Can't we have *one* thread for that? I do not see the point that we copy n' paste our replies between threads. David you give here nearly the same reply as in "Automated formatting of Java files". Is there a reason to treat java seperate from other files?
salu2 On Fri, 2005-08-26 at 16:57 +1000, David Crossley wrote: > Diwaker Gupta wrote: > > When I say XML, I mean all kinds of XML files (XSL, *.fv, *.ft) > > How will the tool know which are xml file-types? > We have a multitude of xml filename extensions. > There is a list on one of the tools in the > "committers" svn repository at relicense/src/insert.pl > > > As with Java, I have committed a "tidied" version of the html2document.xsl > > sample document in the test-whitespace directory. As with Jalopy, Tidy > > (http://tidy.sf.net) is extremely configurable, so if you don't like > > something or have a suggestion, just let me know. > > I think that it is doing too much, e.g. removing the > blank lines before major elements, e.g. <xsl:template> > > We should start with very simple stuff, e.g. just tabs > and trailing whitespace, then gradually add other operations. > However we don't want to get too strict on code style. > > I suggest that we define a list of what we would possibly > want to adrress. Here is a start: > > 1 whitespace at end-of-line > 2 tabs to four-space > 3 indentation > 4 word-wrap for long lines > 5 whitespace between attribute definitions > > We need to at least do 1 and 2 while 5 may be doubtful. > > It seems to be too vigorous with wrapping. Perhaps wider > would be better. > < <xsl:key name="h5s" match="h5" > use="generate-id(preceding-sibling::h4[1])"/> > --- > > <xsl:key name="h5s" match="h5" > > use="generate-id(preceding-sibling::h4[1])" /> > > Also we need to run it on one of the xdocs > in site-author/content/xdocs to see what it does > with word-wrapping for long lines of element content. > > -David -- thorsten "Together we stand, divided we fall!" Hey you (Pink Floyd)
