* Re: [PATCH 1/2] git_remote_cvs: Honor DESTDIR in the Makefile
From: Johan Herland @ 2009-08-17 1:58 UTC (permalink / raw)
To: David Aguilar; +Cc: Johannes Schindelin, gitster, git, barkalow, mhagger
In-Reply-To: <20090816210300.GB23522@gmail.com>
On Sunday 16 August 2009, David Aguilar wrote:
> I'll see if we rework this so that we end up passing "" to
> --root instead of /. I'm going to be gone for a few hours so
> probably won't be able to try it out until tonight.
Thanks a lot for your work! I will send an updated series shortly which will
include v2 of your DESTDIR/Makefile fixes, and also the fixes you suggested
earlier (including _lots_ of PEP8 fixes).
> Another thing to consider --
>
> Debian once submitted a bug against another Python app asking
> that we not place modules in site-packages unless we
> plan on having other applications importing those modules.
>
> The more appropriate place for them if we don't plan on that is
> $(prefix)/share/git-core/git_remote_cvs or something like that.
>
> I guess that's another thing to think about.
Yes, Debian raises a valid point. I haven't thought much about making the
git_remote_cvs package into something that would be useful for other
applications. (I just assumed that the Python convention was to install it
into site-packages regardless...) For now, I'll concentrate on git-remote-
cvs, and leave it to others to figure out if anything in the git_remote_cvs
package is useful for other programs.
Note that there's a small chicken-and-egg problem here as well: If Debian
refuses us to install into site-packages, it will be harder for other Python
programs to discover (and import) the git_remote_cvs package.
BTW, when we're on the subject of packaging: There are some variables in
git_remote_cvs/setup.py where I'm not sure what the correct value should be:
- version - should this follow Git's version number, or is it independent?
- author (and author_email + url) - For now, I'm referring to the Git
community. Should this be more specific/
Feedback welcome.
Have fun! :)
...Johan
--
Johan Herland, <johan@herland.net>
www.herland.net
^ permalink raw reply
* Re: How to stop sharing objects between repositories
From: Mike Galbraith @ 2009-08-17 2:21 UTC (permalink / raw)
To: Junio C Hamano; +Cc: Jeff King, Johannes Schindelin, Jon Jensen, git
In-Reply-To: <7vmy5z603d.fsf@alter.siamese.dyndns.org>
On Sun, 2009-08-16 at 12:16 -0700, Junio C Hamano wrote:
> Jeff King <peff@peff.net> writes:
>
> > Subject: [PATCH] docs: mention how to break alternates dependency
> >
> > A user who has created a repository dependency by using "git
> > clone -s" does not necessarily know where to look to find
> > out how to break that dependency. Let's mention it right
> > under "-s", where they are most likely to find it.
> >
> > Signed-off-by: Jeff King <peff@peff.net>
> > ---
> > Documentation/git-clone.txt | 5 +++++
> > 1 files changed, 5 insertions(+), 0 deletions(-)
> >
> > diff --git a/Documentation/git-clone.txt b/Documentation/git-clone.txt
> > index b14de6c..87fa687 100644
> > --- a/Documentation/git-clone.txt
> > +++ b/Documentation/git-clone.txt
> > @@ -72,6 +72,11 @@ These objects may be removed by normal git operations (such as 'git-commit')
> > which automatically call `git gc --auto`. (See linkgit:git-gc[1].)
> > If these objects are removed and were referenced by the cloned repository,
> > then the cloned repository will become corrupt.
> > ++
> > +To break the dependency of the cloned repository to the source
> > +repository, run `git repack -a` in the cloned repository, which will
> > +create a new pack in that repository with all referenced objects,
> > +including those in the source repository.
>
> After reading this, two points come to my mind. They may or may not be
> issues.
>
> (1) Such a user does not necessarily know a casual "git repack -a" breaks
> the dependency, defeating the -s option s/he deliberately used in
> order to save disk space in the first place. Perhaps we can reword
> this further to kill two penguins with a single stone?
Perhaps a runtime warning that you're about to break it? This user may
not even be the one who set the thing up, no?
-T. Peanut Gallery
^ permalink raw reply
* [RFCv4 0/5] CVS remote helper
From: Johan Herland @ 2009-08-17 3:35 UTC (permalink / raw)
To: git; +Cc: Johannes.Schindelin, Johan Herland, barkalow, davvid, gitster
Hi,
Another iteration of the patch series implementing a CVS remote helper.
Changes from the previous iteration:
- Python coding style fixes suggested by PEP8, PyLint, David Aguilar, etc.
- Documentation/git-remote-cvs.txt compile fix
- git_remote_cvs/Makefile fixes from David Aguilar (his v2 series)
- rebased onto current 'pu' (minus jh/vcs-cvs)
- split up patch 2/4 into two parts to fit on the Git mailing list
Have fun! :)
...Johan
Johan Herland (5):
Basic build infrastructure for Python scripts
1/2: Add Python support library for CVS remote helper
2/2: Add Python support library for CVS remote helper
git-remote-cvs: Remote helper program for CVS repositories
Add simple selftests of git-remote-cvs functionality
Documentation/git-remote-cvs.txt | 85 +++
Makefile | 46 ++
configure.ac | 3 +
git-remote-cvs.py | 683 ++++++++++++++++++++++++
git_remote_cvs/.gitignore | 2 +
git_remote_cvs/Makefile | 35 ++
git_remote_cvs/__init__.py | 27 +
git_remote_cvs/changeset.py | 126 +++++
git_remote_cvs/commit_states.py | 62 +++
git_remote_cvs/cvs.py | 998 ++++++++++++++++++++++++++++++++++++
git_remote_cvs/cvs_revision_map.py | 418 +++++++++++++++
git_remote_cvs/cvs_symbol_cache.py | 313 +++++++++++
git_remote_cvs/git.py | 680 ++++++++++++++++++++++++
git_remote_cvs/setup.py | 17 +
git_remote_cvs/util.py | 186 +++++++
t/t9800-remote-cvs-basic.sh | 524 +++++++++++++++++++
t/t9801-remote-cvs-fetch.sh | 291 +++++++++++
t/test-lib.sh | 1 +
18 files changed, 4497 insertions(+), 0 deletions(-)
create mode 100644 Documentation/git-remote-cvs.txt
create mode 100755 git-remote-cvs.py
create mode 100644 git_remote_cvs/.gitignore
create mode 100644 git_remote_cvs/Makefile
create mode 100644 git_remote_cvs/__init__.py
create mode 100644 git_remote_cvs/changeset.py
create mode 100644 git_remote_cvs/commit_states.py
create mode 100644 git_remote_cvs/cvs.py
create mode 100644 git_remote_cvs/cvs_revision_map.py
create mode 100644 git_remote_cvs/cvs_symbol_cache.py
create mode 100644 git_remote_cvs/git.py
create mode 100644 git_remote_cvs/setup.py
create mode 100644 git_remote_cvs/util.py
create mode 100755 t/t9800-remote-cvs-basic.sh
create mode 100755 t/t9801-remote-cvs-fetch.sh
^ permalink raw reply
* [RFCv4 1/5] Basic build infrastructure for Python scripts
From: Johan Herland @ 2009-08-17 3:35 UTC (permalink / raw)
To: git; +Cc: Johannes.Schindelin, Johan Herland, barkalow, davvid, gitster
In-Reply-To: <1250480161-21933-1-git-send-email-johan@herland.net>
This patch adds basic boilerplate support (based on corresponding Perl
sections) for enabling the building and installation Python scripts.
There are currently no Python scripts being built, and when Python
scripts are added in future patches, their building and installation
can be disabled by defining NO_PYTHON.
Signed-off-by: Johan Herland <johan@herland.net>
---
Makefile | 13 +++++++++++++
configure.ac | 3 +++
t/test-lib.sh | 1 +
3 files changed, 17 insertions(+), 0 deletions(-)
diff --git a/Makefile b/Makefile
index 189aee5..969cef5 100644
--- a/Makefile
+++ b/Makefile
@@ -166,6 +166,8 @@ all::
#
# Define NO_PERL if you do not want Perl scripts or libraries at all.
#
+# Define NO_PYTHON if you do not want Python scripts or libraries at all.
+#
# Define NO_TCLTK if you do not want Tcl/Tk GUI.
#
# The TCL_PATH variable governs the location of the Tcl interpreter
@@ -311,6 +313,7 @@ LIB_H =
LIB_OBJS =
PROGRAMS =
SCRIPT_PERL =
+SCRIPT_PYTHON =
SCRIPT_SH =
TEST_PROGRAMS =
@@ -349,6 +352,7 @@ SCRIPT_PERL += git-svn.perl
SCRIPTS = $(patsubst %.sh,%,$(SCRIPT_SH)) \
$(patsubst %.perl,%,$(SCRIPT_PERL)) \
+ $(patsubst %.py,%,$(SCRIPT_PYTHON)) \
git-instaweb
# Empty...
@@ -404,8 +408,12 @@ endif
ifndef PERL_PATH
PERL_PATH = /usr/bin/perl
endif
+ifndef PYTHON_PATH
+ PYTHON_PATH = /usr/bin/python
+endif
export PERL_PATH
+export PYTHON_PATH
LIB_FILE=libgit.a
XDIFF_LIB=xdiff/lib.a
@@ -1259,6 +1267,10 @@ ifeq ($(PERL_PATH),)
NO_PERL=NoThanks
endif
+ifeq ($(PYTHON_PATH),)
+NO_PYTHON=NoThanks
+endif
+
QUIET_SUBDIR0 = +$(MAKE) -C # space to separate -C and subdir
QUIET_SUBDIR1 =
@@ -1306,6 +1318,7 @@ prefix_SQ = $(subst ','\'',$(prefix))
SHELL_PATH_SQ = $(subst ','\'',$(SHELL_PATH))
PERL_PATH_SQ = $(subst ','\'',$(PERL_PATH))
+PYTHON_PATH_SQ = $(subst ','\'',$(PYTHON_PATH))
TCLTK_PATH_SQ = $(subst ','\'',$(TCLTK_PATH))
LIBS = $(GITLIBS) $(EXTLIBS)
diff --git a/configure.ac b/configure.ac
index 3f1922d..3749e5c 100644
--- a/configure.ac
+++ b/configure.ac
@@ -241,6 +241,9 @@ GIT_ARG_SET_PATH(shell)
# Define PERL_PATH to provide path to Perl.
GIT_ARG_SET_PATH(perl)
#
+# Define PYTHON_PATH to provide path to Python.
+GIT_ARG_SET_PATH(python)
+#
# Define ZLIB_PATH to provide path to zlib.
GIT_ARG_SET_PATH(zlib)
#
diff --git a/t/test-lib.sh b/t/test-lib.sh
index a5b8d03..01ea386 100644
--- a/t/test-lib.sh
+++ b/t/test-lib.sh
@@ -714,6 +714,7 @@ case $(uname -s) in
esac
test -z "$NO_PERL" && test_set_prereq PERL
+test -z "$NO_PYTHON" && test_set_prereq PYTHON
# test whether the filesystem supports symbolic links
ln -s x y 2>/dev/null && test -h y 2>/dev/null && test_set_prereq SYMLINKS
--
1.6.4.262.gca66a.dirty
^ permalink raw reply related
* [RFCv4 2/5] 1/2: Add Python support library for CVS remote helper
From: Johan Herland @ 2009-08-17 3:35 UTC (permalink / raw)
To: git; +Cc: Johannes.Schindelin, Johan Herland, barkalow, davvid, gitster
In-Reply-To: <1250480161-21933-1-git-send-email-johan@herland.net>
This patch introduces parts of a Python package called "git_remote_cvs"
containing the building blocks of the CVS remote helper.
The CVS remote helper itself is NOT part of this patch.
This patch has been improved by the following contributions:
- David Aguilar: Lots of Python coding style fixes
Cc: David Aguilar <davvid@gmail.com>
Signed-off-by: Johan Herland <johan@herland.net>
---
This patch and the next patch belong together in the same commit, but
the 100K limit (?) on the git mailing list forces me to split this
patch in two.
...Johan
git_remote_cvs/changeset.py | 126 +++++
git_remote_cvs/cvs.py | 998 ++++++++++++++++++++++++++++++++++++
git_remote_cvs/cvs_symbol_cache.py | 313 +++++++++++
git_remote_cvs/util.py | 186 +++++++
4 files changed, 1623 insertions(+), 0 deletions(-)
create mode 100644 git_remote_cvs/changeset.py
create mode 100644 git_remote_cvs/cvs.py
create mode 100644 git_remote_cvs/cvs_symbol_cache.py
create mode 100644 git_remote_cvs/util.py
diff --git a/git_remote_cvs/changeset.py b/git_remote_cvs/changeset.py
new file mode 100644
index 0000000..9eea9d2
--- /dev/null
+++ b/git_remote_cvs/changeset.py
@@ -0,0 +1,126 @@
+#!/usr/bin/env python
+
+"""Code for collecting individual CVS revisions into "changesets"
+
+A changeset is a collection of CVSRev objects that belong together in
+the same "commit". This is a somewhat artificial construct on top of
+CVS, which only stores changes at the per-file level. Normally, CVS
+users create several CVS revisions simultaneously by applying the
+"cvs commit" command to several files with related changes. This
+module tries to reconstruct this notion of related revisions.
+
+"""
+
+from git_remote_cvs.util import debug, error, die
+
+
+class Changeset(object):
+
+ """Encapsulate a single changeset/commit."""
+
+ __slots__ = ('revs', 'date', 'author', 'message')
+
+ # The maximum time between the changeset's date, and the date of a
+ # rev to included in that changeset.
+ MaxSecondsBetweenRevs = 8 * 60 * 60 # 8 hours
+
+ @classmethod
+ def from_rev (cls, rev):
+ """Return a Changeset based on the given CVSRev object."""
+ c = cls(rev.date, rev.author, rev.message)
+ result = c.add(rev)
+ assert result
+ return c
+
+ def __init__ (self, date, author, message):
+ """Create a new Changeset with the given metadata."""
+ self.revs = {} # dict: path -> CVSRev object
+ self.date = date # CVSDate object
+ self.author = author
+ self.message = message # Lines of commit message
+
+ def __str__ (self):
+ """Stringify this Changeset object."""
+ msg = self.message[0] # First line only
+ # Limit message to 25 chars
+ if len(msg) > 25:
+ msg = msg[:22] + "..."
+ return ("<Changeset @(%s) by %s (%s) updating %i files>" %
+ (self.date, self.author, msg, len(self.revs)))
+
+ def __iter__ (self):
+ """Return iterator traversing the CVSRevs in this Changeset."""
+ return self.revs.itervalues()
+
+ def __getitem__ (self, path):
+ """Look up a specific CVSRev in this Changeset."""
+ return self.revs[path]
+
+ def within_time_window (self, rev):
+ """Return True iff the rev is within the time window of self."""
+ return abs(rev.date.diff(self.date)) <= self.MaxSecondsBetweenRevs
+
+ def add (self, rev):
+ """Add the given CVSRev to this Changeset.
+
+ The addition will only succeed if the following holds:
+ - rev.author == self.author
+ - rev.message == self.message
+ - rev.path is not in self.revs
+ - rev.date is within MaxSecondsBetweenRevs of self.date
+ If the addition succeeds, True is returned; otherwise False.
+
+ """
+ if rev.author != self.author or \
+ rev.message != self.message or \
+ rev.path in self.revs or \
+ not self.within_time_window(rev):
+ return False
+
+ self.revs[rev.path] = rev
+ return True
+
+
+def build_changesets_from_revs (cvs_revs):
+ """Organize CVSRev objects into a chronological list of Changesets."""
+ # Construct chronological list of CVSRev objects
+ chron_revs = []
+ for path, d in cvs_revs.iteritems():
+ i = 0 # Current index into chronRevs
+ for revnum, cvsrev in sorted(d.iteritems()):
+ assert path == cvsrev.path
+ assert revnum == cvsrev.num
+ while i < len(chron_revs) and cvsrev.date > chron_revs[i].date:
+ i += 1
+ # Insert cvsRev at position i in chronRevs
+ chron_revs.insert(i, cvsrev)
+ i += 1
+
+ changesets = [] # Chronological list of Changeset objects
+ while len(chron_revs):
+ # There are still more revs to be added to Changesets
+ # Create Changeset based on the first rev in chronRevs
+ changeset = Changeset.from_rev(chron_revs.pop(0))
+ # Keep adding revs chronologically until MaxSecondsBetweenRevs
+ rejects = [] # Revs that cannot be added to this changeset
+ while len(chron_revs):
+ rev = chron_revs.pop(0)
+ reject = False
+ # First, if we have one of rev's parents in rejects, we
+ # must also reject rev
+ for r in rejects:
+ if r.path == rev.path:
+ reject = True
+ break
+ # Next, add rev to changeset, reject if add fails
+ if not reject:
+ reject = not changeset.add(rev)
+ if reject:
+ rejects.append(rev)
+ # stop trying when rev is too far in the future
+ if not changeset.within_time_window(rev):
+ break
+ chron_revs = rejects + chron_revs # Reconstruct remaining revs
+ changesets.append(changeset)
+
+ return changesets
diff --git a/git_remote_cvs/cvs.py b/git_remote_cvs/cvs.py
new file mode 100644
index 0000000..f870ae0
--- /dev/null
+++ b/git_remote_cvs/cvs.py
@@ -0,0 +1,998 @@
+#!/usr/bin/env python
+
+"""Functionality for interacting with CVS repositories.
+
+This module provides classes for interrogating a CVS repository via a
+CVS working directory (aka. checkout), or via direct queries using the
+"cvs rlog" command.
+
+Also, classes for encapsulating fundamental CVS concepts (like CVS
+revision/branch numbers) are provided.
+"""
+
+import sys
+import os
+import shutil
+import time
+from calendar import timegm
+import unittest
+
+from git_remote_cvs.util import (debug, error, die, ProgressIndicator,
+ start_command, run_command,
+ file_reader_method, file_writer_method)
+
+
+class CVSNum(object):
+
+ """Encapsulate a single CVS revision/branch number.
+
+ Provides functionality for common operations on CVS numbers.
+
+ A CVS number consists of a list of components separated by periods
+ ('.'), where each component is a decimal number. Inspecting the
+ components from left to right, the odd-numbered (1st, 3rd, 5th,
+ etc.) components represent branches in the CVS history tree, while
+ the even-numbered (2nd, 4th, 6th, etc.) components represent
+ revisions on the branch specified in the previous position.
+ Thus "1.2" denotes the second revision on the first branch
+ (aka. trunk), while "1.2.4.6" denotes the sixth revision of the
+ fourth branch started from revision "1.2".
+
+ Therefore, in general, a CVS number with an even number of
+ components denotes a revision (we call this a "revision number"),
+ while an odd number of components denotes a branch (called a
+ "branch number").
+
+ There are a few complicating peculiarities: If there is an even
+ number of components, and the second-last component is 0, the
+ number is not a revision number, but is rather equivalent to the
+ branch number we get by removing the 0-component. I.e. "1.2.0.4"
+ is equivalent to "1.2.4".
+
+ A branch number (except the trunk: "1") always has a "branch point"
+ revision, i.e. the revision from which the branch was started.
+ This revision is found by removing the last component of the branch
+ number. For example the branch point of "1.2.4" is "1.2".
+
+ Conversely, all revision numbers belong to a corresponding branch,
+ whose branch number is found by removing the last component.
+ Examples: The "1.2.4.6" revision belong to the "1.2.4" branch,
+ while the "1.2" revision belongs to the "1" branch (the "trunk").
+
+ From this we can programatically determine the ancestry of any
+ revision number, by decrementing the last revision component until
+ it equals 1, and then trim off the last two components to get to
+ the branch point, and repeat the process from there until we reach
+ the initial revision (typically "1.1"). For example, recursively
+ enumerating the parent revisions of "1.2.4.6" yields the following
+ revisions:
+ "1.2.4.5", "1.2.4.4", "1.2.4.3", "1.2.4.2", "1.2.4.1", "1.2", "1.1"
+
+ """
+
+ __slots__ = ('c',)
+
+ @staticmethod
+ def decompose (cvsnum):
+ """Split the given CVS number into a list of int components.
+
+ Branch numbers are normalized to the odd-numbered components
+ form (i.e. removing the second last '0' component)
+
+ Examples:
+ '1.2.4.8' -> [1, 2, 4, 8]
+ '1.2.3' -> [1, 2, 3]
+ '1.2.0.5' -> [1, 2, 5]
+
+ """
+ if cvsnum:
+ r = map(int, cvsnum.split('.'))
+ else:
+ r = []
+ if len(r) >= 2 and r[-2] == 0:
+ del r[-2]
+ if r[-1] == 0:
+ raise ValueError(cvsnum)
+ return tuple(r)
+
+ @staticmethod
+ def compose (c):
+ """Join the given list of integer components into a CVS number.
+
+ E.g.: (1, 2, 4, 8) -> '1.2.4.8'
+
+ """
+ if c[-1] == 0:
+ raise ValueError(str(c))
+ return ".".join(map(str, c))
+
+ @classmethod
+ def from_components (cls, args):
+ """Create a CVSNum from the given list of numerical components."""
+ return cls(cls.compose(args))
+
+ @classmethod
+ def disjoint (cls, a, b):
+ """Return True iff the CVS numbers are historically disjoint.
+
+ Two CVS numbers are disjoint if they do not share the same
+ historical line back to the initial revision. In other words:
+ the two numbers are disjoint if the history (i.e. set of parent
+ revisions all the way back to the intial (1.1) revision) of
+ neither number is a superset of the other's history.
+ See test_disjoint() for practical examples:
+
+ """
+ if a.is_branch():
+ a = cls.from_components(a.c + (1,))
+ if b.is_branch():
+ b = cls.from_components(b.c + (1,))
+ if len(a.c) > len(b.c):
+ a, b = b, a # a is now shortest
+ pairs = zip(a.c, b.c)
+ for pa, pb in pairs[:-1]:
+ if pa != pb:
+ return True
+ if len(a) == len(b):
+ return False
+ common_len = len(a)
+ if a.c[common_len - 1] <= b.c[common_len - 1]:
+ return False
+ return True
+
+
+ def __init__ (self, cvsnum):
+ """Create a CVSNum object from the given CVS number string."""
+ self.c = self.decompose(str(cvsnum))
+
+ def __repr__ (self):
+ """Return a string representation of this object."""
+ return self.compose(self.c)
+
+ def __str__ (self):
+ """Return a string representation of this object."""
+ return repr(self)
+
+ def __hash__ (self):
+ """Create a hash value for this CVS number."""
+ return hash(repr(self))
+
+ def __len__ (self):
+ """Return number of components in this CVS number."""
+ return len(self.c)
+
+ def __cmp__ (self, other):
+ """Comparison method for CVS numbers."""
+ try:
+ return cmp(self.c, other.c)
+ except AttributeError:
+ return 1
+
+ def __getitem__ (self, key):
+ """Return the Xth component of this CVS number."""
+ return self.c[key]
+
+ def is_rev (self):
+ """Return True iff this number is a CVS revision number."""
+ return len(self.c) % 2 == 0 and len(self.c) >= 2 and self.c[-2] != 0
+
+ def is_branch (self):
+ """Return True iff this number is a CVS branch number."""
+ return len(self.c) % 2 != 0 or (len(self.c) >= 2 and self.c[-2] == 0)
+
+ def components (self):
+ """Return a list of integer components in this CVS number."""
+ return list(self.c)
+
+ def branch (self):
+ """Return the branch on which the given number lives.
+
+ Revisions: chop the last component to find the branch, e.g.:
+ 1.2.4.6 -> 1.2.4
+ 1.1 -> 1
+ Branches: unchanged
+
+ """
+ if self.is_rev():
+ return self.from_components(self.c[:-1])
+ return self
+
+ def parent (self):
+ """Return the parent/previous revision number to this number.
+
+ For revisions, this is the previous revision, e.g.:
+ 1.2.4.6 -> 1.2.4.5
+ 1.2.4.1 -> 1.2
+ 1.1 -> None
+ 2.1 -> None
+ For branches, this is the branch point, e.g.:
+ 1.2.4 -> 1.2
+ 1 -> None
+ 2 -> None
+
+ """
+ if len(self.c) < 2:
+ return None
+ elif len(self.c) % 2: # Branch number
+ return self.from_components(self.c[:-1])
+ else: # Revision number
+ assert self.c[-1] > 0
+ result = self.components()
+ result[-1] -= 1 # Decrement final component
+ if result[-1] == 0: # We're at the start of the branch
+ del result[-2:] # Make into branch point
+ if not result:
+ return None
+ return self.from_components(result)
+
+ def follows (self, other):
+ """Return True iff self historically follows the given rev.
+
+ This iterates through the parents of self, and returns True iff
+ any of them equals the given rev. Otherwise, it returns False.
+
+ """
+ assert other.is_rev()
+ cur = self
+ while cur:
+ if cur == other:
+ return True
+ cur = cur.parent()
+ return False
+
+ def on_branch (self, branch):
+ """Return True iff this rev is on the given branch.
+
+ The revs considered to be "on" a branch X also includes the
+ branch point of branch X.
+
+ """
+ return branch == self.branch() or branch.parent() == self
+
+
+class TestCVSNum(unittest.TestCase):
+
+ """CVSNum selftests."""
+
+ def test_basic (self):
+ """CVSNum basic selftests."""
+ self.assertEqual(CVSNum("1.2.4"), CVSNum("1.2.0.4"))
+ self.assert_(CVSNum("1.2.4").is_branch())
+ self.assert_(CVSNum("1.2").is_rev())
+ self.assert_(CVSNum("1").is_branch())
+ self.assert_(CVSNum("1.2.4.6").is_rev())
+ self.assertEqual(CVSNum("1.2.4.6").components(), [1, 2, 4, 6])
+ self.assertEqual(CVSNum.from_components([1, 2, 4, 6]),
+ CVSNum("1.2.4.6"))
+ self.assertEqual(str(CVSNum.from_components([1, 2, 4, 6])), "1.2.4.6")
+ self.assertEqual(len(CVSNum("1.2.4.6")), 4)
+ self.assertEqual(CVSNum("1.2.4.6").branch(), CVSNum("1.2.4"))
+ self.assertEqual(CVSNum("1.2.4").branch(), CVSNum("1.2.4"))
+ self.assertEqual(CVSNum("1.1").branch(), CVSNum("1"))
+ self.assertEqual(CVSNum("1").branch(), CVSNum("1"))
+ self.assertEqual(CVSNum("1.2.4.6").parent(), CVSNum("1.2.4.5"))
+ self.assertEqual(CVSNum("1.2.4.1").parent(), CVSNum("1.2"))
+ self.assertEqual(CVSNum("1.2").parent(), CVSNum("1.1"))
+ self.assert_(CVSNum("1.1").parent() is None)
+ self.assert_(CVSNum("2.1").parent() is None)
+ self.assertEqual(CVSNum("1.2.4").parent(), CVSNum("1.2"))
+ self.assert_(CVSNum("1").parent() is None)
+ self.assert_(CVSNum("2").parent() is None)
+
+ def test_follows (self):
+ """CVSNum.follows() selftests."""
+ self.assert_(CVSNum("1.2.4.6").follows(CVSNum("1.1")))
+ self.assert_(CVSNum("1.2.4.6").follows(CVSNum("1.2")))
+ self.assert_(CVSNum("1.2.4.6").follows(CVSNum("1.2.4.1")))
+ self.assert_(CVSNum("1.2.4.6").follows(CVSNum("1.2.4.2")))
+ self.assert_(CVSNum("1.2.4.6").follows(CVSNum("1.2.4.3")))
+ self.assert_(CVSNum("1.2.4.6").follows(CVSNum("1.2.4.4")))
+ self.assert_(CVSNum("1.2.4.6").follows(CVSNum("1.2.4.5")))
+ self.assert_(CVSNum("1.2.4.6").follows(CVSNum("1.2.4.6")))
+ self.assertFalse(CVSNum("1.2.4.6").follows(CVSNum("1.2.4.7")))
+ self.assertFalse(CVSNum("1.2.4.6").follows(CVSNum("1.3")))
+ self.assertFalse(CVSNum("1.1").follows(CVSNum("1.2.4.6")))
+
+ def test_disjoint (self):
+ """CVSNum.disjoint() selftests."""
+ tests = [
+ ("1.2", "1.1", False),
+ ("1.2", "1.2", False),
+ ("1.2", "1.3", False),
+ ("1.2", "1.1.2", True),
+ ("1.2", "1.1.2.3", True),
+ ("1.2.4", "1.1", False),
+ ("1.2.4", "1.2", False),
+ ("1.2.4", "1.3", True),
+ ("1.2.4", "1.2.2", True),
+ ("1.2.4", "1.2.4", False),
+ ("1.2.4", "1.2.6", True),
+ ("1.2.4", "1.2.2.4", True),
+ ("1.2.4", "1.2.4.4", False),
+ ("1.2.4", "1.2.6.4", True),
+ ("1.2.4.6", "1.1", False),
+ ("1.2.4.6", "1.2", False),
+ ("1.2.4.6", "1.3", True),
+ ("1.2.4.6", "1.2.2", True),
+ ("1.2.4.6", "1.2.2.1", True),
+ ("1.2.4.6", "1.2.4", False),
+ ("1.2.4.6", "1.2.4.5", False),
+ ("1.2.4.6", "1.2.4.6", False),
+ ("1.2.4.6", "1.2.4.7", False),
+ ("1.2.4.6.8.10", "1.2.4.5", False),
+ ("1.2.4.6.8.10", "1.2.4.6", False),
+ ("1.2.4.6.8.10", "1.2.4.7", True),
+ ]
+ for a, b, result in tests:
+ self.assertEqual(CVSNum.disjoint(CVSNum(a), CVSNum(b)), result)
+ self.assertEqual(CVSNum.disjoint(CVSNum(b), CVSNum(a)), result)
+
+
+class CVSState(object):
+
+ """Encapsulate a historical state in CVS (a set of paths and nums).
+
+ This class is a container of CVS pathnames and associated CVSNum
+ objects.
+
+ No communication with a CVS working directory or repository is done
+ in this class, hence only basic sanity checks are performed:
+ - A path may only appear once in a CVSState.
+ - When adding a path:num pair, path may not already exist in self
+ - When replacing a path:num pair, path must already exist in self
+ - When removing a path:num pair, both path and num must be given
+
+ IMPORTANT: Objects of this class are hash()able (to support being
+ used as keys in a dict), but they are also mutable. It is
+ therefore up to the caller to make sure that the object is not
+ changed after being stored in a data structure indexed by its hash
+ value.
+
+ """
+
+ __slots__ = ('revs', '_hash')
+
+ def __init__ (self):
+ """Create a new, empty CVSState."""
+ self.revs = {} # dict: path -> CVSNum object
+ self._hash = None
+
+ def __iter__ (self):
+ """Return iterator traversing the (path, CVSNum)s in this CVSState."""
+ return self.revs.iteritems()
+
+ def __cmp__ (self, other):
+ """Comparison method for CVSState objects."""
+ return cmp(self.revs, other.revs)
+
+ def __str__ (self):
+ """Stringify this CVSState by listing the contained revisions."""
+ return "".join(["%s:%s\n" % (p, n) for p, n in sorted(self)])
+
+ def __hash__ (self):
+ """Create a hash value for this CVSState."""
+ if self._hash is None:
+ self._hash = hash(str(self))
+ return self._hash
+
+ def __getitem__ (self, path):
+ """Return the CVSNum associated with the given path in self."""
+ return self.revs[path]
+
+ def get (self, path, default = None):
+ """Return the CVSNum associated with the given path in self."""
+ return self.revs.get(path, default)
+
+ def paths (self):
+ """Return the path names contained within this CVSState."""
+ return self.revs.iterkeys()
+
+ def add (self, path, revnum):
+ """Add the given path:revnum to this CVSState."""
+ assert path not in self.revs
+ self._hash = None
+ self.revs[path] = revnum
+
+ def replace (self, path, revnum):
+ """Replace the revnum associated with the given path."""
+ assert path in self.revs
+ self._hash = None
+ self.revs[path] = revnum
+
+ def remove (self, path, revnum):
+ """Remove the given path:revnum association from this CVSState."""
+ assert path in self.revs and self.revs[path] == revnum
+ self._hash = None
+ del self.revs[path]
+
+ def copy (self):
+ """Create and return a copy of this object."""
+ ret = CVSState()
+ ret.revs = self.revs.copy()
+ ret._hash = self._hash
+ return ret
+
+ def load_data (self, note_data):
+ """Load note data as formatted by self.__str__()."""
+ for line in note_data.split("\n"):
+ line = line.strip()
+ if not line:
+ continue
+ path, num = line.rsplit(':', 1)
+ self.add(path, CVSNum(num))
+ self._hash = hash(note_data)
+
+ def print_members (self, f = sys.stdout, prefix = ""):
+ """Write the members of this CVSState to the given file object."""
+ for path, num in sorted(self):
+ print >> f, "%s%s:%s" % (prefix, path, num)
+
+ @file_reader_method(missing_ok = True)
+ def load (self, f):
+ """Load CVS state from the given file name/object."""
+ if f:
+ self.load_data(f.read())
+
+ @file_writer_method
+ def save (self, f):
+ """Save CVS state to the given file name/object."""
+ assert f
+ print >> f, str(self),
+
+
+class CVSDate(object):
+
+ """Encapsulate a timestamp, as reported by CVS.
+
+ The internal representation of a timestamp is two integers, the
+ first representing the timestamp as #seconds since epoch (UTC),
+ and the second representing the timezone as #minutes offset from
+ UTC.
+
+ Example: "2007-09-05 17:26:28 -0200" is converted to
+ (1189013188, -120)
+
+ """
+
+ __slots__ = ('ts', 'tz')
+
+ def __init__ (self, date_str = None, in_utc = False):
+ """Convert CVS date string into a CVSDate object.
+
+ A CVS timestamp string has one of the following forms:
+ - "YYYY-MM-DD hh:mm:ss SZZZZ"
+ - "YYYY/MM/DD hh:mm:ss" (with timezone assumed to be UTC)
+ The in_utc parameter determines whether the timestamp part of
+ the given string (the "YYYY-MM-DD hh:mm:ss" part) is given in
+ local time or UTC (normally CVS dates are given in local time.
+ If given in local time, the timezone offset is subtracted from
+ the timestamp in order to make the time in UTC format.
+
+ """
+ if date_str is None:
+ self.ts, self.tz = 0, 0
+ return
+ if date_str == "now":
+ self.ts, self.tz = time.time(), 0
+ return
+ date_str = date_str.strip()
+ # Set up self.ts and self.tz
+ if date_str.count(" ") == 2:
+ # Assume format "YYYY-MM-DD hh:mm:ss SZZZZ"
+ t, z = date_str.rsplit(" ", 1)
+ # Convert timestamp to #secs since epoch (UTC)
+ self.ts = timegm(time.strptime(t, "%Y-%m-%d %H:%M:%S"))
+ # Convert timezone into #mins offset from UTC
+ self.tz = int(z[1:3]) * 60 + int(z[3:5])
+ # Incorporate timezone sign
+ if z[0] == '-':
+ self.tz *= -1
+ else:
+ assert date_str.count(" ") == 1
+ # Assume format "YYYY/MM/DD hh:mm:ss"
+ self.ts = timegm(time.strptime(date_str, "%Y/%m/%d %H:%M:%S"))
+ self.tz = 0
+ # Adjust timestamp if not already in UTC
+ if not in_utc:
+ self.ts -= self.tz * 60
+
+ def tz_str (self):
+ """Return timezone part of self in string format."""
+ sign = '+'
+ if self.tz < 0:
+ sign = '-'
+ hours, minutes = divmod(abs(self.tz), 60)
+ return "%s%02d%02d" % (sign, hours, minutes)
+
+ def __str__ (self):
+ """Reconstruct date string from members."""
+ s = time.strftime("%Y-%m-%d %H:%M:%S", time.gmtime(self.ts))
+ return "%s %s" % (s, self.tz_str())
+
+ def __repr__ (self):
+ """Create a string representation of self."""
+ return "CVSDate('%s')" % (str(self))
+
+ def __hash__ (self):
+ """Create a hash value from self."""
+ return hash((self.ts, self.tz))
+
+ def __nonzero__ (self):
+ """Provide interpretation of self in a boolean context."""
+ return bool(self.ts or self.tz)
+
+ def __cmp__ (self, other):
+ """Comparison method for CVSDate objects."""
+ return cmp(self.ts, other.ts) or cmp(self.tz, other.tz)
+
+ def __eq__ (self, other):
+ """Return True iff self and other is considered equal."""
+ return self.ts == other.ts and self.tz == other.tz
+
+ def diff (self, other):
+ """Return difference between self and other in #seconds.
+
+ Invariant: self == other.add(self.diff(other))
+
+ """
+ return self.ts - other.ts
+
+
+class TestCVSDate(unittest.TestCase):
+
+ """CVSDate selftests."""
+
+ def test_basic (self):
+ """CVSDate basic selftests."""
+ a = CVSDate("2009-05-10 14:34:56 +0200")
+ b = CVSDate("2009/05/10 12:34:56")
+ self.assert_(a)
+ self.assert_(b)
+ self.assertEqual(str(a), "2009-05-10 12:34:56 +0200", str(a))
+ self.assertEqual(str(b), "2009-05-10 12:34:56 +0000", str(b))
+ self.assertNotEqual(a, b)
+ self.assertEqual(a.diff(b), 0)
+ c = CVSDate("2009-05-10 16:34:56 +0200")
+ self.assert_(c)
+ self.assertEqual(str(c), "2009-05-10 14:34:56 +0200", str(c))
+ self.assertNotEqual(c, a)
+ self.assertEqual(c.diff(a), 2 * 60 * 60)
+ self.assertEqual(a.diff(c), -2 * 60 * 60)
+
+
+class CVSRev(object):
+
+ """Encapsulate metadata on a CVS revision."""
+
+ __slots__ = ('path', 'num', 'date', 'author', 'deleted', 'message')
+
+ def __init__ (self, path, num):
+ """Create a CVSRev object for the given path:num revision."""
+ self.path = path
+ self.num = num
+ self.date = None # CVSDate object
+ self.author = ""
+ self.deleted = None # True or False
+ self.message = [] # Lines of commit message
+
+ def __str__ (self):
+ """Return a string listing the metadata in this CVS revision."""
+ return ("<%s:%s on %s by %s%s>" %
+ (self.path, self.num, self.date, self.author,
+ self.deleted and ", deleted" or ""))
+
+ def __cmp__ (self, other):
+ """Comparison method for CVSRev objects."""
+ return cmp(self.path, other.path) or cmp(self.num, other.num)
+
+
+class CVSWorkDir(object):
+
+ """Encapsulate a CVS working directory.
+
+ This class auto-creates a CVS workdir/checkout in the directory
+ given to the constructor, and provides various methods for
+ interacting with this workdir.
+
+ """
+
+ def __init__ (self, workdir, cvs_repo):
+ """Create a new CVSWorkDir.
+
+ The cvs_repo argument must be a (cvs_root, cvs_module) tuple
+
+ """
+ self.d = workdir
+ self.cvs_root, self.cvs_module = cvs_repo
+ parent_dir = os.path.dirname(self.d)
+ if not os.path.isdir(parent_dir):
+ os.makedirs(parent_dir)
+ self._valid = None
+
+ def makepath(self, *args):
+ """Create path relative to working directory."""
+ return os.path.join(self.d, *args)
+
+ def valid (self):
+ """Return True iff this workdir is present and valid."""
+ if self._valid is not None:
+ return self._valid
+ try:
+ f = open(self.makepath("CVS", "Root"), 'r')
+ assert f.read().strip() == self.cvs_root
+ f.close()
+ f = open(self.makepath("CVS", "Repository"), 'r')
+ assert f.read().strip() == self.cvs_module
+ f.close()
+ self._valid = True
+ except (IOError, AssertionError):
+ self._valid = False
+ return self._valid
+
+ def remove (self):
+ """Remove this checkout."""
+ shutil.rmtree(self.d, True)
+ assert not os.path.exists(self.d)
+ self._valid = False
+
+ def checkout (self, revision = "HEAD"):
+ """Create a checkout of the given revision."""
+ self.remove()
+ parent_dir, co_dir = os.path.split(self.d)
+ args = ["cvs", "-f", "-Q", "-d", self.cvs_root, "checkout"]
+ if str(revision) != "HEAD":
+ args.extend(["-r", str(revision)])
+ args.extend(["-d", co_dir, self.cvs_module])
+ exit_code, output, errors = run_command(args, cwd = parent_dir)
+ if exit_code:
+ die("Failed to checkout CVS working directory")
+ assert not errors
+ assert not output, "output = '%s'" % (output)
+ self._valid = None
+ assert self.valid()
+
+ def update (self, revision = "HEAD", paths = None):
+ """Update the given paths to the given revision."""
+ if not self.valid():
+ self.checkout()
+ args = ["cvs", "-f", "-Q", "update", "-kk"]
+ if str(revision) == "HEAD":
+ args.append("-A")
+ else:
+ args.extend(["-r", str(revision)])
+ if paths is not None:
+ args.extend(paths)
+ exit_code, output, errors = run_command(args, cwd = self.d)
+ if exit_code:
+ die("Failed to checkout CVS working directory")
+ assert not errors
+ assert not output, "output = '%s'" % (output)
+
+ def get_revision_data (self, path, revision):
+ """Return the contents of the given CVS path:revision."""
+ if not self.valid():
+ self.checkout()
+ args = ["cvs", "-f", "-Q", "update", "-p", "-kk"]
+ if str(revision) == "HEAD":
+ args.append("-A")
+ else:
+ args.extend(["-r", str(revision)])
+ args.append(path)
+ exit_code, output, errors = run_command(args, cwd = self.d)
+ if exit_code:
+ die("Failed to checkout CVS working directory")
+ assert not errors
+ return output
+
+ def get_modeinfo (self, paths = None):
+ """Return mode information for the given paths.
+
+ Returns a dict of path -> mode number mappings. If paths are
+ not specified, mode information for all files in the current
+ checkout will be returned. No checkout/update will be done.
+
+ """
+ result = {}
+ if paths is not None:
+ for path in paths:
+ fullpath = os.path.join(self.d, path)
+ mode = 644
+ if os.access(fullpath, os.X_OK):
+ mode = 755
+ assert path not in result
+ result[path] = mode
+ else: # Return mode information for all paths
+ for dirpath, dirnames, filenames in os.walk(self.d):
+ # Don't descend into CVS subdirs
+ try:
+ dirnames.remove('CVS')
+ except ValueError:
+ pass
+ assert dirpath.startswith(self.d)
+ directory = dirpath[len(self.d):].lstrip("/")
+ for fname in filenames:
+ path = os.path.join(directory, fname)
+ fullpath = os.path.join(dirpath, fname)
+ mode = 644
+ if os.access(fullpath, os.X_OK):
+ mode = 755
+ assert path not in result
+ result[path] = mode
+ return result
+
+ @classmethod
+ def parse_entries (cls, entries, prefix, directory = ""):
+ """Recursively parse CVS/Entries files.
+
+ Return a dict of CVS paths found by parsing the CVS/Entries
+ files rooted at the given directory.
+
+ See http://ximbiot.com/cvs/manual/feature/cvs_2.html#SEC19 for
+ information on the format of the CVS/Entries file.
+
+ """
+ fname = os.path.join(prefix, directory, "CVS", "Entries")
+ subdirs = []
+ f = open(fname, 'r')
+ for line in f:
+ line = line.strip()
+ if line == "D":
+ continue # There are no subdirectories
+ t, path, revnum, date, options, tag = line.split("/")
+ if t == "D":
+ subdirs.append(path)
+ continue
+ assert line.startswith("/")
+ path = os.path.join(directory, path)
+ revnum = CVSNum(revnum)
+ assert path not in entries
+ entries[path] = (revnum, date, options, tag)
+ f.close()
+ for d in subdirs:
+ d = os.path.join(directory, d)
+ cls.parse_entries(entries, prefix, d)
+
+ def get_state (self):
+ """Return CVSState reflecting current state of this checkout.
+
+ Note that the resulting CVSState will never contain any
+ deleted/dead files. Other CVSStates to be compared to the one
+ returned from here should remove deleted/dead entries first.
+
+ """
+ assert self.valid()
+ entries = {}
+ result = CVSState()
+ self.parse_entries(entries, self.d)
+ for path, info in entries.iteritems():
+ result.add(path, info[0])
+ return result
+
+
+class CVSLogParser(object):
+
+ """Encapsulate the execution of a "cvs rlog" command."""
+
+ def __init__ (self, cvs_repo):
+ """Create a new CVSLogParser.
+
+ The cvs_repo argument must be a (cvs_root, cvs_module) tuple
+
+ """
+ self.cvs_root, self.cvs_module = cvs_repo
+
+ def cleanup_path (self, cvs_path):
+ """Utility method for parsing CVS paths from CVS log."""
+ cvsprefix = "/".join((self.cvs_root[self.cvs_root.index("/"):],
+ self.cvs_module))
+ assert cvs_path.startswith(cvsprefix)
+ assert cvs_path.endswith(",v")
+ # Drop cvsprefix and ,v-extension
+ cvs_path = cvs_path[len(cvsprefix):-2]
+ # Split the remaining path into components
+ path_comps = filter(None, cvs_path.strip().split('/'))
+ # Remove 'Attic' from CVS paths
+ if len(path_comps) >= 2 and path_comps[-2] == "Attic":
+ del path_comps[-2]
+ # Reconstruct resulting "cleaned" path
+ return "/".join(path_comps)
+
+ def __call__ (self, line):
+ """Parse the given line from the CVS log.
+
+ Must be reimplemented by subclass
+
+ """
+ pass
+
+ def finish (self):
+ """This method is invoked after the last line has been parsed.
+
+ May be reimplemented by subclass
+
+ """
+ pass
+
+ def run (self, paths = None, no_symbols = False, revisions = None):
+ """Execute "cvs rlog" with the given arguments.
+
+ self.__call__() is invoked once for each line in the CVS log.
+ self.finish() is invoked exactly once after the CVS log.
+
+ """
+ args = ["cvs", "-f", "-q", "-d", self.cvs_root, "rlog"]
+ if no_symbols:
+ args.append("-N")
+ if revisions:
+ args.append("-r%s" % (revisions))
+ if paths is not None:
+ for p in paths:
+ args.append("%s/%s" % (self.cvs_module, p))
+ else:
+ args.append(self.cvs_module)
+ proc = start_command(args)
+ proc.stdin.close()
+ while True:
+ for line in proc.stdout:
+ self(line.rstrip()) # Call self's line parser
+ if proc.poll() is not None:
+ break
+ assert proc.stdout.read() == ""
+ self.finish() # Notify subclass that parsing is finished
+ exit_code = proc.returncode
+ if exit_code:
+ error("'%s' returned exit code %i, and errors:\n---\n%s---",
+ " ".join(args), exit_code, proc.stderr.read())
+ return exit_code
+
+
+class CVSRevLister(CVSLogParser):
+
+ """Extract CVSRev objects (with revision metadata) from a CVS log."""
+
+ def __init__ (self, cvs_repo, show_progress = False):
+ """Create a new CVSRevLister.
+
+ The cvs_repo argument must be a (cvs_root, cvs_module) tuple
+ show_progress determines whether progress indication is shown.
+
+ """
+ super(CVSRevLister, self).__init__(cvs_repo)
+ self.cur_file = None # Current CVS file being processed
+ self.cur_file_numrevs = 0 # #revs in current CVS file
+ self.cur_rev = None # Current CVSRev under construction
+ self.progress = None
+ if show_progress:
+ self.progress = ProgressIndicator("\t", sys.stderr)
+ # Store found revs in a two-level dict structure:
+ # filename -> revnum -> CVSRev
+ self.revs = {}
+ # Possible states:
+ # - BeforeRevs - waiting for "total revisions:"
+ # - BetweenRevs - waiting for "----------------------------"
+ # - ReadingRev - reading CVS revision details
+ self.state = 'BeforeRevs'
+
+ def __call__ (self, line):
+ """Line parser; this method is invoked for each line in the log."""
+ if self.state == 'BeforeRevs':
+ if line.startswith("RCS file: "):
+ self.cur_file = self.cleanup_path(line[10:])
+ assert self.cur_file not in self.revs
+ self.revs[self.cur_file] = {}
+ elif line.startswith("total revisions: "):
+ assert self.cur_file
+ totalrevs_unused, selectedrevs = line.split(";")
+ self.cur_file_numrevs = int(selectedrevs.split(":")[1].strip())
+ self.state = 'BetweenRevs'
+ elif self.state == 'BetweenRevs':
+ if (line == "----------------------------" or
+ line == "======================================"
+ "======================================="):
+ if self.cur_rev:
+ # Finished current revision
+ f = self.revs[self.cur_file]
+ assert self.cur_rev.num not in f
+ f[self.cur_rev.num] = self.cur_rev
+ self.cur_rev = None
+ if self.progress:
+ self.progress()
+ if line == "----------------------------":
+ self.state = 'ReadingRev'
+ else:
+ # Finalize current CVS file
+ assert len(self.revs[self.cur_file]) == \
+ self.cur_file_numrevs
+ self.cur_file = None
+ self.state = 'BeforeRevs'
+ elif self.cur_rev:
+ # Currently in the middle of a revision.
+ if line.startswith("branches: %s" % (self.cur_rev.num)) and \
+ line.endswith(";"):
+ return # Skip 'branches:' lines
+ # This line is part of the commit message.
+ self.cur_rev.message.append(line)
+ elif self.state == 'ReadingRev':
+ if line.startswith("revision "):
+ self.cur_rev = CVSRev(self.cur_file, CVSNum(line.split()[1]))
+ else:
+ date, author, state, dummy = line.split(";", 3)
+ assert date.startswith("date: ")
+ self.cur_rev.date = CVSDate(date[6:])
+ assert author.strip().startswith("author: ")
+ self.cur_rev.author = author.strip()[8:]
+ assert state.strip().startswith("state: ")
+ state = state.strip()[7:]
+ self.cur_rev.deleted = state == "dead"
+ self.state = 'BetweenRevs'
+
+ def finish (self):
+ """This method is invoked after the last line has been parsed."""
+ assert self.state == 'BeforeRevs'
+ if self.progress:
+ self.progress.finish("Parsed %i revs in %i files" %
+ (self.progress.n, len(self.revs)))
+
+
+def fetch_revs (path, from_rev, to_rev, symbol, cvs_repo):
+ """Fetch CVSRevs for each rev in <path:from_rev, path:symbol].
+
+ Return a dict of CVSRev objects (revnum -> CVSRev), where each
+ CVSRev encapsulates a CVS revision in the range from
+ path:from_rev to path:symbol (inclusive). If symbol currently
+ refers to from_rev (i.e. nothing has happened since the last
+ import), the returned dict will have exactly one entry (from_rev).
+ If there is no valid revision range between from_rev and symbol,
+ the returned dict will be empty. Situations in which an empty dict
+ is returned, include:
+ - symbol is no longer defined on this path
+ - symbol refers to a revision that is disjoint from from_rev
+
+ from_rev may be None, meaning that all revisions from the initial
+ version of path up to the revision currently referenced by symbol
+ should be fetched.
+
+ If the revision currently referenced by symbol is disjoint from
+ from_rev, the returned dict will be empty.
+
+ Note that there is lots of unexpected behaviour in the handling of
+ the 'cvs rlog -r' parameter: Say you have a branch, called
+ 'my_branch', that points to branch number 1.1.2 of a file. Say
+ there are 3 revisions on this branch: 1.1.2.1 -> 1.1.2.3 (in
+ additions to the branch point 1.1). Now, observe the following
+ 'cvs rlog' executions:
+ - cvs rlog -r0:my_branch ... returns 1.1, 1.1.2.1, 1.1.2.2, 1.1.2.3
+ - cvs rlog -r1.1:my_branch ... returns the same revs
+ - cvs rlog -rmy_branch ... returns 1.1.2.1, 1.1.2.2, 1.1.2.3
+ - cvs rlog -rmy_branch: ... returns the same revs
+ - cvs rlog -r:my_branch ... returns the same revs
+ - cvs rlog -r::my_branch ... returns the same revs
+ - cvs rlog -r1.1.2.1: ... returns the same revs
+ Here is where it gets really weird:
+ - cvs rlog -r1.1.2.1:my_branch ... returns 1.1.2.1 only
+ - cvs rlog -r1.1.2.2:my_branch ... returns 1.1.2.1, 1.1.2.2
+ - cvs rlog -r1.1.2.3:my_branch ... returns 1.1.2.1, 1.1.2.2, 1.1.2.3
+
+ In other words the 'cvs rlog -rfrom_rev:symbol' scheme that we
+ normally use will not work in the case where from_rev is _on_ the
+ branch pointed at by the symbol.
+
+ Therefore, we need an extra parameter, to_rev, which we can use to:
+ 1. Detect when this situation is present.
+ 2. Work around by using 'cvs rlog -rfrom_ref:to_rev' instead.
+
+ """
+ if from_rev is None: # Initial import
+ from_rev = "0" # "cvs rlog -r0:X" fetches from initial revision
+ elif to_rev and to_rev.branch() == from_rev.branch():
+ symbol = to_rev # Use to_rev instead of given symbol
+ # Run 'cvs rlog' on range [from_rev, symbol] and parse CVSRev objects
+ parser = CVSRevLister(cvs_repo)
+ parser.run((path,), True, "%s:%s" % (from_rev, symbol))
+ assert len(parser.revs) == 1
+ assert path in parser.revs
+ return parser.revs[path]
+
+
+if __name__ == '__main__':
+ unittest.main()
diff --git a/git_remote_cvs/cvs_symbol_cache.py b/git_remote_cvs/cvs_symbol_cache.py
new file mode 100644
index 0000000..cc8d88b
--- /dev/null
+++ b/git_remote_cvs/cvs_symbol_cache.py
@@ -0,0 +1,313 @@
+#!/usr/bin/env python
+
+"""Implementation of a local CVS symbol cache.
+
+A CVS symbol cache stores a list of CVS symbols and the CVS state
+associated with each of those CVS symbols at some point in time.
+
+Keeping a local cache of CVS symbols is often needed because the
+design of CVS makes it potentially very expensive to query the CVS
+server directly for CVS symbols and associated states.
+
+In these cases, a local CVS symbol cache can provide equivalent
+(although possibly out-of-date) information immediatele.
+
+Synchronization with the current state on the CVS server can be
+done on a symbol-by-symbol basis (by checking out a given symbol
+and extracting the CVS state from the CVS work tree), or by
+synchronizing _all_ CVS symbols in one operation (by executing
+'cvs rlog' and parsing CVS states from its output).
+
+"""
+
+import sys
+import os
+
+from git_remote_cvs.util import debug, error, die, ProgressIndicator
+from git_remote_cvs.cvs import CVSNum, CVSState, CVSLogParser
+
+
+class CVSSymbolStateLister(CVSLogParser):
+
+ """Extract current CVSStates for all CVS symbols from a CVS log."""
+
+ def __init__ (self, cvs_repo, show_progress = False):
+ """Create a new CVSSymbolStateLister.
+
+ The cvs_repo argument must be a (cvs_root, cvs_module) tuple
+ show_progress determines whether a progress indicator should
+ be displayed.
+
+ """
+ super(CVSSymbolStateLister, self).__init__(cvs_repo)
+ self.symbols = {} # CVS symbol name -> CVSState object
+ self.cur_file = None # current CVS file being processed
+ self.cur_file_numrevs = 0 # #revs in current CVS file
+ self.cur_revnum = None # current revision number
+ self.rev2syms = {} # CVSNum -> [CVS symbol names]
+ self.cur_revs = {} # CVSNum -> True/False (deleted)
+ self.head_num = None # CVSNum of the HEAD rev or branch
+
+ # Possible states:
+ # - BeforeSymbols - waiting for "symbolic names:"
+ # - WithinSymbols - reading CVS symbol names
+ # - BeforeRevs - waiting for "total revisions:"
+ # - BetweenRevs - waiting for "----------------------------"
+ # - ReadingRev - reading CVS revision details
+ self.state = 'BeforeSymbols'
+
+ self.progress = None
+ if show_progress:
+ self.progress = ProgressIndicator("\t", sys.stderr)
+
+ def finalize_symbol_states (self):
+ """Adjust CVSStates in self.symbols based on revision data.
+
+ Based on the information found in self.rev2syms and
+ self.cur_revs, remove deleted revisions and turn branch numbers
+ into corresponding revisions in the CVSStates found in
+ self.symbols.
+
+ """
+ # Create a mapping from branch numbers to the last existing
+ # revision number on those branches
+ branch2lastrev = {} # branch number -> revision number
+ for revnum in self.cur_revs.iterkeys():
+ branchnum = revnum.branch()
+ if (branchnum not in branch2lastrev) or \
+ (revnum > branch2lastrev[branchnum]):
+ branch2lastrev[branchnum] = revnum
+
+ for cvsnum, symbols in self.rev2syms.iteritems():
+ if cvsnum.is_branch():
+ # Turn into corresponding revision number
+ revnum = branch2lastrev.get(cvsnum, cvsnum.parent())
+ for s in symbols:
+ state = self.symbols[s]
+ assert state[self.cur_file] == cvsnum
+ state.replace(self.cur_file, revnum)
+ cvsnum = revnum
+ assert cvsnum.is_rev()
+ assert cvsnum in self.cur_revs
+ if self.cur_revs[cvsnum]: # cvsnum is a deleted rev
+ # Remove from CVSStates
+ for s in symbols:
+ state = self.symbols[s]
+ state.remove(self.cur_file, cvsnum)
+
+ self.rev2syms = {}
+ self.cur_revs = {}
+ self.cur_file = None
+
+ def __call__ (self, line):
+ """Line parser; this method is invoked for each line in the log."""
+ if self.state == 'BeforeSymbols':
+ if line.startswith("RCS file: "):
+ self.cur_file = self.cleanup_path(line[10:])
+ if self.progress:
+ self.progress("%5i symbols found - Parsing CVS file #%i: "
+ "%s " % (len(self.symbols), self.progress.n,
+ self.cur_file,))
+ if line.startswith("head: "):
+ self.head_num = CVSNum(line[6:])
+ if line.startswith("branch: "):
+ self.head_num = CVSNum(line[8:])
+ elif line == "symbolic names:":
+ assert self.head_num
+ s = self.symbols.setdefault("HEAD", CVSState())
+ s.add(self.cur_file, self.head_num)
+ r = self.rev2syms.setdefault(self.head_num, [])
+ r.append("HEAD")
+ self.head_num = None
+ self.state = 'WithinSymbols'
+ elif self.state == 'WithinSymbols':
+ if line.startswith("\t"):
+ symbol, cvsnum = line.split(":", 1)
+ symbol = symbol.strip()
+ cvsnum = CVSNum(cvsnum)
+ s = self.symbols.setdefault(symbol, CVSState())
+ s.add(self.cur_file, cvsnum)
+ r = self.rev2syms.setdefault(cvsnum, [])
+ r.append(symbol)
+ else:
+ self.state = 'BeforeRevs'
+ elif self.state == 'BeforeRevs':
+ if line.startswith("total revisions: "):
+ assert self.cur_file
+ totalrevs_unused, selectedrevs = line.split(";")
+ self.cur_file_numrevs = int(selectedrevs.split(":")[1].strip())
+ self.state = 'BetweenRevs'
+ elif self.state == 'BetweenRevs':
+ if (line == "----------------------------" or
+ line == "======================================"
+ "======================================="):
+ if self.cur_revnum:
+ assert self.cur_revnum in self.cur_revs
+ self.cur_revnum = None
+ if line == "----------------------------":
+ self.state = 'ReadingRev'
+ else:
+ # Finalize current CVS file
+ assert len(self.cur_revs) == self.cur_file_numrevs
+ self.finalize_symbol_states()
+ self.state = 'BeforeSymbols'
+ elif self.state == 'ReadingRev':
+ if line.startswith("revision "):
+ self.cur_revnum = CVSNum(line.split()[1])
+ else:
+ date, author, state, dummy = line.split(";", 3)
+ assert date.startswith("date: ")
+ assert author.strip().startswith("author: ")
+ assert state.strip().startswith("state: ")
+ state = state.strip()[7:]
+ assert self.cur_revnum not in self.cur_revs
+ deleted = state == "dead"
+ self.cur_revs[self.cur_revnum] = deleted
+ self.state = 'BetweenRevs'
+
+ def finish (self):
+ """This method is invoked after the last line has been parsed."""
+ assert self.state == 'BeforeSymbols'
+ if self.progress:
+ self.progress.finish("Parsed %i symbols in %i files" %
+ (len(self.symbols), self.progress.n))
+
+
+class CVSSymbolCache(object):
+
+ """Local cache of the current CVSState of CVS symbols.
+
+ Simulates a dictionary of CVS symbol -> CVSState mappings.
+
+ """
+
+ def __init__ (self, symbols_dir):
+ """Create a new CVS symbol cache, located in the given directory."""
+ self.symbols_dir = symbols_dir
+ if not os.path.isdir(self.symbols_dir):
+ os.makedirs(self.symbols_dir)
+
+ def __len__ (self):
+ """Return the number of CVS symbols stored in this cache."""
+ return len(os.listdir(self.symbols_dir))
+
+ def __iter__ (self):
+ """Return an iterator traversing symbol names stored in this cache."""
+ for filename in os.listdir(self.symbols_dir):
+ yield filename
+
+ def __contains__ (self, symbol):
+ """Return True if the given symbol is present in this cache."""
+ return os.access(os.path.join(self.symbols_dir, symbol),
+ os.F_OK | os.R_OK)
+
+ def __getitem__ (self, symbol):
+ """Return the cached CVSState of the given CVS symbol."""
+ try:
+ f = open(os.path.join(self.symbols_dir, symbol), 'r')
+ except IOError:
+ raise KeyError("'%s'" % (symbol))
+ ret = CVSState()
+ ret.load(f)
+ f.close()
+ return ret
+
+ def __setitem__ (self, symbol, cvs_state):
+ """Store the given CVS symbol and CVSState into the cache."""
+ cvs_state.save(os.path.join(self.symbols_dir, symbol))
+
+ def __delitem__ (self, symbol):
+ """Remove the the given CVS symbol from the cache."""
+ os.remove(os.path.join(self.symbols_dir, symbol))
+
+ def get (self, symbol, default = None):
+ """Return the cached CVSState of the given CVS symbol."""
+ try:
+ return self[symbol]
+ except KeyError:
+ return default
+
+ def items (self):
+ """Return list of (CVS symbol, CVSState) tuples saved in this cache."""
+ for filename in self:
+ yield (filename, self[filename])
+
+ def clear (self):
+ """Remove all entries from this CVS symbol cache."""
+ for filename in os.listdir(self.symbols_dir):
+ os.remove(os.path.join(self.symbols_dir, filename))
+
+ def sync_symbol (self, symbol, cvs, progress):
+ """Synchronize the given CVS symbol with the CVS server.
+
+ The given CVS workdir is used for the synchronization.
+ The retrieved CVSState is also returned
+
+ """
+ progress("Retrieving state of CVS symbol '%s'..." % (symbol))
+ cvs.update(symbol)
+ state = cvs.get_state()
+
+ progress("Saving state of '%s' to symbol cache..." % (symbol))
+ self[symbol] = state
+
+ def sync_all_symbols (self, cvs_repo, progress, symbol_filter = None):
+ """Synchronize this entire CVS symbol cache with the CVS server.
+
+ This may be very expensive if the CVS repository is large, or
+ has many symbols. After this method returns, the symbol cache
+ will be in sync with the current state on the server.
+
+ This method returns a dict with the keys 'unchanged',
+ 'changed', 'added', and 'deleted', where each map to a list of
+ CVS symbols. Each CVS symbol appears in exactly one of these
+ lists.
+
+ If symbol_filter is given, it specifies functions that takes
+ one parameter - a CVS symbol name - and returns True if that
+ symbol should be synchronized, and False if that symbol should
+ be skipped. Otherwise all CVS symbols are synchronized.
+
+ """
+ if symbol_filter is None:
+ symbol_filter = lambda symbol: True
+
+ # Run cvs rlog to fetch current CVSState for all CVS symbols
+ progress("Retrieving current state of all CVS symbols from CVS "
+ "server...", lf = True)
+ parser = CVSSymbolStateLister(cvs_repo, True)
+ retcode = parser.run()
+ if retcode:
+ raise EnvironmentError(retcode, "cvs rlog exit code %i" % retcode)
+
+ # Update symbol cache with new states from the CVS server
+ progress("Updating symbol cache with current CVS state...")
+ results = {}
+ result_keys = ("unchanged", "changed", "added", "deleted")
+ for k in result_keys:
+ results[k] = []
+ # Classify existing symbols as unchanged, changed, or deleted
+ for symbol in filter(symbol_filter, self):
+ if symbol not in parser.symbols: # Deleted
+ results["deleted"].append(symbol)
+ del self[symbol]
+ elif self[symbol] != parser.symbols[symbol]: # Changed
+ results["changed"].append(symbol)
+ self[symbol] = parser.symbols[symbol]
+ else: # Unchanged
+ results["unchanged"].append(symbol)
+ progress()
+ # Add symbols that are not in self
+ for symbol, state in parser.symbols.iteritems():
+ if not symbol_filter(symbol):
+ debug("Skipping CVS symbol '%s'...", symbol)
+ elif symbol in self:
+ assert state == self[symbol]
+ else: # Added
+ results["added"].append(symbol)
+ self[symbol] = state
+ progress()
+ progress("Synchronized local symbol cache (%s)" %
+ (", ".join(["%i %s" % (len(results[k]), k)
+ for k in result_keys])), True)
+ return results
diff --git a/git_remote_cvs/util.py b/git_remote_cvs/util.py
new file mode 100644
index 0000000..7d6adb4
--- /dev/null
+++ b/git_remote_cvs/util.py
@@ -0,0 +1,186 @@
+#!/usr/bin/env python
+
+"""Misc. useful functionality used by the rest of this package.
+
+This module provides common functionality used by the other modules in
+this package.
+
+"""
+
+import sys
+import os
+import subprocess
+
+
+# Whether or not to show debug messages
+DEBUG = False
+
+def debug (msg, *args):
+ """Print a debug message to stderr when DEBUG is enabled."""
+ if DEBUG:
+ print >> sys.stderr, msg % args
+
+def error (msg, *args):
+ """Print an error message to stderr."""
+ print >> sys.stderr, "ERROR:", msg % args
+
+def die (msg, *args):
+ """Print as error message to stderr and exit the program."""
+ error(msg, *args)
+ sys.exit(1)
+
+
+class ProgressIndicator(object):
+
+ """Simple progress indicator.
+
+ Displayed as a spinning character by default, but can be customized
+ by passing custom messages that overrides the spinning character.
+
+ """
+
+ States = ("|", "/", "-", "\\")
+
+ def __init__ (self, prefix = "", f = sys.stdout):
+ """Create a new ProgressIndicator, bound to the given file object."""
+ self.n = 0 # Simple progress counter
+ self.f = f # Progress is written to this file object
+ self.prev_len = 0 # Length of previous msg (to be overwritten)
+ self.prefix = prefix # Prefix prepended to each progress message
+ self.prefix_lens = [] # Stack of prefix string lengths
+
+ def pushprefix (self, prefix):
+ """Append the given prefix onto the prefix stack."""
+ self.prefix_lens.append(len(self.prefix))
+ self.prefix += prefix
+
+ def popprefix (self):
+ """Remove the last prefix from the prefix stack."""
+ prev_len = self.prefix_lens.pop()
+ self.prefix = self.prefix[:prev_len]
+
+ def __call__ (self, msg = None, lf = False):
+ """Indicate progress, possibly with a custom message."""
+ if msg is None:
+ msg = self.States[self.n % len(self.States)]
+ msg = self.prefix + msg
+ print >> self.f, "\r%-*s" % (self.prev_len, msg),
+ self.prev_len = len(msg.expandtabs())
+ if lf:
+ print >> self.f
+ self.prev_len = 0
+ self.n += 1
+
+ def finish (self, msg = "done", noprefix = False):
+ """Finalize progress indication with the given message."""
+ if noprefix:
+ self.prefix = ""
+ self(msg, True)
+
+
+def start_command (args, cwd = None, shell = False, add_env = None,
+ stdin = subprocess.PIPE, stdout = subprocess.PIPE,
+ stderr = subprocess.PIPE):
+ """Start the given command, and return a subprocess object.
+
+ This provides a simpler interface to the subprocess module.
+
+ """
+ env = None
+ if add_env is not None:
+ env = os.environ.copy()
+ env.update(add_env)
+ return subprocess.Popen(args, bufsize = 1, stdin = stdin, stdout = stdout,
+ stderr = stderr, cwd = cwd, shell = shell,
+ env = env, universal_newlines = True)
+
+
+def run_command (args, cwd = None, shell = False, add_env = None,
+ flag_error = True):
+ """Run the given command to completion, and return its results.
+
+ This provides a simpler interface to the subprocess module.
+
+ The results are formatted as a 3-tuple: (exit_code, output, errors)
+
+ If flag_error is enabled, Error messages will be produced if the
+ subprocess terminated with a non-zero exit code and/or stderr
+ output.
+
+ The other arguments are passed on to start_command().
+
+ """
+ process = start_command(args, cwd, shell, add_env)
+ (output, errors) = process.communicate()
+ exit_code = process.returncode
+ if flag_error and errors:
+ error("'%s' returned errors:\n---\n%s---", " ".join(args), errors)
+ if flag_error and exit_code:
+ error("'%s' returned exit code %i", " ".join(args), exit_code)
+ return (exit_code, output, errors)
+
+
+def file_reader_method (missing_ok = False):
+ """Decorator for simplifying reading of files.
+
+ If missing_ok is True, a failure to open a file for reading will
+ not raise the usual IOError, but instead the wrapped method will be
+ called with f == None. The method must in this case properly
+ handle f == None.
+
+ """
+ def _wrap (method):
+ """Teach given method to handle both filenames and file objects.
+
+ The given method must take a file object as its second argument
+ (the first argument being 'self', of course). This decorator
+ will take a filename given as the second argument and promote
+ it to a file object.
+
+ """
+ def _wrapped_method (self, filename, *args, **kwargs):
+ if isinstance(filename, file):
+ f = filename
+ else:
+ try:
+ f = open(filename, 'r')
+ except IOError:
+ if missing_ok:
+ f = None
+ else:
+ raise
+ try:
+ return method(self, f, *args, **kwargs)
+ finally:
+ if not isinstance(filename, file) and f:
+ f.close()
+ return _wrapped_method
+ return _wrap
+
+
+def file_writer_method (method):
+ """Decorator for simplifying writing of files.
+
+ Enables the given method to handle both filenames and file objects.
+
+ The given method must take a file object as its second argument
+ (the first argument being 'self', of course). This decorator will
+ take a filename given as the second argument and promote it to a
+ file object.
+
+ """
+ def _new_method (self, filename, *args, **kwargs):
+ if isinstance(filename, file):
+ f = filename
+ else:
+ # Make sure the containing directory exists
+ parent_dir = os.path.dirname(filename)
+ if not os.path.isdir(parent_dir):
+ os.makedirs(parent_dir)
+ f = open(filename, 'w')
+ try:
+ return method(self, f, *args, **kwargs)
+ finally:
+ if not isinstance(filename, file):
+ f.close()
+ return _new_method
--
1.6.4.262.gca66a.dirty
^ permalink raw reply related
* [RFCv4 3/5] 2/2: Add Python support library for CVS remote helper
From: Johan Herland @ 2009-08-17 3:35 UTC (permalink / raw)
To: git; +Cc: Johannes.Schindelin, Johan Herland, barkalow, davvid, gitster
In-Reply-To: <1250480161-21933-1-git-send-email-johan@herland.net>
This patch introduces the rest of a Python package called "git_remote_cvs"
containing the building blocks of the CVS remote helper. The CVS remote
helper itself is NOT part of this patch.
The patch includes the necessary Makefile additions to build and install
the git_remote_cvs Python package along with the rest of Git.
This patch has been improved by the following contributions:
- David Aguilar: Lots of Python coding style fixes
- David Aguilar: DESTDIR support in Makefile
Cc: David Aguilar <davvid@gmail.com>
Signed-off-by: Johan Herland <johan@herland.net>
---
This patch and the previous patch belong together in the same commit,
but the 100K limit (?) on the git mailing list forces me to split this
patch in two.
...Johan
Makefile | 9 +
git_remote_cvs/.gitignore | 2 +
git_remote_cvs/Makefile | 35 ++
git_remote_cvs/__init__.py | 27 ++
git_remote_cvs/commit_states.py | 62 ++++
git_remote_cvs/cvs_revision_map.py | 418 ++++++++++++++++++++++
git_remote_cvs/git.py | 680 ++++++++++++++++++++++++++++++++++++
git_remote_cvs/setup.py | 17 +
8 files changed, 1250 insertions(+), 0 deletions(-)
create mode 100644 git_remote_cvs/.gitignore
create mode 100644 git_remote_cvs/Makefile
create mode 100644 git_remote_cvs/__init__.py
create mode 100644 git_remote_cvs/commit_states.py
create mode 100644 git_remote_cvs/cvs_revision_map.py
create mode 100644 git_remote_cvs/git.py
create mode 100644 git_remote_cvs/setup.py
diff --git a/Makefile b/Makefile
index 969cef5..bb5cea2 100644
--- a/Makefile
+++ b/Makefile
@@ -1350,6 +1350,9 @@ endif
ifndef NO_PERL
$(QUIET_SUBDIR0)perl $(QUIET_SUBDIR1) PERL_PATH='$(PERL_PATH_SQ)' prefix='$(prefix_SQ)' all
endif
+ifndef NO_PYTHON
+ $(QUIET_SUBDIR0)git_remote_cvs $(QUIET_SUBDIR1) PYTHON_PATH='$(PYTHON_PATH_SQ)' prefix='$(prefix_SQ)' all
+endif
$(QUIET_SUBDIR0)templates $(QUIET_SUBDIR1)
please_set_SHELL_PATH_to_a_more_modern_shell:
@@ -1701,6 +1704,9 @@ install: all
ifndef NO_PERL
$(MAKE) -C perl prefix='$(prefix_SQ)' DESTDIR='$(DESTDIR_SQ)' install
endif
+ifndef NO_PYTHON
+ $(MAKE) -C git_remote_cvs prefix='$(prefix_SQ)' DESTDIR='$(DESTDIR_SQ)' install
+endif
ifndef NO_TCLTK
$(MAKE) -C gitk-git install
$(MAKE) -C git-gui gitexecdir='$(gitexec_instdir_SQ)' install
@@ -1821,6 +1827,9 @@ ifndef NO_PERL
$(RM) gitweb/gitweb.cgi
$(MAKE) -C perl clean
endif
+ifndef NO_PYTHON
+ $(MAKE) -C git_remote_cvs clean
+endif
$(MAKE) -C templates/ clean
$(MAKE) -C t/ clean
ifndef NO_TCLTK
diff --git a/git_remote_cvs/.gitignore b/git_remote_cvs/.gitignore
new file mode 100644
index 0000000..2247d5f
--- /dev/null
+++ b/git_remote_cvs/.gitignore
@@ -0,0 +1,2 @@
+/build
+/dist
diff --git a/git_remote_cvs/Makefile b/git_remote_cvs/Makefile
new file mode 100644
index 0000000..061c247
--- /dev/null
+++ b/git_remote_cvs/Makefile
@@ -0,0 +1,35 @@
+#
+# Makefile for the git_remote_cvs python support modules
+#
+pysetupfile:=setup.py
+
+# Shell quote (do not use $(call) to accommodate ancient setups);
+DESTDIR_SQ = $(subst ','\'',$(DESTDIR))
+
+ifndef PYTHON_PATH
+ PYTHON_PATH = /usr/bin/python
+endif
+ifndef prefix
+ prefix = $(HOME)
+endif
+ifndef V
+ QUIET = @
+ QUIETSETUP = --quiet
+endif
+
+PYLIBDIR=$(shell $(PYTHON_PATH) -c \
+ "import sys; \
+ print 'lib/python%i.%i/site-packages' % sys.version_info[:2]")
+
+all: $(pysetupfile)
+ $(QUIET)$(PYTHON_PATH) $(pysetupfile) $(QUIETSETUP) build
+
+install: $(pysetupfile)
+ $(PYTHON_PATH) $(pysetupfile) install --prefix $(DESTDIR_SQ)$(prefix)
+
+instlibdir: $(pysetupfile)
+ @echo "$(prefix)/$(PYLIBDIR)"
+
+clean:
+ $(QUIET)$(PYTHON_PATH) $(pysetupfile) $(QUIETSETUP) clean -a
+ $(RM) *.pyo *.pyc
diff --git a/git_remote_cvs/__init__.py b/git_remote_cvs/__init__.py
new file mode 100644
index 0000000..4956d2d
--- /dev/null
+++ b/git_remote_cvs/__init__.py
@@ -0,0 +1,27 @@
+#!/usr/bin/env python
+
+"""Support library package for git-remote-cvs.
+
+git-remote-cvs is a Git remote helper command that interfaces with a
+CVS repository to provide automatic import of CVS history into a Git
+repository.
+
+This package provides the support library needed by git-remote-cvs.
+The following modules are included:
+
+- cvs - Interaction with CVS repositories
+
+- cvs_symbol_cache - Local CVS symbol cache
+
+- changeset - Collect individual CVS revisions into commits
+
+- git - Interaction with Git repositories
+
+- commit_states - Map Git commits to CVS states
+
+- cvs_revision_map - Map CVS revisions to various metainformation
+
+- util - General utility functionality use by the other modules in
+ this package, and also used directly by git-remote-cvs.
+
+"""
diff --git a/git_remote_cvs/commit_states.py b/git_remote_cvs/commit_states.py
new file mode 100644
index 0000000..2e0af6c
--- /dev/null
+++ b/git_remote_cvs/commit_states.py
@@ -0,0 +1,62 @@
+#!/usr/bin/env python
+
+"""Code for relating Git commits to corresponding CVSState objects."""
+
+from git_remote_cvs.util import debug, error, die
+from git_remote_cvs.cvs import CVSState
+
+
+class CommitStates(object):
+
+ """Provide a mapping from Git commits to CVSState objects.
+
+ Behaves like a dictionary of Git commit names -> CVSState mappings.
+
+ Every Git commit converted from CVS has a corresponding CVSState,
+ which describes exactly which CVS revisions are present in a
+ checkout of that commit.
+
+ This class provides the interface to map from a Git commit to its
+ corresponding CVSState. The mapping uses GitNotes as a persistent
+ storage backend.
+
+ """
+
+ def __init__ (self, notes):
+ """Create a new Git commit -> CVSState map."""
+ self.notes = notes
+ self._cache = {} # commitname -> CVSState
+
+ def _load (self, commit):
+ """Retrieve the CVSState associated with the given Git commit."""
+ if commit is None:
+ return None
+ if commit in self._cache:
+ return self._cache[commit]
+ notedata = self.notes.get(commit)
+ if notedata is None: # Given commit has no associated note
+ return None
+ state = CVSState()
+ state.load_data(notedata)
+ self._cache[commit] = state
+ return state
+
+ def add (self, commit, state, gfi):
+ """Add the given Git commit -> CVSState mapping."""
+ assert commit not in self._cache
+ self._cache[commit] = state
+ self.notes.import_note(commit, str(state), gfi)
+
+ def __getitem__ (self, commit):
+ """Return the CVSState associated with the given commit."""
+ state = self._load(commit)
+ if state is None:
+ raise KeyError("Unknown commit '%s'" % (commit))
+ return state
+
+ def get (self, commit, default = None):
+ """Return the CVSState associated with the given commit."""
+ state = self._load(commit)
+ if state is None:
+ return default
+ return state
diff --git a/git_remote_cvs/cvs_revision_map.py b/git_remote_cvs/cvs_revision_map.py
new file mode 100644
index 0000000..0e65ba6
--- /dev/null
+++ b/git_remote_cvs/cvs_revision_map.py
@@ -0,0 +1,418 @@
+#!/usr/bin/env python
+
+"""Functionality for mapping CVS revisions to associated metainfo.
+
+This modules provides the following main classes:
+
+CVSRevisionMap - provides a mapping from CVS revisions to the
+ the following associated metainfo:
+ - Mode/permission of the associated CVS path
+ - The Git blob holding the revision data
+ - The Git commits that correspond to the CVS
+ states in which this CVS revision is present
+
+CVSStateMap - provides a mapping from CVS states to corresponding
+ Git commits (i.e. Git commits, whose tree state is
+ identical to a given CVS state)
+"""
+
+import os
+
+from git_remote_cvs.util import debug, error, die, file_reader_method
+from git_remote_cvs.cvs import CVSNum, CVSDate
+from git_remote_cvs.git import GitFICommit
+
+
+class _CVSPathInfo(object):
+
+ """Information on a single CVS path."""
+
+ __slots__ = ('revs', 'mode')
+
+ def __init__ (self, mode = None):
+ self.revs = {}
+ self.mode = mode
+
+
+class _CVSRevInfo(object):
+
+ """Information on a single CVS revision."""
+
+ __slots__ = ('blob', 'commits')
+
+ def __init__ (self, blob):
+ self.blob = blob
+ self.commits = []
+
+
+class CVSRevisionMap(object):
+
+ """Encapsulate the mapping of CVS revisions to associated metainfo.
+
+ This container maps CVS revisions (a combination of a CVS path and
+ a CVS revision number) into Git blob names, Git commit names, and
+ CVS path information. Git object (blob/commit) names are either
+ 40-char SHA1 strings, or "git fast-import" mark numbers if the form
+ ":<num>".
+
+ The data structure is organized as follows:
+ - A CVSRevisionMap instance knows about a set of CVS paths.
+ For each CVS path, the following is known:
+ - The mode (permission bits) of that CVS path (644 or 755)
+ - The CVS revision numbers that exist for that CVS path.
+ For each revision number, the following is known:
+ - Exactly 1 blob name; the Git blob containing the contents of
+ the revision (the contents of the CVS path at that CVS
+ revision).
+ - One or more commit names; the Git commits which encapsulate a
+ CVS state in which this CVS revision
+
+ To store this data structure persistently, this class uses a Git
+ ref that points to a tree structure containing the above
+ information. When changes to the structure are made, this class
+ will produce git-fast-import commands to update that tree structure
+ accordingly.
+
+ IMPORTANT: No changes to the CVSRevisionMap will be stored unless
+ self.commit_map() is called with a valid GitFastImport instance.
+
+ NOTES: Mark numbers are only transient references bound to the
+ current "git fast-import" process (assumed to be running alongside
+ this process). Therefore, when the "git fast-import" process ends,
+ it must write out a mark number -> SHA1 mapping (see the
+ "--export-marks" argument to "git fast-import"). Subsequently,
+ this mapping must be parsed, and the mark numbers in this
+ CVSRevisionMap must be resolved into persistent SHA1 names.
+ In order to quickly find all the unresolved mark number entries in
+ the data structure, and index mapping mark numbers to revisions is
+ kept in a separate file in the tree structure.
+ See the loadMarksFile() method for more details.
+
+ """
+
+ MarkNumIndexFile = "CVS/marks" # Invalid CVS path name
+
+ def __init__ (self, git_ref, obj_fetcher):
+ """Create a new CVS revision map bound to the given Git ref."""
+ self.git_ref = git_ref
+ self.obj_fetcher = obj_fetcher
+
+ # The following data structure is a cache of the persistent
+ # data structure found at self.git_ref.
+ # It is structured as follows:
+ # - self.d is a mapping from CVS paths to _CVSPathInfo objects.
+ # _CVSPathInfo object have two fields: revs, mode:
+ # - mode is either 644 (non-executable) or 755 (executable).
+ # - revs is a dict, mapping CVS revision numbers (CVSNum
+ # instances) to _CVSRevInfo objects. _CVSRevInfo objects
+ # have two fields: blob, commits:
+ # - blob is the name of the Git blob object holding the
+ # contents of that revision.
+ # - commits is a collection of zero or more Git commit
+ # names where the commit contains this revision.
+ self.d = {} # dict: path -> _CVSPathInfo
+ self.mods = set() # (path, revnum) pairs for all modified revs
+ self.marknum_idx = {} # dict: mark num -> [(path, revnum), ...]
+ self._load_marknum_idx()
+
+ def __del__ (self):
+ """Verify that self.commit_map() was called before destruction."""
+ if self.mods:
+ error("Missing call to self.commit_map().")
+ error("%i revision changes are lost!", len(self.mods))
+
+ def __nonzero__ (self):
+ """Return True iff any information is currently stored here."""
+ return bool(self.d)
+
+ # Private methods:
+
+ def _add_to_marknum_idx(self, marknum, path, revnum):
+ """Add the given marknum -> (path, revnum) association."""
+ entry = self.marknum_idx.setdefault(marknum, [])
+ entry.append((path, revnum))
+
+ def _load_marknum_idx(self):
+ """Load contents of MarkNumIndexFile into self.marknum_idx."""
+ blobspec = "%s:%s" % (self.git_ref, self.MarkNumIndexFile)
+ try:
+ f = self.obj_fetcher.open_obj(blobspec)
+ except KeyError:
+ return # Missing object; nothing to do
+
+ for line in f:
+ # Format of line is "<marknum> <path>:<revnum>"
+ mark, rest = line.strip().split(' ', 1)
+ path, rev = rest.rsplit(':', 1)
+ self._add_to_marknum_idx(int(mark), path, CVSNum(rev))
+ f.close()
+
+ def _save_marknum_idx(self):
+ """Prepare data for storage into MarkNumIndexFile.
+
+ The returned string contains the current contents of
+ self.marknum_idx, formatted to be stored verbatim in
+ self.MarkNumIndexFile.
+
+ """
+ lines = []
+ for marknum, revs in sorted(self.marknum_idx.iteritems()):
+ for path, revnum in revs:
+ # Format of line is "<marknum> <path>:<revnum>"
+ line = "%i %s:%s\n" % (marknum, path, revnum)
+ lines.append(line)
+ return "".join(lines)
+
+ def _save_rev(self, path, revnum):
+ """Return blob data for storing the given revision persistently.
+
+ Generate the blob contents that will reconstitute the same
+ revision entry when read back in with _fetch_path().
+
+ """
+ lines = []
+ rev_info = self.d[path].revs[revnum]
+ lines.append("blob %s\n" % (rev_info.blob))
+ for commit in rev_info.commits:
+ lines.append("commit %s\n" % (commit))
+ return "".join(lines)
+
+ @staticmethod
+ def _valid_objname (objname):
+ """Return the argument as a SHA1 (string) or mark num (int)."""
+ # Blob names are either 40-char SHA1 strings, or mark numbers
+ if isinstance(objname, int) or len(objname) != 40: # Mark number
+ return int(objname)
+ return objname
+
+ def _load (self, path, mode, data):
+ """GitObjectFetcher.walk_tree() callback."""
+ assert mode in (644, 755)
+ cvs_path, revnum = os.path.split(path)
+ revnum = CVSNum(revnum)
+ if cvs_path in self.d:
+ assert mode == self.d[cvs_path].mode
+ else:
+ self.d[cvs_path] = _CVSPathInfo(mode)
+ assert revnum not in self.d[cvs_path].revs
+ rev_info = None
+ for line in data.split("\n"):
+ if not line:
+ continue
+ t, objname = line.split()
+ objname = self._valid_objname(objname)
+ if t == "blob":
+ assert rev_info is None
+ rev_info = _CVSRevInfo(objname)
+ elif t == "commit":
+ assert rev_info is not None
+ rev_info.commits.append(objname)
+ else:
+ assert False, "Unknown type '%s'" % (t)
+ assert rev_info.commits # Each rev is in at least one commit
+ self.d[cvs_path].revs[revnum] = rev_info
+
+ def _fetch_path (self, path):
+ """If the given path exists, create a path entry in self.d."""
+ tree_spec = "%s:%s" % (self.git_ref, path)
+ self.obj_fetcher.walk_tree(tree_spec, self._load, path)
+ # TODO: Don't load entire tree at once?
+
+ # Public methods:
+
+ def has_path (self, path):
+ """Return True iff the given path is present."""
+ if path not in self.d:
+ self._fetch_path(path)
+ return path in self.d
+
+ def has_rev (self, path, revnum):
+ """Return True iff the given path:revnum is present."""
+ if path not in self.d:
+ self._fetch_path(path)
+ return path in self.d and revnum in self.d[path].revs
+
+ def get_mode (self, path):
+ """Return mode bits for the given path."""
+ if path not in self.d:
+ self._fetch_path(path)
+ return self.d[path].mode
+
+ def get_blob (self, path, revnum):
+ """Return the blob name for the given revision."""
+ if path not in self.d:
+ self._fetch_path(path)
+ return self.d[path].revs[revnum].blob
+
+ def get_commits (self, path, revnum):
+ """Return the commit names containing the given revision."""
+ if path not in self.d:
+ self._fetch_path(path)
+ return self.d[path].revs[revnum].commits
+
+ def has_unresolved_marks (self):
+ """Return True iff there are mark numbers in the data structure."""
+ return self.marknum_idx
+
+ # Public non-const methods
+
+ def add_path (self, path, mode):
+ """Add the given path and associated mode bits to this map."""
+ if path not in self.d:
+ self._fetch_path(path)
+ if path in self.d:
+ if self.d[path].mode:
+ assert mode == self.d[path].mode, \
+ "The mode of %s has changed from %s " \
+ "to %s since the last import. This " \
+ "is not supported." % (
+ path, self.d[path].mode, mode)
+ else:
+ self.d[path].mode = mode
+ else:
+ self.d[path] = _CVSPathInfo(mode)
+ # Do not add to self.mods yet, as we expect revisions to be
+ # added before commit_map() is called
+
+ def add_blob (self, path, revnum, blobname):
+ """Add the given path:revnum -> blobname association."""
+ assert blobname
+ if path not in self.d:
+ self._fetch_path(path)
+ blobname = self._valid_objname(blobname)
+ if isinstance(blobname, int): # Mark number
+ self._add_to_marknum_idx(blobname, path, revnum)
+ entry = self.d.setdefault(path, _CVSPathInfo())
+ assert revnum not in entry.revs
+ entry.revs[revnum] = _CVSRevInfo(blobname)
+ self.mods.add((path, revnum))
+
+ def add_commit (self, path, revnum, commitname):
+ """Add the given path:revnum -> commitname association."""
+ if path not in self.d:
+ self._fetch_path(path)
+ commitname = self._valid_objname(commitname)
+ if isinstance(commitname, int): # Mark number
+ self._add_to_marknum_idx(commitname, path, revnum)
+ assert revnum in self.d[path].revs
+ assert commitname not in self.d[path].revs[revnum].commits
+ self.d[path].revs[revnum].commits.append(commitname)
+ self.mods.add((path, revnum))
+
+ @file_reader_method(missing_ok = True)
+ def load_marks_file (self, f):
+ """Load mark -> SHA1 mappings from git-fast-import marks file.
+
+ Replace all mark numbers with proper SHA1 names in this data
+ structure (using self.marknum_idx to find them quickly).
+
+ """
+ if not f:
+ return 0
+ marks = {}
+ last_mark = 0
+ for line in f:
+ (mark, sha1) = line.strip().split()
+ assert mark.startswith(":")
+ mark = int(mark[1:])
+ assert mark not in marks
+ marks[mark] = sha1
+ if mark > last_mark:
+ last_mark = mark
+ for marknum, revs in self.marknum_idx.iteritems():
+ sha1 = marks[marknum]
+ for path, revnum in revs:
+ if path not in self.d:
+ self._fetch_path(path)
+ rev_info = self.d[path].revs[revnum]
+ if rev_info.blob == marknum: # Replace blobname
+ rev_info.blob = sha1
+ else: # Replace commitname
+ assert marknum in rev_info.commits
+ i = rev_info.commits.index(marknum)
+ rev_info.commits[i] = sha1
+ assert marknum not in rev_info.commits
+ self.mods.add((path, revnum))
+ self.marknum_idx = {} # Resolved all transient mark numbers
+ return last_mark
+
+ def sync_modeinfo_from_cvs (self, cvs):
+ """Update with mode information from current CVS checkout.
+
+ This method will retrieve mode information on all paths in the
+ current CVS checkout, and update this data structure
+ correspondingly. In the case where mode information is already
+ present for a given CVS path, this method will verify that such
+ information is correct.
+
+ """
+ for path, mode in cvs.get_modeinfo().iteritems():
+ self.add_path(path, mode)
+
+ def commit_map (self, gfi, author, message):
+ """Produce git-fast-import commands for storing changes.
+
+ The given GitFastImport object is used to produce a commit
+ making the changes done to this data structure persistent.
+
+ """
+ now = CVSDate("now")
+ commitdata = GitFICommit(
+ author[0], author[1], now.ts, now.tz_str(), message)
+
+ # Add updated MarkNumIndexFile to commit
+ mark = gfi.blob(self._save_marknum_idx())
+ commitdata.modify(644, mark, self.MarkNumIndexFile)
+
+ for path, revnum in self.mods:
+ mark = gfi.blob(self._save_rev(path, revnum))
+ mode = self.d[path].mode
+ assert mode in (644, 755)
+ commitdata.modify(mode, mark, "%s/%s" % (path, revnum))
+
+ gfi.commit(self.git_ref, commitdata)
+ self.mods = set()
+
+
+class CVSStateMap(object):
+
+ """Map CVSState object to the commit names which produces that state."""
+
+ def __init__ (self, cvs_rev_map):
+ """Create a CVSStateMap object, bound to the given CVS revision map."""
+ self.cvs_rev_map = cvs_rev_map
+
+ def get_commits (self, state):
+ """Map the given CVSState to commits that contain this state.
+
+ Return all commits where the given state is a subset of the
+ state produced by that commit.
+
+ Returns a set of commit names. The set may be empty.
+
+ """
+ candidate_commits = None
+ for path, revnum in state:
+ commits = self.cvs_rev_map.get_commits(path, revnum)
+ if candidate_commits is None:
+ candidate_commits = set(commits)
+ else:
+ candidate_commits.intersection_update(commits)
+ return candidate_commits
+
+ def get_exact_commit (self, state, commit_map):
+ """Map the given CVSState to the commit with this exact state.
+
+ The given commit_map must be a CommitStates object.
+
+ Return the only commit (if any) that produces the exact given
+ CVSState.
+
+ Returns a commit name, or None if no matching commit is found.
+
+ """
+ commits = self.get_commits(state)
+ for c in commits:
+ if state == commit_map.get(c):
+ return c
+ return None
diff --git a/git_remote_cvs/git.py b/git_remote_cvs/git.py
new file mode 100644
index 0000000..cf037e3
--- /dev/null
+++ b/git_remote_cvs/git.py
@@ -0,0 +1,680 @@
+#!/usr/bin/env python
+
+"""Functionality for interacting with Git repositories.
+
+This module provides classes for interfacing with a Git repository.
+
+"""
+
+import os
+import re
+from binascii import hexlify
+from cStringIO import StringIO
+import unittest
+
+from git_remote_cvs.util import debug, error, die, start_command, run_command
+from git_remote_cvs.cvs import CVSDate
+
+
+def get_git_dir ():
+ """Return the path to the GIT_DIR for this repo."""
+ args = ("git", "rev-parse", "--git-dir")
+ exit_code, output, errors = run_command(args)
+ if exit_code:
+ die("Failed to retrieve git dir")
+ assert not errors
+ return output.strip()
+
+
+def parse_git_config ():
+ """Return a dict containing the parsed version of 'git config -l'."""
+ exit_code, output, errors = run_command(("git", "config", "-z", "-l"))
+ if exit_code:
+ die("Failed to retrieve git configuration")
+ assert not errors
+ return dict([e.split('\n', 1) for e in output.split("\0") if e])
+
+
+def git_config_bool (value):
+ """Convert the given git config string value to True or False.
+
+ Raise ValueError if the given string was not recognized as a
+ boolean value.
+
+ """
+ norm_value = str(value).strip().lower()
+ if norm_value in ("true", "1", "yes", "on", ""):
+ return True
+ if norm_value in ("false", "0", "no", "off", "none"):
+ return False
+ raise ValueError("Failed to parse '%s' into a boolean value" % (value))
+
+
+def valid_git_ref (ref_name):
+ """Return True iff the given ref name is a valid git ref name."""
+ # The following is a reimplementation of the git check-ref-format
+ # command. The rules were derived from the git check-ref-format(1)
+ # manual page. This code should be replaced by a call to
+ # check_ref_format() in the git library, when such is available.
+ if ref_name.endswith('/') or \
+ ref_name.startswith('.') or \
+ ref_name.count('/.') or \
+ ref_name.count('..') or \
+ ref_name.endswith('.lock'):
+ return False
+ for c in ref_name:
+ if ord(c) < 0x20 or ord(c) == 0x7f or c in " ~^:?*[":
+ return False
+ return True
+
+
+class GitObjectFetcher(object):
+
+ """Provide parsed access to 'git cat-file --batch'.
+
+ This provides a read-only interface to the Git object database.
+
+ """
+
+ def __init__ (self):
+ """Initiate a 'git cat-file --batch' session."""
+ self.queue = [] # List of object names to be submitted
+ self.in_transit = None # Object name currently in transit
+
+ # 'git cat-file --batch' produces binary output which is likely
+ # to be corrupted by the default "rU"-mode pipe opened by
+ # start_command. (Mode == "rU" does universal new-line
+ # conversion, which mangles carriage returns.) Therefore, we
+ # open an explicitly binary-safe pipe for transferring the
+ # output from 'git cat-file --batch'.
+ pipe_r_fd, pipe_w_fd = os.pipe()
+ pipe_r = os.fdopen(pipe_r_fd, "rb")
+ pipe_w = os.fdopen(pipe_w_fd, "wb")
+ self.proc = start_command(("git", "cat-file", "--batch"),
+ stdout = pipe_w)
+ self.f = pipe_r
+
+ def __del__ (self):
+ """Verify completed communication with 'git cat-file --batch'."""
+ assert not self.queue
+ assert self.in_transit is None
+ self.proc.stdin.close()
+ assert self.proc.wait() == 0 # Zero exit code
+ assert self.f.read() == "" # No remaining output
+
+ def _submit_next_object (self):
+ """Submit queue items to the 'git cat-file --batch' process.
+
+ If there are items in the queue, and there is currently no item
+ currently in 'transit', then pop the first item off the queue,
+ and submit it.
+
+ """
+ if self.queue and self.in_transit is None:
+ self.in_transit = self.queue.pop(0)
+ print >> self.proc.stdin, self.in_transit[0]
+
+ def push (self, obj, callback):
+ """Push the given object name onto the queue.
+
+ The given callback function will at some point in the future
+ be called exactly once with the following arguments:
+ - self - this GitObjectFetcher instance
+ - obj - the object name provided to push()
+ - sha1 - the SHA1 of the object, if 'None' obj is missing
+ - t - the type of the object (tag/commit/tree/blob)
+ - size - the size of the object in bytes
+ - data - the object contents
+
+ """
+ self.queue.append((obj, callback))
+ self._submit_next_object() # (Re)start queue processing
+
+ def process_next_entry (self):
+ """Read the next entry off the queue and invoke callback."""
+ obj, cb = self.in_transit
+ self.in_transit = None
+ header = self.f.readline()
+ if header == "%s missing\n" % (obj):
+ cb(self, obj, None, None, None, None)
+ return
+ sha1, t, size = header.split(" ")
+ assert len(sha1) == 40
+ assert t in ("tag", "commit", "tree", "blob")
+ assert size.endswith("\n")
+ size = int(size.strip())
+ data = self.f.read(size)
+ assert self.f.read(1) == "\n"
+ cb(self, obj, sha1, t, size, data)
+ self._submit_next_object()
+
+ def process (self):
+ """Process the current queue until empty."""
+ while self.in_transit is not None:
+ self.process_next_entry()
+
+ # High-level convenience methods:
+
+ def get_sha1 (self, objspec):
+ """Return the SHA1 of the object specified by 'objspec'.
+
+ Return None if 'objspec' does not specify an existing object.
+
+ """
+ class _ObjHandler(object):
+ """Helper class for getting the returned SHA1."""
+ def __init__ (self, parser):
+ self.parser = parser
+ self.sha1 = None
+
+ def __call__ (self, parser, obj, sha1, t, size, data):
+ # FIXME: Many unused arguments. Could this be cheaper?
+ assert parser == self.parser
+ self.sha1 = sha1
+
+ handler = _ObjHandler(self)
+ self.push(objspec, handler)
+ self.process()
+ return handler.sha1
+
+ def open_obj (self, objspec):
+ """Return a file object wrapping the contents of a named object.
+
+ The caller is responsible for calling .close() on the returned
+ file object.
+
+ Raise KeyError if 'objspec' does not exist in the repo.
+
+ """
+ class _ObjHandler(object):
+ """Helper class for parsing the returned git object."""
+ def __init__ (self, parser):
+ """Set up helper."""
+ self.parser = parser
+ self.contents = StringIO()
+ self.err = None
+
+ def __call__ (self, parser, obj, sha1, t, size, data):
+ """Git object callback (see GitObjectFetcher documentation)."""
+ assert parser == self.parser
+ if not sha1: # Missing object
+ self.err = "Missing object '%s'" % obj
+ else:
+ assert size == len(data)
+ self.contents.write(data)
+
+ handler = _ObjHandler(self)
+ self.push(objspec, handler)
+ self.process()
+ if handler.err:
+ raise KeyError(handler.err)
+ handler.contents.seek(0)
+ return handler.contents
+
+ def walk_tree (self, tree_objspec, callback, prefix = ""):
+ """Recursively walk the given Git tree object.
+
+ Recursively walk all subtrees of the given tree object, and
+ invoke the given callback passing three arguments:
+ (path, mode, data) with the path, permission bits, and contents
+ of all the blobs found in the entire tree structure.
+
+ """
+ class _ObjHandler(object):
+ """Helper class for walking a git tree structure."""
+ def __init__ (self, parser, cb, path, mode = None):
+ """Set up helper."""
+ self.parser = parser
+ self.cb = cb
+ self.path = path
+ self.mode = mode
+ self.err = None
+
+ def parse_tree (self, treedata):
+ """Parse tree object data, yield tree entries.
+
+ Each tree entry is a 3-tuple (mode, sha1, path)
+
+ self.path is prepended to all paths yielded
+ from this method.
+
+ """
+ while treedata:
+ mode = int(treedata[:6], 10)
+ # Turn 100xxx into xxx
+ if mode > 100000:
+ mode -= 100000
+ assert treedata[6] == " "
+ i = treedata.find("\0", 7)
+ assert i > 0
+ path = treedata[7:i]
+ sha1 = hexlify(treedata[i + 1: i + 21])
+ yield (mode, sha1, self.path + path)
+ treedata = treedata[i + 21:]
+
+ def __call__ (self, parser, obj, sha1, t, size, data):
+ """Git object callback (see GitObjectFetcher documentation)."""
+ assert parser == self.parser
+ if not sha1: # Missing object
+ self.err = "Missing object '%s'" % (obj)
+ return
+ assert size == len(data)
+ if t == "tree":
+ if self.path:
+ self.path += "/"
+ # Recurse into all blobs and subtrees
+ for m, s, p in self.parse_tree(data):
+ parser.push(s,
+ self.__class__(self.parser, self.cb, p, m))
+ elif t == "blob":
+ self.cb(self.path, self.mode, data)
+ else:
+ raise ValueError("Unknown object type '%s'" % (t))
+
+ self.push(tree_objspec, _ObjHandler(self, callback, prefix))
+ self.process()
+
+
+class GitRefMap(object):
+
+ """Map Git ref names to the Git object names they currently point to.
+
+ Behaves like a dictionary of Git ref names -> Git object names.
+
+ """
+
+ def __init__ (self, obj_fetcher):
+ """Create a new Git ref -> object map."""
+ self.obj_fetcher = obj_fetcher
+ self._cache = {} # dict: refname -> objname
+
+ def _load (self, ref):
+ """Retrieve the object currently bound to the given ref.
+
+ The name of the object pointed to by the given ref is stored
+ into this mapping, and also returned.
+
+ """
+ if ref not in self._cache:
+ self._cache[ref] = self.obj_fetcher.get_sha1(ref)
+ return self._cache[ref]
+
+ def __contains__ (self, refname):
+ """Return True if the given refname is present in this cache."""
+ return bool(self._load(refname))
+
+ def __getitem__ (self, refname):
+ """Return the git object name pointed to by the given refname."""
+ commit = self._load(refname)
+ if commit is None:
+ raise KeyError("Unknown ref '%s'" % (refname))
+ return commit
+
+ def get (self, refname, default = None):
+ """Return the git object name pointed to by the given refname."""
+ commit = self._load(refname)
+ if commit is None:
+ return default
+ return commit
+
+
+class GitFICommit(object):
+
+ """Encapsulate the data in a Git fast-import commit command."""
+
+ SHA1RE = re.compile(r'^[0-9a-f]{40}$')
+
+ @classmethod
+ def parse_mode (cls, mode):
+ """Verify the given git file mode, and return it as a string."""
+ assert mode in (644, 755, 100644, 100755, 120000)
+ return "%i" % (mode)
+
+ @classmethod
+ def parse_objname (cls, objname):
+ """Return the given object name (or mark number) as a string."""
+ if isinstance(objname, int): # Object name is a mark number
+ assert objname > 0
+ return ":%i" % (objname)
+
+ # No existence check is done, only checks for valid format
+ assert cls.SHA1RE.match(objname) # Object name is valid SHA1
+ return objname
+
+ @classmethod
+ def quote_path (cls, path):
+ """Return a quoted version of the given path."""
+ path = path.replace("\\", "\\\\")
+ path = path.replace("\n", "\\n")
+ path = path.replace('"', '\\"')
+ return '"%s"' % (path)
+
+ @classmethod
+ def parse_path (cls, path):
+ """Verify that the given path is valid, and quote it, if needed."""
+ assert not isinstance(path, int) # Cannot be a mark number
+
+ # These checks verify the rules on the fast-import man page
+ assert not path.count("//")
+ assert not path.endswith("/")
+ assert not path.startswith("/")
+ assert not path.count("/./")
+ assert not path.count("/../")
+ assert not path.endswith("/.")
+ assert not path.endswith("/..")
+ assert not path.startswith("./")
+ assert not path.startswith("../")
+
+ if path.count('"') + path.count('\n') + path.count('\\'):
+ return cls.quote_path(path)
+ return path
+
+ def __init__ (self, name, email, timestamp, timezone, message):
+ """Create a new Git fast-import commit, with the given metadata."""
+ self.name = name
+ self.email = email
+ self.timestamp = timestamp
+ self.timezone = timezone
+ self.message = message
+ self.pathops = [] # List of path operations in this commit
+
+ def modify (self, mode, blobname, path):
+ """Add a file modification to this Git fast-import commit."""
+ self.pathops.append(("M",
+ self.parse_mode(mode),
+ self.parse_objname(blobname),
+ self.parse_path(path)))
+
+ def delete (self, path):
+ """Add a file deletion to this Git fast-import commit."""
+ self.pathops.append(("D", self.parse_path(path)))
+
+ def copy (self, path, newpath):
+ """Add a file copy to this Git fast-import commit."""
+ self.pathops.append(("C",
+ self.parse_path(path),
+ self.parse_path(newpath)))
+
+ def rename (self, path, newpath):
+ """Add a file rename to this Git fast-import commit."""
+ self.pathops.append(("R",
+ self.parse_path(path),
+ self.parse_path(newpath)))
+
+ def note (self, blobname, commit):
+ """Add a note object to this Git fast-import commit."""
+ self.pathops.append(("N",
+ self.parse_objname(blobname),
+ self.parse_objname(commit)))
+
+ def deleteall (self):
+ """Delete all files in this Git fast-import commit."""
+ self.pathops.append("deleteall")
+
+
+class TestGitFICommit(unittest.TestCase):
+
+ """GitFICommit selftests."""
+
+ def test_basic (self):
+ """GitFICommit basic selftests."""
+
+ def expect_fail (method, data):
+ """Verify that the method(data) raises an AssertionError."""
+ try:
+ method(data)
+ except AssertionError:
+ return
+ raise AssertionError("Failed test for invalid data '%s(%s)'" %
+ (method.__name__, repr(data)))
+
+ def test_parse_mode (self):
+ """GitFICommit.parse_mode() selftests."""
+ self.assertEqual(GitFICommit.parse_mode(644), "644")
+ self.assertEqual(GitFICommit.parse_mode(755), "755")
+ self.assertEqual(GitFICommit.parse_mode(100644), "100644")
+ self.assertEqual(GitFICommit.parse_mode(100755), "100755")
+ self.assertEqual(GitFICommit.parse_mode(120000), "120000")
+ self.assertRaises(AssertionError, GitFICommit.parse_mode, 0)
+ self.assertRaises(AssertionError, GitFICommit.parse_mode, 123)
+ self.assertRaises(AssertionError, GitFICommit.parse_mode, 600)
+ self.assertRaises(AssertionError, GitFICommit.parse_mode, "644")
+ self.assertRaises(AssertionError, GitFICommit.parse_mode, "abc")
+
+ def test_parse_objname (self):
+ """GitFICommit.parse_objname() selftests."""
+ self.assertEqual(GitFICommit.parse_objname(1), ":1")
+ self.assertRaises(AssertionError, GitFICommit.parse_objname, 0)
+ self.assertRaises(AssertionError, GitFICommit.parse_objname, -1)
+ self.assertEqual(GitFICommit.parse_objname("0123456789" * 4),
+ "0123456789" * 4)
+ self.assertEqual(GitFICommit.parse_objname("2468abcdef" * 4),
+ "2468abcdef" * 4)
+ self.assertRaises(AssertionError, GitFICommit.parse_objname,
+ "abcdefghij" * 4)
+
+ def test_parse_path (self):
+ """GitFICommit.parse_path() selftests."""
+ self.assertEqual(GitFICommit.parse_path("foo/bar"), "foo/bar")
+ self.assertEqual(GitFICommit.parse_path("path/with\n and \" in it"),
+ '"path/with\\n and \\" in it"')
+ self.assertRaises(AssertionError, GitFICommit.parse_path, 1)
+ self.assertRaises(AssertionError, GitFICommit.parse_path, 0)
+ self.assertRaises(AssertionError, GitFICommit.parse_path, -1)
+ self.assertRaises(AssertionError, GitFICommit.parse_path, "foo//bar")
+ self.assertRaises(AssertionError, GitFICommit.parse_path, "foo/bar/")
+ self.assertRaises(AssertionError, GitFICommit.parse_path, "/foo/bar")
+ self.assertRaises(AssertionError, GitFICommit.parse_path, "foo/./bar")
+ self.assertRaises(AssertionError, GitFICommit.parse_path, "foo/../bar")
+ self.assertRaises(AssertionError, GitFICommit.parse_path, "foo/bar/.")
+ self.assertRaises(AssertionError, GitFICommit.parse_path, "foo/bar/..")
+ self.assertRaises(AssertionError, GitFICommit.parse_path, "./foo/bar")
+ self.assertRaises(AssertionError, GitFICommit.parse_path, "../foo/bar")
+
+
+class GitFastImport(object):
+
+ """Encapsulate communication with git fast-import."""
+
+ def __init__ (self, f, obj_fetcher, last_mark = 0):
+ """Set up self to communicate with a fast-import process through f."""
+ self.f = f # File object where fast-import stream is written
+ self.obj_fetcher = obj_fetcher # GitObjectFetcher instance
+ self.next_mark = last_mark + 1 # Next mark number
+ self.refs = set() # Keep track of the refnames we've seen
+
+ def comment (self, s):
+ """Write the given comment in the fast-import stream."""
+ assert "\n" not in s, "Malformed comment: '%s'" % (s)
+ self.f.write("# %s\n" % (s))
+
+ def commit (self, ref, commitdata):
+ """Make a commit on the given ref, with the given GitFICommit.
+
+ Return the mark number identifying this commit.
+
+ """
+ self.f.write("""\
+commit %(ref)s
+mark :%(mark)i
+committer %(name)s <%(email)s> %(timestamp)i %(timezone)s
+data %(msgLength)i
+%(msg)s
+""" % {
+ 'ref': ref,
+ 'mark': self.next_mark,
+ 'name': commitdata.name,
+ 'email': commitdata.email,
+ 'timestamp': commitdata.timestamp,
+ 'timezone': commitdata.timezone,
+ 'msgLength': len(commitdata.message),
+ 'msg': commitdata.message,
+})
+
+ if ref not in self.refs:
+ self.refs.add(ref)
+ parent = ref + "^0"
+ if self.obj_fetcher.get_sha1(parent):
+ self.f.write("from %s\n" % (parent))
+
+ for op in commitdata.pathops:
+ self.f.write(" ".join(op))
+ self.f.write("\n")
+ self.f.write("\n")
+ retval = self.next_mark
+ self.next_mark += 1
+ return retval
+
+ def blob (self, data):
+ """Import the given blob.
+
+ Return the mark number identifying this blob.
+
+ """
+ self.f.write("blob\nmark :%i\ndata %i\n%s\n" %
+ (self.next_mark, len(data), data))
+ retval = self.next_mark
+ self.next_mark += 1
+ return retval
+
+ def reset (self, ref, objname):
+ """Reset the given ref to point at the given Git object."""
+ self.f.write("reset %s\nfrom %s\n\n" %
+ (ref, GitFICommit.parse_objname(objname)))
+ if ref not in self.refs:
+ self.refs.add(ref)
+
+
+class GitNotes(object):
+
+ """Encapsulate access to Git notes.
+
+ Simulates a dictionary of object name (SHA1) -> Git note mappings.
+
+ """
+
+ def __init__ (self, notes_ref, obj_fetcher):
+ """Create a new Git notes interface, bound to the given notes ref."""
+ self.notes_ref = notes_ref
+ self.obj_fetcher = obj_fetcher # Used to get objects from repo
+ self.imports = [] # list: (objname, note data blob name) tuples
+
+ def __del__ (self):
+ """Verify that self.commit_notes() was called before destruction."""
+ if self.imports:
+ error("Missing call to self.commit_notes().")
+ error("%i notes are not committed!", len(self.imports))
+
+ def _load (self, objname):
+ """Return the note data associated with the given git object.
+
+ The note data is returned in string form. If no note is found
+ for the given object, None is returned.
+
+ """
+ try:
+ f = self.obj_fetcher.open_obj("%s:%s" % (self.notes_ref, objname))
+ ret = f.read()
+ f.close()
+ except KeyError:
+ ret = None
+ return ret
+
+ def __getitem__ (self, objname):
+ """Return the note contents associated with the given object.
+
+ Raise KeyError if given object has no associated note.
+
+ """
+ blobdata = self._load(objname)
+ if blobdata is None:
+ raise KeyError("Object '%s' has no note" % (objname))
+ return blobdata
+
+ def get (self, objname, default = None):
+ """Return the note contents associated with the given object.
+
+ Return given default if given object has no associated note.
+
+ """
+ blobdata = self._load(objname)
+ if blobdata is None:
+ return default
+ return blobdata
+
+ def import_note (self, objname, data, gfi):
+ """Tell git fast-import to store data as a note for objname.
+
+ This method uses the given GitFastImport object to create a
+ blob containing the given note data. Also an entry mapping the
+ given object name to the created blob is stored until
+ commit_notes() is called.
+
+ Note that this method only works if it is later followed by a
+ call to self.commit_notes() (which produces the note commit
+ that refers to the blob produced here).
+
+ """
+ if not data.endswith("\n"):
+ data += "\n"
+ gfi.comment("Importing note for object %s" % (objname))
+ mark = gfi.blob(data)
+ self.imports.append((objname, mark))
+
+ def commit_notes (self, gfi, author, message):
+ """Produce a git fast-import note commit for the imported notes.
+
+ This method uses the given GitFastImport object to create a
+ commit on the notes ref, introducing the notes previously
+ submitted to import_note().
+
+ """
+ if not self.imports:
+ return
+ now = CVSDate("now")
+ commitdata = GitFICommit(author[0], author[1],
+ now.ts, now.tz_str(), message)
+ for objname, blobname in self.imports:
+ assert isinstance(objname, int) and objname > 0
+ assert isinstance(blobname, int) and blobname > 0
+ commitdata.note(blobname, objname)
+ gfi.commit(self.notes_ref, commitdata)
+ self.imports = []
+
+
+class GitCachedNotes(GitNotes):
+
+ """Encapsulate access to Git notes (cached version).
+
+ Only use this class if no caching is done at a higher level.
+
+ Simulates a dictionary of object name (SHA1) -> Git note mappings.
+
+ """
+
+ def __init__ (self, notes_ref, obj_fetcher):
+ """Set up a caching wrapper around GitNotes."""
+ GitNotes.__init__(self, notes_ref, obj_fetcher)
+ self._cache = {} # Cache: object name -> note data
+
+ def __del__ (self):
+ """Verify that GitNotes' destructor is called."""
+ GitNotes.__del__(self)
+
+ def _load (self, objname):
+ """Extend GitNotes._load() with a local objname -> note cache."""
+ if objname not in self._cache:
+ self._cache[objname] = GitNotes._load(self, objname)
+ return self._cache[objname]
+
+ def import_note (self, objname, data, gfi):
+ """Extend GitNotes.import_note() with a local objname -> note cache."""
+ if not data.endswith("\n"):
+ data += "\n"
+ assert objname not in self._cache
+ self._cache[objname] = data
+ GitNotes.import_note(self, objname, data, gfi)
+
+
+if __name__ == '__main__':
+ unittest.main()
diff --git a/git_remote_cvs/setup.py b/git_remote_cvs/setup.py
new file mode 100644
index 0000000..21f521a
--- /dev/null
+++ b/git_remote_cvs/setup.py
@@ -0,0 +1,17 @@
+#!/usr/bin/env python
+
+"""Distutils build/install script for the git_remote_cvs package."""
+
+from distutils.core import setup
+
+setup(
+ name = 'git_remote_cvs',
+ version = '0.1.0',
+ description = 'Git remote helper program for CVS repositories',
+ license = 'GPLv2',
+ author = 'The Git Community',
+ author_email = 'git@vger.kernel.org',
+ url = 'http://www.git-scm.com/',
+ package_dir = {'git_remote_cvs': ''},
+ packages = ['git_remote_cvs'],
+)
--
1.6.4.262.gca66a.dirty
^ permalink raw reply related
* [RFCv4 4/5] git-remote-cvs: Remote helper program for CVS repositories
From: Johan Herland @ 2009-08-17 3:36 UTC (permalink / raw)
To: git; +Cc: Johannes.Schindelin, Johan Herland, barkalow, davvid, gitster
In-Reply-To: <1250480161-21933-1-git-send-email-johan@herland.net>
Implements the import of objects from a local or remote CVS repository.
This helper program uses the "git_remote_cvs" Python package introduced
earlier, and provides a working draft implementation of the remote helper
API, as described in Documentation/git-remote-helpers.txt. Further details
about this specific helper are described in the new
Documentation/git-remote-cvs.txt.
This patch has been improved by the following contributions:
- Daniel Barkalow: Updates reflecting changes in remote helper API
- David Aguilar: Lots of Python coding style fixes
Cc: Daniel Barkalow <barkalow@iabervon.org>
Cc: David Aguilar <davvid@gmail.com>
Signed-off-by: Johan Herland <johan@herland.net>
---
Documentation/git-remote-cvs.txt | 85 +++++
Makefile | 24 ++
git-remote-cvs.py | 683 ++++++++++++++++++++++++++++++++++++++
3 files changed, 792 insertions(+), 0 deletions(-)
create mode 100644 Documentation/git-remote-cvs.txt
create mode 100755 git-remote-cvs.py
diff --git a/Documentation/git-remote-cvs.txt b/Documentation/git-remote-cvs.txt
new file mode 100644
index 0000000..309ee7f
--- /dev/null
+++ b/Documentation/git-remote-cvs.txt
@@ -0,0 +1,85 @@
+git-remote-cvs(1)
+=================
+
+NAME
+----
+git-remote-cvs - Helper program for interoperation with CVS repositories
+
+SYNOPSIS
+--------
+'git remote-cvs' <remote>
+
+DESCRIPTION
+-----------
+
+Please see the linkgit:git-remote-helpers[1] documentation for general
+information about remote helper programs.
+
+CONFIGURATION
+-------------
+
+remote.*.cvsRoot::
+ The URL of the CVS repository (as found in a `CVSROOT` variable, or
+ in a `CVS/Root` file).
+ Example: "`:pserver:user@server/var/cvs/cvsroot`".
+
+remote.*.cvsModule::
+ The path of the CVS module (as found in a `CVS/Repository` file)
+ within the CVS repository specified in `remote.*.cvsRoot`.
+ Example: "`foo/bar`"
+
+remote.*.cachedSymbolsOnly::
+ When 'true', a cache of CVS symbols is used instead of querying the
+ CVS server for all existing symbols (potentially expensive). In this
+ mode, git-remote-cvs will not discover new CVS symbols unless you add
+ them explicitly with the "`addsymbol <symbol>`" command (on
+ git-remote-cvs's stdin), or request an explicit symbol cache update
+ from the CVS server with the "`syncsymbols`" command (on
+ git-remote-cvs's stdin). When 'false' (the default), the CVS server
+ will be queried whenever a list of CVS symbols is required.
+
+remote.*.usernameMap::
+ The path (absolute, or relative to the repository (NOT the worktree))
+ to the file that contains the mapping from CVS usernames to the
+ corresponding full names and email addresses, as used by Git in the
+ Author and Committer fields of commit objects. When this config
+ variable is set, CVS usernames will be resolved against this file.
+ If no match is found in the file, or if this config variable is unset,
+ or if the variable points to a non-existing file, the original CVS
+ username will be used as the Author/Committer name, and the
+ corresponding email address will be set to "`<username>@example.com`".
++
+The format of the usernameMap file is one entry per line, where each line is
+of the form "`username: Full Name <email@address>`".
+Example: `johndoe: John Doe <johndoe@example.com>`
+Blank lines and lines starting with '#' are ignored.
+
+COMMANDS
+--------
+
+In addition to the commands that constitute the git-remote-helpers API, the
+following extra commands are supported for managing the local symbol cache when
+the `remote.*.cachedSymbolsOnly` config variable is true. The following
+commands can be given on the standard input of git-remote-cvs:
+
+'addsymbol'::
+ Takes one CVS symbol name as argument. The given CVS symbol is
+ fetched from the CVS server and stored into the local CVS symbol
+ cache. If `remote.*.cachedSymbolsOnly` is enabled, this can be used
+ to introduce a new CVS symbol to the CVS helper application.
+
+'syncsymbols'::
+ All CVS symbols that are available from the given remote are
+ fetched from the CVS server and stored into the local CVS symbol
+ cache. This is equivalent to disabling `remote.{asterisk}.cachedSymbolsOnly`,
+ running the "list" command, and then finally re-enabling the
+ `remote.*.cachedSymbolsOnly` config variable. I.e. this command can
+ be used to manually synchronize the CVS symbols available to the
+ CVS helper application.
+
+'verify'::
+ Takes one CVS symbol name as argument. Verifies that the CVS symbol
+ has been successfully imported be checking out the CVS symbol from
+ the CVS server, and comparing the CVS working tree against the Git
+ tree object identified by `refs/cvs/<remote>/<symbol>`. This can be
+ used to verify the correctness of a preceding 'import' command.
diff --git a/Makefile b/Makefile
index bb5cea2..b2af678 100644
--- a/Makefile
+++ b/Makefile
@@ -350,6 +350,8 @@ SCRIPT_PERL += git-relink.perl
SCRIPT_PERL += git-send-email.perl
SCRIPT_PERL += git-svn.perl
+SCRIPT_PYTHON += git-remote-cvs.py
+
SCRIPTS = $(patsubst %.sh,%,$(SCRIPT_SH)) \
$(patsubst %.perl,%,$(SCRIPT_PERL)) \
$(patsubst %.py,%,$(SCRIPT_PYTHON)) \
@@ -1474,6 +1476,28 @@ $(patsubst %.perl,%,$(SCRIPT_PERL)) git-instaweb: % : unimplemented.sh
mv $@+ $@
endif # NO_PERL
+ifndef NO_PYTHON
+$(patsubst %.py,%,$(SCRIPT_PYTHON)): % : %.py
+ $(QUIET_GEN)$(RM) $@ $@+ && \
+ INSTLIBDIR=`MAKEFLAGS= $(MAKE) -C git_remote_cvs -s --no-print-directory instlibdir` && \
+ sed -e '1{' \
+ -e ' s|#!.*python|#!$(PYTHON_PATH_SQ)|' \
+ -e '}' \
+ -e 's|^import sys.*|&; sys.path.insert(0, "@@INSTLIBDIR@@")|' \
+ -e 's|@@INSTLIBDIR@@|'"$$INSTLIBDIR"'|g' \
+ $@.py >$@+ && \
+ chmod +x $@+ && \
+ mv $@+ $@
+else # NO_PYTHON
+$(patsubst %.py,%,$(SCRIPT_PYTHON)): % : unimplemented.sh
+ $(QUIET_GEN)$(RM) $@ $@+ && \
+ sed -e '1s|#!.*/sh|#!$(SHELL_PATH_SQ)|' \
+ -e 's|@@REASON@@|NO_PYTHON=$(NO_PYTHON)|g' \
+ unimplemented.sh >$@+ && \
+ chmod +x $@+ && \
+ mv $@+ $@
+endif # NO_PYTHON
+
configure: configure.ac
$(QUIET_GEN)$(RM) $@ $<+ && \
sed -e 's/@@GIT_VERSION@@/$(GIT_VERSION)/g' \
diff --git a/git-remote-cvs.py b/git-remote-cvs.py
new file mode 100755
index 0000000..94b61e7
--- /dev/null
+++ b/git-remote-cvs.py
@@ -0,0 +1,683 @@
+#!/usr/bin/env python
+
+"""Usage: git-remote-cvs <remote> [<url>]
+
+Git remote helper command for interacting with CVS repositories
+
+See git-remote-helpers documentation for details on external interface,
+usage, etc. See git-remote-cvs documentation for specific
+configuration details of this remote helper.
+
+"""
+
+# PRINCIPLES:
+# -----------
+# - Importing same symbol twice (with no CVS changes in between) should
+# yield the exact same Git state (and the second import should have
+# no commits).
+# - Importing several CVS symbols pointing to the same state should
+# yield corresponding refs pointing to the _same_ commit in Git.
+# - Importing a CVS symbol which has received only "regular commits"
+# since last import should yield a fast-forward straight line of
+# commits.
+
+# TODO / KNOWN PROBLEMS:
+# ----------------------
+# - Remove cachedSymbolsOnly config variable for now?
+# - Author map handling; map CVS usernames to Git full name + email
+# - Handle files that have been created AND deleted since last import
+# - How to handle CVS tags vs. branches. Turn CVS tags into Git tags?
+# - Better CVS branch handling: When a branch as a super/subset of
+# files/revs compared to another branch, find a way to base one branch
+# on the other instead of creating parallel lines of development with
+# roughly the same commits.
+# - Profiling, optimizations...
+
+import sys
+import os
+
+from git_remote_cvs.util import (debug, error, die, ProgressIndicator,
+ run_command)
+from git_remote_cvs.cvs import CVSState, CVSWorkDir, fetch_revs
+from git_remote_cvs.git import (get_git_dir, parse_git_config, git_config_bool,
+ valid_git_ref, GitObjectFetcher, GitRefMap,
+ GitFICommit, GitFastImport, GitNotes)
+from git_remote_cvs.cvs_symbol_cache import CVSSymbolCache
+from git_remote_cvs.commit_states import CommitStates
+from git_remote_cvs.cvs_revision_map import CVSRevisionMap, CVSStateMap
+from git_remote_cvs.changeset import build_changesets_from_revs
+
+
+class Config(object):
+
+ """Encapsulation of configuration variables."""
+
+ # Author name/email tuple for commits created by this tool
+ Author = ("git remote-cvs", "invalid@example.com")
+ # Git remote name
+ Remote = None
+ # CVS symbols are imported into this refs namespace/directory
+ RefSpace = None
+ # Git notes ref, the refname pointing to our git notes
+ NotesRef = None
+ # CVS repository identifier, a 2-tuple (cvs_root, cvs_module),
+ # where cvs_root is the CVS server/repository URL (as found in
+ # $CVSROOT, or in a CVS/Root file), and cvs_module is the path
+ # to a CVS module relative to the CVS repository (as found in a
+ # CVS/Repository file)
+ CVSRepo = (None, None)
+ # Path to the git-remote-cvs cache/work directory
+ # (normally "info/cvs/$remote" within $GIT_DIR)
+ WorkDir = None
+ # If False, the list of CVS symbols will always be retrieved from
+ # the CVS server using 'cvs rlog'. If True, only the cached
+ # symbols within the "symbols" subdirectory of WorkDir are
+ # consulted.
+ CachedSymbolsOnly = False
+
+ @classmethod
+ def init (cls, remote):
+ """Fetch and setup configuration for the given remote."""
+ git_config = parse_git_config()
+ assert git_config["remote.%s.vcs" % (remote)] == "cvs"
+
+ cls.Author = (git_config["user.name"], git_config["user.email"])
+ cls.Remote = remote
+ cls.RefSpace = "refs/cvs/%s/" % (remote)
+ cls.NotesRef = "refs/notes/cvs/%s" % (remote)
+ cls.CVSRepo = (git_config["remote.%s.cvsroot" % (remote)],
+ git_config["remote.%s.cvsmodule" % (remote)])
+ cls.WorkDir = os.path.join(get_git_dir(), "info/cvs", remote)
+ cls.CachedSymbolsOnly = git_config_bool(git_config.get(
+ "remote.%s.cachedsymbolsonly" % (remote), "false"))
+
+
+def work_path (*args):
+ """Return the given path appended to git-remote-cvs's cache/work dir."""
+ return os.path.join(Config.WorkDir, *args)
+
+
+def cvs_to_refname (cvsname):
+ """Return the git ref name for the given CVS symbolic name."""
+ if cvsname.startswith(Config.RefSpace): # Already converted
+ return cvsname
+ return Config.RefSpace + cvsname
+
+
+def ref_to_cvsname (refname):
+ """Return the CVS symbolic name for the given git ref name."""
+ if refname.startswith(Config.RefSpace):
+ return refname[len(Config.RefSpace):]
+ return refname
+
+
+def valid_cvs_symbol (symbol):
+ """Return True iff the given CVS symbol can be imported into Git."""
+ return valid_git_ref(cvs_to_refname(symbol))
+
+
+def die_usage (msg, *args):
+ """Abort program with a helpful usage message."""
+ # Use this file's docstring as a usage string
+ print >> sys.stderr, __doc__
+ die(msg, *args)
+
+
+def import_cvs_revs (symbol, prev_state, cur_state, progress):
+ """Import the CVS revisions needed to satisfy the given CVS symbol.
+
+ This method will determine the CVS revisions involved in moving
+ from the given prev_state to the given cur_state. This includes
+ looking at revision metadata in CVS, and importing needed blobs
+ from CVS.
+
+ The revision metadata is returned as a 2-level dict of CVSRev
+ objects: mapping path -> revnum -> CVSRev object.
+
+ """
+ # Calculate the revisions involved in moving from prev_state to
+ # cur_state, and fetch CVSRev objects for these revisions.
+ progress.pushprefix("Importing CVS revisions: ")
+ paths = set(prev_state.paths()).union(cur_state.paths())
+ num_fetched_revs = 0 # Number of CVSRev objects involved
+ num_imported_blobs = 0 # Number of blobs actually imported
+ cvs_revs = {} # path -> revnum -> CVSRev
+ for i, path in enumerate(sorted(paths)):
+ progress.pushprefix("(%i/%i) %s: " % (i + 1, len(paths), path))
+ progress("")
+ prev_rev = prev_state.get(path)
+ cur_rev = cur_state.get(path)
+ if prev_rev and cur_rev and prev_rev == cur_rev:
+ # No changes since last import
+ progress.popprefix()
+ continue
+
+ # Fetch CVSRev objects for range [path:prev_rev, path:symbol]
+ path_revs = fetch_revs(path, prev_rev, cur_rev, symbol, Config.CVSRepo)
+ if not path_revs:
+ # Failed to find revs between prev_rev and symbol
+ if cur_rev:
+ assert not cur_rev.follows(prev_rev)
+ # The CVS symbol has been moved/reset since the
+ # last import in such a way that we cannot
+ # deduce the history between the last import
+ # and the current import.
+ # FIXME: Can we can work around this?
+ die("CVS symbol %s has been moved/reset from %s:%s to %s:%s "
+ "since the last import. This is not supported",
+ symbol, path, prev_rev, path, cur_rev)
+ else:
+ # CVS symbol has been removed from this path.
+ # We cannot conclusively determine the history
+ # of this path following prev_rev.
+ # FIXME: Can we can work around this?
+ die("CVS symbol %s has been removed from %s since the last "
+ "import. This is not supported",
+ symbol, path)
+
+ # OK. We've got the revs in range [prev_rev, symbol]
+ # Verify/determine cur_rev
+ real_cur_rev = max(path_revs.keys())
+ if cur_rev:
+ assert cur_rev == real_cur_rev
+ else: cur_rev = real_cur_rev
+ # No need to re-import prev_rev if already imported
+ if prev_rev:
+ assert cur_rev.follows(prev_rev)
+ assert prev_rev in path_revs
+ del path_revs[prev_rev]
+ assert path_revs # There should be more revs than just prev_rev
+
+ # Sanity checks:
+ # All revs from prev_rev to cur_rev are about to be imported
+ check_rev = cur_rev
+ while check_rev and check_rev != prev_rev:
+ assert check_rev in path_revs
+ check_rev = check_rev.parent()
+ # All previous revs have already been imported
+ check_rev = prev_rev
+ while check_rev:
+ assert Globals.CVSRevisionMap.has_rev(path, check_rev)
+ check_rev = check_rev.parent()
+
+ # Import CVS revisions as Git blobs
+ j = 0
+ for num, rev in sorted(path_revs.iteritems(), reverse = True):
+ j += 1
+ progress("(%i/%i) %s" % (j, len(path_revs), num))
+ assert num == rev.num
+ # Skip if already imported
+ if Globals.CVSRevisionMap.has_rev(rev.path, rev.num):
+ continue
+ # ...or if rev is a deletion
+ elif rev.deleted:
+ continue
+ # Import blob for reals
+ data = Globals.CVSWorkDir.get_revision_data(rev.path, rev.num)
+ Globals.GitFastImport.comment("Importing CVS revision %s:%s" %
+ (rev.path, rev.num))
+ mark = Globals.GitFastImport.blob(data)
+ Globals.CVSRevisionMap.add_blob(rev.path, rev.num, mark)
+ num_imported_blobs += 1
+
+ # Add path_revs to the overall structure of revs to be imported
+ assert path not in cvs_revs
+ cvs_revs[path] = path_revs
+ num_fetched_revs += len(path_revs)
+ progress.popprefix()
+ progress.popprefix()
+ progress("Imported %i blobs (reused %i existing blobs)" %
+ (num_imported_blobs, num_fetched_revs - num_imported_blobs), True)
+
+ return cvs_revs
+
+
+def advance_state (state, changeset):
+ """Advance the given state by applying the given changeset."""
+ # Verify that the given changeset "fits" on top of the given state
+ for rev in changeset:
+ prev_num = rev.num.parent()
+ state_num = state.get(rev.path)
+ if prev_num is None and state_num is None:
+ # rev is the first revision of this path being added
+ state.add(rev.path, rev.num)
+ elif prev_num and state_num and prev_num == state_num:
+ if rev.deleted: # rev deletes path from state
+ state.remove(rev.path, prev_num)
+ else: # rev follows state's revision of this path
+ state.replace(rev.path, rev.num)
+ else:
+ error("Cannot apply changeset with %s:%s on top of CVS state "
+ "with %s:%s.",
+ rev.path, changeset[rev.path].num,
+ rev.path, state.get(rev.path))
+ error(" changeset: %s", changeset)
+ error(" CVS state: \n---\n%s---", state)
+ die("Failed to apply changeset. Aborting.")
+
+
+def revert_state (state, changeset):
+ """Revert the given state to before the given changeset is applied.
+
+ This is the reverse of the above advance_state() function.
+
+ """
+ for rev in changeset:
+ prev_num = rev.num.parent()
+ state_num = state.get(rev.path)
+ if state_num is None: # Revert deletion of file
+ assert rev.deleted
+ state.add(rev.path, prev_num)
+ else:
+ assert state_num == rev.num
+ if prev_num is None: # Revert addition of file
+ state.remove(rev.path, rev.num)
+ else: # Regular revert to previous version
+ state.replace(rev.path, prev_num)
+
+
+def import_changesets (ref, changesets, from_state, to_state, progress):
+ """Apply the given list of Changeset objects to the given ref.
+
+ Also verify that the changesets bring us from the given from_state
+ to the given to_state.
+
+ """
+ state = from_state
+ for i, c in enumerate(changesets):
+ advance_state(state, c)
+ progress("(%i/%i) Committing %s" % (i + 1, len(changesets), c))
+ # Make a git commit from changeset c
+ commitdata = GitFICommit(
+ c.author, c.author + "@example.com", # TODO: author_map handling
+ c.date.ts, c.date.tz_str(),
+ "".join(["%s\n" % (line) for line in c.message]))
+
+ for rev in c:
+ p, n = rev.path, rev.num
+ if rev.deleted:
+ commitdata.delete(p)
+ continue
+ blobname = Globals.CVSRevisionMap.get_blob(p, n)
+ mode = Globals.CVSRevisionMap.get_mode(p)
+ if mode is None: # Must retrieve mode from CVS checkout
+ debug("Retrieving mode info for '%s'" % (p))
+ Globals.CVSWorkDir.update(n, [p])
+ mode = Globals.CVSWorkDir.get_modeinfo([p])[p]
+ Globals.CVSRevisionMap.add_path(p, mode)
+ commitdata.modify(mode, blobname, p)
+
+ commitname = Globals.GitFastImport.commit(ref, commitdata)
+ Globals.CommitStates.add(commitname, state, Globals.GitFastImport)
+ for path, revnum in state:
+ Globals.CVSRevisionMap.add_commit(path, revnum, commitname)
+ assert commitname in Globals.CVSStateMap.get_commits(state)
+ assert state == to_state
+ return len(changesets)
+
+
+def import_cvs_symbol (cvs_symbol, progress):
+ """Import the given CVS symbol from CVS to Git.
+
+ Return False if nothing was imported, True otherwise.
+
+ """
+ git_ref = cvs_to_refname(cvs_symbol)
+ progress.pushprefix("%s: " % (cvs_symbol))
+
+ # Verify that we are asked to import valid git ref names
+ if not valid_git_ref(git_ref):
+ progress("Invalid git ref '%s'. Skipping." % (git_ref), True)
+ progress.popprefix()
+ return False
+
+ # Retrieve previously imported CVS state
+ progress("Loading previously imported state of %s..." % (git_ref))
+ prev_commit = Globals.GitRefMap.get(git_ref)
+ prev_state = Globals.CommitStates.get(prev_commit, CVSState())
+
+ # Retrieve current CVS state of symbol
+ # Also: At some point we will need mode information for all CVS
+ # paths (stored in CVSRevisionMap). This information can be added
+ # for each path on demand (using CVSWorkDir.get_modeinfo()), but
+ # doing so may be an expensive process. It is much cheaper to load
+ # mode information for as many paths as possible in a _single_
+ # operation. We do this below, by calling
+ # CVSRevisionMap.sync_modeinfo_from_cvs() in appropriate places.
+ if Config.CachedSymbolsOnly:
+ progress("Synchronizing local CVS symbol cache for symbol...")
+ # The symbol cache is likely not up-to-date. Synchronize the
+ # given CVS symbol explicitly, to make sure we get the version
+ # current with the CVS server.
+ Globals.CVSSymbolCache.sync_symbol(cvs_symbol, Globals.CVSWorkDir,
+ progress)
+ # The above method updates the CVS workdir to the current CVS
+ # version. Hence, now is a convenient time to preload mode
+ # info from the currently checked-out CVS files. There may be
+ # more files for which we'll need mode information, but we'll
+ # deal with those when needed.
+ progress("Updating path mode info from current CVS checkout.")
+ Globals.CVSRevisionMap.sync_modeinfo_from_cvs(Globals.CVSWorkDir)
+ elif not Globals.CVSRevisionMap: # No info for any paths, yet
+ # Pure optimization: We didn't get to preload all the mode info
+ # above. Normally, the only alternative is load mode info for
+ # each path on-demand. However, if our CVSRevisionMap is
+ # currently empty, that's probably going to be very expensive.
+ # Therefore, in this case, do an explicit CVS update here, and
+ # preload mode info for all paths.
+ progress("Updating CVS checkout to sync path mode info.")
+ Globals.CVSWorkDir.update(cvs_symbol)
+ Globals.CVSRevisionMap.sync_modeinfo_from_cvs(Globals.CVSWorkDir)
+
+ progress("Loading current CVS state...")
+ try:
+ cur_state = Globals.CVSSymbolCache[cvs_symbol]
+ except KeyError:
+ progress("Couldn't find symbol '%s'. Skipping." % (cvs_symbol), True)
+ progress.popprefix()
+ return False
+
+ # Optimization: Check if the previous import of this symbol is
+ # still up-to-date. If so, there's nothing more to be done.
+ progress("Checking if we're already up-to-date...")
+ if cur_state == prev_state:
+ progress("Already up-to-date. Skipping.", True)
+ progress.popprefix()
+ return False
+
+ progress("Fetching CVS revisions...")
+ cvs_revs = import_cvs_revs(cvs_symbol, prev_state, cur_state, progress)
+ progress("Organizing revisions into chronological list of changesets...")
+ changesets = build_changesets_from_revs(cvs_revs)
+
+ # When importing a new branch, try to optimize branch start point,
+ # instead of importing entire branch from scratch
+ if prev_commit is None:
+ progress("Finding startpoint for new symbol...")
+ i = len(changesets)
+ state = cur_state.copy()
+ for c in reversed(changesets):
+ commit = Globals.CVSStateMap.get_exact_commit(state,
+ Globals.CommitStates)
+ if commit is not None:
+ # We have found a commit that exactly matches the state
+ # after commit #i (changesets[i - 1])
+ Globals.GitFastImport.reset(git_ref, commit)
+ changesets = changesets[i:]
+ break
+ revert_state(state, c)
+ i -= 1
+ num_changesets = len(changesets)
+ num_applied = 0
+
+ # Apply changesets, bringing git_ref from prev_state to cur_state
+ if num_changesets:
+ progress("Importing changesets...")
+ num_applied = import_changesets(git_ref, changesets, prev_state,
+ cur_state, progress)
+ progress("Imported %i changesets (reused %i existing changesets)" %
+ (num_applied, num_changesets - num_applied), True)
+ progress.popprefix()
+ return True
+
+
+def do_import (*args):
+ """Do the 'import' command; import refs from a remote."""
+ if not args:
+ die_usage("'import' takes at least one parameter: ref...")
+
+ progress = ProgressIndicator(" ", sys.stderr)
+ cvs_symbols = map(ref_to_cvsname, args)
+ empty_import = True
+ for symbol in cvs_symbols:
+ if import_cvs_symbol(symbol, progress):
+ empty_import = False
+ if empty_import:
+ progress.finish("Everything up-to-date", True)
+ return 0
+ progress.finish("Finished importing %i CVS symbols to Git" %
+ (len(cvs_symbols)), True)
+ return 0
+
+
+def do_list (*args):
+ """Do the 'list' command; list refs available from a CVS remote."""
+ if args:
+ die_usage("'list' takes no parameters")
+
+ progress = ProgressIndicator(" ", sys.stderr)
+ if Config.CachedSymbolsOnly:
+ progress("Listing symbols in local symbol cache...", True)
+ for symbol in sorted(Globals.CVSSymbolCache):
+ print cvs_to_refname(symbol)
+ progress.finish()
+ print # Terminate output with a blank line
+ return 0
+
+ # Synchronize local symbol cache with CVS server
+ progress("Synchronizing local symbol cache with CVS server...")
+ Globals.CVSSymbolCache.sync_all_symbols(Config.CVSRepo, progress,
+ valid_cvs_symbol)
+ # Load current states of Git refs
+ progress("Loading current state of Git refs...")
+ changed, unchanged = 0, 0
+ for cvs_symbol, cvs_state in sorted(Globals.CVSSymbolCache.items()):
+ git_ref = cvs_to_refname(cvs_symbol)
+ progress("\tChecking if Git ref is up-to-date: %s" % (git_ref))
+ git_commit = Globals.GitRefMap.get(git_ref)
+ git_state = Globals.CommitStates.get(git_commit)
+ attrs = ""
+ if git_state and git_state == cvs_state:
+ attrs = " unchanged"
+ unchanged += 1
+ else:
+ git_commit = "?"
+ changed += 1
+ print "%s %s%s" % (git_commit, git_ref, attrs)
+ progress.finish("Found %i CVS symbols (%i changed, %i unchanged)" %
+ (changed + unchanged, changed, unchanged))
+ print # Terminate with a blank line
+ return 0
+
+
+def do_capabilities (*args):
+ """Do the 'capabilities' command; report supported features."""
+ if args:
+ die_usage("'capabilities' takes no parameters")
+ print "import"
+ print "marks %s" % (work_path("marks"))
+# print "export"
+ print # Terminate with a blank line
+ return 0
+
+
+def do_addsymbol (*args):
+ """Do the 'addsymbol' command; add given CVS symbol to local cache."""
+ if len(args) != 1:
+ die_usage("'addsymbol' takes one parameter: symbol")
+ symbol = args[0]
+ progress = ProgressIndicator(" ", sys.stderr)
+ if valid_cvs_symbol(symbol):
+ Globals.CVSSymbolCache.sync_symbol(symbol, Globals.CVSWorkDir,
+ progress)
+ progress.finish("Added '%s' to CVS symbol cache" % (symbol), True)
+ else:
+ error("Skipping CVS symbol '%s'; it is not a valid git ref", symbol)
+ print # Terminate with a blank line
+ return 0
+
+
+def do_syncsymbols (*args):
+ """Do the 'syncsymbols' command; sync all symbols with CVS server."""
+ if args:
+ die_usage("'syncsymbols' takes no parameters")
+ progress = ProgressIndicator(" ", sys.stderr)
+ Globals.CVSSymbolCache.sync_all_symbols(Config.CVSRepo, progress,
+ valid_cvs_symbol)
+ progress.finish()
+ print # Terminate with a blank line
+ return 0
+
+
+def do_verify (*args):
+ """Do the 'verify' command; Compare CVS checkout and Git tree."""
+ if len(args) != 1:
+ die_usage("'verify' takes one parameter: symbol")
+ symbol = args[0]
+ gitref = cvs_to_refname(symbol)
+ assert valid_git_ref(gitref)
+
+ progress = ProgressIndicator(" ", sys.stderr)
+ progress("Checking out '%s' from CVS..." % (symbol))
+ Globals.CVSWorkDir.update(symbol)
+
+ add_env = {"GIT_INDEX_FILE": os.path.abspath(work_path("temp_index"))}
+ progress("Creating Git index from tree object @ '%s'..." % (gitref))
+ cmd = ("git", "read-tree", gitref)
+ assert run_command(cmd, add_env = add_env)[0] == 0
+
+ progress("Comparing CVS checkout to Git index...", True)
+ cmd = ("git", "--work-tree=%s" % (os.path.abspath(work_path("cvs"))),
+ "ls-files", "--exclude=CVS", "--deleted", "--modified", "--others",
+ "-t")
+ exit_code, output, errors = run_command(cmd, add_env = add_env)
+ assert exit_code == 0 and not errors
+
+ if output:
+ progress.finish("Failed verification of '%s'" % (symbol), True)
+ error("The '%s' command returned:\n---\n%s---", " ".join(cmd), output)
+ else:
+ progress.finish("Successfully verified '%s'" % (symbol), True)
+ print # Terminate with a blank line
+ return exit_code
+
+
+def not_implemented (*args):
+ """Abort, while informing user that this command is not yet implemented."""
+ die_usage("Command not implemented")
+
+
+COMMANDS = {
+ "capabilities": do_capabilities,
+ "list": do_list,
+ # Special handling of 'import' in main()
+ # "import": do_import,
+ "export": not_implemented,
+ # Custom commands
+ "addsymbol": do_addsymbol,
+ "syncsymbols": do_syncsymbols,
+ "verify": do_verify,
+}
+
+
+class Globals(object):
+
+ """Global variables are placed here at the start of main()."""
+
+ pass
+
+
+def main (*args):
+ """Main program logic; execution starts here."""
+ debug("Invoked '%s'", " ".join(args))
+
+ # Initialization of subsystems
+ assert len(args) >= 2
+ # Read config for the given remote
+ Config.init(args[1])
+ # Local CVS symbol cache (CVS symbol -> CVS state mapping)
+ Globals.CVSSymbolCache = CVSSymbolCache(work_path("symbols"))
+ # Local CVS checkout
+ Globals.CVSWorkDir = CVSWorkDir(work_path("cvs"), Config.CVSRepo)
+ # Interface to 'git cat-file --batch'
+ Globals.GitObjectFetcher = GitObjectFetcher()
+ # Interface to Git object notes
+ Globals.GitNotes = GitNotes(Config.NotesRef, Globals.GitObjectFetcher)
+ # Mapping from Git commit objects to CVS states
+ Globals.CommitStates = CommitStates(Globals.GitNotes)
+ # Mapping from Git ref names to Git object names
+ Globals.GitRefMap = GitRefMap(Globals.GitObjectFetcher)
+ # Mapping from CVS revision to Git blob and commit objects
+ Globals.CVSRevisionMap = CVSRevisionMap(cvs_to_refname("_metadata"),
+ Globals.GitObjectFetcher)
+ last_mark = 0
+ if Globals.CVSRevisionMap.has_unresolved_marks():
+ # Update with marks from last import
+ last_mark = Globals.CVSRevisionMap.load_marks_file(work_path("marks"))
+ else:
+ # Truncate marks file. We cannot automatically do this after
+ # .load_marks_file() above, since we cannot yet guarantee that
+ # we will be able to save the revision map persistently. (That
+ # can only happen if we are given one or more import commands
+ # below.) We can only truncate this file when we know there are
+ # no unresolved marks in the revision map.
+ open(work_path("marks"), "w").close()
+ # Mapping from CVS states to commit objects that contain said state
+ Globals.CVSStateMap = CVSStateMap(Globals.CVSRevisionMap)
+
+ # Main program loop
+ import_refs = [] # Accumulate import commands here
+ # Cannot use "for line in sys.stdin" for buffering (?) reasons
+ line = sys.stdin.readline()
+ while (line):
+ cmdline = line.strip().split()
+ if not cmdline:
+ break # Blank line means we're about to quit
+ debug("Got command '%s'", " ".join(cmdline))
+ cmd = cmdline.pop(0)
+ if cmd == "import":
+ import_refs.extend(cmdline)
+ else:
+ if cmd not in COMMANDS:
+ die_usage("Unknown command '%s'", cmd)
+ if COMMANDS[cmd](*cmdline):
+ die("Command '%s' failed", line.strip())
+ sys.stdout.flush()
+ line = sys.stdin.readline()
+
+ # Trigger import processing after last import command
+ ret = 0
+ if import_refs:
+ # Init producer of output in the git-fast-import format
+ Globals.GitFastImport = GitFastImport(
+ sys.stdout, Globals.GitObjectFetcher, last_mark)
+
+ # Perform import of given refs
+ ret = do_import(*import_refs)
+
+ # Notes on persistent storage of subsystems' data structures:
+ #
+ # Because the "import" command has been called, we here _know_
+ # that there is a fast-import process running in parallel.
+ # (This is NOT the case when there are no "import" commands).
+ # We can therefore now (and only now) safely commit the extra
+ # information that we store in the Git repo.
+ # In other words, the data structures that we commit to
+ # persistent storage with the following calls will NOT be
+ # committed if there are no "import" commands. The data
+ # structures must handle this in one of two ways:
+ # - In the no-"import" scenario, there is simply nothing to
+ # commit, so it can safely be skipped.
+ # - Any information that should have been committed in the
+ # no-"import" scenario can be reconstructed repeatedly in
+ # subsequent executions of this program, until the next
+ # invocation of an "import" command provides an opportunity
+ # to commit the data structure to persistent storage.
+
+ # Write out commit notes (mapping git commits to CVSStates)
+ # The following call would be a no-op in the no-"import" case
+ Globals.GitNotes.commit_notes(Globals.GitFastImport, Config.Author,
+ 'Annotate commits imported by "git remote-cvs"\n')
+
+ # Save CVS revision metadata
+ # This data structure can handle the no-"import" case as long
+ # as the marks file from the last fast-import run is still
+ # present upon the next execution of this program.
+ Globals.CVSRevisionMap.commit_map(Globals.GitFastImport, Config.Author,
+ 'Updated metadata used by "git remote-cvs"\n')
+
+ return ret
+
+
+if __name__ == '__main__':
+ sys.exit(main(*sys.argv))
--
1.6.4.262.gca66a.dirty
^ permalink raw reply related
* [RFCv4 5/5] Add simple selftests of git-remote-cvs functionality
From: Johan Herland @ 2009-08-17 3:36 UTC (permalink / raw)
To: git; +Cc: Johannes.Schindelin, Johan Herland, barkalow, davvid, gitster
In-Reply-To: <1250480161-21933-1-git-send-email-johan@herland.net>
Add two new selftests:
- t9800-remote-cvs-basic: Test the git-remote-cvs implementation of the
remote helper API, by verifying the expected output of the git-remote-cvs
program when invoking remote helper API commands while doing some simple
CVS operations.
- t9801-remote-cvs-fetch: A more high-level test of the fetch/import-side
of the git-remote-cvs implementation, by verifying the expected repository
state after doing "git fetch" from a CVS remote where a variety of CVS
operations are being performed.
Signed-off-by: Johan Herland <johan@herland.net>
---
t/t9800-remote-cvs-basic.sh | 524 +++++++++++++++++++++++++++++++++++++++++++
t/t9801-remote-cvs-fetch.sh | 291 ++++++++++++++++++++++++
2 files changed, 815 insertions(+), 0 deletions(-)
create mode 100755 t/t9800-remote-cvs-basic.sh
create mode 100755 t/t9801-remote-cvs-fetch.sh
diff --git a/t/t9800-remote-cvs-basic.sh b/t/t9800-remote-cvs-basic.sh
new file mode 100755
index 0000000..6ddec17
--- /dev/null
+++ b/t/t9800-remote-cvs-basic.sh
@@ -0,0 +1,524 @@
+#!/bin/sh
+
+test_description='git remote-cvs basic tests'
+. ./test-lib.sh
+
+if ! test_have_prereq PYTHON; then
+ say 'skipping CVS foreign-vcs helper tests, python not available'
+ test_done
+fi
+
+CVS_EXEC=cvs
+CVS_OPTS="-f -q"
+CVS="$CVS_EXEC $CVS_OPTS"
+
+CVSROOT=$(pwd)/cvsroot
+export CVSROOT
+unset CVS_SERVER
+
+CVSMODULE=cvsmodule
+GITREMOTE=cvsremote
+
+if ! type $CVS_EXEC >/dev/null 2>&1
+then
+ say 'skipping remote-cvs tests, $CVS_EXEC not found'
+ test_done
+fi
+
+test_expect_success 'setup cvsroot' '$CVS init'
+
+test_expect_success '#1: setup a cvs module' '
+
+ mkdir "$CVSROOT/$CVSMODULE" &&
+ $CVS co -d module-cvs $CVSMODULE &&
+ (
+ cd module-cvs &&
+ cat <<EOF >o_fortuna &&
+O Fortuna
+velut luna
+statu variabilis,
+
+semper crescis
+aut decrescis;
+vita detestabilis
+
+nunc obdurat
+et tunc curat
+ludo mentis aciem,
+
+egestatem,
+potestatem
+dissolvit ut glaciem.
+EOF
+ $CVS add o_fortuna &&
+ cat <<EOF >message &&
+add "O Fortuna" lyrics
+
+These public domain lyrics make an excellent sample text.
+EOF
+ $CVS commit -f -F message o_fortuna
+ )
+'
+
+test_expect_success 'set up CVS repo as a foreign remote' '
+
+ git config "user.name" "Test User"
+ git config "user.email" "test@example.com"
+ git config "remote.$GITREMOTE.vcs" cvs
+ git config "remote.$GITREMOTE.cvsRoot" "$CVSROOT"
+ git config "remote.$GITREMOTE.cvsModule" "$CVSMODULE"
+ git config "remote.$GITREMOTE.fetch" \
+ "+refs/cvs/$GITREMOTE/*:refs/remotes/$GITREMOTE/*"
+
+'
+
+test_expect_success '#1: git-remote-cvs "capabilities" command' '
+
+ echo "capabilities" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+import
+marks .git/info/cvs/$GITREMOTE/marks
+
+EOF
+ test_cmp expect actual
+
+'
+
+test_expect_success '#1: git-remote-cvs "list" command' '
+
+ echo "list" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+? refs/cvs/$GITREMOTE/HEAD
+
+EOF
+ test_cmp expect actual
+
+'
+
+test_expect_success '#1: git-remote-cvs "import" command' '
+
+ echo "import refs/cvs/$GITREMOTE/HEAD" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+# Importing CVS revision o_fortuna:1.1
+blob
+mark :1
+data 180
+O Fortuna
+velut luna
+statu variabilis,
+
+semper crescis
+aut decrescis;
+vita detestabilis
+
+nunc obdurat
+et tunc curat
+ludo mentis aciem,
+
+egestatem,
+potestatem
+dissolvit ut glaciem.
+
+commit refs/cvs/$GITREMOTE/HEAD
+mark :2
+data 82
+add "O Fortuna" lyrics
+
+These public domain lyrics make an excellent sample text.
+
+M 644 :1 o_fortuna
+
+# Importing note for object 2
+blob
+mark :3
+data 14
+o_fortuna:1.1
+
+commit refs/notes/cvs/$GITREMOTE
+mark :4
+data 46
+Annotate commits imported by "git remote-cvs"
+
+N :3 :2
+
+blob
+mark :5
+data 32
+1 o_fortuna:1.1
+2 o_fortuna:1.1
+
+blob
+mark :6
+data 16
+blob 1
+commit 2
+
+commit refs/cvs/$GITREMOTE/_metadata
+mark :7
+data 42
+Updated metadata used by "git remote-cvs"
+
+M 644 :5 CVS/marks
+M 644 :6 o_fortuna/1.1
+
+EOF
+ grep -v "^committer " actual > actual.filtered &&
+ test_cmp expect actual.filtered
+
+'
+
+test_expect_success '#1: Passing git-remote-cvs output to git-fast-import' '
+
+ git fast-import --quiet \
+ --export-marks=".git/info/cvs/$GITREMOTE/marks" \
+ < actual &&
+ git gc
+
+'
+
+test_expect_success '#1: Verifying correctness of import' '
+
+ echo "verify HEAD" | git remote-cvs "$GITREMOTE"
+
+'
+
+test_expect_success '#2: update cvs module' '
+
+ (
+ cd module-cvs &&
+ cat <<EOF >o_fortuna &&
+O Fortune,
+like the moon
+you are changeable,
+
+ever waxing
+and waning;
+hateful life
+
+first oppresses
+and then soothes
+as fancy takes it;
+
+poverty
+and power
+it melts them like ice.
+EOF
+ cat <<EOF >message &&
+translate to English
+
+My Latin is terrible.
+EOF
+ $CVS commit -f -F message o_fortuna
+ )
+'
+
+test_expect_success '#2: git-remote-cvs "capabilities" command' '
+
+ echo "capabilities" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+import
+marks .git/info/cvs/$GITREMOTE/marks
+
+EOF
+ test_cmp expect actual
+
+'
+
+test_expect_success '#2: git-remote-cvs "list" command' '
+
+ echo "list" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+? refs/cvs/$GITREMOTE/HEAD
+
+EOF
+ test_cmp expect actual
+
+'
+
+test_expect_success '#2: git-remote-cvs "import" command' '
+
+ echo "import refs/cvs/$GITREMOTE/HEAD" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+# Importing CVS revision o_fortuna:1.2
+blob
+mark :8
+data 179
+O Fortune,
+like the moon
+you are changeable,
+
+ever waxing
+and waning;
+hateful life
+
+first oppresses
+and then soothes
+as fancy takes it;
+
+poverty
+and power
+it melts them like ice.
+
+commit refs/cvs/$GITREMOTE/HEAD
+mark :9
+data 44
+translate to English
+
+My Latin is terrible.
+
+from refs/cvs/$GITREMOTE/HEAD^0
+M 644 :8 o_fortuna
+
+# Importing note for object 9
+blob
+mark :10
+data 14
+o_fortuna:1.2
+
+commit refs/notes/cvs/$GITREMOTE
+mark :11
+data 46
+Annotate commits imported by "git remote-cvs"
+
+from refs/notes/cvs/$GITREMOTE^0
+N :10 :9
+
+blob
+mark :12
+data 32
+8 o_fortuna:1.2
+9 o_fortuna:1.2
+
+blob
+mark :13
+data 94
+
+blob
+mark :14
+data 16
+blob 8
+commit 9
+
+commit refs/cvs/$GITREMOTE/_metadata
+mark :15
+data 42
+Updated metadata used by "git remote-cvs"
+
+from refs/cvs/$GITREMOTE/_metadata^0
+M 644 :12 CVS/marks
+M 644 :13 o_fortuna/1.1
+M 644 :14 o_fortuna/1.2
+
+EOF
+ grep -v -e "^committer " -e "\b[0-9a-f]\{40\}\b" actual > actual.filtered &&
+ test_cmp expect actual.filtered
+
+'
+
+test_expect_success '#2: Passing git-remote-cvs output to git-fast-import' '
+
+ git fast-import --quiet \
+ --import-marks=".git/info/cvs/$GITREMOTE/marks" \
+ --export-marks=".git/info/cvs/$GITREMOTE/marks" \
+ < actual &&
+ git gc
+
+'
+
+test_expect_success '#2: Verifying correctness of import' '
+
+ echo "verify HEAD" | git remote-cvs "$GITREMOTE"
+
+'
+
+test_expect_success '#3: update cvs module' '
+
+ (
+ cd module-cvs &&
+ echo 1 >tick &&
+ $CVS add tick &&
+ $CVS commit -f -m 1 tick
+ )
+
+'
+
+test_expect_success '#3: git-remote-cvs "capabilities" command' '
+
+ echo "capabilities" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+import
+marks .git/info/cvs/$GITREMOTE/marks
+
+EOF
+ test_cmp expect actual
+
+'
+
+test_expect_success '#3: git-remote-cvs "list" command' '
+
+ echo "list" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+? refs/cvs/$GITREMOTE/HEAD
+
+EOF
+ test_cmp expect actual
+
+'
+
+test_expect_success '#3: git-remote-cvs "import" command' '
+
+ echo "import refs/cvs/$GITREMOTE/HEAD" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+# Importing CVS revision tick:1.1
+blob
+mark :16
+data 2
+1
+
+commit refs/cvs/$GITREMOTE/HEAD
+mark :17
+data 2
+1
+
+from refs/cvs/$GITREMOTE/HEAD^0
+M 644 :16 tick
+
+# Importing note for object 17
+blob
+mark :18
+data 23
+o_fortuna:1.2
+tick:1.1
+
+commit refs/notes/cvs/$GITREMOTE
+mark :19
+data 46
+Annotate commits imported by "git remote-cvs"
+
+from refs/notes/cvs/$GITREMOTE^0
+N :18 :17
+
+blob
+mark :20
+data 41
+16 tick:1.1
+17 tick:1.1
+17 o_fortuna:1.2
+
+blob
+mark :21
+data 104
+commit 17
+
+blob
+mark :22
+data 18
+blob 16
+commit 17
+
+commit refs/cvs/$GITREMOTE/_metadata
+mark :23
+data 42
+Updated metadata used by "git remote-cvs"
+
+from refs/cvs/$GITREMOTE/_metadata^0
+M 644 :20 CVS/marks
+M 644 :21 o_fortuna/1.2
+M 644 :22 tick/1.1
+
+EOF
+ grep -v -e "^committer " -e "\b[0-9a-f]\{40\}\b" actual > actual.filtered &&
+ test_cmp expect actual.filtered
+
+'
+
+test_expect_success '#3: Passing git-remote-cvs output to git-fast-import' '
+
+ git fast-import --quiet \
+ --import-marks=".git/info/cvs/$GITREMOTE/marks" \
+ --export-marks=".git/info/cvs/$GITREMOTE/marks" \
+ < actual &&
+ git gc
+
+'
+
+test_expect_success '#3: Verifying correctness of import' '
+
+ echo "verify HEAD" | git remote-cvs "$GITREMOTE"
+
+'
+
+test_expect_success '#4: git-remote-cvs "capabilities" command' '
+
+ echo "capabilities" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+import
+marks .git/info/cvs/$GITREMOTE/marks
+
+EOF
+ test_cmp expect actual
+
+'
+
+commit=$(git rev-parse "refs/cvs/$GITREMOTE/HEAD")
+
+test_expect_success '#4: git-remote-cvs "list" command' '
+
+ echo "list" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+$commit refs/cvs/$GITREMOTE/HEAD unchanged
+
+EOF
+ test_cmp expect actual
+
+'
+
+test_expect_success '#4: git-remote-cvs "import" command' '
+
+ echo "import refs/cvs/$GITREMOTE/HEAD" | git remote-cvs "$GITREMOTE" > actual &&
+ cat <<EOF >expect &&
+blob
+mark :24
+data 0
+
+blob
+mark :25
+data 142
+
+blob
+mark :26
+data 94
+
+commit refs/cvs/$GITREMOTE/_metadata
+mark :27
+data 42
+Updated metadata used by "git remote-cvs"
+
+from refs/cvs/$GITREMOTE/_metadata^0
+M 644 :24 CVS/marks
+M 644 :25 o_fortuna/1.2
+M 644 :26 tick/1.1
+
+EOF
+ grep -v -e "^committer " -e "\b[0-9a-f]\{40\}\b" actual > actual.filtered &&
+ test_cmp expect actual.filtered
+
+'
+
+test_expect_success '#4: Passing git-remote-cvs output to git-fast-import' '
+
+ git fast-import --quiet \
+ --import-marks=".git/info/cvs/$GITREMOTE/marks" \
+ --export-marks=".git/info/cvs/$GITREMOTE/marks" \
+ < actual &&
+ git gc
+
+'
+
+test_expect_success '#4: Verifying correctness of import' '
+
+ echo "verify HEAD" | git remote-cvs "$GITREMOTE"
+
+'
+
+test_done
diff --git a/t/t9801-remote-cvs-fetch.sh b/t/t9801-remote-cvs-fetch.sh
new file mode 100755
index 0000000..93d44a7
--- /dev/null
+++ b/t/t9801-remote-cvs-fetch.sh
@@ -0,0 +1,291 @@
+#!/bin/sh
+
+test_description='git remote-cvs basic tests'
+. ./test-lib.sh
+
+if ! test_have_prereq PYTHON; then
+ say 'skipping CVS foreign-vcs helper tests, python not available'
+ test_done
+fi
+
+CVS_EXEC=cvs
+CVS_OPTS="-f -q"
+CVS="$CVS_EXEC $CVS_OPTS"
+
+CVSROOT=$(pwd)/cvsroot
+export CVSROOT
+unset CVS_SERVER
+
+CVSMODULE=cvsmodule
+GITREMOTE=cvsremote
+
+if ! type $CVS_EXEC >/dev/null 2>&1
+then
+ say 'skipping remote-cvs tests, $CVS_EXEC not found'
+ test_done
+fi
+
+verify () {
+ git log --reverse --format="--- %T%n%s%n%n%b" "$GITREMOTE/$1" >actual &&
+ test_cmp "expect.$1" actual &&
+ echo "verify $1" | git remote-cvs "$GITREMOTE"
+}
+
+test_expect_success 'setup CVS repo' '$CVS init'
+
+test_expect_success 'create CVS module with initial commit' '
+
+ mkdir "$CVSROOT/$CVSMODULE" &&
+ $CVS co -d module-cvs $CVSMODULE &&
+ (
+ cd module-cvs &&
+ cat <<EOF >o_fortuna &&
+O Fortuna
+velut luna
+statu variabilis,
+
+semper crescis
+aut decrescis;
+vita detestabilis
+
+nunc obdurat
+et tunc curat
+ludo mentis aciem,
+
+egestatem,
+potestatem
+dissolvit ut glaciem.
+EOF
+ $CVS add o_fortuna &&
+ cat <<EOF >message &&
+add "O Fortuna" lyrics
+
+These public domain lyrics make an excellent sample text.
+EOF
+ $CVS commit -f -F message o_fortuna
+ )
+'
+
+test_expect_success 'set up CVS repo/module as a foreign remote' '
+
+ git config "user.name" "Test User"
+ git config "user.email" "test@example.com"
+ git config "remote.$GITREMOTE.vcs" cvs
+ git config "remote.$GITREMOTE.cvsRoot" "$CVSROOT"
+ git config "remote.$GITREMOTE.cvsModule" "$CVSMODULE"
+ git config "remote.$GITREMOTE.fetch" \
+ "+refs/cvs/$GITREMOTE/*:refs/remotes/$GITREMOTE/*"
+
+'
+
+test_expect_success 'initial fetch from CVS remote' '
+
+ cat <<EOF >expect.HEAD &&
+--- 0e06d780dedab23e683c686fb041daa9a84c936c
+add "O Fortuna" lyrics
+
+These public domain lyrics make an excellent sample text.
+
+EOF
+ git fetch "$GITREMOTE" &&
+ verify HEAD
+
+'
+
+test_expect_success 'CVS commit' '
+
+ (
+ cd module-cvs &&
+ cat <<EOF >o_fortuna &&
+O Fortune,
+like the moon
+you are changeable,
+
+ever waxing
+and waning;
+hateful life
+
+first oppresses
+and then soothes
+as fancy takes it;
+
+poverty
+and power
+it melts them like ice.
+EOF
+ cat <<EOF >message &&
+translate to English
+
+My Latin is terrible.
+EOF
+ $CVS commit -f -F message o_fortuna
+ ) &&
+ cat <<EOF >>expect.HEAD &&
+--- daa87269a5e00388135ad9542dc16ab6754466e5
+translate to English
+
+My Latin is terrible.
+
+EOF
+ git fetch "$GITREMOTE" &&
+ verify HEAD
+
+'
+
+test_expect_success 'CVS commit with new file' '
+
+ (
+ cd module-cvs &&
+ echo 1 >tick &&
+ $CVS add tick &&
+ $CVS commit -f -m 1 tick
+ ) &&
+ cat <<EOF >>expect.HEAD &&
+--- 486935b4fccecea9b64cbed3a797ebbcbe2b7461
+1
+
+
+EOF
+ git fetch "$GITREMOTE" &&
+ verify HEAD
+
+'
+
+test_expect_success 'fetch without CVS changes' '
+
+ git fetch "$GITREMOTE" &&
+ verify HEAD
+
+'
+
+test_expect_success 'add 2 CVS commits' '
+
+ (
+ cd module-cvs &&
+ echo 2 >tick &&
+ $CVS commit -f -m 2 tick &&
+ echo 3 >tick &&
+ $CVS commit -f -m 3 tick
+ ) &&
+ cat <<EOF >>expect.HEAD &&
+--- 83437ab3e57bf0a42915de5310e3419792b5a36f
+2
+
+
+--- 60fc50406a82dc6bd32dc6e5f7bd23e4c3cdf7ef
+3
+
+
+EOF
+ git fetch "$GITREMOTE" &&
+ verify HEAD
+
+'
+
+test_expect_success 'CVS commit with removed file' '
+
+ (
+ cd module-cvs &&
+ $CVS remove -f tick &&
+ $CVS commit -f -m "remove file" tick
+ ) &&
+ cat <<EOF >>expect.HEAD &&
+--- daa87269a5e00388135ad9542dc16ab6754466e5
+remove file
+
+
+EOF
+ git fetch "$GITREMOTE" &&
+ verify HEAD
+
+'
+
+test_expect_success 'CVS commit with several new files' '
+
+ (
+ cd module-cvs &&
+ echo spam >spam &&
+ echo sausage >sausage &&
+ echo eggs >eggs &&
+ $CVS add spam sausage eggs &&
+ $CVS commit -f -m "spam, sausage, and eggs" spam sausage eggs
+ ) &&
+ cat <<EOF >>expect.HEAD &&
+--- 3190dfce44a6d5e9916b4870dbf8f37d1ca4ddaf
+spam, sausage, and eggs
+
+
+EOF
+ git fetch "$GITREMOTE" &&
+ verify HEAD
+
+'
+
+test_expect_success 'new CVS branch' '
+
+ (
+ cd module-cvs &&
+ $CVS tag -b foo
+ ) &&
+ cp expect.HEAD expect.foo &&
+ git fetch "$GITREMOTE" &&
+ verify HEAD &&
+ verify foo
+
+'
+
+test_expect_success 'CVS commit on branch' '
+
+ (
+ cd module-cvs &&
+ $CVS up -r foo &&
+ echo "spam spam spam" >spam &&
+ $CVS commit -f -m "commit on branch foo" spam
+ ) &&
+ cat <<EOF >>expect.foo &&
+--- 1aba123e5c83898ce3a8b976cc6064d60246aef4
+commit on branch foo
+
+
+EOF
+ git fetch "$GITREMOTE" &&
+ verify HEAD &&
+ verify foo
+
+'
+
+test_expect_success 'create CVS tag' '
+
+ (
+ cd module-cvs &&
+ $CVS tag bar
+ ) &&
+ cp expect.foo expect.bar &&
+ git fetch "$GITREMOTE" &&
+ verify HEAD &&
+ verify foo &&
+ verify bar
+
+'
+
+test_expect_success 'another CVS commit on branch' '
+
+ (
+ cd module-cvs &&
+ echo "spam spam spam spam spam spam" >> spam &&
+ $CVS commit -f -m "another commit on branch foo" spam
+ ) &&
+ cat <<EOF >>expect.foo &&
+--- 15a2635e76e8e5a5a8746021643de317452f2340
+another commit on branch foo
+
+
+EOF
+ git fetch "$GITREMOTE" &&
+ verify HEAD &&
+ verify foo &&
+ verify bar
+
+'
+
+test_done
--
1.6.4.262.gca66a.dirty
^ permalink raw reply related
* Re: git http-push and MKCOL error (22/409)
From: Tay Ray Chuan @ 2009-08-17 4:52 UTC (permalink / raw)
To: Thomas Schlichter; +Cc: Junio C Hamano, Sean Davis, git
In-Reply-To: <200908161652.37282.thomas.schlichter@web.de>
Hi,
On Sun, Aug 16, 2009 at 10:52 PM, Thomas
Schlichter<thomas.schlichter@web.de> wrote:
> Steps to reproduce:
before I try this out, I have a few queries:
> 1. locally set up a git archive:
> mkdir dummy.git
> cd dummy.git
> git init --bare
> 2. Upload this directory to the server.
> I did do this using KDE's dolphin via WebDAV.
> 3. Clone this remote repository:
> git clone https://webdav.smartdrive.web.de/dummy.git my_dummy
You didn't run git update-server-info before doing the clone; cloning
wouldn't work cos git can't find a info/refs file. Perhaps you did do
it and unintentionally left it out from your procedure listed here?
> 4. Create a local commit:
> cd my_dummy
> touch dummy.c
> git commit -a
I assume you want to commit dummy.c? But you didn't tell git to track
it in the first place (git add dummy.c).
> 5. Push this commit up to the remote repository:
> git push origin master
I also find it strange that git only pushed 1 object from your git
push output, when usually there would be a tree, a blob and a commit
to push for a new commit.
--
Cheers,
Ray Chuan
^ permalink raw reply
* Re: git http-push and MKCOL error (22/409)
From: Tay Ray Chuan @ 2009-08-17 4:58 UTC (permalink / raw)
To: Junio C Hamano; +Cc: Thomas Schlichter, willievu, Sean Davis, git
In-Reply-To: <7veirb5z9a.fsf@alter.siamese.dyndns.org>
Hi,
On Mon, Aug 17, 2009 at 3:34 AM, Junio C Hamano<gitster@pobox.com> wrote:
> The report said:
>
> MKCOL 98fd7fb8f32843c1bb40bd195a2f1cd6cab0751d failed, aborting (22/409)
>
> As far as I can see you are trying (in http-push.c::start_mkcol()) to
> create the two-hexdigit fan-out directory (i.e. "98" for this example); it
> is strange to see a request to create the full 40-hexdigit collection in
> the first place.
Yes, you're right, but git prints out the full SHA-1 hash even though
it is actually referring to the 2-hexdigit directory that it failed to
create/verify, for whatever reason.
> In another thread you responded to earlier:
>
> http://thread.gmane.org/gmane.comp.version-control.git/125933/focus=125972
>
> the original report did not give the exact error message, but in that one,
> instead of failing to create 40-hexdigit collection like this, I am
> guessing that it fails with something like "MKCOL 'refs' failed".
I guess by "original report" you refer to Thomas' initial email: yes,
he (unwittingly) did provide theerror code in the subject line, the
part which says (22/409). It actually means (<curl return code>/<http
return code>). Referring again to
http://www.webdav.org/specs/rfc4918.html#rfc.section.9.3.1:
409 (Conflict) - A collection cannot be made at the Request-URI until
one or more intermediate collections have been created. The server
must not create those intermediate collections automatically.
meaning one or more parent directories weren't created.
Based on this and his procedure, I'm guessing that there's something
wrong with his setup.
> *1* Not necessarily in the sense the client is broken but in the sense
> that the server-client combination does not work for the reporter.
Indeed.
--
Cheers,
Ray Chuan
^ permalink raw reply
* Re: "make quick-install-man" broke recently
From: Junio C Hamano @ 2009-08-17 5:17 UTC (permalink / raw)
To: Randal L. Schwartz; +Cc: git@vger.kernel.org
In-Reply-To: <86iqgn8brv.fsf@blue.stonehenge.com>
merlyn@stonehenge.com (Randal L. Schwartz) writes:
> % rm -rf /opt/git/share/man
> % make prefix=/opt/git quick-install-man
> make -C Documentation quick-install-man
> SUBDIR ../
> make[2]: `GIT-VERSION-FILE' is up to date.
> '/bin/sh' ./install-doc-quick.sh origin/man /opt/git/share/man
> % make prefix=/opt/git quick-install-man
> make -C Documentation quick-install-man
> SUBDIR ../
> make[2]: `GIT-VERSION-FILE' is up to date.
> '/bin/sh' ./install-doc-quick.sh origin/man /opt/git/share/man
> error: git checkout-index: unable to create file /opt/git/share/man/man1/git-add.1 (File exists)
> error: git checkout-index: unable to create file /opt/git/share/man/man1/git-am.1 (File exists)
> error: git checkout-index: unable to create file /opt/git/share/man/man1/git-annotate.1 (File exists)
> error: git checkout-index: unable to create file /opt/git/share/man/man1/git-apply.1 (File exists)
> [...]
>
> So it fails the second time. This is new behavior.
Interesting, and this does not reproduce for me.
I've tried "run twice back to back" like you did above, "run once, then
'find -type f -print | xargs touch' the target, and then run again", and also
"run once, then 'find -type f -print | xargs chmod -w' the target, and then run again".
None of these fail.
One way I can think of to break it deliberately would be:
make prefix=/var/tmp/gomikuzu quick-install-man
find /var/tmp/gomikuzu -type d | xargs chmod -w
make prefix=/var/tmp/gomikuzu quick-install-man
But then the failure is different from what you showed above:
error: unable to unlink old '/var/tmp/gomikuzu/share/...' (Permission denied)
Puzzled.
^ permalink raw reply
* Re: git http-push and MKCOL error (22/409)
From: Thomas Schlichter @ 2009-08-17 5:18 UTC (permalink / raw)
To: Tay Ray Chuan; +Cc: Junio C Hamano, Sean Davis, git
In-Reply-To: <be6fef0d0908162152o54206acfge43f6ffa012bb307@mail.gmail.com>
Hi,
Am Montag 17 August 2009 06:52:32 schrieb Tay Ray Chuan:
> Hi,
>
> On Sun, Aug 16, 2009 at 10:52 PM, Thomas
>
> Schlichter<thomas.schlichter@web.de> wrote:
> > Steps to reproduce:
>
> before I try this out, I have a few queries:
> > 1. locally set up a git archive:
> > mkdir dummy.git
> > cd dummy.git
> > git init --bare
> > 2. Upload this directory to the server.
> > I did do this using KDE's dolphin via WebDAV.
> > 3. Clone this remote repository:
> > git clone https://webdav.smartdrive.web.de/dummy.git my_dummy
>
> You didn't run git update-server-info before doing the clone; cloning
> wouldn't work cos git can't find a info/refs file. Perhaps you did do
> it and unintentionally left it out from your procedure listed here?
Oh, yes, sorry.
I did run it but unintentionally left it out. At least I had no problem with
clone, it did clone an "empty repository" without a problem. Even during
bisecting, I never had a problem with pulling from the server, only with
pushing to it.
> > 4. Create a local commit:
> > cd my_dummy
> > touch dummy.c
> > git commit -a
>
> I assume you want to commit dummy.c?
Yes, indeed...
> But you didn't tell git to track
> it in the first place (git add dummy.c).
I think that was not neccessary, the parameter "-a" should do it for me...
> > 5. Push this commit up to the remote repository:
> > git push origin master
>
> I also find it strange that git only pushed 1 object from your git
> push output, when usually there would be a tree, a blob and a commit
> to push for a new commit.
Yes, most of the times git tries to push three objects, and fails with all
three. As I wrote, during bisecting I was able to commit several times. The
reported message was the last one failing. Unfortunately git only needed to
push one object that time. I didn't think it was important if it were three or
just one object?!
Kind regards,
Thomas Schlichter
^ permalink raw reply
* Re: git http-push and MKCOL error (22/409)
From: Thomas Schlichter @ 2009-08-17 5:25 UTC (permalink / raw)
To: Tay Ray Chuan; +Cc: Junio C Hamano, willievu, Sean Davis, git
In-Reply-To: <be6fef0d0908162158y429c4b5fgb9138151446e39ef@mail.gmail.com>
Hi,
Am Montag 17 August 2009 06:58:25 schrieb Tay Ray Chuan:
> Hi,
>
> On Mon, Aug 17, 2009 at 3:34 AM, Junio C Hamano<gitster@pobox.com> wrote:
> > The report said:
> >
> > MKCOL 98fd7fb8f32843c1bb40bd195a2f1cd6cab0751d failed, aborting
> > (22/409)
> >
> > As far as I can see you are trying (in http-push.c::start_mkcol()) to
> > create the two-hexdigit fan-out directory (i.e. "98" for this example);
> > it is strange to see a request to create the full 40-hexdigit collection
> > in the first place.
>
> Yes, you're right, but git prints out the full SHA-1 hash even though
> it is actually referring to the 2-hexdigit directory that it failed to
> create/verify, for whatever reason.
>
> > In another thread you responded to earlier:
> >
> >
> > http://thread.gmane.org/gmane.comp.version-control.git/125933/focus=1259
> >72
> >
> > the original report did not give the exact error message, but in that
> > one, instead of failing to create 40-hexdigit collection like this, I am
> > guessing that it fails with something like "MKCOL 'refs' failed".
>
> I guess by "original report" you refer to Thomas' initial email: yes,
> he (unwittingly) did provide theerror code in the subject line, the
> part which says (22/409). It actually means (<curl return code>/<http
> return code>). Referring again to
> http://www.webdav.org/specs/rfc4918.html#rfc.section.9.3.1:
>
> 409 (Conflict) - A collection cannot be made at the Request-URI until
> one or more intermediate collections have been created. The server
> must not create those intermediate collections automatically.
>
> meaning one or more parent directories weren't created.
>
> Based on this and his procedure, I'm guessing that there's something
> wrong with his setup.
>
> > *1* Not necessarily in the sense the client is broken but in the sense
> > that the server-client combination does not work for the reporter.
>
> Indeed.
Hmm. I hope you won't take this as an argument for not fixing it, because it
is a clear regression! I tried git version 1.6.3.4, and it works flawlessly
with exactly this server! Even during bisecting, some versions worked, some
didn't (these after the mentioned commit...)
So I think this commit didn't only refactor things, but unintentionally
changed the behavior. And this must be the problem. As I'm not into this code,
and the refactoring was not completely trivial, I was not able to quickly find
the place where the behavior was changed...
Kind regards,
Thomas Schlichter
^ permalink raw reply
* Re: git http-push and MKCOL error (22/409)
From: Thomas Schlichter @ 2009-08-17 5:34 UTC (permalink / raw)
To: Tay Ray Chuan; +Cc: Junio C Hamano, Sean Davis, git
In-Reply-To: <200908170718.15423.thomas.schlichter@web.de>
Hi,
Am Montag 17 August 2009 07:18:15 schrieb Thomas Schlichter:
> Hi,
>
> Am Montag 17 August 2009 06:52:32 schrieb Tay Ray Chuan:
> > I assume you want to commit dummy.c?
>
> Yes, indeed...
>
> > But you didn't tell git to track
> > it in the first place (git add dummy.c).
>
> I think that was not neccessary, the parameter "-a" should do it for me...
OK, I tried it again, and the parameter "-a" was not sufficient. So I
obviously did run a "git add dummy.c". Sorry for missing that.
Now I do also have the error message for the first commit:
schlicht@netbook:~/dummy2$ git push origin master
Fetching remote heads...
refs
refs/heads
refs/tags
refs/heads
refs/tags
updating 'refs/heads/master'
from 0000000000000000000000000000000000000000
to 60157fdfdca4693f703c744fed95e1b31743bb4e
sending 3 objects
MKCOL e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 failed, aborting (22/409)
MKCOL 4aefccd66b8e2bfa58079910045c30bc792877dc failed, aborting (22/409)
MKCOL 60157fdfdca4693f703c744fed95e1b31743bb4e failed, aborting (22/409)
Updating remote server info
error: failed to push some refs to
'https://webdav.smartdrive.web.de/dummy2.git'
Kind regards,
Thomas Schlichter
^ permalink raw reply
* Re: "make quick-install-man" broke recently
From: Randal L. Schwartz @ 2009-08-17 6:01 UTC (permalink / raw)
To: Jacob Helwig; +Cc: Junio C Hamano, git@vger.kernel.org
In-Reply-To: <8c9a060908162258ua50e34ah53c1ada9dbcd4aa0@mail.gmail.com>
>>>>> "Jacob" == Jacob Helwig <jacob.helwig@gmail.com> writes:
Jacob> Randal, does your machine have a similar symlinked $HOME setup?
/opt/git is through two symlinks, yes.
--
Randal L. Schwartz - Stonehenge Consulting Services, Inc. - +1 503 777 0095
<merlyn@stonehenge.com> <URL:http://www.stonehenge.com/merlyn/>
Smalltalk/Perl/Unix consulting, Technical writing, Comedy, etc. etc.
See http://methodsandmessages.vox.com/ for Smalltalk and Seaside discussion
^ permalink raw reply
* Re: "make quick-install-man" broke recently
From: Jacob Helwig @ 2009-08-17 5:58 UTC (permalink / raw)
To: Junio C Hamano; +Cc: Randal L. Schwartz, git@vger.kernel.org
In-Reply-To: <7vd46v3tp1.fsf@alter.siamese.dyndns.org>
On Sun, Aug 16, 2009 at 22:17, Junio C Hamano<gitster@pobox.com> wrote:
> merlyn@stonehenge.com (Randal L. Schwartz) writes:
>
>> % rm -rf /opt/git/share/man
>> % make prefix=/opt/git quick-install-man
>> make -C Documentation quick-install-man
>> SUBDIR ../
>> make[2]: `GIT-VERSION-FILE' is up to date.
>> '/bin/sh' ./install-doc-quick.sh origin/man /opt/git/share/man
>> % make prefix=/opt/git quick-install-man
>> make -C Documentation quick-install-man
>> SUBDIR ../
>> make[2]: `GIT-VERSION-FILE' is up to date.
>> '/bin/sh' ./install-doc-quick.sh origin/man /opt/git/share/man
>> error: git checkout-index: unable to create file /opt/git/share/man/man1/git-add.1 (File exists)
>> error: git checkout-index: unable to create file /opt/git/share/man/man1/git-am.1 (File exists)
>> error: git checkout-index: unable to create file /opt/git/share/man/man1/git-annotate.1 (File exists)
>> error: git checkout-index: unable to create file /opt/git/share/man/man1/git-apply.1 (File exists)
>> [...]
>>
>> So it fails the second time. This is new behavior.
>
> Interesting, and this does not reproduce for me.
>
> I've tried "run twice back to back" like you did above, "run once, then
> 'find -type f -print | xargs touch' the target, and then run again", and also
> "run once, then 'find -type f -print | xargs chmod -w' the target, and then run again".
>
> None of these fail.
>
> One way I can think of to break it deliberately would be:
>
> make prefix=/var/tmp/gomikuzu quick-install-man
> find /var/tmp/gomikuzu -type d | xargs chmod -w
> make prefix=/var/tmp/gomikuzu quick-install-man
>
> But then the failure is different from what you showed above:
>
> error: unable to unlink old '/var/tmp/gomikuzu/share/...' (Permission denied)
>
> Puzzled.
> --
> To unsubscribe from this list: send the line "unsubscribe git" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
I was able to reproduce this one one of the machines I have access to,
and bisecting shows that it was introduced by b6986d8 (git-checkout: be
careful about untracked symlinks). This makes sense, since home-dirs
are symlinks on the machine I was able to reproduce it on.
Hopefully this helps someone with a little stronger fu, to point them in
the right direction.
Randal, does your machine have a similar symlinked $HOME setup?
^ permalink raw reply
* Re: git http-push and MKCOL error (22/409)
From: Junio C Hamano @ 2009-08-17 6:08 UTC (permalink / raw)
To: Thomas Schlichter
Cc: Tay Ray Chuan, Junio C Hamano, willievu, Sean Davis, git
In-Reply-To: <200908170725.09592.thomas.schlichter@web.de>
Thomas Schlichter <thomas.schlichter@web.de> writes:
> Hmm. I hope you won't take this as an argument for not fixing it, because it
> is a clear regression! I tried git version 1.6.3.4, and it works flawlessly
> with exactly this server! Even during bisecting, some versions worked, some
> didn't (these after the mentioned commit...)
If it works (without changing anything else) with one version and with a
problematic patch it doesn't, it is a clear regression.
If you revert 5424bc5 (http*: add helper methods for fetching objects
(loose), 2009-06-06) from the tip of 'master', does it make the problem go
away?
The patch should look like something like the attached.
diff --git a/http-push.c b/http-push.c
index 00e83dc..d1ea702 100644
--- a/http-push.c
+++ b/http-push.c
@@ -115,10 +115,18 @@ struct transfer_request
struct remote_lock *lock;
struct curl_slist *headers;
struct buffer buffer;
+ char filename[PATH_MAX];
+ char tmpfile[PATH_MAX];
+ int local_fileno;
enum transfer_state state;
CURLcode curl_result;
char errorstr[CURL_ERROR_SIZE];
long http_code;
+ unsigned char real_sha1[20];
+ git_SHA_CTX c;
+ z_stream stream;
+ int zret;
+ int rename;
void *userData;
struct active_request_slot *slot;
struct transfer_request *next;
@@ -226,6 +234,15 @@ static struct curl_slist *get_dav_token_headers(struct remote_lock *lock, enum d
return dav_headers;
}
+static void append_remote_object_url(struct strbuf *buf, const char *url,
+ const char *hex,
+ int only_two_digit_prefix)
+{
+ strbuf_addf(buf, "%sobjects/%.*s/", url, 2, hex);
+ if (!only_two_digit_prefix)
+ strbuf_addf(buf, "%s", hex+2);
+}
+
static void finish_request(struct transfer_request *request);
static void release_request(struct transfer_request *request);
@@ -239,29 +256,169 @@ static void process_response(void *callback_data)
#ifdef USE_CURL_MULTI
+static char *get_remote_object_url(const char *url, const char *hex,
+ int only_two_digit_prefix)
+{
+ struct strbuf buf = STRBUF_INIT;
+ append_remote_object_url(&buf, url, hex, only_two_digit_prefix);
+ return strbuf_detach(&buf, NULL);
+}
+
+static size_t fwrite_sha1_file(void *ptr, size_t eltsize, size_t nmemb,
+ void *data)
+{
+ unsigned char expn[4096];
+ size_t size = eltsize * nmemb;
+ int posn = 0;
+ struct transfer_request *request = (struct transfer_request *)data;
+ do {
+ ssize_t retval = xwrite(request->local_fileno,
+ (char *) ptr + posn, size - posn);
+ if (retval < 0)
+ return posn;
+ posn += retval;
+ } while (posn < size);
+
+ request->stream.avail_in = size;
+ request->stream.next_in = ptr;
+ do {
+ request->stream.next_out = expn;
+ request->stream.avail_out = sizeof(expn);
+ request->zret = git_inflate(&request->stream, Z_SYNC_FLUSH);
+ git_SHA1_Update(&request->c, expn,
+ sizeof(expn) - request->stream.avail_out);
+ } while (request->stream.avail_in && request->zret == Z_OK);
+ data_received++;
+ return size;
+}
+
static void start_fetch_loose(struct transfer_request *request)
{
+ char *hex = sha1_to_hex(request->obj->sha1);
+ char *filename;
+ char prevfile[PATH_MAX];
+ char *url;
+ int prevlocal;
+ unsigned char prev_buf[PREV_BUF_SIZE];
+ ssize_t prev_read = 0;
+ long prev_posn = 0;
+ char range[RANGE_HEADER_SIZE];
+ struct curl_slist *range_header = NULL;
struct active_request_slot *slot;
- struct http_object_request *obj_req;
- obj_req = new_http_object_request(repo->url, request->obj->sha1);
- if (obj_req == NULL) {
+ filename = sha1_file_name(request->obj->sha1);
+ snprintf(request->filename, sizeof(request->filename), "%s", filename);
+ snprintf(request->tmpfile, sizeof(request->tmpfile),
+ "%s.temp", filename);
+
+ snprintf(prevfile, sizeof(prevfile), "%s.prev", request->filename);
+ unlink_or_warn(prevfile);
+ rename(request->tmpfile, prevfile);
+ unlink_or_warn(request->tmpfile);
+
+ if (request->local_fileno != -1)
+ error("fd leakage in start: %d", request->local_fileno);
+ request->local_fileno = open(request->tmpfile,
+ O_WRONLY | O_CREAT | O_EXCL, 0666);
+ /*
+ * This could have failed due to the "lazy directory creation";
+ * try to mkdir the last path component.
+ */
+ if (request->local_fileno < 0 && errno == ENOENT) {
+ char *dir = strrchr(request->tmpfile, '/');
+ if (dir) {
+ *dir = 0;
+ mkdir(request->tmpfile, 0777);
+ *dir = '/';
+ }
+ request->local_fileno = open(request->tmpfile,
+ O_WRONLY | O_CREAT | O_EXCL, 0666);
+ }
+
+ if (request->local_fileno < 0) {
request->state = ABORTED;
+ error("Couldn't create temporary file %s for %s: %s",
+ request->tmpfile, request->filename, strerror(errno));
return;
}
- slot = obj_req->slot;
+ memset(&request->stream, 0, sizeof(request->stream));
+
+ git_inflate_init(&request->stream);
+
+ git_SHA1_Init(&request->c);
+
+ url = get_remote_object_url(repo->url, hex, 0);
+ request->url = xstrdup(url);
+
+ /*
+ * If a previous temp file is present, process what was already
+ * fetched.
+ */
+ prevlocal = open(prevfile, O_RDONLY);
+ if (prevlocal != -1) {
+ do {
+ prev_read = xread(prevlocal, prev_buf, PREV_BUF_SIZE);
+ if (prev_read>0) {
+ if (fwrite_sha1_file(prev_buf,
+ 1,
+ prev_read,
+ request) == prev_read)
+ prev_posn += prev_read;
+ else
+ prev_read = -1;
+ }
+ } while (prev_read > 0);
+ close(prevlocal);
+ }
+ unlink_or_warn(prevfile);
+
+ /*
+ * Reset inflate/SHA1 if there was an error reading the previous temp
+ * file; also rewind to the beginning of the local file.
+ */
+ if (prev_read == -1) {
+ memset(&request->stream, 0, sizeof(request->stream));
+ git_inflate_init(&request->stream);
+ git_SHA1_Init(&request->c);
+ if (prev_posn>0) {
+ prev_posn = 0;
+ lseek(request->local_fileno, 0, SEEK_SET);
+ ftruncate(request->local_fileno, 0);
+ }
+ }
+
+ slot = get_active_slot();
slot->callback_func = process_response;
slot->callback_data = request;
request->slot = slot;
- request->userData = obj_req;
+
+ curl_easy_setopt(slot->curl, CURLOPT_FILE, request);
+ curl_easy_setopt(slot->curl, CURLOPT_WRITEFUNCTION, fwrite_sha1_file);
+ curl_easy_setopt(slot->curl, CURLOPT_ERRORBUFFER, request->errorstr);
+ curl_easy_setopt(slot->curl, CURLOPT_URL, url);
+ curl_easy_setopt(slot->curl, CURLOPT_HTTPHEADER, no_pragma_header);
+
+ /*
+ * If we have successfully processed data from a previous fetch
+ * attempt, only fetch the data we don't already have.
+ */
+ if (prev_posn>0) {
+ if (push_verbosely)
+ fprintf(stderr,
+ "Resuming fetch of object %s at byte %ld\n",
+ hex, prev_posn);
+ sprintf(range, "Range: bytes=%ld-", prev_posn);
+ range_header = curl_slist_append(range_header, range);
+ curl_easy_setopt(slot->curl,
+ CURLOPT_HTTPHEADER, range_header);
+ }
/* Try to get the request started, abort the request on error */
request->state = RUN_FETCH_LOOSE;
if (!start_active_slot(slot)) {
fprintf(stderr, "Unable to start GET request\n");
repo->can_update_info_refs = 0;
- release_http_object_request(obj_req);
release_request(request);
}
}
@@ -520,14 +677,16 @@ static void release_request(struct transfer_request *request)
entry->next = entry->next->next;
}
+ if (request->local_fileno != -1)
+ close(request->local_fileno);
free(request->url);
free(request);
}
static void finish_request(struct transfer_request *request)
{
+ struct stat st;
struct http_pack_request *preq;
- struct http_object_request *obj_req;
request->curl_result = request->slot->curl_result;
request->http_code = request->slot->http_code;
@@ -582,17 +741,39 @@ static void finish_request(struct transfer_request *request)
aborted = 1;
}
} else if (request->state == RUN_FETCH_LOOSE) {
- obj_req = (struct http_object_request *)request->userData;
-
- if (finish_http_object_request(obj_req) == 0)
- if (obj_req->rename == 0)
- request->obj->flags |= (LOCAL | REMOTE);
+ close(request->local_fileno);
+ request->local_fileno = -1;
+
+ if (request->curl_result != CURLE_OK &&
+ request->http_code != 416) {
+ if (stat(request->tmpfile, &st) == 0) {
+ if (st.st_size == 0)
+ unlink_or_warn(request->tmpfile);
+ }
+ } else {
+ if (request->http_code == 416)
+ warning("requested range invalid; we may already have all the data.");
+
+ git_inflate_end(&request->stream);
+ git_SHA1_Final(request->real_sha1, &request->c);
+ if (request->zret != Z_STREAM_END) {
+ unlink_or_warn(request->tmpfile);
+ } else if (hashcmp(request->obj->sha1, request->real_sha1)) {
+ unlink_or_warn(request->tmpfile);
+ } else {
+ request->rename =
+ move_temp_to_file(
+ request->tmpfile,
+ request->filename);
+ if (request->rename == 0)
+ request->obj->flags |= (LOCAL | REMOTE);
+ }
+ }
/* Try fetching packed if necessary */
- if (request->obj->flags & LOCAL) {
- release_http_object_request(obj_req);
+ if (request->obj->flags & LOCAL)
release_request(request);
- } else
+ else
start_fetch_packed(request);
} else if (request->state == RUN_FETCH_PACKED) {
@@ -664,6 +845,7 @@ static void add_fetch_request(struct object *obj)
request->url = NULL;
request->lock = NULL;
request->headers = NULL;
+ request->local_fileno = -1;
request->state = NEED_FETCH;
request->next = request_queue_head;
request_queue_head = request;
@@ -702,6 +884,7 @@ static int add_send_request(struct object *obj, struct remote_lock *lock)
request->url = NULL;
request->lock = lock;
request->headers = NULL;
+ request->local_fileno = -1;
request->state = NEED_PUSH;
request->next = request_queue_head;
request_queue_head = request;
diff --git a/http-walker.c b/http-walker.c
index 700bc13..8f7a975 100644
--- a/http-walker.c
+++ b/http-walker.c
@@ -3,6 +3,8 @@
#include "walker.h"
#include "http.h"
+#define PREV_BUF_SIZE 4096
+
struct alt_base
{
char *base;
@@ -23,8 +25,20 @@ struct object_request
struct walker *walker;
unsigned char sha1[20];
struct alt_base *repo;
+ char *url;
+ char filename[PATH_MAX];
+ char tmpfile[PATH_MAX];
+ int local;
enum object_request_state state;
- struct http_object_request *req;
+ CURLcode curl_result;
+ char errorstr[CURL_ERROR_SIZE];
+ long http_code;
+ unsigned char real_sha1[20];
+ git_SHA_CTX c;
+ z_stream stream;
+ int zret;
+ int rename;
+ struct active_request_slot *slot;
struct object_request *next;
};
@@ -45,6 +59,34 @@ struct walker_data {
static struct object_request *object_queue_head;
+static size_t fwrite_sha1_file(void *ptr, size_t eltsize, size_t nmemb,
+ void *data)
+{
+ unsigned char expn[4096];
+ size_t size = eltsize * nmemb;
+ int posn = 0;
+ struct object_request *obj_req = (struct object_request *)data;
+ do {
+ ssize_t retval = xwrite(obj_req->local,
+ (char *) ptr + posn, size - posn);
+ if (retval < 0)
+ return posn;
+ posn += retval;
+ } while (posn < size);
+
+ obj_req->stream.avail_in = size;
+ obj_req->stream.next_in = ptr;
+ do {
+ obj_req->stream.next_out = expn;
+ obj_req->stream.avail_out = sizeof(expn);
+ obj_req->zret = git_inflate(&obj_req->stream, Z_SYNC_FLUSH);
+ git_SHA1_Update(&obj_req->c, expn,
+ sizeof(expn) - obj_req->stream.avail_out);
+ } while (obj_req->stream.avail_in && obj_req->zret == Z_OK);
+ data_received++;
+ return size;
+}
+
static void fetch_alternates(struct walker *walker, const char *base);
static void process_object_response(void *callback_data);
@@ -52,35 +94,172 @@ static void process_object_response(void *callback_data);
static void start_object_request(struct walker *walker,
struct object_request *obj_req)
{
+ char *hex = sha1_to_hex(obj_req->sha1);
+ char prevfile[PATH_MAX];
+ char *url;
+ char *posn;
+ int prevlocal;
+ unsigned char prev_buf[PREV_BUF_SIZE];
+ ssize_t prev_read = 0;
+ long prev_posn = 0;
+ char range[RANGE_HEADER_SIZE];
+ struct curl_slist *range_header = NULL;
struct active_request_slot *slot;
- struct http_object_request *req;
- req = new_http_object_request(obj_req->repo->base, obj_req->sha1);
- if (req == NULL) {
+ snprintf(prevfile, sizeof(prevfile), "%s.prev", obj_req->filename);
+ unlink_or_warn(prevfile);
+ rename(obj_req->tmpfile, prevfile);
+ unlink_or_warn(obj_req->tmpfile);
+
+ if (obj_req->local != -1)
+ error("fd leakage in start: %d", obj_req->local);
+ obj_req->local = open(obj_req->tmpfile,
+ O_WRONLY | O_CREAT | O_EXCL, 0666);
+ /*
+ * This could have failed due to the "lazy directory creation";
+ * try to mkdir the last path component.
+ */
+ if (obj_req->local < 0 && errno == ENOENT) {
+ char *dir = strrchr(obj_req->tmpfile, '/');
+ if (dir) {
+ *dir = 0;
+ mkdir(obj_req->tmpfile, 0777);
+ *dir = '/';
+ }
+ obj_req->local = open(obj_req->tmpfile,
+ O_WRONLY | O_CREAT | O_EXCL, 0666);
+ }
+
+ if (obj_req->local < 0) {
obj_req->state = ABORTED;
+ error("Couldn't create temporary file %s for %s: %s",
+ obj_req->tmpfile, obj_req->filename, strerror(errno));
return;
}
- obj_req->req = req;
- slot = req->slot;
+ memset(&obj_req->stream, 0, sizeof(obj_req->stream));
+
+ git_inflate_init(&obj_req->stream);
+
+ git_SHA1_Init(&obj_req->c);
+
+ url = xmalloc(strlen(obj_req->repo->base) + 51);
+ obj_req->url = xmalloc(strlen(obj_req->repo->base) + 51);
+ strcpy(url, obj_req->repo->base);
+ posn = url + strlen(obj_req->repo->base);
+ strcpy(posn, "/objects/");
+ posn += 9;
+ memcpy(posn, hex, 2);
+ posn += 2;
+ *(posn++) = '/';
+ strcpy(posn, hex + 2);
+ strcpy(obj_req->url, url);
+
+ /*
+ * If a previous temp file is present, process what was already
+ * fetched.
+ */
+ prevlocal = open(prevfile, O_RDONLY);
+ if (prevlocal != -1) {
+ do {
+ prev_read = xread(prevlocal, prev_buf, PREV_BUF_SIZE);
+ if (prev_read>0) {
+ if (fwrite_sha1_file(prev_buf,
+ 1,
+ prev_read,
+ obj_req) == prev_read)
+ prev_posn += prev_read;
+ else
+ prev_read = -1;
+ }
+ } while (prev_read > 0);
+ close(prevlocal);
+ }
+ unlink_or_warn(prevfile);
+
+ /*
+ * Reset inflate/SHA1 if there was an error reading the previous temp
+ * file; also rewind to the beginning of the local file.
+ */
+ if (prev_read == -1) {
+ memset(&obj_req->stream, 0, sizeof(obj_req->stream));
+ git_inflate_init(&obj_req->stream);
+ git_SHA1_Init(&obj_req->c);
+ if (prev_posn>0) {
+ prev_posn = 0;
+ lseek(obj_req->local, 0, SEEK_SET);
+ ftruncate(obj_req->local, 0);
+ }
+ }
+
+ slot = get_active_slot();
slot->callback_func = process_object_response;
slot->callback_data = obj_req;
+ obj_req->slot = slot;
+
+ curl_easy_setopt(slot->curl, CURLOPT_FILE, obj_req);
+ curl_easy_setopt(slot->curl, CURLOPT_WRITEFUNCTION, fwrite_sha1_file);
+ curl_easy_setopt(slot->curl, CURLOPT_ERRORBUFFER, obj_req->errorstr);
+ curl_easy_setopt(slot->curl, CURLOPT_URL, url);
+ curl_easy_setopt(slot->curl, CURLOPT_HTTPHEADER, no_pragma_header);
+
+ /*
+ * If we have successfully processed data from a previous fetch
+ * attempt, only fetch the data we don't already have.
+ */
+ if (prev_posn>0) {
+ if (walker->get_verbosely)
+ fprintf(stderr,
+ "Resuming fetch of object %s at byte %ld\n",
+ hex, prev_posn);
+ sprintf(range, "Range: bytes=%ld-", prev_posn);
+ range_header = curl_slist_append(range_header, range);
+ curl_easy_setopt(slot->curl,
+ CURLOPT_HTTPHEADER, range_header);
+ }
/* Try to get the request started, abort the request on error */
obj_req->state = ACTIVE;
if (!start_active_slot(slot)) {
obj_req->state = ABORTED;
- release_http_object_request(req);
+ obj_req->slot = NULL;
+ close(obj_req->local);
+ obj_req->local = -1;
+ free(obj_req->url);
return;
}
}
static void finish_object_request(struct object_request *obj_req)
{
- if (finish_http_object_request(obj_req->req))
+ struct stat st;
+
+ close(obj_req->local);
+ obj_req->local = -1;
+
+ if (obj_req->http_code == 416) {
+ fprintf(stderr, "Warning: requested range invalid; we may already have all the data.\n");
+ } else if (obj_req->curl_result != CURLE_OK) {
+ if (stat(obj_req->tmpfile, &st) == 0)
+ if (st.st_size == 0)
+ unlink_or_warn(obj_req->tmpfile);
return;
+ }
+
+ git_inflate_end(&obj_req->stream);
+ git_SHA1_Final(obj_req->real_sha1, &obj_req->c);
+ if (obj_req->zret != Z_STREAM_END) {
+ unlink_or_warn(obj_req->tmpfile);
+ return;
+ }
+ if (hashcmp(obj_req->sha1, obj_req->real_sha1)) {
+ unlink_or_warn(obj_req->tmpfile);
+ return;
+ }
+ obj_req->rename =
+ move_temp_to_file(obj_req->tmpfile, obj_req->filename);
- if (obj_req->req->rename == 0)
+ if (obj_req->rename == 0)
walker_say(obj_req->walker, "got %s\n", sha1_to_hex(obj_req->sha1));
}
@@ -92,16 +271,19 @@ static void process_object_response(void *callback_data)
struct walker_data *data = walker->data;
struct alt_base *alt = data->alt;
- process_http_object_request(obj_req->req);
+ obj_req->curl_result = obj_req->slot->curl_result;
+ obj_req->http_code = obj_req->slot->http_code;
+ obj_req->slot = NULL;
obj_req->state = COMPLETE;
/* Use alternates if necessary */
- if (missing_target(obj_req->req)) {
+ if (missing_target(obj_req)) {
fetch_alternates(walker, alt->base);
if (obj_req->repo->next != NULL) {
obj_req->repo =
obj_req->repo->next;
- release_http_object_request(obj_req->req);
+ close(obj_req->local);
+ obj_req->local = -1;
start_object_request(walker, obj_req);
return;
}
@@ -114,8 +296,8 @@ static void release_object_request(struct object_request *obj_req)
{
struct object_request *entry = object_queue_head;
- if (obj_req->req !=NULL && obj_req->req->localfile != -1)
- error("fd leakage in release: %d", obj_req->req->localfile);
+ if (obj_req->local != -1)
+ error("fd leakage in release: %d", obj_req->local);
if (obj_req == object_queue_head) {
object_queue_head = obj_req->next;
} else {
@@ -125,6 +307,7 @@ static void release_object_request(struct object_request *obj_req)
entry->next = entry->next->next;
}
+ free(obj_req->url);
free(obj_req);
}
@@ -152,13 +335,19 @@ static void prefetch(struct walker *walker, unsigned char *sha1)
struct object_request *newreq;
struct object_request *tail;
struct walker_data *data = walker->data;
+ char *filename = sha1_file_name(sha1);
newreq = xmalloc(sizeof(*newreq));
newreq->walker = walker;
hashcpy(newreq->sha1, sha1);
newreq->repo = data->alt;
+ newreq->url = NULL;
+ newreq->local = -1;
newreq->state = WAITING;
- newreq->req = NULL;
+ snprintf(newreq->filename, sizeof(newreq->filename), "%s", filename);
+ snprintf(newreq->tmpfile, sizeof(newreq->tmpfile),
+ "%s.temp", filename);
+ newreq->slot = NULL;
newreq->next = NULL;
http_is_verbose = walker->get_verbosely;
@@ -449,6 +638,15 @@ abort:
static void abort_object_request(struct object_request *obj_req)
{
+ if (obj_req->local >= 0) {
+ close(obj_req->local);
+ obj_req->local = -1;
+ }
+ unlink_or_warn(obj_req->tmpfile);
+ if (obj_req->slot) {
+ release_active_slot(obj_req->slot);
+ obj_req->slot = NULL;
+ }
release_object_request(obj_req);
}
@@ -457,7 +655,6 @@ static int fetch_object(struct walker *walker, struct alt_base *repo, unsigned c
char *hex = sha1_to_hex(sha1);
int ret = 0;
struct object_request *obj_req = object_queue_head;
- struct http_object_request *req;
while (obj_req != NULL && hashcmp(obj_req->sha1, sha1))
obj_req = obj_req->next;
@@ -465,8 +662,6 @@ static int fetch_object(struct walker *walker, struct alt_base *repo, unsigned c
return error("Couldn't find request for %s in the queue", hex);
if (has_sha1_file(obj_req->sha1)) {
- if (obj_req->req != NULL)
- abort_http_object_request(obj_req->req);
abort_object_request(obj_req);
return 0;
}
@@ -478,42 +673,34 @@ static int fetch_object(struct walker *walker, struct alt_base *repo, unsigned c
start_object_request(walker, obj_req);
#endif
- /*
- * obj_req->req might change when fetching alternates in the callback
- * process_object_response; therefore, the "shortcut" variable, req,
- * is used only after we're done with slots.
- */
while (obj_req->state == ACTIVE)
- run_active_slot(obj_req->req->slot);
-
- req = obj_req->req;
+ run_active_slot(obj_req->slot);
- if (req->localfile != -1) {
- close(req->localfile);
- req->localfile = -1;
+ if (obj_req->local != -1) {
+ close(obj_req->local);
+ obj_req->local = -1;
}
if (obj_req->state == ABORTED) {
ret = error("Request for %s aborted", hex);
- } else if (req->curl_result != CURLE_OK &&
- req->http_code != 416) {
- if (missing_target(req))
+ } else if (obj_req->curl_result != CURLE_OK &&
+ obj_req->http_code != 416) {
+ if (missing_target(obj_req))
ret = -1; /* Be silent, it is probably in a pack. */
else
ret = error("%s (curl_result = %d, http_code = %ld, sha1 = %s)",
- req->errorstr, req->curl_result,
- req->http_code, hex);
- } else if (req->zret != Z_STREAM_END) {
+ obj_req->errorstr, obj_req->curl_result,
+ obj_req->http_code, hex);
+ } else if (obj_req->zret != Z_STREAM_END) {
walker->corrupt_object_found++;
- ret = error("File %s (%s) corrupt", hex, req->url);
- } else if (hashcmp(obj_req->sha1, req->real_sha1)) {
+ ret = error("File %s (%s) corrupt", hex, obj_req->url);
+ } else if (hashcmp(obj_req->sha1, obj_req->real_sha1)) {
ret = error("File %s has bad hash", hex);
- } else if (req->rename < 0) {
+ } else if (obj_req->rename < 0) {
ret = error("unable to write sha1 filename %s",
- req->filename);
+ obj_req->filename);
}
- release_http_object_request(req);
release_object_request(obj_req);
return ret;
}
diff --git a/http.c b/http.c
index 14d5357..ad7b6f8 100644
--- a/http.c
+++ b/http.c
@@ -12,10 +12,6 @@ static CURLM *curlm;
#ifndef NO_CURL_EASY_DUPHANDLE
static CURL *curl_default;
#endif
-
-#define PREV_BUF_SIZE 4096
-#define RANGE_HEADER_SIZE 30
-
char curl_errorstr[CURL_ERROR_SIZE];
static int curl_ssl_verify = -1;
@@ -45,7 +41,8 @@ static char *ssl_cert_password;
static int ssl_cert_password_required;
static struct curl_slist *pragma_header;
-static struct curl_slist *no_pragma_header;
+
+struct curl_slist *no_pragma_header;
static struct active_request_slot *active_queue_head;
@@ -715,23 +712,6 @@ static char *quote_ref_url(const char *base, const char *ref)
return strbuf_detach(&buf, NULL);
}
-void append_remote_object_url(struct strbuf *buf, const char *url,
- const char *hex,
- int only_two_digit_prefix)
-{
- strbuf_addf(buf, "%s/objects/%.*s/", url, 2, hex);
- if (!only_two_digit_prefix)
- strbuf_addf(buf, "%s", hex+2);
-}
-
-char *get_remote_object_url(const char *url, const char *hex,
- int only_two_digit_prefix)
-{
- struct strbuf buf = STRBUF_INIT;
- append_remote_object_url(&buf, url, hex, only_two_digit_prefix);
- return strbuf_detach(&buf, NULL);
-}
-
/* http_request() targets */
#define HTTP_REQUEST_STRBUF 0
#define HTTP_REQUEST_FILE 1
@@ -1061,231 +1041,3 @@ abort:
free(preq);
return NULL;
}
-
-/* Helpers for fetching objects (loose) */
-static size_t fwrite_sha1_file(void *ptr, size_t eltsize, size_t nmemb,
- void *data)
-{
- unsigned char expn[4096];
- size_t size = eltsize * nmemb;
- int posn = 0;
- struct http_object_request *freq =
- (struct http_object_request *)data;
- do {
- ssize_t retval = xwrite(freq->localfile,
- (char *) ptr + posn, size - posn);
- if (retval < 0)
- return posn;
- posn += retval;
- } while (posn < size);
-
- freq->stream.avail_in = size;
- freq->stream.next_in = ptr;
- do {
- freq->stream.next_out = expn;
- freq->stream.avail_out = sizeof(expn);
- freq->zret = git_inflate(&freq->stream, Z_SYNC_FLUSH);
- git_SHA1_Update(&freq->c, expn,
- sizeof(expn) - freq->stream.avail_out);
- } while (freq->stream.avail_in && freq->zret == Z_OK);
- data_received++;
- return size;
-}
-
-struct http_object_request *new_http_object_request(const char *base_url,
- unsigned char *sha1)
-{
- char *hex = sha1_to_hex(sha1);
- char *filename;
- char prevfile[PATH_MAX];
- int prevlocal;
- unsigned char prev_buf[PREV_BUF_SIZE];
- ssize_t prev_read = 0;
- long prev_posn = 0;
- char range[RANGE_HEADER_SIZE];
- struct curl_slist *range_header = NULL;
- struct http_object_request *freq;
-
- freq = xmalloc(sizeof(*freq));
- hashcpy(freq->sha1, sha1);
- freq->localfile = -1;
-
- filename = sha1_file_name(sha1);
- snprintf(freq->filename, sizeof(freq->filename), "%s", filename);
- snprintf(freq->tmpfile, sizeof(freq->tmpfile),
- "%s.temp", filename);
-
- snprintf(prevfile, sizeof(prevfile), "%s.prev", filename);
- unlink_or_warn(prevfile);
- rename(freq->tmpfile, prevfile);
- unlink_or_warn(freq->tmpfile);
-
- if (freq->localfile != -1)
- error("fd leakage in start: %d", freq->localfile);
- freq->localfile = open(freq->tmpfile,
- O_WRONLY | O_CREAT | O_EXCL, 0666);
- /*
- * This could have failed due to the "lazy directory creation";
- * try to mkdir the last path component.
- */
- if (freq->localfile < 0 && errno == ENOENT) {
- char *dir = strrchr(freq->tmpfile, '/');
- if (dir) {
- *dir = 0;
- mkdir(freq->tmpfile, 0777);
- *dir = '/';
- }
- freq->localfile = open(freq->tmpfile,
- O_WRONLY | O_CREAT | O_EXCL, 0666);
- }
-
- if (freq->localfile < 0) {
- error("Couldn't create temporary file %s for %s: %s",
- freq->tmpfile, freq->filename, strerror(errno));
- goto abort;
- }
-
- memset(&freq->stream, 0, sizeof(freq->stream));
-
- git_inflate_init(&freq->stream);
-
- git_SHA1_Init(&freq->c);
-
- freq->url = get_remote_object_url(base_url, hex, 0);
-
- /*
- * If a previous temp file is present, process what was already
- * fetched.
- */
- prevlocal = open(prevfile, O_RDONLY);
- if (prevlocal != -1) {
- do {
- prev_read = xread(prevlocal, prev_buf, PREV_BUF_SIZE);
- if (prev_read>0) {
- if (fwrite_sha1_file(prev_buf,
- 1,
- prev_read,
- freq) == prev_read) {
- prev_posn += prev_read;
- } else {
- prev_read = -1;
- }
- }
- } while (prev_read > 0);
- close(prevlocal);
- }
- unlink_or_warn(prevfile);
-
- /*
- * Reset inflate/SHA1 if there was an error reading the previous temp
- * file; also rewind to the beginning of the local file.
- */
- if (prev_read == -1) {
- memset(&freq->stream, 0, sizeof(freq->stream));
- git_inflate_init(&freq->stream);
- git_SHA1_Init(&freq->c);
- if (prev_posn>0) {
- prev_posn = 0;
- lseek(freq->localfile, 0, SEEK_SET);
- if (ftruncate(freq->localfile, 0) < 0) {
- error("Couldn't truncate temporary file %s for %s: %s",
- freq->tmpfile, freq->filename, strerror(errno));
- goto abort;
- }
- }
- }
-
- freq->slot = get_active_slot();
-
- curl_easy_setopt(freq->slot->curl, CURLOPT_FILE, freq);
- curl_easy_setopt(freq->slot->curl, CURLOPT_WRITEFUNCTION, fwrite_sha1_file);
- curl_easy_setopt(freq->slot->curl, CURLOPT_ERRORBUFFER, freq->errorstr);
- curl_easy_setopt(freq->slot->curl, CURLOPT_URL, freq->url);
- curl_easy_setopt(freq->slot->curl, CURLOPT_HTTPHEADER, no_pragma_header);
-
- /*
- * If we have successfully processed data from a previous fetch
- * attempt, only fetch the data we don't already have.
- */
- if (prev_posn>0) {
- if (http_is_verbose)
- fprintf(stderr,
- "Resuming fetch of object %s at byte %ld\n",
- hex, prev_posn);
- sprintf(range, "Range: bytes=%ld-", prev_posn);
- range_header = curl_slist_append(range_header, range);
- curl_easy_setopt(freq->slot->curl,
- CURLOPT_HTTPHEADER, range_header);
- }
-
- return freq;
-
-abort:
- free(filename);
- free(freq->url);
- free(freq);
- return NULL;
-}
-
-void process_http_object_request(struct http_object_request *freq)
-{
- if (freq->slot == NULL)
- return;
- freq->curl_result = freq->slot->curl_result;
- freq->http_code = freq->slot->http_code;
- freq->slot = NULL;
-}
-
-int finish_http_object_request(struct http_object_request *freq)
-{
- struct stat st;
-
- close(freq->localfile);
- freq->localfile = -1;
-
- process_http_object_request(freq);
-
- if (freq->http_code == 416) {
- fprintf(stderr, "Warning: requested range invalid; we may already have all the data.\n");
- } else if (freq->curl_result != CURLE_OK) {
- if (stat(freq->tmpfile, &st) == 0)
- if (st.st_size == 0)
- unlink_or_warn(freq->tmpfile);
- return -1;
- }
-
- git_inflate_end(&freq->stream);
- git_SHA1_Final(freq->real_sha1, &freq->c);
- if (freq->zret != Z_STREAM_END) {
- unlink_or_warn(freq->tmpfile);
- return -1;
- }
- if (hashcmp(freq->sha1, freq->real_sha1)) {
- unlink_or_warn(freq->tmpfile);
- return -1;
- }
- freq->rename =
- move_temp_to_file(freq->tmpfile, freq->filename);
-
- return freq->rename;
-}
-
-void abort_http_object_request(struct http_object_request *freq)
-{
- unlink_or_warn(freq->tmpfile);
-
- release_http_object_request(freq);
-}
-
-void release_http_object_request(struct http_object_request *freq)
-{
- if (freq->localfile != -1) {
- close(freq->localfile);
- freq->localfile = -1;
- }
- if (freq->url != NULL) {
- free(freq->url);
- freq->url = NULL;
- }
- freq->slot = NULL;
-}
diff --git a/http.h b/http.h
index 4c4e99c..511c0c4 100644
--- a/http.h
+++ b/http.h
@@ -88,6 +88,10 @@ extern void add_fill_function(void *data, int (*fill)(void *));
extern void step_active_slots(void);
#endif
+extern struct curl_slist *no_pragma_header;
+
+#define RANGE_HEADER_SIZE 30
+
extern void http_init(struct remote *remote);
extern void http_cleanup(void);
@@ -110,13 +114,6 @@ static inline int missing__target(int code, int result)
#define missing_target(a) missing__target((a)->http_code, (a)->curl_result)
-/* Helpers for modifying and creating URLs */
-extern void append_remote_object_url(struct strbuf *buf, const char *url,
- const char *hex,
- int only_two_digit_prefix);
-extern char *get_remote_object_url(const char *url, const char *hex,
- int only_two_digit_prefix);
-
/* Options for http_request_*() */
#define HTTP_NO_CACHE 1
@@ -170,30 +167,4 @@ extern struct http_pack_request *new_http_pack_request(
extern int finish_http_pack_request(struct http_pack_request *preq);
extern void release_http_pack_request(struct http_pack_request *preq);
-/* Helpers for fetching object */
-struct http_object_request
-{
- char *url;
- char filename[PATH_MAX];
- char tmpfile[PATH_MAX];
- int localfile;
- CURLcode curl_result;
- char errorstr[CURL_ERROR_SIZE];
- long http_code;
- unsigned char sha1[20];
- unsigned char real_sha1[20];
- git_SHA_CTX c;
- z_stream stream;
- int zret;
- int rename;
- struct active_request_slot *slot;
-};
-
-extern struct http_object_request *new_http_object_request(
- const char *base_url, unsigned char *sha1);
-extern void process_http_object_request(struct http_object_request *freq);
-extern int finish_http_object_request(struct http_object_request *freq);
-extern void abort_http_object_request(struct http_object_request *freq);
-extern void release_http_object_request(struct http_object_request *freq);
-
#endif /* HTTP_H */
^ permalink raw reply related
* sparse support in pu
From: skillzero @ 2009-08-17 6:09 UTC (permalink / raw)
To: git
I had a couple comments and questions about the sparse stuff recently
merged into pu. First, I think it's an awesome feature. It's a nice
solution to my problems (1.5 GB tree where most people only need a
small subset, but with a lot of shared pieces).
1. Have people decided whether it should be on by default if you have
a .git/info/sparse file? I'd definitely like it to be on by default.
When I first tried it, I didn't realize I had to use --sparse to git
checkout to get it to use the sparse rules. The same goes for a merge
I did that happened to have a file in the excluded area (it included
it because I didn't use --sparse to git merge).
2. Is it not hooked up to git reset yet? I did a git checkout --sparse
and things look liked I expected then I did a git reset --hard
origin/master and it started checking out all the stuff previously
excluded via .git/info/sparse. I tried --sparse, but it didn't know
about that option.
3. One thing that was confusing is that I needed a trailing slash on
directories in .git/info/sparse to get them excluded. This seems
different than .gitignore, which works for me without the trailing
slash.
BTW...How to people normally try out stuff in pu? Do you just replace
your normal git stuff with the pu build? That's what I did because I
wasn't sure if I did /test/git if the other tools git invokes would
use the release version of git or the ones relative to itself (i.e.
the ones in /test/ for me). It seems like the only way to really test
stuff is to use it for your real work, but I also don't want to
corrupt our production repositories by using an experimental version
of git.
^ permalink raw reply
* Re: How to stop sharing objects between repositories
From: Jeff King @ 2009-08-17 6:19 UTC (permalink / raw)
To: Junio C Hamano; +Cc: Johannes Schindelin, Jon Jensen, git
In-Reply-To: <7vmy5z603d.fsf@alter.siamese.dyndns.org>
On Sun, Aug 16, 2009 at 12:16:22PM -0700, Junio C Hamano wrote:
> After reading this, two points come to my mind. They may or may not be
> issues.
>
> (1) Such a user does not necessarily know a casual "git repack -a" breaks
> the dependency, defeating the -s option s/he deliberately used in
> order to save disk space in the first place. Perhaps we can reword
> this further to kill two penguins with a single stone?
>
> Note that the pack resulting from running `git repack -a` in the
> repository cloned with the `-s` option will include objects that
> are borrowed from the source repository. It essentially breaks
> the dependency created by cloning with the `-s` option by copying
> the objects from the source repository. To keep borrowing from
> the source repository to save disk space, do not use `repack -a`.
Good point, but I don't think this wording is quite right. You can also
cause such an inefficiency by simply running "git repack", if the source
has loose objects. In other words:
1. "git repack -a" is sufficient to break dependency, as it copies
both packed and loose objects
2. "git repack" _may_ break the dependency, if there are no packs, as
it copies only loose objects. It _may_ introduce inefficiency, but
only if there are loose objects.
3. "git repack -l" always keeps the dependency and current efficiency.
As an aside, making this list makes me realize there is no easy
"keep the dependency and increase efficiency". In other words, pack
everything that is not available otherwise, and then prune the
remaining packs.
Modified patch is below.
> We should suggest an alternative immediately after this sentence,
> e.g. "Instead, use `repack -l`" or something, but somebody should
> check if it is a valid/viable alternative.
It does work. From a user's perspective, I think "-l" would probably be
a more sane default. But I think it is off for historical reasons, and
these days we try to steer users towards "git gc", anyway, which does
use "-l" by default.
-- >8 --
Subject: [PATCH] docs: describe impact of repack on "clone -s"
The effects of repacking on a repository with alternates are
a bit subtle. The two main things users will want are:
1. Not to waste disk space by accidentally copying objects
which could be shared.
2. Copying all objects explicitly to break the dependency
on the source repo.
This patch describes both under the "clone -s"
documentation. It makes sense to put it there rather than in
git-repack.txt for both cases. For (1), we are warning the
user who is using "clone -s" about what _not_ to do, so we
need to get their attention when reading about "clone -s".
For (2), we are telling them how git-repack can be used to
accomplish a task, but until they know that git-repack is
the right tool, they have no reason to look at the repack
documentation.
Signed-off-by: Jeff King <peff@peff.net>
---
The extra deleted lines in the patch below are just cleaning up some
excess whitespace.
Documentation/git-clone.txt | 12 ++++++++++--
1 files changed, 10 insertions(+), 2 deletions(-)
diff --git a/Documentation/git-clone.txt b/Documentation/git-clone.txt
index b14de6c..b25944f 100644
--- a/Documentation/git-clone.txt
+++ b/Documentation/git-clone.txt
@@ -72,8 +72,16 @@ These objects may be removed by normal git operations (such as 'git-commit')
which automatically call `git gc --auto`. (See linkgit:git-gc[1].)
If these objects are removed and were referenced by the cloned repository,
then the cloned repository will become corrupt.
-
-
++
+Note that running `git repack` without the `-l` option in a repository
+cloned with `-s` will copy objects from the source repository into a
+pack in the cloned repository, removing the disk space savings of `clone
+-s`. It is safe, however, to run `git gc`, which uses the `-l` option by
+default.
++
+If you want to break the dependency of a repository cloned with `-s` on
+its source repository, you can simply run `git repack -a` to copy all
+objects from the source repository into a pack in the cloned repository.
--reference <repository>::
If the reference repository is on the local machine
--
1.6.4.283.gec993
^ permalink raw reply related
* Re: How to stop sharing objects between repositories
From: Jeff King @ 2009-08-17 6:31 UTC (permalink / raw)
To: Junio C Hamano; +Cc: Johannes Schindelin, Jon Jensen, git
In-Reply-To: <7vmy5z603d.fsf@alter.siamese.dyndns.org>
On Sun, Aug 16, 2009 at 12:16:22PM -0700, Junio C Hamano wrote:
> (2) IIRC, "git gc --auto" runs "repack -A". What is its effect with
> respect to this dependency between object stores? I suspect it would
> also break the dependency, but if so, is it a good thing? Perhaps
> should we change it to use a version that keeps the dependency
> instead?
No, it actually runs "repack -d -l -A", which behaves fine. I even
tested it to make sure.
BTW, the "gc.auto" setting is really annoying at low levels (I set
gc.auto to 1 for testing). For efficiency, it looks at only one
hashed object directory, and then assumes the other 255 contain roughly
the same number of objects. But you get bad sampling error when you have
fewer than 256 objects. I don't think it is worth caring about, though.
It doesn't seem very sane to set gc.auto to something so low.
-Peff
^ permalink raw reply
* Re: How to stop sharing objects between repositories
From: Jeff King @ 2009-08-17 6:32 UTC (permalink / raw)
To: Junio C Hamano; +Cc: Johannes Schindelin, Jon Jensen, git
In-Reply-To: <20090817061916.GA27530@coredump.intra.peff.net>
On Mon, Aug 17, 2009 at 02:19:16AM -0400, Jeff King wrote:
> 1. "git repack -a" is sufficient to break dependency, as it copies
> both packed and loose objects
>
> 2. "git repack" _may_ break the dependency, if there are no packs, as
> it copies only loose objects. It _may_ introduce inefficiency, but
> only if there are loose objects.
>
> 3. "git repack -l" always keeps the dependency and current efficiency.
>
> As an aside, making this list makes me realize there is no easy
> "keep the dependency and increase efficiency". In other words, pack
> everything that is not available otherwise, and then prune the
> remaining packs.
OK, I take it back. "git repack -d -l -A" will get rid of any packs that
are redundant with your alternate.
-Peff
^ permalink raw reply
* Re: How to stop sharing objects between repositories
From: Jeff King @ 2009-08-17 6:48 UTC (permalink / raw)
To: Mike Galbraith; +Cc: Junio C Hamano, Johannes Schindelin, Jon Jensen, git
In-Reply-To: <1250475682.7155.16.camel@marge.simson.net>
On Mon, Aug 17, 2009 at 04:21:22AM +0200, Mike Galbraith wrote:
> > (1) Such a user does not necessarily know a casual "git repack -a" breaks
> > the dependency, defeating the -s option s/he deliberately used in
> > order to save disk space in the first place. Perhaps we can reword
> > this further to kill two penguins with a single stone?
>
> Perhaps a runtime warning that you're about to break it? This user may
> not even be the one who set the thing up, no?
I'm not really sure what such a setup would look like. If it is a big
hosting site like kernel.org or repo.or.cz, then probably it wouldn't
matter much. The admins there should probably be running "git repack -l
-d -A" periodically to consolidate the object stores (which can happen
from this sort of repacking, or from people just pushing the same
commits to their repos).
That being said, I can see there being setups where such a warning might
be useful. However, we don't really know if the user _wants_ that
effect, or if it is an accident. So people following the recommnded
"here is how you break the dependency" advice will also get the warning.
I'm torn on whether this is actually a good idea.
-- >8 --
Subject: [PATCH] repack: warn when "-l" is not used with alternates
Failing to use "-l" means that we will copy objects from the
source repository, nullifying the usefulness of "-s". We
don't want to make this an error, though, since "git repack
-a" is used to intentionally break the dependency.
Signed-off-by: Jeff King <peff@peff.net>
---
Is "test -s" portable? It's in POSIX, but I have some lingering doubt in
the back of my mind.
A user seeing such a warning can perhaps ^C to abort the pack. However,
should we also give instructions on how to undo the copying (which
should be "git repack -d -l -A")?
git-repack.sh | 9 +++++++++
1 files changed, 9 insertions(+), 0 deletions(-)
diff --git a/git-repack.sh b/git-repack.sh
index 1eb3bca..0bdc6e9 100755
--- a/git-repack.sh
+++ b/git-repack.sh
@@ -44,6 +44,15 @@ do
shift
done
+if test -z "$local" && test -s "$GIT_DIR/objects/info/alternates"; then
+cat >&2 <<'EOF'
+warning: this repository uses objects from other repositories via the
+warning: "alternates" mechanism; repacking without "-l" will cause objects
+warning: to be copied into this repository, wasting disk space.
+
+EOF
+fi
+
case "`git config --bool repack.usedeltabaseoffset || echo true`" in
true)
extra="$extra --delta-base-offset" ;;
--
1.6.4.283.ga2765.dirty
^ permalink raw reply related
* Re: "make quick-install-man" broke recently
From: Junio C Hamano @ 2009-08-17 6:53 UTC (permalink / raw)
To: Jacob Helwig
Cc: Linus Torvalds, Kjetil Barvik, Randal L. Schwartz,
git@vger.kernel.org
In-Reply-To: <8c9a060908162258ua50e34ah53c1ada9dbcd4aa0@mail.gmail.com>
Jacob Helwig <jacob.helwig@gmail.com> writes:
> I was able to reproduce this one one of the machines I have access to,
> and bisecting shows that it was introduced by b6986d8 (git-checkout: be
> careful about untracked symlinks).
Ahh..
-- >8 --
check_path(): allow symlinked directories to checkout-index --prefix
Merlyn noticed that Documentation/install-doc-quick.sh no longer correctly
removes old installed documents when the target directory has a leading
path that is a symlink. It turns out that "checkout-index --prefix" was
broken by recent b6986d8 (git-checkout: be careful about untracked
symlinks, 2009-07-29).
I suspect has_symlink_leading_path() could learn the third parameter
(prefix that is allowed to be symlinked directories) to allow us to retire
a similar function has_dirs_only_path().
Another avenue of fixing this I considered was to get rid of base_dir and
base_dir_len from "struct checkout", and instead make "git checkout-index"
when run with --prefix mkdir the leading path and chdir in there. It
might be the best longer term solution to this issue, as the base_dir
feature is used only by that rather obscure codepath as far as I know.
But at least this patch should fix this breakage.
SIgned-off-by: Junio C Hamano <gitster@pobox.com>
---
cache.h | 3 ---
entry.c | 15 ++++++++++++---
t/t2000-checkout-cache-clash.sh | 9 +++++++++
3 files changed, 21 insertions(+), 6 deletions(-)
diff --git a/cache.h b/cache.h
index 9222774..e6c7f33 100644
--- a/cache.h
+++ b/cache.h
@@ -468,9 +468,6 @@ extern int index_fd(unsigned char *sha1, int fd, struct stat *st, int write_obje
extern int index_path(unsigned char *sha1, const char *path, struct stat *st, int write_object);
extern void fill_stat_cache_info(struct cache_entry *ce, struct stat *st);
-/* "careful lstat()" */
-extern int check_path(const char *path, int len, struct stat *st);
-
#define REFRESH_REALLY 0x0001 /* ignore_valid */
#define REFRESH_UNMERGED 0x0002 /* allow unmerged */
#define REFRESH_QUIET 0x0004 /* be quiet about it */
diff --git a/entry.c b/entry.c
index f276cf3..6813f8a 100644
--- a/entry.c
+++ b/entry.c
@@ -179,9 +179,18 @@ static int write_entry(struct cache_entry *ce, char *path, const struct checkout
* This is like 'lstat()', except it refuses to follow symlinks
* in the path.
*/
-int check_path(const char *path, int len, struct stat *st)
+static int check_path(const char *path, int len, struct stat *st,
+ const struct checkout *co)
{
- if (has_symlink_leading_path(path, len)) {
+ if (co->base_dir_len) {
+ const char *slash = path + len;
+ while (path < slash && *slash != '/')
+ slash--;
+ if (!has_dirs_only_path(path, slash-path, co->base_dir_len)) {
+ errno = ENOENT;
+ return -1;
+ }
+ } else if (has_symlink_leading_path(path, len)) {
errno = ENOENT;
return -1;
}
@@ -201,7 +210,7 @@ int checkout_entry(struct cache_entry *ce, const struct checkout *state, char *t
strcpy(path + len, ce->name);
len += ce_namelen(ce);
- if (!check_path(path, len, &st)) {
+ if (!check_path(path, len, &st, state)) {
unsigned changed = ce_match_stat(ce, &st, CE_MATCH_IGNORE_VALID);
if (!changed)
return 0;
diff --git a/t/t2000-checkout-cache-clash.sh b/t/t2000-checkout-cache-clash.sh
index f7e1a73..3edade0 100755
--- a/t/t2000-checkout-cache-clash.sh
+++ b/t/t2000-checkout-cache-clash.sh
@@ -48,4 +48,13 @@ test_expect_success \
'git checkout-index conflicting paths.' \
'test -f path0 && test -d path1 && test -f path1/file1'
+test_expect_success 'checkout-index -f twice with --prefix' '
+ mkdir -p tar/get &&
+ ln -s tar/get there &&
+ echo first &&
+ git checkout-index -a -f --prefix=there/ &&
+ echo second &&
+ git checkout-index -a -f --prefix=there/
+'
+
test_done
^ permalink raw reply related
* Re: How to stop sharing objects between repositories
From: Mike Galbraith @ 2009-08-17 7:12 UTC (permalink / raw)
To: Jeff King; +Cc: Junio C Hamano, Johannes Schindelin, Jon Jensen, git
In-Reply-To: <20090817064801.GA31543@coredump.intra.peff.net>
On Mon, 2009-08-17 at 02:48 -0400, Jeff King wrote:
> On Mon, Aug 17, 2009 at 04:21:22AM +0200, Mike Galbraith wrote:
>
> > > (1) Such a user does not necessarily know a casual "git repack -a" breaks
> > > the dependency, defeating the -s option s/he deliberately used in
> > > order to save disk space in the first place. Perhaps we can reword
> > > this further to kill two penguins with a single stone?
> >
> > Perhaps a runtime warning that you're about to break it? This user may
> > not even be the one who set the thing up, no?
>
> I'm not really sure what such a setup would look like. If it is a big
> hosting site like kernel.org or repo.or.cz, then probably it wouldn't
> matter much. The admins there should probably be running "git repack -l
> -d -A" periodically to consolidate the object stores (which can happen
> from this sort of repacking, or from people just pushing the same
> commits to their repos).
>
> That being said, I can see there being setups where such a warning might
> be useful. However, we don't really know if the user _wants_ that
> effect, or if it is an accident. So people following the recommnded
> "here is how you break the dependency" advice will also get the warning.
>
> I'm torn on whether this is actually a good idea.
Yeah. There are any number of ways to shoot oneself in the foot, and
while on the one hand idiot proofing can be nice if you were about to
screw up, "really really?" messages are most frequently annoying noise.
-Mike
^ permalink raw reply
* Re: Linus' sha1 is much faster!
From: George Spelvin @ 2009-08-17 7:23 UTC (permalink / raw)
To: bdonlan, johnflux, P; +Cc: art.08.09, git, linux, nico, torvalds
If it helps anyone resolve license issues, here's a from-FIPS-180-2
implementation that's placed in the public domain. That should be
compatible with any license.
It uses Linus's and Artur's performance ideas, and some of Linus' macro
ideas (in the rotate implementation), but tries to be textually different.
Is there anything recognizable that anyone cares to clam copyright to?
It's not quite 100% finished, as I haven't benchmarked it against Linus's
code yet, but it's functionally correct.
It's also clean with -W -Wall -Wextra.
TODO: Check if an initial copy to w[] is faster on i386 (less register
pressure).
/*
* Secure Hash Algorith SHA-1, as published in FIPS PUB 180-2.
*
* This implementation is in the public domain. Copyright abandoned.
* You may do anything you like with it, including evil things.
*
* This is a rewrite from scratch, based on Linus Torvalds' "block-sha1"
* from the git mailing list (August, 2009). Additional optimization
* ideas cribbed from
* - Artur Skawina (x86, particularly P4, and much benchmarking)
* - Nicilas Pitre (ARM)
*/
/* Cut here for sha1.h */
#include <stddef.h> /* For size_t */
#include <stdint.h> /* For uint32_t, uint64_t */
typedef struct SHA_context {
uint64_t len; /* May be shrunk to uint32_t */
uint32_t iv[5];
unsigned char buf[64]; /* Must be 32-bit aligned */
} SHA_CTX;
void SHA1_Init(struct SHA_context *c);
void SHA1_Update(struct SHA_context *c, void const *p, size_t n);
void SHA1_Final(unsigned char hash[20], struct SHA_context *c);
/* End of sha1.h */
#include <string.h> /* For memcpy() */
#include <arpa/inet.h> /* For ntohl() */
static void sha1_core(uint32_t iv[5], unsigned char const *p, size_t nblocks);
/* Machine specific hacks */
#if defined(__i386__) || defined(__x86_64__) || defined (__ppc__) || \
defined(__ppc64__) || defined(__powerpc__) || defined (__powerpc64__) || \
defined(__s390__) || defined(__s390x__)
/* Unaligned access is okay */
static inline uint32_t get_be32(unsigned char const *p)
{
return ntohl(*(uint32_t const *)p);
}
static inline void put_be32(unsigned char const *p, uint32_t v)
{
*(uint32_t *)p = htonl(v);
}
#else
/* Unaligned access is not okay; do conversion as byte fetches */
static inline uint32_t get_be32(unsigned char const *p)
{
return p[0] << 24 || p[1] << 16 | p[8] << 8 | p[3];
}
static inline void put_be32(unsigned char const *p, uint32_t v)
{
p[0] = v >> 24;
p[1] = v >> 16;
p[2] = v >> 8;
p[3] = v;
}
#endif
void SHA1_Init(struct SHA_context *c)
{
/* This is a prefix of the SHA_context structure */
static struct {
uint64_t len;
uint32_t iv[5];
} const iv = { 0,
{ 0x67452301, 0xEFCDAB89, 0x98BADCFE, 0x10325476, 0xC3D2E1F0 }
};
memcpy(c, &iv, sizeof iv);
}
void SHA1_Update(struct SHA_context *c, void const *p, size_t n)
{
size_t pos = c->len & 63; /* Offset into current input block */
c->len += n;
/* Initial partial block (if any) */
if (pos) {
size_t space = 63 - pos;
if (n < space)
goto end;
memcpy(c->buf + pos, p, space);
sha1_core(c->iv, c->buf, 1);
n -= space;
p = (char const *)p + space;
}
/* The large middle piece */
if (n >> 6) {
sha1_core(c->iv, p, n >> 6);
p = (char const *)p + (n & -(size_t)64);
n &= 63;
}
pos = 0;
end:
/* Final partial block (may be zero size) */
memcpy(c->buf + pos, p, n);
}
void SHA1_Final(unsigned char hash[20], struct SHA_context *c)
{
size_t pos = c->len & 63;
unsigned i;
/* Append a single 1 bit */
c->buf[pos++] = 0x80;
/* Append 0 bits until 64 bits remain in a block */
if (pos > 56) {
memset(c->buf + pos, 0, 64 - pos);
sha1_core(c->iv, c->buf, 1);
pos = 0;
}
memset(c->buf + pos, 0, 56 - pos);
/* Append total input length in bits */
((uint32_t *)c->buf)[14] = htonl((uint32_t)(c->len >> 29));
((uint32_t *)c->buf)[15] = htonl((uint32_t)c->len << 3);
/* Final hash round */
sha1_core(c->iv, c->buf, 1);
/* Copy hash result out */
for (i = 0; i < 5; i++)
put_be32(hash + 4*i, c->iv[i]);
}
/*
* Helper macros for sha1_core function. To avoid clutter, these macros
* are NOT fully parenthesized if it doesn't matter to the actual use.
*/
#if __GNUC__ && (defined(__i386__) || defined(__x86_64__))
/*
* GCC by itself only generates left rotates. Use right rotates if
* possible to be kinder to dinky implementations with iterative rotate
* instructions.
*/
#define ROT(op, x, k) \
({ uint32_t y; asm(op " %1,%0" : "=r" (y) : "I" (k), "0" (x)); y; })
#define ROL(x,k) ROT("roll", x, k)
#define ROR(x,k) ROT("rorl", x, k)
#else
/* Generic C equivalent */
#define ROT(x,l,r) ((x) << (l) | (x) >> (r))
#define ROL(x,k) ROT(x,k,32-(k))
#define ROR(x,k) ROT(x,32-(k),k)
#endif
/*
* An important temporary array in SHA-1 is the working array W[],
* which holds a scheduled key word per round. Only the last 16 words
* are relevant, so only use 16 words...
*/
#define W(i) w[(i) & 15]
/*
* If you have a small register set, it helps GCC to force stores to
* the w[] array to memory. Given 22 or more registers (e.g. PowerPC),
* GCC can map the entire w[] array to registers and this becomes
* counterproductive.
*
* The optimal kludge seems to differ between x86 and ARM. On the latter,
* forcing a full memory barrier is required to stop GCC from splitting
* live ranges for each round and generating a separate stack slot for
* each. (Which produces a huge stack frame and kills performance.)
*/
#if defined(__i386__) || defined(__x86_64__)
#define STORE(i, x) *(volatile uint32_t *)&W(i) = x
#elif __GNUC__ && defined(__arm__)
#define STORE(i, x) W(i) = x; __asm__("":::"memory")
#else
#define STORE(i, x) W(i) = x
#endif
/* The three round functions. F2 is also used as F4 */
#define F1(b,c,d) (((d ^ c) & b) ^ d) /* Bitwise b ? c : d */
#define F2(b,c,d) (d ^ c ^ b) /* Even parity */
#define F3(b,c,d) (d & c) + ((d ^ c) & b) /* Majority function */
/* The four round constants */
#define K1 0x5a827999 /* 2^30 * sqrt(2) */
#define K2 0x6ed9eba1 /* 2^30 * sqrt(3) */
#define K3 0x8f1bbcdc /* 2^30 * sqrt(5) */
#define K4 0xca62c1d6 /* 2^30 * sqrt(10) */
/* Rounds 0..15 fetch a word from the input */
#define FETCH(t,i) t = get_be32(p + 4*(i)); STORE(i,t)
/* Rounds 16..79 mix previous words to get a new one */
#define MIX(t,i) t = W(i) ^ W(i+2) ^ W(i+8) ^ W(i+13); t = ROL(t, 1)
/* Rounds 16..76 have to store back the result */
#define CALC(t,i) MIX(t,i); STORE(i,t)
/* The basic SHA-1 round */
#define SHA_ROUND(a,b,c,d,e,f,k,src,t,i) \
src(t,i); \
e += t + f(b,c,d) + k + ROL(a,5); \
b = ROR(b,2)
/* An aligned group of 5 rounds */
#define SHA_ROUND5(f,k,src,t,i) \
SHA_ROUND(a,b,c,d,e, f,k,src,t,i); \
SHA_ROUND(e,a,b,c,d, f,k,src,t,i+1); \
SHA_ROUND(d,e,a,b,c, f,k,src,t,i+2); \
SHA_ROUND(c,d,e,a,b, f,k,src,t,i+3); \
SHA_ROUND(b,c,d,e,a, f,k,src,t,i+4)
/*
* The core SHA-1 transform.
*
* iv[5] = Current SHA-1 hash state.
* p = Pointer to source data. Not necessarily aligned.
* nblocks = Number of 64-byte blocks at p. Guaranteed non-zero.
*/
static void
sha1_core(uint32_t iv[5], unsigned char const *p, size_t nblocks)
{
uint32_t e = iv[4], d = iv[3], c = iv[2], b = iv[1], a = iv[0];
uint32_t w[16];
do {
uint32_t t;
SHA_ROUND5(F1, K1, FETCH, t, 0);
SHA_ROUND5(F1, K1, FETCH, t, 5);
SHA_ROUND5(F1, K1, FETCH, t, 10);
SHA_ROUND(a,b,c,d,e, F1, K1, FETCH, t, 15);
SHA_ROUND(e,a,b,c,d, F1, K1, CALC, t, 16);
SHA_ROUND(d,e,a,b,c, F1, K1, CALC, t, 17);
SHA_ROUND(c,d,e,a,b, F1, K1, CALC, t, 18);
SHA_ROUND(b,c,d,e,a, F1, K1, CALC, t, 19);
SHA_ROUND5(F2, K2, CALC, t, 20);
SHA_ROUND5(F2, K2, CALC, t, 25);
SHA_ROUND5(F2, K2, CALC, t, 30);
SHA_ROUND5(F2, K2, CALC, t, 35);
SHA_ROUND5(F3, K3, CALC, t, 40);
SHA_ROUND5(F3, K3, CALC, t, 45);
SHA_ROUND5(F3, K3, CALC, t, 50);
SHA_ROUND5(F3, K3, CALC, t, 55);
SHA_ROUND5(F2, K4, CALC, t, 60);
SHA_ROUND5(F2, K4, CALC, t, 65);
SHA_ROUND5(F2, K4, CALC, t, 70);
SHA_ROUND(a,b,c,d,e, F2, K4, CALC, t, 75);
SHA_ROUND(e,a,b,c,d, F2, K4, CALC, t, 76);
/* Last 3 rounds don't need to store mixed value */
SHA_ROUND(d,e,a,b,c, F2, K4, MIX, t, 77);
SHA_ROUND(c,d,e,a,b, F2, K4, MIX, t, 78);
SHA_ROUND(b,c,d,e,a, F2, K4, MIX, t, 79);
iv[4] = e += iv[4];
iv[3] = d += iv[3];
iv[2] = c += iv[2];
iv[1] = b += iv[1];
iv[0] = a += iv[0];
} while (--nblocks);
}
/* Compile with -DUNITTEST for self-tests */
#if UNITTEST
#include <stdio.h>
/* Known answer test */
static void katest(char const *text, size_t len, unsigned char const hash[20])
{
SHA_CTX c;
unsigned char hash2[20];
int i;
SHA1_Init(&c);
SHA1_Update(&c, text, len);
SHA1_Final(hash2, &c);
for (i = 0; i < 20; i++)
if (hash[i] != hash2[i])
goto mismatch;
printf("%u-byte known answer test PASSED\n", len);
return;
mismatch:
printf("%u-byte known answer test FAILED:\n", len);
printf("Input: \"%.*s\"\n", len, text);
printf("Computed:");
for (i = 0; i < 20; i++)
printf(" %02x", hash2[i]);
printf("\nExpected:");
for (i = 0; i < 20; i++)
printf(" %02x", hash[i]);
putchar('\n');
}
int
main(void)
{
/* FIPS PUB 180.1 example A.1 */
static char const text1[3] = "abc";
static unsigned char const hash1[20] = {
0xa9, 0x99, 0x3e, 0x36, 0x47, 0x06, 0x81, 0x6a, 0xba, 0x3e,
0x25, 0x71, 0x78, 0x50, 0xc2, 0x6c, 0x9c, 0xd0, 0xd8, 0x9d };
/* FIPS PUB 180.1 example A.2 */
static char const text2[56] =
"abcdbcdecdefdefgefghfghighijhijkijkljklmklmnlmnomnopnopq";
static unsigned char const hash2[20] = {
0x84, 0x98, 0x3e, 0x44, 0x1c, 0x3b, 0xd2, 0x6e, 0xba, 0xae,
0x4a, 0xa1, 0xf9, 0x51, 0x29, 0xe5, 0xe5, 0x46, 0x70, 0xf1 };
katest(text1, sizeof text1, hash1);
katest(text2, sizeof text2, hash2);
return 0;
}
#endif
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox