- Use State monad for [Result] passing around?
Needs to have IO too. mtl?
- Parallel HTTP fetch for separate domains
- Add PDF support
- Better --only-failures output
- Cache length option should accept units: s, m, h, d - default to s
- More detailed HttpException errors
- Should be able to detect links without http:// or https:// prefixes
- Should detect if internet connection is down
- Should show line number of found links
- Can't parse links containing single quote marks
- URLs containing unicode characters won't parse
- Fetching message sometimes goes to more than one line, so next line doesn't replace it (saveCursor/restoreCursor?)
When a line is longer than the available width
- Is an invalid URL an error?
Should an invalid URL count as an error? The parser shouldn't really pick up invalid URLs. But if it looks like one and fails then it is probably worth high-lighting.
- Edge case: shouldn't delay if only a single URL being checked
- Basic link parsing
- Basic link checking
- Should output file that problem occured in
- Should return correct exit code
- Should store recently tested links in .brokdb file
- Option for cache length
- Should space out requests
- Option for interval between requests
- Should be able to ignore specific URLs
List of URL prefixes
- --help command
- Use attoparsec
Already used by HTTP-conduit, so may as well use it
- Fixed InvalidURLException crash
- Fixed issue with HEAD request returning a 404
- Sees
https://andhttp://as valid URLs - Sees
https://*as a valid URL - Checks the same URL multiple times if in different files