AntiPattern Name: Parsing HTML with regular expressions
Type: Development
Problem: Extracting information from HTML text, such as elements or groups of elements with a specific property.
Context: Programmer understands how to use regular expressions and sees that HTML looks fairly regular and easy to parse using this tool.
Causes:
- Using a formated page as a data source instead of seeking a proper data source.
- Not having taken a formal class in parser construction.
- Not reading the formal HTML specification.
Supposed Solution: Write a half-baked regular expression that works for specific test cases.
Resulting Context: Regular expression may work for most cases, but fails on oddly structured but valid HTML or broken HTML, both of which are all-too-common on the Wild Wild Web.
Common Resulting problems:
- The parser breaks if the HTML code is changed at all.
- HTML code remaining in the parsed result.
- Inability to cope with tricky nestling structures.
- Failure to correctly decode entities.
RefactoredSolution:
Find a data source that was designed to be used as a data source. Alternatively, use an existing parser of the correct complexity.
See also:
ParsingHtml,
http://stackoverflow.com/a/1732454/699224 for another take on this anti-pattern
CategoryDevelopmentAntiPattern