git/rev-list.c

720 lines
16 KiB
C
Raw Normal View History

#include "cache.h"
#include "refs.h"
#include "tag.h"
#include "commit.h"
#include "tree.h"
#include "blob.h"
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
#include "epoch.h"
2005-10-21 04:25:09 +00:00
#include "diff.h"
#include "revision.h"
/* bits #0 and #1 in revision.h */
#define COUNTED (1u << 2)
#define SHOWN (1u << 3)
#define TREECHANGE (1u << 4)
#define TMP_MARK (1u << 5) /* for isolated cases; clean after use */
static const char rev_list_usage[] =
"git-rev-list [OPTION] <commit-id>... [ -- paths... ]\n"
" limiting output:\n"
" --max-count=nr\n"
" --max-age=epoch\n"
" --min-age=epoch\n"
" --sparse\n"
" --no-merges\n"
" --remove-empty\n"
" --all\n"
" ordering output:\n"
" --merge-order [ --show-breaks ]\n"
" --topo-order\n"
" --date-order\n"
" formatting output:\n"
" --parents\n"
" --objects | --objects-edge\n"
" --unpacked\n"
" --header | --pretty\n"
" --abbrev=nr | --no-abbrev\n"
" special purpose:\n"
" --bisect"
;
struct rev_info revs;
static int unpacked = 0;
static int bisect_list = 0;
static int verbose_header = 0;
static int abbrev = DEFAULT_ABBREV;
static int show_parents = 0;
static int hdr_termination = 0;
static const char *commit_prefix = "";
static enum cmit_fmt commit_format = CMIT_FMT_RAW;
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
static int merge_order = 0;
static int show_breaks = 0;
static int stop_traversal = 0;
static int no_merges = 0;
static void show_commit(struct commit *commit)
{
commit->object.flags |= SHOWN;
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
if (show_breaks) {
commit_prefix = "| ";
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
if (commit->object.flags & DISCONTINUITY) {
commit_prefix = "^ ";
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
} else if (commit->object.flags & BOUNDARY) {
commit_prefix = "= ";
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
}
}
printf("%s%s", commit_prefix, sha1_to_hex(commit->object.sha1));
if (show_parents) {
struct commit_list *parents = commit->parents;
while (parents) {
struct object *o = &(parents->item->object);
parents = parents->next;
if (o->flags & TMP_MARK)
continue;
printf(" %s", sha1_to_hex(o->sha1));
o->flags |= TMP_MARK;
}
/* TMP_MARK is a general purpose flag that can
* be used locally, but the user should clean
* things up after it is done with them.
*/
for (parents = commit->parents;
parents;
parents = parents->next)
parents->item->object.flags &= ~TMP_MARK;
}
if (commit_format == CMIT_FMT_ONELINE)
putchar(' ');
else
putchar('\n');
if (verbose_header) {
static char pretty_header[16384];
pretty_print_commit(commit_format, commit, ~0, pretty_header, sizeof(pretty_header), abbrev);
printf("%s%c", pretty_header, hdr_termination);
}
fflush(stdout);
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
}
static int rewrite_one(struct commit **pp)
{
for (;;) {
struct commit *p = *pp;
if (p->object.flags & (TREECHANGE | UNINTERESTING))
return 0;
if (!p->parents)
return -1;
*pp = p->parents->item;
}
}
static void rewrite_parents(struct commit *commit)
{
struct commit_list **pp = &commit->parents;
while (*pp) {
struct commit_list *parent = *pp;
if (rewrite_one(&parent->item) < 0) {
*pp = parent->next;
continue;
}
pp = &parent->next;
}
}
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
static int filter_commit(struct commit * commit)
{
if (stop_traversal && (commit->object.flags & BOUNDARY))
return STOP;
if (commit->object.flags & (UNINTERESTING|SHOWN))
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
return CONTINUE;
if (revs.min_age != -1 && (commit->date > revs.min_age))
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
return CONTINUE;
if (revs.max_age != -1 && (commit->date < revs.max_age)) {
stop_traversal=1;
return CONTINUE;
}
if (no_merges && (commit->parents && commit->parents->next))
return CONTINUE;
if (revs.paths && revs.dense) {
if (!(commit->object.flags & TREECHANGE))
return CONTINUE;
rewrite_parents(commit);
}
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
return DO;
}
static int process_commit(struct commit * commit)
{
int action=filter_commit(commit);
if (action == STOP) {
return STOP;
}
if (action == CONTINUE) {
return CONTINUE;
}
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
if (revs.max_count != -1 && !revs.max_count--)
return STOP;
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
show_commit(commit);
return CONTINUE;
}
static struct object_list **process_blob(struct blob *blob,
struct object_list **p,
struct name_path *path,
const char *name)
{
struct object *obj = &blob->object;
if (!revs.blob_objects)
return p;
if (obj->flags & (UNINTERESTING | SEEN))
return p;
obj->flags |= SEEN;
return add_object(obj, p, path, name);
}
static struct object_list **process_tree(struct tree *tree,
struct object_list **p,
struct name_path *path,
const char *name)
{
struct object *obj = &tree->object;
struct tree_entry_list *entry;
struct name_path me;
if (!revs.tree_objects)
return p;
if (obj->flags & (UNINTERESTING | SEEN))
return p;
if (parse_tree(tree) < 0)
die("bad tree object %s", sha1_to_hex(obj->sha1));
obj->flags |= SEEN;
p = add_object(obj, p, path, name);
me.up = path;
me.elem = name;
me.elem_len = strlen(name);
entry = tree->entries;
tree->entries = NULL;
while (entry) {
struct tree_entry_list *next = entry->next;
if (entry->directory)
p = process_tree(entry->item.tree, p, &me, entry->name);
else
p = process_blob(entry->item.blob, p, &me, entry->name);
free(entry);
entry = next;
}
return p;
}
static void show_commit_list(struct commit_list *list)
{
struct object_list *objects = NULL, **p = &objects, *pending;
while (list) {
struct commit *commit = pop_most_recent_commit(&list, SEEN);
p = process_tree(commit->tree, p, NULL, "");
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
if (process_commit(commit) == STOP)
break;
}
for (pending = revs.pending_objects; pending; pending = pending->next) {
struct object *obj = pending->item;
const char *name = pending->name;
if (obj->flags & (UNINTERESTING | SEEN))
continue;
if (obj->type == tag_type) {
obj->flags |= SEEN;
p = add_object(obj, p, NULL, name);
continue;
}
if (obj->type == tree_type) {
p = process_tree((struct tree *)obj, p, NULL, name);
continue;
}
if (obj->type == blob_type) {
p = process_blob((struct blob *)obj, p, NULL, name);
continue;
}
die("unknown pending object %s (%s)", sha1_to_hex(obj->sha1), name);
}
while (objects) {
/* An object with name "foo\n0000000..." can be used to
* confuse downstream git-pack-objects very badly.
*/
const char *ep = strchr(objects->name, '\n');
if (ep) {
printf("%s %.*s\n", sha1_to_hex(objects->item->sha1),
(int) (ep - objects->name),
objects->name);
}
else
printf("%s %s\n", sha1_to_hex(objects->item->sha1), objects->name);
objects = objects->next;
}
}
static int everybody_uninteresting(struct commit_list *orig)
{
struct commit_list *list = orig;
while (list) {
struct commit *commit = list->item;
list = list->next;
if (commit->object.flags & UNINTERESTING)
continue;
return 0;
}
return 1;
}
/*
* This is a truly stupid algorithm, but it's only
* used for bisection, and we just don't care enough.
*
* We care just barely enough to avoid recursing for
* non-merge entries.
*/
static int count_distance(struct commit_list *entry)
{
int nr = 0;
while (entry) {
struct commit *commit = entry->item;
struct commit_list *p;
if (commit->object.flags & (UNINTERESTING | COUNTED))
break;
if (!revs.paths || (commit->object.flags & TREECHANGE))
nr++;
commit->object.flags |= COUNTED;
p = commit->parents;
entry = p;
if (p) {
p = p->next;
while (p) {
nr += count_distance(p);
p = p->next;
}
}
}
return nr;
}
static void clear_distance(struct commit_list *list)
{
while (list) {
struct commit *commit = list->item;
commit->object.flags &= ~COUNTED;
list = list->next;
}
}
static struct commit_list *find_bisection(struct commit_list *list)
{
int nr, closest;
struct commit_list *p, *best;
nr = 0;
p = list;
while (p) {
if (!revs.paths || (p->item->object.flags & TREECHANGE))
nr++;
p = p->next;
}
closest = 0;
best = list;
for (p = list; p; p = p->next) {
int distance;
if (revs.paths && !(p->item->object.flags & TREECHANGE))
continue;
distance = count_distance(p);
clear_distance(list);
if (nr - distance < distance)
distance = nr - distance;
if (distance > closest) {
best = p;
closest = distance;
}
}
if (best)
best->next = NULL;
return best;
}
static void mark_edge_parents_uninteresting(struct commit *commit)
{
struct commit_list *parents;
for (parents = commit->parents; parents; parents = parents->next) {
struct commit *parent = parents->item;
if (!(parent->object.flags & UNINTERESTING))
continue;
mark_tree_uninteresting(parent->tree);
if (revs.edge_hint && !(parent->object.flags & SHOWN)) {
parent->object.flags |= SHOWN;
printf("-%s\n", sha1_to_hex(parent->object.sha1));
}
}
}
static void mark_edges_uninteresting(struct commit_list *list)
{
for ( ; list; list = list->next) {
struct commit *commit = list->item;
if (commit->object.flags & UNINTERESTING) {
mark_tree_uninteresting(commit->tree);
continue;
}
mark_edge_parents_uninteresting(commit);
}
}
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
#define TREE_SAME 0
#define TREE_NEW 1
#define TREE_DIFFERENT 2
static int tree_difference = TREE_SAME;
2005-10-21 04:25:09 +00:00
static void file_add_remove(struct diff_options *options,
int addremove, unsigned mode,
const unsigned char *sha1,
const char *base, const char *path)
{
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
int diff = TREE_DIFFERENT;
/*
* Is it an add of a new file? It means that
* the old tree didn't have it at all, so we
* will turn "TREE_SAME" -> "TREE_NEW", but
* leave any "TREE_DIFFERENT" alone (and if
* it already was "TREE_NEW", we'll keep it
* "TREE_NEW" of course).
*/
if (addremove == '+') {
diff = tree_difference;
if (diff != TREE_SAME)
return;
diff = TREE_NEW;
}
tree_difference = diff;
2005-10-21 04:25:09 +00:00
}
static void file_change(struct diff_options *options,
unsigned old_mode, unsigned new_mode,
const unsigned char *old_sha1,
const unsigned char *new_sha1,
const char *base, const char *path)
{
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
tree_difference = TREE_DIFFERENT;
2005-10-21 04:25:09 +00:00
}
static struct diff_options diff_opt = {
.recursive = 1,
.add_remove = file_add_remove,
.change = file_change,
};
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
static int compare_tree(struct tree *t1, struct tree *t2)
{
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
if (!t1)
return TREE_NEW;
if (!t2)
return TREE_DIFFERENT;
tree_difference = TREE_SAME;
if (diff_tree_sha1(t1->object.sha1, t2->object.sha1, "", &diff_opt) < 0)
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
return TREE_DIFFERENT;
return tree_difference;
}
static int same_tree_as_empty(struct tree *t1)
{
int retval;
void *tree;
struct tree_desc empty, real;
if (!t1)
return 0;
tree = read_object_with_reference(t1->object.sha1, "tree", &real.size, NULL);
if (!tree)
return 0;
real.buf = tree;
empty.buf = "";
empty.size = 0;
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
tree_difference = 0;
retval = diff_tree(&empty, &real, "", &diff_opt);
free(tree);
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
return retval >= 0 && !tree_difference;
}
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
static void try_to_simplify_commit(struct commit *commit)
2005-10-21 04:25:09 +00:00
{
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
struct commit_list **pp, *parent;
2005-10-21 04:25:09 +00:00
if (!commit->tree)
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
return;
2005-10-21 04:25:09 +00:00
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
if (!commit->parents) {
if (!same_tree_as_empty(commit->tree))
commit->object.flags |= TREECHANGE;
return;
}
pp = &commit->parents;
while ((parent = *pp) != NULL) {
2005-10-21 04:25:09 +00:00
struct commit *p = parent->item;
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
if (p->object.flags & UNINTERESTING) {
pp = &parent->next;
continue;
}
2005-10-21 04:25:09 +00:00
parse_commit(p);
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
switch (compare_tree(p->tree, commit->tree)) {
case TREE_SAME:
parent->next = NULL;
commit->parents = parent;
return;
case TREE_NEW:
if (revs.remove_empty_trees && same_tree_as_empty(p->tree)) {
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
*pp = parent->next;
continue;
}
/* fallthrough */
case TREE_DIFFERENT:
pp = &parent->next;
2005-10-21 04:25:09 +00:00
continue;
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
}
die("bad tree compare for commit %s", sha1_to_hex(commit->object.sha1));
2005-10-21 04:25:09 +00:00
}
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
commit->object.flags |= TREECHANGE;
2005-10-21 04:25:09 +00:00
}
static void add_parents_to_list(struct commit *commit, struct commit_list **list)
{
struct commit_list *parent = commit->parents;
/*
* If the commit is uninteresting, don't try to
* prune parents - we want the maximal uninteresting
* set.
*
* Normally we haven't parsed the parent
* yet, so we won't have a parent of a parent
* here. However, it may turn out that we've
* reached this commit some other way (where it
* wasn't uninteresting), in which case we need
* to mark its parents recursively too..
*/
if (commit->object.flags & UNINTERESTING) {
while (parent) {
struct commit *p = parent->item;
parent = parent->next;
parse_commit(p);
p->object.flags |= UNINTERESTING;
if (p->parents)
mark_parents_uninteresting(p);
if (p->object.flags & SEEN)
continue;
p->object.flags |= SEEN;
insert_by_date(p, list);
}
return;
}
/*
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
* Ok, the commit wasn't uninteresting. Try to
* simplify the commit history and find the parent
* that has no differences in the path set if one exists.
2005-10-21 04:25:09 +00:00
*/
if (revs.paths)
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
try_to_simplify_commit(commit);
2005-10-21 04:25:09 +00:00
rev-list: stop when the file disappears The one thing I've considered doing (I really should) is to add a "stop when you don't find the file" option to "git-rev-list". This patch does some of the work towards that: it removes the "parent" thing when the file disappears, so a "git annotate" could do do something like git-rev-list --remove-empty --parents HEAD -- "$filename" and it would get a good graph that stops when the filename disappears (it's not perfect though: it won't remove all the unintersting commits). It also simplifies the logic of finding tree differences a bit, at the cost of making it a tad less efficient. The old logic was two-phase: it would first simplify _only_ merges tree as it traversed the tree, and then simplify the linear parts of the remainder independently. That was pretty optimal from an efficiency standpoint because it avoids doing any comparisons that we can see are unnecessary, but it made it much harder to understand than it really needed to be. The new logic is a lot more straightforward, and compares the trees as it traverses the graph (ie everything is a single phase). That makes it much easier to stop graph traversal at any point where a file disappears. As an example, let's say that you have a git repository that has had a file called "A" some time in the past. That file gets renamed to B, and then gets renamed back again to A. The old "git-rev-list" would show two commits: the commit that renames B to A (because it changes A) _and_ as its parent the commit that renames A to B (because it changes A). With the new --remove-empty flag, git-rev-list will show just the commit that renames B to A as the "root" commit, and stop traversal there (because that's what you want for "annotate" - you want to stop there, and for every "root" commit you then separately see if it really is a new file, or if the paths history disappeared because it was renamed from some other file). With this patch, you should be able to basically do a "poor mans 'git annotate'" with a fairly simple loop: push("HEAD", "$filename") while (revision,filename = pop()) { for each i in $(git-rev-list --parents --remove-empty $revision -- "$filename") pseudo-parents($i) = git-rev-list parents for that line if (pseudo-parents($i) is non-empty) { show diff of $i against pseudo-parents continue } /* See if the _real_ parents of $i had a rename */ parent($i) = real-parent($i) if (find-rename in $parent($i)->$i) push $parent($i), "old-name" } which should be doable in perl or something (doing stacks in shell is just too painful to be worth it, so I'm not going to do this). Anybody want to try? Linus
2006-01-18 22:47:30 +00:00
parent = commit->parents;
2005-10-21 04:25:09 +00:00
while (parent) {
struct commit *p = parent->item;
parent = parent->next;
parse_commit(p);
if (p->object.flags & SEEN)
continue;
p->object.flags |= SEEN;
insert_by_date(p, list);
}
}
static struct commit_list *limit_list(struct commit_list *list)
{
struct commit_list *newlist = NULL;
struct commit_list **p = &newlist;
while (list) {
2005-10-21 04:25:09 +00:00
struct commit_list *entry = list;
struct commit *commit = list->item;
struct object *obj = &commit->object;
2005-10-21 04:25:09 +00:00
list = list->next;
free(entry);
if (revs.max_age != -1 && (commit->date < revs.max_age))
obj->flags |= UNINTERESTING;
if (unpacked && has_sha1_pack(obj->sha1))
obj->flags |= UNINTERESTING;
2005-10-21 04:25:09 +00:00
add_parents_to_list(commit, &list);
if (obj->flags & UNINTERESTING) {
mark_parents_uninteresting(commit);
if (everybody_uninteresting(list))
break;
continue;
}
if (revs.min_age != -1 && (commit->date > revs.min_age))
continue;
p = &commit_list_insert(commit, p)->next;
}
if (revs.tree_objects)
mark_edges_uninteresting(newlist);
if (bisect_list)
newlist = find_bisection(newlist);
return newlist;
}
2005-10-21 04:25:09 +00:00
int main(int argc, const char **argv)
{
struct commit_list *list;
int i, limited = 0;
argc = setup_revisions(argc, argv, &revs);
for (i = 1 ; i < argc; i++) {
2005-10-21 04:25:09 +00:00
const char *arg = argv[i];
/* accept -<digit>, like traditilnal "head" */
if ((*arg == '-') && isdigit(arg[1])) {
revs.max_count = atoi(arg + 1);
continue;
}
if (!strcmp(arg, "-n")) {
if (++i >= argc)
die("-n requires an argument");
revs.max_count = atoi(argv[i]);
continue;
}
if (!strncmp(arg,"-n",2)) {
revs.max_count = atoi(arg + 2);
continue;
}
if (!strcmp(arg, "--header")) {
verbose_header = 1;
continue;
}
if (!strcmp(arg, "--no-abbrev")) {
abbrev = 0;
continue;
}
if (!strncmp(arg, "--abbrev=", 9)) {
abbrev = strtoul(arg + 9, NULL, 10);
if (abbrev && abbrev < MINIMUM_ABBREV)
abbrev = MINIMUM_ABBREV;
else if (40 < abbrev)
abbrev = 40;
continue;
}
if (!strncmp(arg, "--pretty", 8)) {
commit_format = get_commit_format(arg+8);
verbose_header = 1;
hdr_termination = '\n';
if (commit_format == CMIT_FMT_ONELINE)
commit_prefix = "";
else
commit_prefix = "commit ";
continue;
}
if (!strncmp(arg, "--no-merges", 11)) {
no_merges = 1;
continue;
}
if (!strcmp(arg, "--parents")) {
show_parents = 1;
continue;
}
if (!strcmp(arg, "--bisect")) {
bisect_list = 1;
continue;
}
if (!strcmp(arg, "--unpacked")) {
unpacked = 1;
limited = 1;
continue;
}
if (!strcmp(arg, "--merge-order")) {
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
merge_order = 1;
continue;
}
if (!strcmp(arg, "--show-breaks")) {
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
show_breaks = 1;
continue;
}
usage(rev_list_usage);
}
list = revs.commits;
if (list)
limited = 1;
if (revs.topo_order)
limited = 1;
if (!list &&
(!(revs.tag_objects||revs.tree_objects||revs.blob_objects) && !revs.pending_objects))
usage(rev_list_usage);
if (revs.paths) {
limited = 1;
diff_tree_setup_paths(revs.paths);
}
if (revs.max_age != -1 || revs.min_age != -1)
limited = 1;
[PATCH] Avoid wasting memory in git-rev-list As pointed out on the list, git-rev-list can use a lot of memory. One low-hanging fruit is to free the commit buffer for commits that we parse. By default, parse_commit() will save away the buffer, since a lot of cases do want it, and re-reading it continually would be unnecessary. However, in many cases the buffer isn't actually necessary and saving it just wastes memory. We could just free the buffer ourselves, but especially in git-rev-list, we actually end up using the helper functions that automatically add parent commits to the commit lists, so we don't actually control the commit parsing directly. Instead, just make this behaviour of "parse_commit()" a global flag. Maybe this is a bit tasteless, but it's very simple, and it makes a noticable difference in memory usage. Before the change: [torvalds@g5 linux]$ /usr/bin/time git-rev-list v2.6.12..HEAD > /dev/null 0.26user 0.02system 0:00.28elapsed 99%CPU (0avgtext+0avgdata 0maxresident)k 0inputs+0outputs (0major+3714minor)pagefaults 0swaps after the change: [torvalds@g5 linux]$ /usr/bin/time git-rev-list v2.6.12..HEAD > /dev/null 0.26user 0.00system 0:00.27elapsed 100%CPU (0avgtext+0avgdata 0maxresident)k 0inputs+0outputs (0major+2433minor)pagefaults 0swaps note how the minor faults have decreased from 3714 pages to 2433 pages. That's all due to the fewer anonymous pages allocated to hold the comment buffers and their metadata. Signed-off-by: Linus Torvalds <torvalds@osdl.org> Signed-off-by: Junio C Hamano <junkio@cox.net>
2005-09-15 21:43:17 +00:00
save_commit_buffer = verbose_header;
track_object_refs = 0;
[PATCH] Avoid wasting memory in git-rev-list As pointed out on the list, git-rev-list can use a lot of memory. One low-hanging fruit is to free the commit buffer for commits that we parse. By default, parse_commit() will save away the buffer, since a lot of cases do want it, and re-reading it continually would be unnecessary. However, in many cases the buffer isn't actually necessary and saving it just wastes memory. We could just free the buffer ourselves, but especially in git-rev-list, we actually end up using the helper functions that automatically add parent commits to the commit lists, so we don't actually control the commit parsing directly. Instead, just make this behaviour of "parse_commit()" a global flag. Maybe this is a bit tasteless, but it's very simple, and it makes a noticable difference in memory usage. Before the change: [torvalds@g5 linux]$ /usr/bin/time git-rev-list v2.6.12..HEAD > /dev/null 0.26user 0.02system 0:00.28elapsed 99%CPU (0avgtext+0avgdata 0maxresident)k 0inputs+0outputs (0major+3714minor)pagefaults 0swaps after the change: [torvalds@g5 linux]$ /usr/bin/time git-rev-list v2.6.12..HEAD > /dev/null 0.26user 0.00system 0:00.27elapsed 100%CPU (0avgtext+0avgdata 0maxresident)k 0inputs+0outputs (0major+2433minor)pagefaults 0swaps note how the minor faults have decreased from 3714 pages to 2433 pages. That's all due to the fewer anonymous pages allocated to hold the comment buffers and their metadata. Signed-off-by: Linus Torvalds <torvalds@osdl.org> Signed-off-by: Junio C Hamano <junkio@cox.net>
2005-09-15 21:43:17 +00:00
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
if (!merge_order) {
sort_by_date(&list);
if (list && !limited && revs.max_count == 1 &&
!revs.tag_objects && !revs.tree_objects && !revs.blob_objects) {
show_commit(list->item);
return 0;
}
if (limited)
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
list = limit_list(list);
if (revs.topo_order)
sort_in_topological_order(&list, revs.lifo);
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
show_commit_list(list);
} else {
#ifndef NO_OPENSSL
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
if (sort_list_in_merge_order(list, &process_commit)) {
die("merge order sort failed\n");
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
}
#else
die("merge order sort unsupported, OpenSSL not linked");
#endif
[PATCH] Modify git-rev-list to linearise the commit history in merge order. This patch linearises the GIT commit history graph into merge order which is defined by invariants specified in Documentation/git-rev-list.txt. The linearisation produced by this patch is superior in an objective sense to that produced by the existing git-rev-list implementation in that the linearisation produced is guaranteed to have the minimum number of discontinuities, where a discontinuity is defined as an adjacent pair of commits in the output list which are not related in a direct child-parent relationship. With this patch a graph like this: a4 --- | \ \ | b4 | |/ | | a3 | | | | | a2 | | | | c3 | | | | | c2 | b3 | | | /| | b2 | | | c1 | | / | b1 a1 | | | a0 | | / root Sorts like this: = a4 | c3 | c2 | c1 ^ b4 | b3 | b2 | b1 ^ a3 | a2 | a1 | a0 = root Instead of this: = a4 | c3 ^ b4 | a3 ^ c2 ^ b3 ^ a2 ^ b2 ^ c1 ^ a1 ^ b1 ^ a0 = root A test script, t/t6000-rev-list.sh, includes a test which demonstrates that the linearisation produced by --merge-order has less discontinuities than the linearisation produced by git-rev-list without the --merge-order flag specified. To see this, do the following: cd t ./t6000-rev-list.sh cd trash cat actual-default-order cat actual-merge-order The existing behaviour of git-rev-list is preserved, by default. To obtain the modified behaviour, specify --merge-order or --merge-order --show-breaks on the command line. This version of the patch has been tested on the git repository and also on the linux-2.6 repository and has reasonable performance on both - ~50-100% slower than the original algorithm. This version of the patch has incorporated a functional equivalent of the Linus' output limiting algorithm into the merge-order algorithm itself. This operates per the notes associated with Linus' commit 337cb3fb8da45f10fe9a0c3cf571600f55ead2ce. This version has incorporated Linus' feedback regarding proposed changes to rev-list.c. (see: [PATCH] Factor out filtering in rev-list.c) This version has improved the way sort_first_epoch marks commits as uninteresting. For more details about this change, refer to Documentation/git-rev-list.txt and http://blackcubes.dyndns.org/epoch/. Signed-off-by: Jon Seymour <jon.seymour@gmail.com> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-06 15:39:40 +00:00
}
return 0;
}