I was using TagSoup 0.8 with great success. On upgrading to 0.9 I have
Thanks Neil, Using Network.HTTP worked. However something else I have just run into concerns some web pages that start with: <?xml version="1.0" encoding="iso-8859-1"?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> I get the following bad result: TagText "HTTP/1.1 200 OK\r\nContent-Type: text/html\r\nLast-Modified: Tue, 27 Oct 2009 19:30:40 GMT\r\nETag: \"6f248cf73b57ca1:25e2\"\r\nDate: Sun, 23 May 2010 22:46:41 GMT\r\nTransfer-Encoding: chunked\r\nConnection: close\r\nConnection: Transfer-Encoding\r\n\r\n00004000\r\n\255\254<\NUL?\NULx\NULm\NULl\NUL \NULv\NULe\NULr\NULs\NULi\NULo\NULn\NUL=\NUL\"\NUL1\NUL.\NUL0\NUL\"\NUL \NULe\NULn\NULc\NULo\NULd\NULi\NULn\NULg\NUL=\NUL\"\NULi\NULs\NULo\NUL-\NUL8 \NUL8\NUL5\NUL9\NUL-\NUL1\NUL\"\NUL etc etc Is this an easy thing to fix? I've started to look over the code. -----Original Message----- From: Neil Mitchell [mailto:ndmitchell@gmail.com] Sent: Wednesday, May 19, 2010 12:19 PM To: Ralph Hodgson Cc: Daniel Fischer; haskell-cafe@haskell.org; Don Stewart Subject: Re: [Haskell-cafe] TagSoup 0.9 Hi Ralph, this error:
TQ\TagSoup\TagSoupExtensions.lhs:29:17:
`Tag' is not applied to enough type arguments
Expected kind `*', but `Tag' has kind `* -> *'
In the type synonym declaration for `Bundle'
Failed, modules loaded: TQ.Common.TextAndListHandling.
My change notes have this being a change between 0.6 and 0.8. As Malcolm says, any old uses of "Tag" should become "Tag String". The reason is that Tag is now parameterised, and you can use Tag ByteString etc. However, I should point out that Tag ByteString won't be any faster than Tag String in this version (it's in the future work pile).
Forgot to add: I now need to understand the following warnings on this
line "> import Text.HTML.Download":
Everyone's comments have been right. I previously included Text.HTML.Download so that it was easy to test tagsoup against the web. Since I first wrote that snippet the HTTP downloading libraries have improved substantially, so people should use those in favour of the version in tagsoup - you'll be able to connect to more websites in more reliable ways, go through proxies etc. I don't intend to remove the Download module any time soon, but I will do eventually. Thanks, Neil