mark the post answered your question as the answer, and mark other helpful posts as helpful, so they will appear differently to other users who are visiting your thread for the same problem.
Visual C++ MVP

Answer 8

This is seems to be the correct answer. However, I see some people are currently using Regex to parse HTML document. If we could not use Regular Expression, so can XPath be a better choice since I only need to parse a specific HTML element?

Answer 9

>However, I see some people are currently using Regex to parse HTML document.

Either they don't care about total correctness or their input have limited complexity. For example if the page structure is known and have limited depth.

>so can XPath be a better choice

No parse performance has nothing to do with XPath. XPath is a query language, if you are able to use it, then the parsing is already done and you have got XML as result.

---

---

The following is signature, not part of post
Please mark the post answered your question as the answer, and mark other helpful posts as helpful, so they will appear differently to other users who are visiting your thread for the same problem.
Visual C++ MVP

Answer 10

Thank you very much for your answer.