On 2/2/06 7:45 AM, "David Abrahams" <[EMAIL PROTECTED]> wrote:
> Joel de Guzman <[EMAIL PROTECTED]> writes: > >> David Abrahams wrote: >> >>>> At any rate, while it is easy to just shut off the post-processor >>>> as Eric is probably doing now, please, please don't. It is our >>>> only ticket to true regression testing. >>> I don't see the relationship between postprocessing and regression >>> testing. >> >> The way to test the validity of quickbook is to test its output >> against a stable benchmark. How do you test the equivalence of >> two xml files? For example: >> >> <tag>stuff</tag> >> >> and >> >> <tag> >> stuff >> </tag> >> >> are essentially equivalent in the POV of XML. XML can only be >> reliably tested for equivalence if you normalize it into a >> canonical form. That is the main purpose of the post processor. > > Seems to me that normalization could be part of the testing system, > rather than *forced* to be part of the qbk generation process every > time. I would argue the other way: why isn't QuickBook output normalized as it's generated? It's the "be conservative in what you send" principle. But for the "be liberal in what you receive" principle, we could use normalization within testing. There are third-party tools that can normalize XML output as part of pretty-printing. Examples are xmllint (part of the "libxml" toolkit at <http://xmlsoft.org>) and xmlformat (at <http://www.kitebird.com/software/xmlformat/>). > > In any case, got a workaround for me? I want to make progress on > those docs! -- Daryle Walker Mac, Internet, and Video Game Junkie darylew AT hotmail DOT com ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642 _______________________________________________ Boost-docs mailing list [email protected] Unsubscribe and other administrative requests: https://lists.sourceforge.net/lists/listinfo/boost-docs
