BOOK: a heuristic anomaly-detection script

That's my grand name for a few "grep"s hacked together.

When I see or hear of an anomaly in the Subversion book source XML files, like unwanted spaces before </para> (thanks to Chris Pepper for pointing this one out), I always think there are likely to be more of them, and I see if I can find a way to detect them automatically. The attached BASH shell script "svn-book-anomalies.sh" is what I have hacked together over the last year or so. It employs horribly inexact methods of detection and of filtering out the XML mark-up from the real text - but it works well enough to find a few mistakes.

In particular, this finds several more spaces before </para>, and several places where double quote characters are used instead of the correct <quote>, and several occurrences of plain "svn" where it should be "Subversion", "<command>svn</command>", etc.

Please feel free to use it and/or improve it. If you can make it filter out the mark-up properly with 'xsltproc' or whatever, that would be great. Just don't check it in to the repository in its present, brittle, ugly form.

- Julian

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org