]>
Commit | Line | Data |
---|---|---|
6467c213 ER |
1 | Summary: Utility to clone CVS repositories over the cvspserver interface |
2 | Name: cvsclone | |
3 | Version: 0.1 | |
4 | Release: 1 | |
5 | License: GPL v2 | |
6 | Source0: http://repo.or.cz/w/cvsclone.git/snapshot/558950ab442bc0551c8c16f8d3d6bc972818a81d.tar.gz#/%{name}.tgz | |
91cd312d | 7 | # Source0-md5: f678e9cd37f4d2f85466a4e817a108c2 |
6467c213 ER |
8 | Group: Development/Languages |
9 | URL: https://samba.org/ftp/tridge/rtc/cvsclone.l | |
10 | BuildRequires: sed >= 4.0 | |
11 | BuildRoot: %{tmpdir}/%{name}-%{version}-root-%(id -u -n) | |
12 | ||
13 | %description | |
14 | As git cvsimport is pretty slow over the wire (and as cvs2git can do a | |
15 | substantially better job if there are a lot of branches), it is often | |
16 | better to "clone" a whole CVS repository. Enter cvsclone, written by | |
17 | Peter Backes. | |
18 | ||
19 | You can find a few touchups by yours truly here, allowing to pipe the | |
20 | output of "cvs rlog" into a file, editing the file to fix errors (such | |
21 | as geniuses who put the cvs log of moved files into a commit message) | |
22 | and use that file as input instead. | |
23 | ||
24 | %prep | |
25 | %setup -qn %{name} | |
26 | ||
27 | %{__sed} -i -e ' | |
28 | s,gcc,$(CC), | |
29 | s,-g,$(CFLAGS), | |
30 | ' Makefile | |
31 | ||
32 | %build | |
33 | %{__make} \ | |
34 | CC="%{__cc}" \ | |
35 | CFLAGS="%{rpmcflags}" | |
36 | ||
37 | %install | |
38 | rm -rf $RPM_BUILD_ROOT | |
39 | install -d $RPM_BUILD_ROOT%{_bindir} | |
40 | install -p cvsclone $RPM_BUILD_ROOT%{_bindir} | |
41 | ||
42 | %clean | |
43 | rm -rf $RPM_BUILD_ROOT | |
44 | ||
45 | %files | |
46 | %defattr(644,root,root,755) | |
47 | %attr(755,root,root) %{_bindir}/cvsclone |