To check an URL like http://www.myhomepage.org/ it is enough to execute linkchecker http://www.myhomepage.org/. This will check the complete domain of www.myhomepage.org recursively. All links pointing outside of the domain are also checked for validity.
For more options, read the man page linkchecker(1) or execute linkchecker -h.
All URLs have to pass a preliminary syntax test. Minor quoting mistakes will issue a warning, all other invalid syntax issues are errors. After the syntax check passes, the URL is queued for connection checking. All connection check types are described below.
HTTP links (http:, https:)
After connecting to the given HTTP server the given path or query is requested. All redirections are followed, and if user/password is given it will be used as authorization when necessary. Permanently moved pages issue a warning. All final HTTP status codes other than 2xx are errors.
Local files (file:)
A regular, readable file that can be opened is valid. A readable directory is also valid. All other files, for example device files, unreadable or non-existing files are errors.
File contents are checked for recursion.
Mail links (mailto:)
A mailto: link eventually resolves to a list of email addresses. If one address fails, the whole list will fail. For each mail address we check the following things:
FTP links (ftp:)
For FTP links we do:
Gopher links (gopher:)
We try to send the given selector (or query) to the gopher server.
Telnet links (telnet:)
We try to connect and if user/password are given, login to the given telnet server.
NNTP links (news:, snews:, nntp)
We try to connect to the given NNTP server. If a news group or article is specified, try to request it from the server.
Ignored links (javascript:, etc.)
An ignored link will only print a warning. No further checking will be made.
Here is a complete list of recognized, but ignored links. The most prominent of them should be JavaScript links.
Recursion occurs on HTML files, Opera bookmark files and directories. Note that the directory recursion reads all files in that directory, not just a subset like index.htm*.
Q: LinkChecker produced an error, but my web page is ok with Netscape/IE/Opera/... Is this a bug in LinkChecker?
A: Please check your web pages first. Are they really ok? Use a syntax highlighting editor. Use HTML Tidy. Check if you are using a proxy which produces the error.
Q: I still get an error, but the page is definitely ok.
A: Some servers deny access of automated tools (also called robots) like LinkChecker. This is not a bug in LinkChecker but rather a policy by the webmaster running the website you are checking. It might even be possible for a website to send robots different web pages than normal browsers.
Q: How can I tell LinkChecker which proxy to use?
A: LinkChecker works transparently with proxies. In a Unix or Windows environment, set the http_proxy, https_proxy, ftp_proxy or gopher_proxy environment variables to a URL that identifies the proxy server before starting LinkChecker. For example
$ http_proxy="http://www.someproxy.com:3128" $ export http_proxy
In a Macintosh environment, LinkChecker will retrieve proxy information from Internet Config.
Q: The link "mailto:john@company.com?subject=Hello John" is reported as an error.
A: You have to quote special characters (e.g. spaces) in the subject field. The correct link should be "mailto:...?subject=Hello%20John" Unfortunately browsers like IE and Netscape do not enforce this.
Q: Has LinkChecker JavaScript support?
A: No, it never will. If your page is not working without JS then your web design is broken. Use PHP or Zope or ASP for dynamic content, and use JavaScript just as an addon for your web pages.
Q: I don't get this --extern/--intern stuff.
A: When it comes to checking there are three types of URLs. Note that local files are also represented als URLs (ie file://). So local files can be external URLs.
LinkChecker provides four options which affect URLs to fall in one of those three categories: --intern, --extern, --extern-strict-all and --denyallow. By default all URLs are internal. With --extern you specify what URLs are external. With --intern you specify what URLs are internal. Now imagine you have both --extern and --intern. What happens when an URL matches both patterns? Or when it matches none? In this situation the --denyallow option specifies the order in which we match the URL. By default it is internal/external, with --denyallow the order is external/internal. Either way, the first match counts, and if none matches, the last checked category is the category for the URL. Finally, with --extern-strict-all all external URLs are strict.
Oh, and just to boggle your mind: you can have more than one external regular expression in a config file and for each of those expressions you can specify if those matched external URLs should be strict or not.
An example. We don't want to check mailto urls. Then its -i'!^mailto:'. The '!' negates an expression. With --extern-strictall, we don't even connect to any mail hosts.
Another example. We check our site www.mycompany.com, don't recurse into external links point outside from our site and want to ignore links to hollowood.com and hullabulla.com completely. This can only be done with a configuration entry like
[filtering] extern1=hollowood.com 1 extern2=hullabulla.com 1 # the 1 means strict external ie don't even connect
and the command linkchecker --intern=www.mycompany.com www.mycompany.com
Q: Is LinkCheckers cookie feature insecure?
A: Cookies can not store more information as is in the HTTP request itself, so you are not giving away any more system information. After storing however, the cookies are sent out to the server on request. Not to every server, but only to the one who the cookie originated from! This could be used to "track" subsequent requests to this server, and this is what some people annoys (including me). Cookies are only stored in memory. After LinkChecker finishes, they are lost. So the tracking is restricted to the checking time. The cookie feature is disabled as default.
Q: I want to have my own logging class. How can I use it in LinkChecker?
A: Currently, only a Python API lets you define new logging classes. Define your own logging class as a subclass of StandardLogger or any other logging class in the log module. Then call the addLogger function in Config.Configuration to register your new Logger. After this append a new Logging instance to the fileoutput.
import linkcheck, MyLogger
log_format = 'mylog'
log_args = {'fileoutput': log_format, 'filename': 'foo.txt'}
cfg = linkcheck.configuration.Configuration()
cfg.logger_add(log_format, MyLogger.MyLogger)
cfg['fileoutput'].append(cfg.logger_new(log_format, log_args))
Q: LinkChecker does not ignore anchor references on caching.
Q: Some links with anchors are getting checked twice.
A: This is not a bug. It is common practice to believe that if an URL ABC#anchor1 works then ABC#anchor2 works too. That is not specified anywhere and I have seen server-side scripts that fail on some anchors and not on others. This is the reason for always checking URLs with different anchors. If you really want to disable this, use the --no-anchor-caching option.
Q: I see LinkChecker gets a /robots.txt file for every site it checks. What is that about?
A: LinkChecker follows the robots.txt exclusion standard. To avoid misuse of LinkChecker, you cannot turn this feature off. See the Web Robot pages and the Spidering report for more info.
Q: Ctrl-C does not stop LinkChecker immediately. Why is that so?
A: The Python interpreter has to wait for all threads to finish, and this means waiting for all open sockets to close. The default timeout for sockets is 30 seconds, hence the delay. You can change the default socket timeout with the --timeout option.
Q: How do I print unreachable/dead documents of my website with LinkChecker?
A: No can do. This would require file system access to your web repository and access to your web server configuration.
You can instead store the linkchecker results in a database and look for missing files.
Q: How do I check HTML/XML syntax with LinkChecker?
A: No can do. Use the HTML Tidy program.