-------------------------------- Some notes about using html2text -------------------------------- 1. HTTP support The original html2text doesn't support any complicated HTTP queries and answers. The Debian version of html2text doesn't provide http support at all. However, you can easily operate by using the curl or wget packages: curl -s http://www.server.org/aaa/bbb/ccc.html | html2text wget http://www.server.org/aaa/bbb/ccc.html -O- | html2text Using wget or curl for downloading packages allows you to use: - proxy servers; - https; - ftp; - some IPv6 support; - and any other downloading feature that wget or curl supports. Please don't submit bugs about direct HTTP support; use the approach mentioned above instead. 2. Input recoding The original html2text lacks support for recognizing encoding of html file. However, the Debian version has it and even has it turned on by default. Use the new '-nometa' option to restore the original behavior. To ensure html2text processes your document right, check that one of following cases is true: - the document has 'META HTTP-EQUIV' section with right encoding; - the document is encoded in ISO_8859-1; - the document is encoded in US ASCII and one of '-ascii' or '-utf8' options is supplied; - the document is encoded in UTF-8 and '-utf8' option is supplied. 3. Output recoding The original html2text doesn't have support for output recoding. However, the Debian version does. Note that this recoding will be done only if output is a terminal. html2text recodes output into the current user's locale charset, stored in LC_CTYPE. This will work in most cases. If you need to recode output to some other charset, you have to specify LC_CTYPE=<locale>.<charset> html2text <options> If you have LC_ALL set, you have to substitute LC_ALL for LC_CTYPE in the example above. 4. Backspaces Debian version of html2text now does recoding, which may use multi-byte encodings. Since backspace-producing code does not work with multi-byte encodings (at least currently), producing of backspaces is disabled. -- Eugene V. Lyubimkin
Generated by dwww version 1.15 on Fri May 24 05:44:37 CEST 2024.