Documentation is in main as ceb28b67204e. Since this apparently works with the
internal cvsps, marking this as resolved.

I have sent a patch with documentation to the mailing list, see
[PATCH] convert: add documentation for CVS source

frank: is there some easy-to-find documentation that points to the built-in
cvsps, explains what problems it solves and how to enable it? I had one user ask
me some questions about this, and it would be nice if I had an easy answer.

cvsps.py now in crew

On Sat, Apr 19, 2008 at 6:02 PM, Frank Kingswood <mercurial-bugs@selenic.com>
wrote:

>
> Frank Kingswood <frank@kingswood-consulting.co.uk> added the comment:
>
> Thanks for your feedback. That was the final nudge needed to make me put
> the caching code into cvsps.py.  A new patch implementing builtin cvsps
> is now available:
>

I applied your patches through to

 http://www.selenic.com/pipermail/mercurial-devel/2008-April/006013.html

to http://selenic.com/repo/hg 6567:24dbf3a32b18

(which applied cleanly), and hg convert builtin worked cleanly for me. Null
incremental converts run at less than a second now, which is great.

Thanks,

Dickon

Thanks for your feedback. That was the final nudge needed to make me put 
the caching code into cvsps.py.  A new patch implementing builtin cvsps 
is now available:

http://www.selenic.com/pipermail/mercurial-devel/2008-April/005936.html

Frank

I did a make local of mercurial 1.0 with the patch from
http://www.selenic.com/pipermail/mercurial-devel/2008-April/005743.html applied. 


First I tried "~/src/mercurial-int-cvsps/hg convert v5 v5-hg-int" (without
touching my .hgrc) and got:

invalid rev format censored_2004_02_19
connecting to /local/dr/cvsroot
scanning source...
sorting...
converting...

and an exit code zero. (The word censored was actually a short lower case
alphabetic name, but unfortunately the name of a client I can't quote here :)).

I then tried:

~/src/mercurial-int-cvsps/hg convert --config convert.cvsps=builtin v5 v5-hg-int

and this completed with a reasonable looking Mercurial repository, though I've
only had a cursory glace at it. Subsequent reruns of the command managed to spot
  additional commits to CVS and copy them across, but I note incremental runs
are generally taking about 160 seconds. Incremental runs of Mercurial 1.0
convert take less than one second with no changes pending, and about 45 seconds
when a few changes need to be synthesised into changesets.

Dickon Reed wrote:
> New submission from Dickon Reed <dickon@cantab.net>:
> 
> This is with Mercurial 1.0, cvsps 2.2b1. I do not get any problem with cvsps 2.1.

Can you try the builtin cvsps?
A patch to add a builtin cvsps is on the mercurial-devel list here:
http://www.selenic.com/pipermail/mercurial-devel/2008-April/005743.html

Frank

This is with Mercurial 1.0, cvsps 2.2b1. I do not get any problem with cvsps 2.1.

[dr@runbench-dev dr]$ nice /usr/local/bin/hg convert v5
assuming destination v5-hg
initializing destination v5-hg repository
connecting to /local/dr/cvsroot/cvsroot
scanning source...
** unknown exception encountered, details follow
** report bug details to http://www.selenic.com/mercurial/bts
** or mercurial@selenic.com
** Mercurial Distributed SCM (version 1.0)
Traceback (most recent call last):
  File "/usr/local/bin/hg", line 20, in ?
    mercurial.dispatch.run()
  File "/usr/local/lib/python2.4/site-packages/mercurial/dispatch.py", line 20, 
in run
    sys.exit(dispatch(sys.argv[1:]))
  File "/usr/local/lib/python2.4/site-packages/mercurial/dispatch.py", line 29, 
in dispatch
    return _runcatch(u, args)
  File "/usr/local/lib/python2.4/site-packages/mercurial/dispatch.py", line 45, 
in _runcatch
    return _dispatch(ui, args)
  File "/usr/local/lib/python2.4/site-packages/mercurial/dispatch.py", line 364,
 in _dispatch
    ret = _runcommand(ui, options, cmd, d)
  File "/usr/local/lib/python2.4/site-packages/mercurial/dispatch.py", line 417,
 in _runcommand
    return checkargs()
  File "/usr/local/lib/python2.4/site-packages/mercurial/dispatch.py", line 373,
 in checkargs
    return cmdfunc()
  File "/usr/local/lib/python2.4/site-packages/mercurial/dispatch.py", line 358,
 in <lambda>
    d = lambda: func(ui, *args, **cmdoptions)
  File "/usr/local/lib/python2.4/site-packages/hgext/convert/__init__.py", line 
127, in convert
    return convcmd.convert(ui, src, dest, revmapfile, **opts)
  File "/usr/local/lib/python2.4/site-packages/hgext/convert/convcmd.py", line 3
53, in convert
    c.convert()
  File "/usr/local/lib/python2.4/site-packages/hgext/convert/convcmd.py", line 2
78, in convert
    parents = self.walktree(heads)
  File "/usr/local/lib/python2.4/site-packages/hgext/convert/convcmd.py", line 1
01, in walktree
    commit = self.cachecommit(n)
  File "/usr/local/lib/python2.4/site-packages/hgext/convert/convcmd.py", line 2
17, in cachecommit
    commit = self.source.getcommit(rev)
  File "/usr/local/lib/python2.4/site-packages/hgext/convert/cvs.py", line 306, 
in getcommit
    return self.changeset[rev]
KeyError: 'bad'