From 7f193c204a4fcde76ef37978cd47e3de63aba66c Mon Sep 17 00:00:00 2001 From: Michael Schwendt Date: Tue, 19 Feb 2013 12:12:30 +0100 Subject: [PATCH] Merge "remove-newlines" (from alt tags) patch (Debian #299027). - Include html2text script as python-html2script. - Minor spec cleanup. --- aaronsw-html2text-remove-newlines.patch | 11 +++++++++ python-html2text.spec | 33 ++++++++++++++----------- 2 files changed, 30 insertions(+), 14 deletions(-) create mode 100644 aaronsw-html2text-remove-newlines.patch diff --git a/aaronsw-html2text-remove-newlines.patch b/aaronsw-html2text-remove-newlines.patch new file mode 100644 index 0000000..54f8b85 --- /dev/null +++ b/aaronsw-html2text-remove-newlines.patch @@ -0,0 +1,11 @@ +diff -Nur aaronsw-html2text-c368b14-orig/html2text.py aaronsw-html2text-c368b14/html2text.py +--- aaronsw-html2text-c368b14-orig/html2text.py 2012-01-07 16:01:09.000000000 +0100 ++++ aaronsw-html2text-c368b14/html2text.py 2013-02-19 11:53:09.609134393 +0100 +@@ -479,6 +479,7 @@ + if has_key(attrs, 'src'): + attrs['href'] = attrs['src'] + alt = attrs.get('alt', '') ++ alt = re.sub('\n', ' ', alt) + if self.inline_links: + self.o("![") + self.o(alt) diff --git a/python-html2text.spec b/python-html2text.spec index 88694c0..077a6d8 100644 --- a/python-html2text.spec +++ b/python-html2text.spec @@ -1,17 +1,17 @@ -# sitelib for noarch packages, sitearch for others (remove the unneeded one) %{!?python_sitelib: %define python_sitelib %(%{__python} -c "from distutils.sysconfig import get_python_lib; print get_python_lib()")} %global tardirname aaronsw-html2text-c368b14 Name: python-html2text Version: 3.200.3 -Release: 3%{?dist} -Summary: Converts a page of HTML into clean, easy-to-read plain ASCII text +Release: 4%{?dist} +Summary: Converts a page of HTML into plain ASCII text Group: Development/Languages License: GPLv3 URL: http://www.aaronsw.com/2002/html2text/ #Source0: https://github.com/aaronsw/html2text/tarball/%{version}/%{name}-%{version}.tar.gz +# git clone git://github.com/aaronsw/html2text.git aaronsw-html2text-c368b14 ; cd aaronsw-html2text-c368b14 ; git checkout c368b14 Source0: aaronsw-html2text-3.02-77-gc368b14.tar.gz BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-root-%(%{__id_u} -n) @@ -20,8 +20,12 @@ BuildArch: noarch BuildRequires: python BuildRequires: python-setuptools-devel +# Debian bug #299027 +Patch0: aaronsw-html2text-remove-newlines.patch + + %description -html2text is a Python script that convers a page of HTML into clean, +html2text is a Python script that converts a page of HTML into clean, easy-to-read plain ASCII text. Better yet, that ASCII also happens to be valid Markdown (a text-to-HTML format). @@ -29,34 +33,35 @@ Also known as: THE ASCIINATOR, html to text, htm to txt, htm2txt, ... %prep %setup -q -n %{tardirname} +%patch0 -p1 -b .remove-newlines %build -echo Nothing to build %install -rm -rf $RPM_BUILD_ROOT %{__python} setup.py install --root $RPM_BUILD_ROOT -# TODO: conflicts with package' html2text' obviously -rm $RPM_BUILD_ROOT/usr/bin/html2text +# new script in 3.200.3 : conflicts with package' html2text' obviously +mv $RPM_BUILD_ROOT/usr/bin/html2text $RPM_BUILD_ROOT/usr/bin/%{name} + %check cd test PYTHONPATH=${RPM_BUILD_ROOT}%{python_sitelib} %{__python} run_tests.py -%clean -rm -rf $RPM_BUILD_ROOT - - %files -%defattr(-,root,root,-) +/usr/bin/%{name} %{python_sitelib}/* %doc COPYING %changelog +* Tue Feb 19 2013 Michael Schwendt - 3.200.3-4 +- Merge "remove-newlines" (from alt tags) patch (Debian #299027). +- Include html2text script as python-html2script. +- Minor spec cleanup. + * Thu Feb 14 2013 Fedora Release Engineering - 3.200.3-3 - Rebuilt for https://fedoraproject.org/wiki/Fedora_19_Mass_Rebuild @@ -96,7 +101,7 @@ rm -rf $RPM_BUILD_ROOT * Sat Nov 29 2008 Ignacio Vazquez-Abrams - 2.34-2.1 - Rebuild for Python 2.6 -* Sat Oct 09 2008 Thorsten Leemhuis - 2.34-1 +* Sat Oct 11 2008 Thorsten Leemhuis - 2.34-1 - update to 2.34 * Sat Sep 27 2008 Thorsten Leemhuis - 2.33-1