ANNOUNCE: Haddock version 0.6

There is a new release of Haddock, version 0.6. This version has several improvements over the previous release; the full list is below. Source and binary distributions can be obtained from Haddock's homepage: http://www.haskell.org/haddock/ Enjoy! Simon Changes in version 0.6: * Implicit parameters, zip comprehensions and 'mdo' are now supported by the parser. * Some lexical bugs were fixed, mainly concerning literal paragraphs next to non-literal paragraphs. * New options: --gen-index, --use-index, --gen-contents and --use-contents, for generting a combined index and contents for several libraries from the appropriate .haddock files. * There is now one index rather than separate Function/Constructor and Type/Class indices. Where a name refers to several entities, these are listed separately in the index. * New option: -v, elicits more verbose warnings. Some warnings are now turned off by default; Haddock should be a little less noisy in general. * Markup for definition lists has been added. See the documentation for details. * New option: --package for setting the package name. The package is listed alongside each module name in the combined contents. * Entities which are re-exported from another package now at least leave a hyperlink in the generated HTML, rather than nothing at all. * Some fixes for bugs which caused incorrect hyperlinks in the generated HTML. In particular, instances should now be linked correctly. * Some aesthetic improvements to the generated HTML.

On Tue, 11 Nov 2003 11:01:00 -0000, "Simon Marlow"
There is a new release of Haddock, version 0.6.
Are interface files binary-incompatible with those of previous releases? With a Cygwin-compiled 0.6, I get an error: Fail: end of file Action: Data.Binary.getWord8 while trying to --read-interface a GHC 6.01 .haddock file (for example, ghc/doc/html/haskell98/haskell98.haddock). OTOH, haddock --html --dump-interface=doc\M1.haddock M1.hs haddock --html --read-interface=doc\,doc\M1.haddock M2.hs seems to work. And BTW, why it is necessary to write --read-interface=path,full-filename, i.e., why --read-interface=c:\ghc\doc\html\haskell98,c:\ghc\doc\html\haskell98.haddock instead of --read-interface=c:\ghc\doc\html\haskell98,haskell98.haddock ? Thanks, /L/e/k/t/u

On Tue, 11 Nov 2003 11:01:00 -0000, "Simon Marlow"
wrote: There is a new release of Haddock, version 0.6.
Are interface files binary-incompatible with those of previous releases?
With a Cygwin-compiled 0.6, I get an error:
Fail: end of file Action: Data.Binary.getWord8
Yes, they're binary incompatible I'm afraid.
And BTW, why it is necessary to write --read-interface=path,full-filename, i.e., why
--read-interface=c:\ghc\doc\html\haskell98,c:\ghc\doc\html\has kell98.haddock
The path is the path to the HTML files, and the filename is the name of the .haddock file. They can be completely different, e.g. --read-interface=http://www.haskell.org/ghc/docs/latest/html/base,/usr/local/share/ghc-6.0.1/... Cheers, Simon

On Wed, 12 Nov 2003 10:07:58 -0000
"Simon Marlow"
Yes, they're binary incompatible I'm afraid.
Oh, well... :)
The path is the path to the HTML files, and the filename is the name of the .haddock file. They can be completely different, e.g.
Of course. Silly me. Thanks, Juanma

In article
There is a new release of Haddock, version 0.6. This version has several improvements over the previous release; the full list is below.
Excellent! I was stuck on implicit parameters for awhile. I found one issue in 0.6 though, it seems to choke on explicit kind declarations, for instance: newtype Constant (m :: * -> *) a = MkConstant a deriving (Eq); -- Ashley Yakeley, Seattle WA
participants (4)
-
Ashley Yakeley
-
Juanma Barranquero
-
Juanma Barranquero
-
Simon Marlow