On 1/21/2011 7:13 AM, John Maddock wrote:
I'm not sure if we're doing reviews of tools - we haven't in the past - but I thought it would be good to get feedback anyway...
Auto-index is a tool for semi-automating the task of adding an index to quickbook/boostbook/doxygen/docbook documentation. Auto-index acts at the docbook level to scan for terms in a concordance script-file, but can also scan headers for classes, typedefs, functions and macros to index and a lot more besides. Auto-index supports multiple indexes (for example class, function, macro and typedef indexes out of the box), and can either delegate the actual index formatting to the XSL stylesheets, or produce the index itself (faster, also better looking for HTML indexes).
Code is in the sandbox under tools/auto_index, documentation can be viewed online at: http://svn.boost.org/svn/boost/sandbox/tools/auto_index/doc/html/index .html
Currently the math, type_traits and spirit libraries are using this tool in trunk to produce indexes - take a look at those libraries to see the sort of indexes produced.
I am for this review also. I would like to use indexing in my libraries in the sandbox but have not done it because it would require the end-user to download your library in order to regenerate the docs. If auto-index were part of Boost it would be easier for potential Boost libraries in the sandbox to use it.
Your users don't need to have auto_index installed. You can have a jamfile which looks at the command line options bjam was invoked with and use auto_index only if explicitly requested (i.e. bjam --enable-index). See the spirit docs how it can be solved. Regards Hartmut --------------- http://boost-spirit.com