LinkLint

LinkLint is the automated link checker for the Library Web. LinkLint checks each directory/file on the Library once a month. LinkLint will send one of the following email notification messages to Web Workers and Data Owners:

1. LinkLint didn't find any problems in a directory/file.
2. LinkLint found errors, and links to the error report(s) are included in the email notification message.

If applicable, review LinkLint reports in the following order:

1. urlfailF.html

This is the broken link report for remote URLs. urlfailX.html contains the same data, but is presented in what I believe is a funky and unhelpful way. But you make the call; check either urlfailF.html or urlfailX.html first.

2. errorF.html

This is the broken link report for local URLs. errorX.html contains the same data, but is also presented in what I believe is a funky and unhelpful way. Check either errorF.html or errorX.html next.

3. errorAX.html

This is the broken link report for internal links (a link to section of the same web page). errorA.html contains the same data, but it doesn't identify the name of the file that contains the broken link - very unhelpful!

4. urlmoved.html

As of 11/18/03 urlmoved.html is a tricky report to decipher. urlmoved.html is not cross-referenced -- it doesn't tell you the name of the HTML file that contains the moved URL. If you use Dreamweaver as your web authoring tool you can globally search your HTML files to find out which ones contain moved URLs. If you use some other web authoring tool, you can use Verity to find out which of your HTML files contain the links that have moved (instructions at www.lib.berkeley.edu/digicoll/libraryweb/veritylinks.html).

Here are a couple of things to keep in mind when you edit an HTML file based upon the information in a urlmoved.html report:

• Don't change the URL of a licensed resource that grants access via an IP or password authentication server. LinkLint will report these kinds of URLs as moved, but you don't want to change any links to post-authenticated URLs, or the URLs will no longer work.
• Don't change links from http://pathfinder.berkeley.edu/ to http://sunsite2.berkeley.edu:8000/
• LinkLint wants to see an ending "/" at the end of a link to a directory or a top level website. For example, if LinkLint finds the following link to the Library Website: http://www.lib.berkeley.edu it will report it as moved to http://www.lib.berkeley.edu/. You can greatly reduce the number of "moved" URLs in a urlmoved.html report if you remember to include the ending "/" at the end of a link to a directory, or top level website.


5. urlhost.html

Ignore urlhost.html. The problems reported in urlhost.html also show up in urlfailF.html.

I'm hopeful that programming will be able to remedy some of the LinkLint report problems and redundancies mentioned above. If you have have questions/comments/suggestions about LinkLint, please contact webman!