All of lore.kernel.org
 help / color / mirror / Atom feed
From: Tor Arvid Lund <torarvid@gmail.com>
To: Ian Wienand <ianw@vmware.com>
Cc: "git@vger.kernel.org" <git@vger.kernel.org>
Subject: Re: [PATCH] Support different branch layouts in git-p4
Date: Tue, 8 Feb 2011 00:27:12 +0100	[thread overview]
Message-ID: <AANLkTimGKc4MTwb=AnZ_Bv1EGS7yfgrFupxBOVVSm4s8@mail.gmail.com> (raw)
In-Reply-To: <4D4F3738.7010603@vmware.com>

On Mon, Feb 7, 2011 at 1:05 AM, Ian Wienand <ianw@vmware.com> wrote:
> On 04/02/11 16:37, Tor Arvid Lund wrote:
>> For starters, I don't think that I like git-p4 being taught to solve
>> problems that seem to be caused by a poor/unfortunate perforce layout.
>
> I do think this //depot/project/branch type layout is pretty typical,
> although I admit I don't have a lot of experience with alternative p4
> setups.

You may be right, although I suspect that
//depot/department/project/branch may be equally typical. At my
$dayjob, we have gone through several "reorganizations" of the
perforce layouts. I'm the guy that never likes any of them ;)

>> A solution which I think would work well for everyone, is if files
>> would be placed according to the right-hand patterns in the
>> client-spec.
>
> I did consider this at first.  My only issue is that it is a bit
> confusing to use the client spec for filtering (and in this case
> re-writing), but not for actually selecting the depots to clone, which
> I still need to replicate on the command line.  However that is a much
> larger change.
>
> What do you think of this one?

In general, me thinks me likes it :-)

(and it turned out much smaller than I would have originally guessed)

I should probably mention that I haven't tested your patch at all. I
will have a pretty rough week at work, so it would be great if anyone
else feels like chiming in on this one... But I have some quick
observations below:

> ---
>
> diff --git a/contrib/fast-import/git-p4 b/contrib/fast-import/git-p4
> index 04ce7e3..eb9620c 100755
> --- a/contrib/fast-import/git-p4
> +++ b/contrib/fast-import/git-p4
> @@ -878,6 +878,7 @@ class P4Sync(Command):
>         self.cloneExclude = []
>         self.useClientSpec = False
>         self.clientSpecDirs = []
> +        self.clientName = None
>
>         if gitConfig("git-p4.syncFromOrigin") == "false":
>             self.syncWithOrigin = False
> @@ -910,6 +911,22 @@ class P4Sync(Command):
>         return files
>
>     def stripRepoPath(self, path, prefixes):
> +        if self.useClientSpec:
> +
> +            # if using the client spec, we use the output directory
> +            # specified in the client.  For example, a view
> +            #   //depot/foo/branch/... //client/branch/foo/...
> +            # will end up putting all foo/branch files into
> +            #  branch/foo/
> +            for val in self.clientSpecDirs:
> +                if path.startswith(val[0]):
> +                    # replace the depot path with the client path
> +                    path = path.replace(val[0], val[1][1])
> +                    # now strip out the client (//client/...)
> +                    path = re.sub("^(//[^/]+/)", '', path)
> +                    # the rest is all path
> +                    return path
> +
>         if self.keepRepoPath:
>             prefixes = [re.sub("^(//[^/]+/).*", r'\1', prefixes[0])]
>
> @@ -1032,7 +1049,7 @@ class P4Sync(Command):
>             includeFile = True
>             for val in self.clientSpecDirs:
>                 if f['path'].startswith(val[0]):
> -                    if val[1] <= 0:
> +                    if val[1][0] <= 0:
>                         includeFile = False
>                     break
>
> @@ -1474,20 +1491,36 @@ class P4Sync(Command):
>         temp = {}
>         for entry in specList:
>             for k,v in entry.iteritems():
> +                if k.startswith("Client"):
> +                    self.clientName = v
> +
>                 if k.startswith("View"):
>                     if v.startswith('"'):
>                         start = 1
>                     else:
>                         start = 0
>                     index = v.find("...")
> +
> +                    # save the "client view"; i.e the RHS of the view
> +                    # line that tells the client where to put the
> +                    # files for this view.
> +                    cv = v[index+4:] # +4 to remove previous '... '

This feels less robust than what we might want. Isn't the format of a
client-spec line either:

-?//depot/path[/...]\s+//client/path[/...]\n

or

-?"//depot/path with spaces/path[/...]"\s+"//client/path with spaces/path[/...]"

.. where -? means an optional '-' char, and \s+ is
'whatever-length-and-kind-of-whitespace'. I'm just guessing from
memory regarding these patterns, but assuming that the section
separator is exactly the string '... ' seems risky, no? :)

> +                    cv_index = cv.find("...")
> +                    cv=cv[:cv_index]

What if a line doesn't end with "..." ? Maybe add an "if cv_index >= 0"

> +
> +                    # now save the view; +index means included, -index
> +                    # means it should be filtered out.
>                     v = v[start:index]
>                     if v.startswith("-"):
>                         v = v[1:]
> -                        temp[v] = -len(v)
> +                        include = -len(v)
>                     else:
> -                        temp[v] = len(v)
> +                        include = len(v)
> +
> +                    temp[v] = (include, cv)
> +
>         self.clientSpecDirs = temp.items()
> -        self.clientSpecDirs.sort( lambda x, y: abs( y[1] ) - abs( x[1] ) )
> +        self.clientSpecDirs.sort( lambda x, y: abs( y[1][0] ) - abs( x[1][0] ) )
>
>     def run(self, args):
>         self.depotPaths = []
> diff --git a/contrib/fast-import/git-p4.txt b/contrib/fast-import/git-p4.txt
> index 49b3359..e09da44 100644
> --- a/contrib/fast-import/git-p4.txt
> +++ b/contrib/fast-import/git-p4.txt
> @@ -191,6 +191,11 @@ git-p4.useclientspec
>
>   git config [--global] git-p4.useclientspec false
>
> +The P4CLIENT environment variable should be correctly set for p4 to be
> +able to find the relevant client.  This client spec will be used to
> +both filter the files cloned by git and set the directory layout as
> +specified in the client (this implies --keep-path style semantics).
> +
>  Implementation Details...
>  =========================
>
>

Aight. I need some sleep now. Nice work so far, Ian! :)

    -- Tor Arvid

  reply	other threads:[~2011-02-07 23:27 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-02-01 22:59 [PATCH] Support different branch layouts in git-p4 Ian Wienand
2011-02-05  0:37 ` Tor Arvid Lund
2011-02-07  0:05   ` Ian Wienand
2011-02-07 23:27     ` Tor Arvid Lund [this message]
2011-02-09  3:46       ` Ian Wienand
2011-02-10 13:43         ` Pete Wyckoff
2011-02-08  1:22     ` Pete Wyckoff

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='AANLkTimGKc4MTwb=AnZ_Bv1EGS7yfgrFupxBOVVSm4s8@mail.gmail.com' \
    --to=torarvid@gmail.com \
    --cc=git@vger.kernel.org \
    --cc=ianw@vmware.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.