git/builtin-apply.c

2894 lines
70 KiB
C
Raw Normal View History

/*
* apply.c
*
* Copyright (C) Linus Torvalds, 2005
*
* This applies patches on top of some (arbitrary) version of the SCM.
*
*/
#include "cache.h"
#include "cache-tree.h"
#include "quote.h"
#include "blob.h"
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
#include "delta.h"
#include "builtin.h"
/*
* --check turns on checking that the working tree matches the
* files that are being modified, but doesn't apply the patch
* --stat does just a diffstat, and doesn't actually apply
* --numstat does numeric diffstat, and doesn't actually apply
* --index-info shows the old and new index info for paths if available.
* --index updates the cache as well.
* --cached updates only the cache without ever touching the working tree.
*/
static const char *prefix;
static int prefix_length = -1;
static int newfd = -1;
static int unidiff_zero;
static int p_value = 1;
static int p_value_known;
static int check_index;
static int update_index;
static int cached;
static int diffstat;
static int numstat;
static int summary;
static int check;
static int apply = 1;
static int apply_in_reverse;
static int apply_with_reject;
static int apply_verbosely;
static int no_add;
static const char *fake_ancestor;
static int line_termination = '\n';
static unsigned long p_context = ULONG_MAX;
static const char apply_usage[] =
"git-apply [--stat] [--numstat] [--summary] [--check] [--index] [--cached] [--apply] [--no-add] [--index-info] [--allow-binary-replacement] [--reverse] [--reject] [--verbose] [-z] [-pNUM] [-CNUM] [--whitespace=<nowarn|warn|error|error-all|strip>] <patch>...";
static enum whitespace_eol {
nowarn_whitespace,
warn_on_whitespace,
error_on_whitespace,
strip_whitespace,
} new_whitespace = warn_on_whitespace;
static int whitespace_error;
static int squelch_whitespace_errors = 5;
static int applied_after_fixing_ws;
static const char *patch_input_file;
static void parse_whitespace_option(const char *option)
{
if (!option) {
new_whitespace = warn_on_whitespace;
return;
}
if (!strcmp(option, "warn")) {
new_whitespace = warn_on_whitespace;
return;
}
if (!strcmp(option, "nowarn")) {
new_whitespace = nowarn_whitespace;
return;
}
if (!strcmp(option, "error")) {
new_whitespace = error_on_whitespace;
return;
}
if (!strcmp(option, "error-all")) {
new_whitespace = error_on_whitespace;
squelch_whitespace_errors = 0;
return;
}
if (!strcmp(option, "strip")) {
new_whitespace = strip_whitespace;
return;
}
die("unrecognized whitespace option '%s'", option);
}
static void set_default_whitespace_mode(const char *whitespace_option)
{
if (!whitespace_option && !apply_default_whitespace) {
new_whitespace = (apply
? warn_on_whitespace
: nowarn_whitespace);
}
}
/*
* For "diff-stat" like behaviour, we keep track of the biggest change
* we've seen, and the longest filename. That allows us to do simple
* scaling.
*/
static int max_change, max_len;
/*
* Various "current state", notably line numbers and what
* file (and how) we're patching right now.. The "is_xxxx"
* things are flags, where -1 means "don't know yet".
*/
static int linenr = 1;
/*
* This represents one "hunk" from a patch, starting with
* "@@ -oldpos,oldlines +newpos,newlines @@" marker. The
* patch text is pointed at by patch, and its byte length
* is stored in size. leading and trailing are the number
* of context lines.
*/
struct fragment {
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
unsigned long leading, trailing;
unsigned long oldpos, oldlines;
unsigned long newpos, newlines;
const char *patch;
int size;
int rejected;
struct fragment *next;
};
/*
* When dealing with a binary patch, we reuse "leading" field
* to store the type of the binary hunk, either deflated "delta"
* or deflated "literal".
*/
#define binary_patch_method leading
#define BINARY_DELTA_DEFLATED 1
#define BINARY_LITERAL_DEFLATED 2
struct patch {
char *new_name, *old_name, *def_name;
unsigned int old_mode, new_mode;
int is_new, is_delete; /* -1 = unknown, 0 = false, 1 = true */
int rejected;
unsigned long deflate_origlen;
int lines_added, lines_deleted;
int score;
unsigned int is_toplevel_relative:1;
unsigned int inaccurate_eof:1;
unsigned int is_binary:1;
unsigned int is_copy:1;
unsigned int is_rename:1;
struct fragment *fragments;
char *result;
size_t resultsize;
char old_sha1_prefix[41];
char new_sha1_prefix[41];
struct patch *next;
};
static void say_patch_name(FILE *output, const char *pre, struct patch *patch, const char *post)
{
fputs(pre, output);
if (patch->old_name && patch->new_name &&
strcmp(patch->old_name, patch->new_name)) {
quote_c_style(patch->old_name, NULL, output, 0);
fputs(" => ", output);
quote_c_style(patch->new_name, NULL, output, 0);
} else {
const char *n = patch->new_name;
if (!n)
n = patch->old_name;
quote_c_style(n, NULL, output, 0);
}
fputs(post, output);
}
#define CHUNKSIZE (8192)
#define SLOP (16)
static void read_patch_file(struct strbuf *sb, int fd)
{
if (strbuf_read(sb, fd, 0) < 0)
die("git-apply: read returned %s", strerror(errno));
/*
* Make sure that we have some slop in the buffer
* so that we can do speculative "memcmp" etc, and
* see to it that it is NUL-filled.
*/
strbuf_grow(sb, SLOP);
memset(sb->buf + sb->len, 0, SLOP);
}
static unsigned long linelen(const char *buffer, unsigned long size)
{
unsigned long len = 0;
while (size--) {
len++;
if (*buffer++ == '\n')
break;
}
return len;
}
static int is_dev_null(const char *str)
{
return !memcmp("/dev/null", str, 9) && isspace(str[9]);
}
#define TERM_SPACE 1
#define TERM_TAB 2
static int name_terminate(const char *name, int namelen, int c, int terminate)
{
if (c == ' ' && !(terminate & TERM_SPACE))
return 0;
if (c == '\t' && !(terminate & TERM_TAB))
return 0;
return 1;
}
static char *find_name(const char *line, char *def, int p_value, int terminate)
{
int len;
const char *start = line;
if (*line == '"') {
struct strbuf name;
/* Proposed "new-style" GNU patch/diff format; see
* http://marc.theaimsgroup.com/?l=git&m=112927316408690&w=2
*/
strbuf_init(&name, 0);
if (!unquote_c_style(&name, line, NULL)) {
char *cp;
for (cp = name.buf; p_value; p_value--) {
cp = strchr(cp, '/');
if (!cp)
break;
cp++;
}
if (cp) {
/* name can later be freed, so we need
* to memmove, not just return cp
*/
strbuf_remove(&name, 0, cp - name.buf);
free(def);
return strbuf_detach(&name, NULL);
}
}
strbuf_release(&name);
}
for (;;) {
char c = *line;
if (isspace(c)) {
if (c == '\n')
break;
if (name_terminate(start, line-start, c, terminate))
break;
}
line++;
if (c == '/' && !--p_value)
start = line;
}
if (!start)
return def;
len = line - start;
if (!len)
return def;
/*
* Generally we prefer the shorter name, especially
* if the other one is just a variation of that with
* something else tacked on to the end (ie "file.orig"
* or "file~").
*/
if (def) {
int deflen = strlen(def);
if (deflen < len && !strncmp(start, def, deflen))
return def;
free(def);
}
return xmemdupz(start, len);
}
static int count_slashes(const char *cp)
{
int cnt = 0;
char ch;
while ((ch = *cp++))
if (ch == '/')
cnt++;
return cnt;
}
/*
* Given the string after "--- " or "+++ ", guess the appropriate
* p_value for the given patch.
*/
static int guess_p_value(const char *nameline)
{
char *name, *cp;
int val = -1;
if (is_dev_null(nameline))
return -1;
name = find_name(nameline, NULL, 0, TERM_SPACE | TERM_TAB);
if (!name)
return -1;
cp = strchr(name, '/');
if (!cp)
val = 0;
else if (prefix) {
/*
* Does it begin with "a/$our-prefix" and such? Then this is
* very likely to apply to our directory.
*/
if (!strncmp(name, prefix, prefix_length))
val = count_slashes(prefix);
else {
cp++;
if (!strncmp(cp, prefix, prefix_length))
val = count_slashes(prefix) + 1;
}
}
free(name);
return val;
}
/*
* Get the name etc info from the --/+++ lines of a traditional patch header
*
* FIXME! The end-of-filename heuristics are kind of screwy. For existing
* files, we can happily check the index for a match, but for creating a
* new file we should try to match whatever "patch" does. I have no idea.
*/
static void parse_traditional_patch(const char *first, const char *second, struct patch *patch)
{
char *name;
first += 4; /* skip "--- " */
second += 4; /* skip "+++ " */
if (!p_value_known) {
int p, q;
p = guess_p_value(first);
q = guess_p_value(second);
if (p < 0) p = q;
if (0 <= p && p == q) {
p_value = p;
p_value_known = 1;
}
}
if (is_dev_null(first)) {
patch->is_new = 1;
patch->is_delete = 0;
name = find_name(second, NULL, p_value, TERM_SPACE | TERM_TAB);
patch->new_name = name;
} else if (is_dev_null(second)) {
patch->is_new = 0;
patch->is_delete = 1;
name = find_name(first, NULL, p_value, TERM_SPACE | TERM_TAB);
patch->old_name = name;
} else {
name = find_name(first, NULL, p_value, TERM_SPACE | TERM_TAB);
name = find_name(second, name, p_value, TERM_SPACE | TERM_TAB);
patch->old_name = patch->new_name = name;
}
if (!name)
die("unable to find filename in patch at line %d", linenr);
}
static int gitdiff_hdrend(const char *line, struct patch *patch)
{
return -1;
}
/*
* We're anal about diff header consistency, to make
* sure that we don't end up having strange ambiguous
* patches floating around.
*
* As a result, gitdiff_{old|new}name() will check
* their names against any previous information, just
* to make sure..
*/
static char *gitdiff_verify_name(const char *line, int isnull, char *orig_name, const char *oldnew)
{
if (!orig_name && !isnull)
return find_name(line, NULL, p_value, TERM_TAB);
if (orig_name) {
int len;
const char *name;
char *another;
name = orig_name;
len = strlen(name);
if (isnull)
die("git-apply: bad git-diff - expected /dev/null, got %s on line %d", name, linenr);
another = find_name(line, NULL, p_value, TERM_TAB);
if (!another || memcmp(another, name, len))
die("git-apply: bad git-diff - inconsistent %s filename on line %d", oldnew, linenr);
free(another);
return orig_name;
}
else {
/* expect "/dev/null" */
if (memcmp("/dev/null", line, 9) || line[9] != '\n')
die("git-apply: bad git-diff - expected /dev/null on line %d", linenr);
return NULL;
}
}
static int gitdiff_oldname(const char *line, struct patch *patch)
{
patch->old_name = gitdiff_verify_name(line, patch->is_new, patch->old_name, "old");
return 0;
}
static int gitdiff_newname(const char *line, struct patch *patch)
{
patch->new_name = gitdiff_verify_name(line, patch->is_delete, patch->new_name, "new");
return 0;
}
static int gitdiff_oldmode(const char *line, struct patch *patch)
{
patch->old_mode = strtoul(line, NULL, 8);
return 0;
}
static int gitdiff_newmode(const char *line, struct patch *patch)
{
patch->new_mode = strtoul(line, NULL, 8);
return 0;
}
static int gitdiff_delete(const char *line, struct patch *patch)
{
patch->is_delete = 1;
patch->old_name = patch->def_name;
return gitdiff_oldmode(line, patch);
}
static int gitdiff_newfile(const char *line, struct patch *patch)
{
patch->is_new = 1;
patch->new_name = patch->def_name;
return gitdiff_newmode(line, patch);
}
static int gitdiff_copysrc(const char *line, struct patch *patch)
{
patch->is_copy = 1;
patch->old_name = find_name(line, NULL, 0, 0);
return 0;
}
static int gitdiff_copydst(const char *line, struct patch *patch)
{
patch->is_copy = 1;
patch->new_name = find_name(line, NULL, 0, 0);
return 0;
}
static int gitdiff_renamesrc(const char *line, struct patch *patch)
{
patch->is_rename = 1;
patch->old_name = find_name(line, NULL, 0, 0);
return 0;
}
static int gitdiff_renamedst(const char *line, struct patch *patch)
{
patch->is_rename = 1;
patch->new_name = find_name(line, NULL, 0, 0);
return 0;
}
static int gitdiff_similarity(const char *line, struct patch *patch)
{
if ((patch->score = strtoul(line, NULL, 10)) == ULONG_MAX)
patch->score = 0;
return 0;
}
static int gitdiff_dissimilarity(const char *line, struct patch *patch)
{
if ((patch->score = strtoul(line, NULL, 10)) == ULONG_MAX)
patch->score = 0;
return 0;
}
static int gitdiff_index(const char *line, struct patch *patch)
{
/* index line is N hexadecimal, "..", N hexadecimal,
* and optional space with octal mode.
*/
const char *ptr, *eol;
int len;
ptr = strchr(line, '.');
if (!ptr || ptr[1] != '.' || 40 < ptr - line)
return 0;
len = ptr - line;
memcpy(patch->old_sha1_prefix, line, len);
patch->old_sha1_prefix[len] = 0;
line = ptr + 2;
ptr = strchr(line, ' ');
eol = strchr(line, '\n');
if (!ptr || eol < ptr)
ptr = eol;
len = ptr - line;
if (40 < len)
return 0;
memcpy(patch->new_sha1_prefix, line, len);
patch->new_sha1_prefix[len] = 0;
if (*ptr == ' ')
patch->new_mode = patch->old_mode = strtoul(ptr+1, NULL, 8);
return 0;
}
/*
* This is normal for a diff that doesn't change anything: we'll fall through
* into the next diff. Tell the parser to break out.
*/
static int gitdiff_unrecognized(const char *line, struct patch *patch)
{
return -1;
}
static const char *stop_at_slash(const char *line, int llen)
{
int i;
for (i = 0; i < llen; i++) {
int ch = line[i];
if (ch == '/')
return line + i;
}
return NULL;
}
/* This is to extract the same name that appears on "diff --git"
* line. We do not find and return anything if it is a rename
* patch, and it is OK because we will find the name elsewhere.
* We need to reliably find name only when it is mode-change only,
* creation or deletion of an empty file. In any of these cases,
* both sides are the same name under a/ and b/ respectively.
*/
static char *git_header_name(char *line, int llen)
{
const char *name;
const char *second = NULL;
size_t len;
line += strlen("diff --git ");
llen -= strlen("diff --git ");
if (*line == '"') {
const char *cp;
struct strbuf first;
struct strbuf sp;
strbuf_init(&first, 0);
strbuf_init(&sp, 0);
if (unquote_c_style(&first, line, &second))
goto free_and_fail1;
/* advance to the first slash */
cp = stop_at_slash(first.buf, first.len);
/* we do not accept absolute paths */
if (!cp || cp == first.buf)
goto free_and_fail1;
strbuf_remove(&first, 0, cp + 1 - first.buf);
/* second points at one past closing dq of name.
* find the second name.
*/
while ((second < line + llen) && isspace(*second))
second++;
if (line + llen <= second)
goto free_and_fail1;
if (*second == '"') {
if (unquote_c_style(&sp, second, NULL))
goto free_and_fail1;
cp = stop_at_slash(sp.buf, sp.len);
if (!cp || cp == sp.buf)
goto free_and_fail1;
/* They must match, otherwise ignore */
if (strcmp(cp + 1, first.buf))
goto free_and_fail1;
strbuf_release(&sp);
return strbuf_detach(&first, NULL);
}
/* unquoted second */
cp = stop_at_slash(second, line + llen - second);
if (!cp || cp == second)
goto free_and_fail1;
cp++;
if (line + llen - cp != first.len + 1 ||
memcmp(first.buf, cp, first.len))
goto free_and_fail1;
return strbuf_detach(&first, NULL);
free_and_fail1:
strbuf_release(&first);
strbuf_release(&sp);
return NULL;
}
/* unquoted first name */
name = stop_at_slash(line, llen);
if (!name || name == line)
return NULL;
name++;
/* since the first name is unquoted, a dq if exists must be
* the beginning of the second name.
*/
for (second = name; second < line + llen; second++) {
if (*second == '"') {
struct strbuf sp;
const char *np;
strbuf_init(&sp, 0);
if (unquote_c_style(&sp, second, NULL))
goto free_and_fail2;
np = stop_at_slash(sp.buf, sp.len);
if (!np || np == sp.buf)
goto free_and_fail2;
np++;
len = sp.buf + sp.len - np;
if (len < second - name &&
!strncmp(np, name, len) &&
isspace(name[len])) {
/* Good */
strbuf_remove(&sp, 0, np - sp.buf);
return strbuf_detach(&sp, NULL);
}
free_and_fail2:
strbuf_release(&sp);
return NULL;
}
}
/*
* Accept a name only if it shows up twice, exactly the same
* form.
*/
for (len = 0 ; ; len++) {
switch (name[len]) {
default:
continue;
case '\n':
return NULL;
case '\t': case ' ':
second = name+len;
for (;;) {
char c = *second++;
if (c == '\n')
return NULL;
if (c == '/')
break;
}
if (second[len] == '\n' && !memcmp(name, second, len)) {
return xmemdupz(name, len);
}
}
}
return NULL;
}
/* Verify that we recognize the lines following a git header */
static int parse_git_header(char *line, int len, unsigned int size, struct patch *patch)
{
unsigned long offset;
/* A git diff has explicit new/delete information, so we don't guess */
patch->is_new = 0;
patch->is_delete = 0;
/*
* Some things may not have the old name in the
* rest of the headers anywhere (pure mode changes,
* or removing or adding empty files), so we get
* the default name from the header.
*/
patch->def_name = git_header_name(line, len);
line += len;
size -= len;
linenr++;
for (offset = len ; size > 0 ; offset += len, size -= len, line += len, linenr++) {
static const struct opentry {
const char *str;
int (*fn)(const char *, struct patch *);
} optable[] = {
{ "@@ -", gitdiff_hdrend },
{ "--- ", gitdiff_oldname },
{ "+++ ", gitdiff_newname },
{ "old mode ", gitdiff_oldmode },
{ "new mode ", gitdiff_newmode },
{ "deleted file mode ", gitdiff_delete },
{ "new file mode ", gitdiff_newfile },
{ "copy from ", gitdiff_copysrc },
{ "copy to ", gitdiff_copydst },
{ "rename old ", gitdiff_renamesrc },
{ "rename new ", gitdiff_renamedst },
{ "rename from ", gitdiff_renamesrc },
{ "rename to ", gitdiff_renamedst },
{ "similarity index ", gitdiff_similarity },
{ "dissimilarity index ", gitdiff_dissimilarity },
{ "index ", gitdiff_index },
{ "", gitdiff_unrecognized },
};
int i;
len = linelen(line, size);
if (!len || line[len-1] != '\n')
break;
for (i = 0; i < ARRAY_SIZE(optable); i++) {
const struct opentry *p = optable + i;
int oplen = strlen(p->str);
if (len < oplen || memcmp(p->str, line, oplen))
continue;
if (p->fn(line + oplen, patch) < 0)
return offset;
break;
}
}
return offset;
}
static int parse_num(const char *line, unsigned long *p)
{
char *ptr;
if (!isdigit(*line))
return 0;
*p = strtoul(line, &ptr, 10);
return ptr - line;
}
static int parse_range(const char *line, int len, int offset, const char *expect,
unsigned long *p1, unsigned long *p2)
{
int digits, ex;
if (offset < 0 || offset >= len)
return -1;
line += offset;
len -= offset;
digits = parse_num(line, p1);
if (!digits)
return -1;
offset += digits;
line += digits;
len -= digits;
*p2 = 1;
if (*line == ',') {
digits = parse_num(line+1, p2);
if (!digits)
return -1;
offset += digits+1;
line += digits+1;
len -= digits+1;
}
ex = strlen(expect);
if (ex > len)
return -1;
if (memcmp(line, expect, ex))
return -1;
return offset + ex;
}
/*
* Parse a unified diff fragment header of the
* form "@@ -a,b +c,d @@"
*/
static int parse_fragment_header(char *line, int len, struct fragment *fragment)
{
int offset;
if (!len || line[len-1] != '\n')
return -1;
/* Figure out the number of lines in a fragment */
offset = parse_range(line, len, 4, " +", &fragment->oldpos, &fragment->oldlines);
offset = parse_range(line, len, offset, " @@", &fragment->newpos, &fragment->newlines);
return offset;
}
static int find_header(char *line, unsigned long size, int *hdrsize, struct patch *patch)
{
unsigned long offset, len;
patch->is_toplevel_relative = 0;
patch->is_rename = patch->is_copy = 0;
patch->is_new = patch->is_delete = -1;
patch->old_mode = patch->new_mode = 0;
patch->old_name = patch->new_name = NULL;
for (offset = 0; size > 0; offset += len, size -= len, line += len, linenr++) {
unsigned long nextlen;
len = linelen(line, size);
if (!len)
break;
/* Testing this early allows us to take a few shortcuts.. */
if (len < 6)
continue;
/*
* Make sure we don't find any unconnected patch fragments.
* That's a sign that we didn't find a header, and that a
* patch has become corrupted/broken up.
*/
if (!memcmp("@@ -", line, 4)) {
struct fragment dummy;
if (parse_fragment_header(line, len, &dummy) < 0)
continue;
die("patch fragment without header at line %d: %.*s",
linenr, (int)len-1, line);
}
if (size < len + 6)
break;
/*
* Git patch? It might not have a real patch, just a rename
* or mode change, so we handle that specially
*/
if (!memcmp("diff --git ", line, 11)) {
int git_hdr_len = parse_git_header(line, len, size, patch);
if (git_hdr_len <= len)
continue;
if (!patch->old_name && !patch->new_name) {
if (!patch->def_name)
die("git diff header lacks filename information (line %d)", linenr);
patch->old_name = patch->new_name = patch->def_name;
}
patch->is_toplevel_relative = 1;
*hdrsize = git_hdr_len;
return offset;
}
/** --- followed by +++ ? */
if (memcmp("--- ", line, 4) || memcmp("+++ ", line + len, 4))
continue;
/*
* We only accept unified patches, so we want it to
* at least have "@@ -a,b +c,d @@\n", which is 14 chars
* minimum
*/
nextlen = linelen(line + len, size - len);
if (size < nextlen + 14 || memcmp("@@ -", line + len + nextlen, 4))
continue;
/* Ok, we'll consider it a patch */
parse_traditional_patch(line, line+len, patch);
*hdrsize = len + nextlen;
linenr += 2;
return offset;
}
return -1;
}
static void check_whitespace(const char *line, int len)
{
const char *err = "Adds trailing whitespace";
int seen_space = 0;
int i;
/*
* We know len is at least two, since we have a '+' and we
* checked that the last character was a '\n' before calling
* this function. That is, an addition of an empty line would
* check the '+' here. Sneaky...
*/
if (isspace(line[len-2]))
goto error;
/*
* Make sure that there is no space followed by a tab in
* indentation.
*/
err = "Space in indent is followed by a tab";
for (i = 1; i < len; i++) {
if (line[i] == '\t') {
if (seen_space)
goto error;
}
else if (line[i] == ' ')
seen_space = 1;
else
break;
}
return;
error:
whitespace_error++;
if (squelch_whitespace_errors &&
squelch_whitespace_errors < whitespace_error)
;
else
fprintf(stderr, "%s.\n%s:%d:%.*s\n",
err, patch_input_file, linenr, len-2, line+1);
}
/*
* Parse a unified diff. Note that this really needs to parse each
* fragment separately, since the only way to know the difference
* between a "---" that is part of a patch, and a "---" that starts
* the next patch is to look at the line counts..
*/
static int parse_fragment(char *line, unsigned long size, struct patch *patch, struct fragment *fragment)
{
int added, deleted;
int len = linelen(line, size), offset;
unsigned long oldlines, newlines;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
unsigned long leading, trailing;
offset = parse_fragment_header(line, len, fragment);
if (offset < 0)
return -1;
oldlines = fragment->oldlines;
newlines = fragment->newlines;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
leading = 0;
trailing = 0;
/* Parse the thing.. */
line += len;
size -= len;
linenr++;
added = deleted = 0;
for (offset = len;
0 < size;
offset += len, size -= len, line += len, linenr++) {
if (!oldlines && !newlines)
break;
len = linelen(line, size);
if (!len || line[len-1] != '\n')
return -1;
switch (*line) {
default:
return -1;
case '\n': /* newer GNU diff, an empty context line */
case ' ':
oldlines--;
newlines--;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
if (!deleted && !added)
leading++;
trailing++;
break;
case '-':
if (apply_in_reverse &&
new_whitespace != nowarn_whitespace)
check_whitespace(line, len);
deleted++;
oldlines--;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
trailing = 0;
break;
case '+':
if (!apply_in_reverse &&
new_whitespace != nowarn_whitespace)
check_whitespace(line, len);
added++;
newlines--;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
trailing = 0;
break;
/* We allow "\ No newline at end of file". Depending
* on locale settings when the patch was produced we
* don't know what this line looks like. The only
* thing we do know is that it begins with "\ ".
* Checking for 12 is just for sanity check -- any
* l10n of "\ No newline..." is at least that long.
*/
case '\\':
if (len < 12 || memcmp(line, "\\ ", 2))
return -1;
break;
}
}
if (oldlines || newlines)
return -1;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
fragment->leading = leading;
fragment->trailing = trailing;
/* If a fragment ends with an incomplete line, we failed to include
* it in the above loop because we hit oldlines == newlines == 0
* before seeing it.
*/
if (12 < size && !memcmp(line, "\\ ", 2))
offset += linelen(line, size);
patch->lines_added += added;
patch->lines_deleted += deleted;
if (0 < patch->is_new && oldlines)
return error("new file depends on old contents");
if (0 < patch->is_delete && newlines)
return error("deleted file still has contents");
return offset;
}
static int parse_single_patch(char *line, unsigned long size, struct patch *patch)
{
unsigned long offset = 0;
unsigned long oldlines = 0, newlines = 0, context = 0;
struct fragment **fragp = &patch->fragments;
while (size > 4 && !memcmp(line, "@@ -", 4)) {
struct fragment *fragment;
int len;
fragment = xcalloc(1, sizeof(*fragment));
len = parse_fragment(line, size, patch, fragment);
if (len <= 0)
die("corrupt patch at line %d", linenr);
fragment->patch = line;
fragment->size = len;
oldlines += fragment->oldlines;
newlines += fragment->newlines;
context += fragment->leading + fragment->trailing;
*fragp = fragment;
fragp = &fragment->next;
offset += len;
line += len;
size -= len;
}
/*
* If something was removed (i.e. we have old-lines) it cannot
* be creation, and if something was added it cannot be
* deletion. However, the reverse is not true; --unified=0
* patches that only add are not necessarily creation even
* though they do not have any old lines, and ones that only
* delete are not necessarily deletion.
*
* Unfortunately, a real creation/deletion patch do _not_ have
* any context line by definition, so we cannot safely tell it
* apart with --unified=0 insanity. At least if the patch has
* more than one hunk it is not creation or deletion.
*/
if (patch->is_new < 0 &&
(oldlines || (patch->fragments && patch->fragments->next)))
patch->is_new = 0;
if (patch->is_delete < 0 &&
(newlines || (patch->fragments && patch->fragments->next)))
patch->is_delete = 0;
if (!unidiff_zero || context) {
/* If the user says the patch is not generated with
* --unified=0, or if we have seen context lines,
* then not having oldlines means the patch is creation,
* and not having newlines means the patch is deletion.
*/
if (patch->is_new < 0 && !oldlines) {
patch->is_new = 1;
patch->old_name = NULL;
}
if (patch->is_delete < 0 && !newlines) {
patch->is_delete = 1;
patch->new_name = NULL;
}
}
if (0 < patch->is_new && oldlines)
die("new file %s depends on old contents", patch->new_name);
if (0 < patch->is_delete && newlines)
die("deleted file %s still has contents", patch->old_name);
if (!patch->is_delete && !newlines && context)
fprintf(stderr, "** warning: file %s becomes empty but "
"is not deleted\n", patch->new_name);
return offset;
}
static inline int metadata_changes(struct patch *patch)
{
return patch->is_rename > 0 ||
patch->is_copy > 0 ||
patch->is_new > 0 ||
patch->is_delete ||
(patch->old_mode && patch->new_mode &&
patch->old_mode != patch->new_mode);
}
static char *inflate_it(const void *data, unsigned long size,
unsigned long inflated_size)
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
{
z_stream stream;
void *out;
int st;
memset(&stream, 0, sizeof(stream));
stream.next_in = (unsigned char *)data;
stream.avail_in = size;
stream.next_out = out = xmalloc(inflated_size);
stream.avail_out = inflated_size;
inflateInit(&stream);
st = inflate(&stream, Z_FINISH);
if ((st != Z_STREAM_END) || stream.total_out != inflated_size) {
free(out);
return NULL;
}
return out;
}
static struct fragment *parse_binary_hunk(char **buf_p,
unsigned long *sz_p,
int *status_p,
int *used_p)
{
/* Expect a line that begins with binary patch method ("literal"
* or "delta"), followed by the length of data before deflating.
* a sequence of 'length-byte' followed by base-85 encoded data
* should follow, terminated by a newline.
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
*
* Each 5-byte sequence of base-85 encodes up to 4 bytes,
* and we would limit the patch line to 66 characters,
* so one line can fit up to 13 groups that would decode
* to 52 bytes max. The length byte 'A'-'Z' corresponds
* to 1-26 bytes, and 'a'-'z' corresponds to 27-52 bytes.
*/
int llen, used;
unsigned long size = *sz_p;
char *buffer = *buf_p;
int patch_method;
unsigned long origlen;
char *data = NULL;
int hunk_size = 0;
struct fragment *frag;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
llen = linelen(buffer, size);
used = llen;
*status_p = 0;
if (!prefixcmp(buffer, "delta ")) {
patch_method = BINARY_DELTA_DEFLATED;
origlen = strtoul(buffer + 6, NULL, 10);
}
else if (!prefixcmp(buffer, "literal ")) {
patch_method = BINARY_LITERAL_DEFLATED;
origlen = strtoul(buffer + 8, NULL, 10);
}
else
return NULL;
linenr++;
buffer += llen;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
while (1) {
int byte_length, max_byte_length, newsize;
llen = linelen(buffer, size);
used += llen;
linenr++;
if (llen == 1) {
/* consume the blank line */
buffer++;
size--;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
break;
}
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
/* Minimum line is "A00000\n" which is 7-byte long,
* and the line length must be multiple of 5 plus 2.
*/
if ((llen < 7) || (llen-2) % 5)
goto corrupt;
max_byte_length = (llen - 2) / 5 * 4;
byte_length = *buffer;
if ('A' <= byte_length && byte_length <= 'Z')
byte_length = byte_length - 'A' + 1;
else if ('a' <= byte_length && byte_length <= 'z')
byte_length = byte_length - 'a' + 27;
else
goto corrupt;
/* if the input length was not multiple of 4, we would
* have filler at the end but the filler should never
* exceed 3 bytes
*/
if (max_byte_length < byte_length ||
byte_length <= max_byte_length - 4)
goto corrupt;
newsize = hunk_size + byte_length;
data = xrealloc(data, newsize);
if (decode_85(data + hunk_size, buffer + 1, byte_length))
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
goto corrupt;
hunk_size = newsize;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
buffer += llen;
size -= llen;
}
frag = xcalloc(1, sizeof(*frag));
frag->patch = inflate_it(data, hunk_size, origlen);
if (!frag->patch)
goto corrupt;
free(data);
frag->size = origlen;
*buf_p = buffer;
*sz_p = size;
*used_p = used;
frag->binary_patch_method = patch_method;
return frag;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
corrupt:
free(data);
*status_p = -1;
error("corrupt binary patch at line %d: %.*s",
linenr-1, llen-1, buffer);
return NULL;
}
static int parse_binary(char *buffer, unsigned long size, struct patch *patch)
{
/* We have read "GIT binary patch\n"; what follows is a line
* that says the patch method (currently, either "literal" or
* "delta") and the length of data before deflating; a
* sequence of 'length-byte' followed by base-85 encoded data
* follows.
*
* When a binary patch is reversible, there is another binary
* hunk in the same format, starting with patch method (either
* "literal" or "delta") with the length of data, and a sequence
* of length-byte + base-85 encoded data, terminated with another
* empty line. This data, when applied to the postimage, produces
* the preimage.
*/
struct fragment *forward;
struct fragment *reverse;
int status;
int used, used_1;
forward = parse_binary_hunk(&buffer, &size, &status, &used);
if (!forward && !status)
/* there has to be one hunk (forward hunk) */
return error("unrecognized binary patch at line %d", linenr-1);
if (status)
/* otherwise we already gave an error message */
return status;
reverse = parse_binary_hunk(&buffer, &size, &status, &used_1);
if (reverse)
used += used_1;
else if (status) {
/* not having reverse hunk is not an error, but having
* a corrupt reverse hunk is.
*/
free((void*) forward->patch);
free(forward);
return status;
}
forward->next = reverse;
patch->fragments = forward;
patch->is_binary = 1;
return used;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
}
static int parse_chunk(char *buffer, unsigned long size, struct patch *patch)
{
int hdrsize, patchsize;
int offset = find_header(buffer, size, &hdrsize, patch);
if (offset < 0)
return offset;
patchsize = parse_single_patch(buffer + offset + hdrsize, size - offset - hdrsize, patch);
if (!patchsize) {
static const char *binhdr[] = {
"Binary files ",
"Files ",
NULL,
};
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
static const char git_binary[] = "GIT binary patch\n";
int i;
int hd = hdrsize + offset;
unsigned long llen = linelen(buffer + hd, size - hd);
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
if (llen == sizeof(git_binary) - 1 &&
!memcmp(git_binary, buffer + hd, llen)) {
int used;
linenr++;
used = parse_binary(buffer + hd + llen,
size - hd - llen, patch);
if (used)
patchsize = used + llen;
else
patchsize = 0;
}
else if (!memcmp(" differ\n", buffer + hd + llen - 8, 8)) {
for (i = 0; binhdr[i]; i++) {
int len = strlen(binhdr[i]);
if (len < size - hd &&
!memcmp(binhdr[i], buffer + hd, len)) {
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
linenr++;
patch->is_binary = 1;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
patchsize = llen;
break;
}
}
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
}
/* Empty patch cannot be applied if it is a text patch
* without metadata change. A binary patch appears
* empty to us here.
*/
if ((apply || check) &&
(!patch->is_binary && !metadata_changes(patch)))
die("patch with only garbage at line %d", linenr);
}
return offset + hdrsize + patchsize;
}
#define swap(a,b) myswap((a),(b),sizeof(a))
#define myswap(a, b, size) do { \
unsigned char mytmp[size]; \
memcpy(mytmp, &a, size); \
memcpy(&a, &b, size); \
memcpy(&b, mytmp, size); \
} while (0)
static void reverse_patches(struct patch *p)
{
for (; p; p = p->next) {
struct fragment *frag = p->fragments;
swap(p->new_name, p->old_name);
swap(p->new_mode, p->old_mode);
swap(p->is_new, p->is_delete);
swap(p->lines_added, p->lines_deleted);
swap(p->old_sha1_prefix, p->new_sha1_prefix);
for (; frag; frag = frag->next) {
swap(frag->newpos, frag->oldpos);
swap(frag->newlines, frag->oldlines);
}
}
}
static const char pluses[] = "++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++";
static const char minuses[]= "----------------------------------------------------------------------";
static void show_stats(struct patch *patch)
{
struct strbuf qname;
char *cp = patch->new_name ? patch->new_name : patch->old_name;
int max, add, del;
strbuf_init(&qname, 0);
quote_c_style(cp, &qname, NULL, 0);
/*
* "scale" the filename
*/
max = max_len;
if (max > 50)
max = 50;
if (qname.len > max) {
cp = strchr(qname.buf + qname.len + 3 - max, '/');
if (!cp)
cp = qname.buf + qname.len + 3 - max;
strbuf_splice(&qname, 0, cp - qname.buf, "...", 3);
}
if (patch->is_binary) {
printf(" %-*s | Bin\n", max, qname.buf);
strbuf_release(&qname);
return;
}
printf(" %-*s |", max, qname.buf);
strbuf_release(&qname);
/*
* scale the add/delete
*/
max = max + max_change > 70 ? 70 - max : max_change;
add = patch->lines_added;
del = patch->lines_deleted;
if (max_change > 0) {
int total = ((add + del) * max + max_change / 2) / max_change;
add = (add * max + max_change / 2) / max_change;
del = total - add;
}
printf("%5d %.*s%.*s\n", patch->lines_added + patch->lines_deleted,
add, pluses, del, minuses);
}
static int read_old_data(struct stat *st, const char *path, struct strbuf *buf)
{
switch (st->st_mode & S_IFMT) {
case S_IFLNK:
strbuf_grow(buf, st->st_size);
if (readlink(path, buf->buf, st->st_size) != st->st_size)
return -1;
strbuf_setlen(buf, st->st_size);
return 0;
case S_IFREG:
if (strbuf_read_file(buf, path, st->st_size) != st->st_size)
return error("unable to open or read %s", path);
convert_to_git(path, buf->buf, buf->len, buf);
return 0;
default:
return -1;
}
}
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
static int find_offset(const char *buf, unsigned long size, const char *fragment, unsigned long fragsize, int line, int *lines)
{
int i;
unsigned long start, backwards, forwards;
if (fragsize > size)
return -1;
start = 0;
if (line > 1) {
unsigned long offset = 0;
i = line-1;
while (offset + fragsize <= size) {
if (buf[offset++] == '\n') {
start = offset;
if (!--i)
break;
}
}
}
/* Exact line number? */
if ((start + fragsize <= size) &&
!memcmp(buf + start, fragment, fragsize))
return start;
/*
* There's probably some smart way to do this, but I'll leave
* that to the smart and beautiful people. I'm simple and stupid.
*/
backwards = start;
forwards = start;
for (i = 0; ; i++) {
unsigned long try;
int n;
/* "backward" */
if (i & 1) {
if (!backwards) {
if (forwards + fragsize > size)
break;
continue;
}
do {
--backwards;
} while (backwards && buf[backwards-1] != '\n');
try = backwards;
} else {
while (forwards + fragsize <= size) {
if (buf[forwards++] == '\n')
break;
}
try = forwards;
}
if (try + fragsize > size)
continue;
if (memcmp(buf + try, fragment, fragsize))
continue;
n = (i >> 1)+1;
if (i & 1)
n = -n;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
*lines = n;
return try;
}
/*
* We should start searching forward and backward.
*/
return -1;
}
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
static void remove_first_line(const char **rbuf, int *rsize)
{
const char *buf = *rbuf;
int size = *rsize;
unsigned long offset;
offset = 0;
while (offset <= size) {
if (buf[offset++] == '\n')
break;
}
*rsize = size - offset;
*rbuf = buf + offset;
}
static void remove_last_line(const char **rbuf, int *rsize)
{
const char *buf = *rbuf;
int size = *rsize;
unsigned long offset;
offset = size - 1;
while (offset > 0) {
if (buf[--offset] == '\n')
break;
}
*rsize = offset + 1;
}
static int apply_line(char *output, const char *patch, int plen)
{
/* plen is number of bytes to be copied from patch,
* starting at patch+1 (patch[0] is '+'). Typically
* patch[plen] is '\n', unless this is the incomplete
* last line.
*/
int i;
int add_nl_to_tail = 0;
int fixed = 0;
int last_tab_in_indent = -1;
int last_space_in_indent = -1;
int need_fix_leading_space = 0;
char *buf;
if ((new_whitespace != strip_whitespace) || !whitespace_error ||
*patch != '+') {
memcpy(output, patch + 1, plen);
return plen;
}
if (1 < plen && isspace(patch[plen-1])) {
if (patch[plen] == '\n')
add_nl_to_tail = 1;
plen--;
while (0 < plen && isspace(patch[plen]))
plen--;
fixed = 1;
}
for (i = 1; i < plen; i++) {
char ch = patch[i];
if (ch == '\t') {
last_tab_in_indent = i;
if (0 <= last_space_in_indent)
need_fix_leading_space = 1;
}
else if (ch == ' ')
last_space_in_indent = i;
else
break;
}
buf = output;
if (need_fix_leading_space) {
int consecutive_spaces = 0;
/* between patch[1..last_tab_in_indent] strip the
* funny spaces, updating them to tab as needed.
*/
for (i = 1; i < last_tab_in_indent; i++, plen--) {
char ch = patch[i];
if (ch != ' ') {
consecutive_spaces = 0;
*output++ = ch;
} else {
consecutive_spaces++;
if (consecutive_spaces == 8) {
*output++ = '\t';
consecutive_spaces = 0;
}
}
}
fixed = 1;
i = last_tab_in_indent;
}
else
i = 1;
memcpy(output, patch + i, plen);
if (add_nl_to_tail)
output[plen++] = '\n';
if (fixed)
applied_after_fixing_ws++;
return output + plen - buf;
}
static int apply_one_fragment(struct strbuf *buf, struct fragment *frag, int inaccurate_eof)
{
int match_beginning, match_end;
const char *patch = frag->patch;
int offset, size = frag->size;
char *old = xmalloc(size);
char *new = xmalloc(size);
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
const char *oldlines, *newlines;
int oldsize = 0, newsize = 0;
int new_blank_lines_at_end = 0;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
unsigned long leading, trailing;
int pos, lines;
while (size > 0) {
char first;
int len = linelen(patch, size);
int plen;
int added_blank_line = 0;
if (!len)
break;
/*
* "plen" is how much of the line we should use for
* the actual patch data. Normally we just remove the
* first character on the line, but if the line is
* followed by "\ No newline", then we also remove the
* last one (which is the newline, of course).
*/
plen = len-1;
if (len < size && patch[len] == '\\')
plen--;
first = *patch;
if (apply_in_reverse) {
if (first == '-')
first = '+';
else if (first == '+')
first = '-';
}
switch (first) {
case '\n':
/* Newer GNU diff, empty context line */
if (plen < 0)
/* ... followed by '\No newline'; nothing */
break;
old[oldsize++] = '\n';
new[newsize++] = '\n';
break;
case ' ':
case '-':
memcpy(old + oldsize, patch + 1, plen);
oldsize += plen;
if (first == '-')
break;
/* Fall-through for ' ' */
case '+':
if (first != '+' || !no_add) {
int added = apply_line(new + newsize, patch,
plen);
newsize += added;
if (first == '+' &&
added == 1 && new[newsize-1] == '\n')
added_blank_line = 1;
}
break;
case '@': case '\\':
/* Ignore it, we already handled it */
break;
default:
if (apply_verbosely)
error("invalid start of line: '%c'", first);
return -1;
}
if (added_blank_line)
new_blank_lines_at_end++;
else
new_blank_lines_at_end = 0;
patch += len;
size -= len;
}
if (inaccurate_eof && oldsize > 0 && old[oldsize - 1] == '\n' &&
newsize > 0 && new[newsize - 1] == '\n') {
oldsize--;
newsize--;
}
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
oldlines = old;
newlines = new;
leading = frag->leading;
trailing = frag->trailing;
/*
* If we don't have any leading/trailing data in the patch,
* we want it to match at the beginning/end of the file.
*
* But that would break if the patch is generated with
* --unified=0; sane people wouldn't do that to cause us
* trouble, but we try to please not so sane ones as well.
*/
if (unidiff_zero) {
match_beginning = (!leading && !frag->oldpos);
match_end = 0;
}
else {
match_beginning = !leading && (frag->oldpos == 1);
match_end = !trailing;
}
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
lines = 0;
pos = frag->newpos;
for (;;) {
offset = find_offset(buf->buf, buf->len,
oldlines, oldsize, pos, &lines);
if (match_end && offset + oldsize != buf->len)
offset = -1;
if (match_beginning && offset)
offset = -1;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
if (offset >= 0) {
if (new_whitespace == strip_whitespace &&
(buf->len - oldsize - offset == 0)) /* end of file? */
newsize -= new_blank_lines_at_end;
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
/* Warn if it was necessary to reduce the number
* of context lines.
*/
if ((leading != frag->leading) ||
(trailing != frag->trailing))
fprintf(stderr, "Context reduced to (%ld/%ld)"
" to apply fragment at %d\n",
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
leading, trailing, pos + lines);
strbuf_splice(buf, offset, oldsize, newlines, newsize);
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
offset = 0;
break;
}
/* Am I at my context limits? */
if ((leading <= p_context) && (trailing <= p_context))
break;
if (match_beginning || match_end) {
match_beginning = match_end = 0;
continue;
}
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
/* Reduce the number of context lines
* Reduce both leading and trailing if they are equal
* otherwise just reduce the larger context.
*/
if (leading >= trailing) {
remove_first_line(&oldlines, &oldsize);
remove_first_line(&newlines, &newsize);
pos--;
leading--;
}
if (trailing > leading) {
remove_last_line(&oldlines, &oldsize);
remove_last_line(&newlines, &newsize);
trailing--;
}
}
if (offset && apply_verbosely)
error("while searching for:\n%.*s", oldsize, oldlines);
free(old);
free(new);
return offset;
}
static int apply_binary_fragment(struct strbuf *buf, struct patch *patch)
{
struct fragment *fragment = patch->fragments;
unsigned long len;
void *dst;
/* Binary patch is irreversible without the optional second hunk */
if (apply_in_reverse) {
if (!fragment->next)
return error("cannot reverse-apply a binary patch "
"without the reverse hunk to '%s'",
patch->new_name
? patch->new_name : patch->old_name);
fragment = fragment->next;
}
switch (fragment->binary_patch_method) {
case BINARY_DELTA_DEFLATED:
dst = patch_delta(buf->buf, buf->len, fragment->patch,
fragment->size, &len);
if (!dst)
return -1;
/* XXX patch_delta NUL-terminates */
strbuf_attach(buf, dst, len, len + 1);
return 0;
case BINARY_LITERAL_DEFLATED:
strbuf_reset(buf);
strbuf_add(buf, fragment->patch, fragment->size);
return 0;
}
return -1;
}
static int apply_binary(struct strbuf *buf, struct patch *patch)
{
const char *name = patch->old_name ? patch->old_name : patch->new_name;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
unsigned char sha1[20];
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
/* For safety, we require patch index line to contain
* full 40-byte textual SHA1 for old and new, at least for now.
*/
if (strlen(patch->old_sha1_prefix) != 40 ||
strlen(patch->new_sha1_prefix) != 40 ||
get_sha1_hex(patch->old_sha1_prefix, sha1) ||
get_sha1_hex(patch->new_sha1_prefix, sha1))
return error("cannot apply binary patch to '%s' "
"without full index line", name);
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
if (patch->old_name) {
/* See if the old one matches what the patch
* applies to.
*/
hash_sha1_file(buf->buf, buf->len, blob_type, sha1);
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
if (strcmp(sha1_to_hex(sha1), patch->old_sha1_prefix))
return error("the patch applies to '%s' (%s), "
"which does not match the "
"current contents.",
name, sha1_to_hex(sha1));
}
else {
/* Otherwise, the old one must be empty. */
if (buf->len)
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
return error("the patch applies to an empty "
"'%s' but it is not empty", name);
}
get_sha1_hex(patch->new_sha1_prefix, sha1);
if (is_null_sha1(sha1)) {
strbuf_release(buf);
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
return 0; /* deletion patch */
}
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
if (has_sha1_file(sha1)) {
/* We already have the postimage */
enum object_type type;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
unsigned long size;
char *result;
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
result = read_sha1_file(sha1, &type, &size);
if (!result)
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
return error("the necessary postimage %s for "
"'%s' cannot be read",
patch->new_sha1_prefix, name);
/* XXX read_sha1_file NUL-terminates */
strbuf_attach(buf, result, size, size + 1);
} else {
/* We have verified buf matches the preimage;
* apply the patch data to it, which is stored
* in the patch->fragments->{patch,size}.
*/
if (apply_binary_fragment(buf, patch))
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
return error("binary patch does not apply to '%s'",
name);
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
/* verify that the result matches */
hash_sha1_file(buf->buf, buf->len, blob_type, sha1);
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
if (strcmp(sha1_to_hex(sha1), patch->new_sha1_prefix))
return error("binary patch to '%s' creates incorrect result (expecting %s, got %s)",
name, patch->new_sha1_prefix, sha1_to_hex(sha1));
}
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
return 0;
}
static int apply_fragments(struct strbuf *buf, struct patch *patch)
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
{
struct fragment *frag = patch->fragments;
const char *name = patch->old_name ? patch->old_name : patch->new_name;
if (patch->is_binary)
return apply_binary(buf, patch);
binary patch. This adds "binary patch" to the diff output and teaches apply what to do with them. On the diff generation side, traditionally, we said "Binary files differ\n" without giving anything other than the preimage and postimage object name on the index line. This was good enough for applying a patch generated from your own repository (very useful while rebasing), because the postimage would be available in such a case. However, this was not useful when the recipient of such a patch via e-mail were to apply it, even if the preimage was available. This patch allows the diff to generate "binary" patch when operating under --full-index option. The binary patch follows the usual extended git diff headers, and looks like this: "GIT binary patch\n" <length byte><data>"\n" ... "\n" Each line is prefixed with a "length-byte", whose value is upper or lowercase alphabet that encodes number of bytes that the data on the line decodes to (1..52 -- 'A' means 1, 'B' means 2, ..., 'Z' means 26, 'a' means 27, ...). <data> is 1 or more groups of 5-byte sequence, each of which encodes up to 4 bytes in base85 encoding. Because 52 / 4 * 5 = 65 and we have the length byte, an output line is capped to 66 characters. The payload is the same diff-delta as we use in the packfiles. On the consumption side, git-apply now can decode and apply the binary patch when --allow-binary-replacement is given, the diff was generated with --full-index, and the receiving repository has the preimage blob, which is the same condition as it always required when accepting an "Binary files differ\n" patch. Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-05-05 07:51:44 +08:00
while (frag) {
if (apply_one_fragment(buf, frag, patch->inaccurate_eof)) {
error("patch failed: %s:%ld", name, frag->oldpos);
if (!apply_with_reject)
return -1;
frag->rejected = 1;
}
frag = frag->next;
}
return 0;
}
static int read_file_or_gitlink(struct cache_entry *ce, struct strbuf *buf)
{
if (!ce)
return 0;
if (S_ISGITLINK(ntohl(ce->ce_mode))) {
strbuf_grow(buf, 100);
strbuf_addf(buf, "Subproject commit %s\n", sha1_to_hex(ce->sha1));
} else {
enum object_type type;
unsigned long sz;
char *result;
result = read_sha1_file(ce->sha1, &type, &sz);
if (!result)
return -1;
/* XXX read_sha1_file NUL-terminates */
strbuf_attach(buf, result, sz, sz + 1);
}
return 0;
}
static int apply_data(struct patch *patch, struct stat *st, struct cache_entry *ce)
{
struct strbuf buf;
strbuf_init(&buf, 0);
if (cached) {
if (read_file_or_gitlink(ce, &buf))
return error("read of %s failed", patch->old_name);
} else if (patch->old_name) {
if (S_ISGITLINK(patch->old_mode)) {
if (ce) {
read_file_or_gitlink(ce, &buf);
} else {
/*
* There is no way to apply subproject
* patch without looking at the index.
*/
patch->fragments = NULL;
}
} else {
if (read_old_data(st, patch->old_name, &buf))
return error("read of %s failed", patch->old_name);
}
}
if (apply_fragments(&buf, patch) < 0)
return -1; /* note with --reject this succeeds. */
patch->result = strbuf_detach(&buf, &patch->resultsize);
if (0 < patch->is_delete && patch->resultsize)
return error("removal patch leaves file contents");
return 0;
}
apply: do not get confused by symlinks in the middle HPA noticed that git-rebase fails when changes involve symlinks in the middle of the hierarchy. Consider: * The tree state before the patch is applied has arch/x86_64/boot as a symlink pointing at ../i386/boot/ * The patch tries to remove arch/x86_64/boot symlink, and create bunch of files there: .gitignore, Makefile, etc. git-apply tries to be careful while applying patches; it never touches the working tree until it is convinced that the patch would apply cleanly. One of the check it does is that when it knows a path is going to be created by the patch, it runs lstat() on the path to make sure it does not exist. This leads to a false alarm. Because we do not touch the working tree before all the check passes, when we try to make sure that arch/x86_64/boot/.gitignore does not exist yet, we haven't removed the arch/x86_64/boot symlink. The lstat() check ends up seeing arch/i386/boot/.gitignore through the yet-to-be-removed symlink, and says "Hey, you already have a file there, but what you fed me is a patch to create a new file. I am not going to clobber what you have in the working tree." We have similar checks to see a file we are going to modify does exist and match the preimage of the diff, which is done by directly opening and reading the file. For a file we are going to delete, we make sure that it does exist and matches what is going to be removed (a removal patch records the full preimage, so we check what you have in your working tree matches it in full -- otherwise we would risk losing your local changes), which again is done by directly opening and reading the file. These checks need to be adjusted so that they are not fooled by symlinks in the middle. - To make sure something does not exist, first lstat(). If it does not exist, it does not, so be happy. If it _does_, we might be getting fooled by a symlink in the middle, so break leading paths and see if there are symlinks involved. When we are checking for a path a/b/c/d, if any of a, a/b, a/b/c is a symlink, then a/b/c/d does _NOT_ exist, for the purpose of our test. This would fix this particular case you saw, and would not add extra overhead in the usual case. - To make sure something already exists, first lstat(). If it does not exist, barf (up to this, we already do). Even if it does seem to exist, we might be getting fooled by a symlink in the middle, so make sure leading paths are not symlinks. This would make the normal codepath much more expensive for deep trees, which is a bit worrisome. This patch implements the first side of the check "making sure it does not exist". The latter "making sure it exists" check is not done yet, so applying the patch in reverse would still fail, but we have to start from somewhere. Signed-off-by: Junio C Hamano <junkio@cox.net>
2007-05-12 13:26:08 +08:00
static int check_to_create_blob(const char *new_name, int ok_if_exists)
{
struct stat nst;
if (!lstat(new_name, &nst)) {
if (S_ISDIR(nst.st_mode) || ok_if_exists)
return 0;
/*
* A leading component of new_name might be a symlink
* that is going to be removed with this patch, but
* still pointing at somewhere that has the path.
* In such a case, path "new_name" does not exist as
* far as git is concerned.
*/
if (has_symlink_leading_path(new_name, NULL))
return 0;
return error("%s: already exists in working directory", new_name);
}
else if ((errno != ENOENT) && (errno != ENOTDIR))
return error("%s: %s", new_name, strerror(errno));
return 0;
}
static int verify_index_match(struct cache_entry *ce, struct stat *st)
{
if (S_ISGITLINK(ntohl(ce->ce_mode))) {
if (!S_ISDIR(st->st_mode))
return -1;
return 0;
}
return ce_match_stat(ce, st, 1);
}
static int check_patch(struct patch *patch, struct patch *prev_patch)
{
struct stat st;
const char *old_name = patch->old_name;
const char *new_name = patch->new_name;
const char *name = old_name ? old_name : new_name;
struct cache_entry *ce = NULL;
int ok_if_exists;
patch->rejected = 1; /* we will drop this after we succeed */
/*
* Make sure that we do not have local modifications from the
* index when we are looking at the index. Also make sure
* we have the preimage file to be patched in the work tree,
* unless --cached, which tells git to apply only in the index.
*/
if (old_name) {
int stat_ret = 0;
unsigned st_mode = 0;
if (!cached)
stat_ret = lstat(old_name, &st);
if (check_index) {
int pos = cache_name_pos(old_name, strlen(old_name));
if (pos < 0)
return error("%s: does not exist in index",
old_name);
ce = active_cache[pos];
if (stat_ret < 0) {
struct checkout costate;
if (errno != ENOENT)
return error("%s: %s", old_name,
strerror(errno));
/* checkout */
costate.base_dir = "";
costate.base_dir_len = 0;
costate.force = 0;
costate.quiet = 0;
costate.not_new = 0;
costate.refresh_cache = 1;
if (checkout_entry(ce,
&costate,
NULL) ||
lstat(old_name, &st))
return -1;
}
if (!cached && verify_index_match(ce, &st))
return error("%s: does not match index",
old_name);
if (cached)
st_mode = ntohl(ce->ce_mode);
} else if (stat_ret < 0)
return error("%s: %s", old_name, strerror(errno));
if (!cached)
st_mode = ntohl(ce_mode_from_stat(ce, st.st_mode));
if (patch->is_new < 0)
patch->is_new = 0;
if (!patch->old_mode)
patch->old_mode = st_mode;
if ((st_mode ^ patch->old_mode) & S_IFMT)
return error("%s: wrong type", old_name);
if (st_mode != patch->old_mode)
fprintf(stderr, "warning: %s has type %o, expected %o\n",
old_name, st_mode, patch->old_mode);
}
if (new_name && prev_patch && 0 < prev_patch->is_delete &&
!strcmp(prev_patch->old_name, new_name))
/* A type-change diff is always split into a patch to
* delete old, immediately followed by a patch to
* create new (see diff.c::run_diff()); in such a case
* it is Ok that the entry to be deleted by the
* previous patch is still in the working tree and in
* the index.
*/
ok_if_exists = 1;
else
ok_if_exists = 0;
if (new_name &&
((0 < patch->is_new) | (0 < patch->is_rename) | patch->is_copy)) {
if (check_index &&
cache_name_pos(new_name, strlen(new_name)) >= 0 &&
!ok_if_exists)
return error("%s: already exists in index", new_name);
if (!cached) {
apply: do not get confused by symlinks in the middle HPA noticed that git-rebase fails when changes involve symlinks in the middle of the hierarchy. Consider: * The tree state before the patch is applied has arch/x86_64/boot as a symlink pointing at ../i386/boot/ * The patch tries to remove arch/x86_64/boot symlink, and create bunch of files there: .gitignore, Makefile, etc. git-apply tries to be careful while applying patches; it never touches the working tree until it is convinced that the patch would apply cleanly. One of the check it does is that when it knows a path is going to be created by the patch, it runs lstat() on the path to make sure it does not exist. This leads to a false alarm. Because we do not touch the working tree before all the check passes, when we try to make sure that arch/x86_64/boot/.gitignore does not exist yet, we haven't removed the arch/x86_64/boot symlink. The lstat() check ends up seeing arch/i386/boot/.gitignore through the yet-to-be-removed symlink, and says "Hey, you already have a file there, but what you fed me is a patch to create a new file. I am not going to clobber what you have in the working tree." We have similar checks to see a file we are going to modify does exist and match the preimage of the diff, which is done by directly opening and reading the file. For a file we are going to delete, we make sure that it does exist and matches what is going to be removed (a removal patch records the full preimage, so we check what you have in your working tree matches it in full -- otherwise we would risk losing your local changes), which again is done by directly opening and reading the file. These checks need to be adjusted so that they are not fooled by symlinks in the middle. - To make sure something does not exist, first lstat(). If it does not exist, it does not, so be happy. If it _does_, we might be getting fooled by a symlink in the middle, so break leading paths and see if there are symlinks involved. When we are checking for a path a/b/c/d, if any of a, a/b, a/b/c is a symlink, then a/b/c/d does _NOT_ exist, for the purpose of our test. This would fix this particular case you saw, and would not add extra overhead in the usual case. - To make sure something already exists, first lstat(). If it does not exist, barf (up to this, we already do). Even if it does seem to exist, we might be getting fooled by a symlink in the middle, so make sure leading paths are not symlinks. This would make the normal codepath much more expensive for deep trees, which is a bit worrisome. This patch implements the first side of the check "making sure it does not exist". The latter "making sure it exists" check is not done yet, so applying the patch in reverse would still fail, but we have to start from somewhere. Signed-off-by: Junio C Hamano <junkio@cox.net>
2007-05-12 13:26:08 +08:00
int err = check_to_create_blob(new_name, ok_if_exists);
if (err)
return err;
}
if (!patch->new_mode) {
if (0 < patch->is_new)
patch->new_mode = S_IFREG | 0644;
else
patch->new_mode = patch->old_mode;
}
}
if (new_name && old_name) {
int same = !strcmp(old_name, new_name);
if (!patch->new_mode)
patch->new_mode = patch->old_mode;
if ((patch->old_mode ^ patch->new_mode) & S_IFMT)
return error("new mode (%o) of %s does not match old mode (%o)%s%s",
patch->new_mode, new_name, patch->old_mode,
same ? "" : " of ", same ? "" : old_name);
}
if (apply_data(patch, &st, ce) < 0)
return error("%s: patch does not apply", name);
patch->rejected = 0;
return 0;
}
static int check_patch_list(struct patch *patch)
{
struct patch *prev_patch = NULL;
int err = 0;
for (prev_patch = NULL; patch ; patch = patch->next) {
if (apply_verbosely)
say_patch_name(stderr,
"Checking patch ", patch, "...\n");
err |= check_patch(patch, prev_patch);
prev_patch = patch;
}
return err;
}
/* This function tries to read the sha1 from the current index */
static int get_current_sha1(const char *path, unsigned char *sha1)
{
int pos;
if (read_cache() < 0)
return -1;
pos = cache_name_pos(path, strlen(path));
if (pos < 0)
return -1;
hashcpy(sha1, active_cache[pos]->sha1);
return 0;
}
/* Build an index that contains the just the files needed for a 3way merge */
static void build_fake_ancestor(struct patch *list, const char *filename)
{
struct patch *patch;
struct index_state result = { 0 };
int fd;
/* Once we start supporting the reverse patch, it may be
* worth showing the new sha1 prefix, but until then...
*/
for (patch = list; patch; patch = patch->next) {
const unsigned char *sha1_ptr;
unsigned char sha1[20];
struct cache_entry *ce;
const char *name;
name = patch->old_name ? patch->old_name : patch->new_name;
if (0 < patch->is_new)
continue;
else if (get_sha1(patch->old_sha1_prefix, sha1))
/* git diff has no index line for mode/type changes */
if (!patch->lines_added && !patch->lines_deleted) {
if (get_current_sha1(patch->new_name, sha1) ||
get_current_sha1(patch->old_name, sha1))
die("mode change for %s, which is not "
"in current HEAD", name);
sha1_ptr = sha1;
} else
die("sha1 information is lacking or useless "
"(%s).", name);
else
sha1_ptr = sha1;
ce = make_cache_entry(patch->old_mode, sha1_ptr, name, 0, 0);
if (add_index_entry(&result, ce, ADD_CACHE_OK_TO_ADD))
die ("Could not add %s to temporary index", name);
}
fd = open(filename, O_WRONLY | O_CREAT, 0666);
if (fd < 0 || write_index(&result, fd) || close(fd))
die ("Could not write temporary index to %s", filename);
discard_index(&result);
}
static void stat_patch_list(struct patch *patch)
{
int files, adds, dels;
for (files = adds = dels = 0 ; patch ; patch = patch->next) {
files++;
adds += patch->lines_added;
dels += patch->lines_deleted;
show_stats(patch);
}
printf(" %d files changed, %d insertions(+), %d deletions(-)\n", files, adds, dels);
}
static void numstat_patch_list(struct patch *patch)
{
for ( ; patch; patch = patch->next) {
const char *name;
name = patch->new_name ? patch->new_name : patch->old_name;
if (patch->is_binary)
printf("-\t-\t");
else
printf("%d\t%d\t", patch->lines_added, patch->lines_deleted);
write_name_quoted(name, stdout, line_termination);
}
}
static void show_file_mode_name(const char *newdelete, unsigned int mode, const char *name)
{
if (mode)
printf(" %s mode %06o %s\n", newdelete, mode, name);
else
printf(" %s %s\n", newdelete, name);
}
static void show_mode_change(struct patch *p, int show_name)
{
if (p->old_mode && p->new_mode && p->old_mode != p->new_mode) {
if (show_name)
printf(" mode change %06o => %06o %s\n",
p->old_mode, p->new_mode, p->new_name);
else
printf(" mode change %06o => %06o\n",
p->old_mode, p->new_mode);
}
}
static void show_rename_copy(struct patch *p)
{
const char *renamecopy = p->is_rename ? "rename" : "copy";
const char *old, *new;
/* Find common prefix */
old = p->old_name;
new = p->new_name;
while (1) {
const char *slash_old, *slash_new;
slash_old = strchr(old, '/');
slash_new = strchr(new, '/');
if (!slash_old ||
!slash_new ||
slash_old - old != slash_new - new ||
memcmp(old, new, slash_new - new))
break;
old = slash_old + 1;
new = slash_new + 1;
}
/* p->old_name thru old is the common prefix, and old and new
* through the end of names are renames
*/
if (old != p->old_name)
printf(" %s %.*s{%s => %s} (%d%%)\n", renamecopy,
(int)(old - p->old_name), p->old_name,
old, new, p->score);
else
printf(" %s %s => %s (%d%%)\n", renamecopy,
p->old_name, p->new_name, p->score);
show_mode_change(p, 0);
}
static void summary_patch_list(struct patch *patch)
{
struct patch *p;
for (p = patch; p; p = p->next) {
if (p->is_new)
show_file_mode_name("create", p->new_mode, p->new_name);
else if (p->is_delete)
show_file_mode_name("delete", p->old_mode, p->old_name);
else {
if (p->is_rename || p->is_copy)
show_rename_copy(p);
else {
if (p->score) {
printf(" rewrite %s (%d%%)\n",
p->new_name, p->score);
show_mode_change(p, 0);
}
else
show_mode_change(p, 1);
}
}
}
}
static void patch_stats(struct patch *patch)
{
int lines = patch->lines_added + patch->lines_deleted;
if (lines > max_change)
max_change = lines;
if (patch->old_name) {
int len = quote_c_style(patch->old_name, NULL, NULL, 0);
if (!len)
len = strlen(patch->old_name);
if (len > max_len)
max_len = len;
}
if (patch->new_name) {
int len = quote_c_style(patch->new_name, NULL, NULL, 0);
if (!len)
len = strlen(patch->new_name);
if (len > max_len)
max_len = len;
}
}
static void remove_file(struct patch *patch, int rmdir_empty)
{
if (update_index) {
if (remove_file_from_cache(patch->old_name) < 0)
die("unable to remove %s from index", patch->old_name);
}
if (!cached) {
if (S_ISGITLINK(patch->old_mode)) {
if (rmdir(patch->old_name))
warning("unable to remove submodule %s",
patch->old_name);
} else if (!unlink(patch->old_name) && rmdir_empty) {
char *name = xstrdup(patch->old_name);
char *end = strrchr(name, '/');
while (end) {
*end = 0;
if (rmdir(name))
break;
end = strrchr(name, '/');
}
free(name);
}
}
}
static void add_index_file(const char *path, unsigned mode, void *buf, unsigned long size)
{
struct stat st;
struct cache_entry *ce;
int namelen = strlen(path);
unsigned ce_size = cache_entry_size(namelen);
if (!update_index)
return;
ce = xcalloc(1, ce_size);
memcpy(ce->name, path, namelen);
ce->ce_mode = create_ce_mode(mode);
ce->ce_flags = htons(namelen);
if (S_ISGITLINK(mode)) {
const char *s = buf;
if (get_sha1_hex(s + strlen("Subproject commit "), ce->sha1))
die("corrupt patch for subproject %s", path);
} else {
if (!cached) {
if (lstat(path, &st) < 0)
die("unable to stat newly created file %s",
path);
fill_stat_cache_info(ce, &st);
}
if (write_sha1_file(buf, size, blob_type, ce->sha1) < 0)
die("unable to create backing store for newly created file %s", path);
}
if (add_cache_entry(ce, ADD_CACHE_OK_TO_ADD) < 0)
die("unable to add cache entry for %s", path);
}
static int try_create_file(const char *path, unsigned int mode, const char *buf, unsigned long size)
{
int fd;
struct strbuf nbuf;
if (S_ISGITLINK(mode)) {
struct stat st;
if (!lstat(path, &st) && S_ISDIR(st.st_mode))
return 0;
return mkdir(path, 0777);
}
if (has_symlinks && S_ISLNK(mode))
/* Although buf:size is counted string, it also is NUL
* terminated.
*/
return symlink(buf, path);
fd = open(path, O_CREAT | O_EXCL | O_WRONLY, (mode & 0100) ? 0777 : 0666);
if (fd < 0)
return -1;
strbuf_init(&nbuf, 0);
if (convert_to_working_tree(path, buf, size, &nbuf)) {
size = nbuf.len;
buf = nbuf.buf;
}
write_or_die(fd, buf, size);
strbuf_release(&nbuf);
if (close(fd) < 0)
die("closing file %s: %s", path, strerror(errno));
return 0;
}
/*
* We optimistically assume that the directories exist,
* which is true 99% of the time anyway. If they don't,
* we create them and try again.
*/
static void create_one_file(char *path, unsigned mode, const char *buf, unsigned long size)
{
if (cached)
return;
if (!try_create_file(path, mode, buf, size))
return;
if (errno == ENOENT) {
if (safe_create_leading_directories(path))
return;
if (!try_create_file(path, mode, buf, size))
return;
}
if (errno == EEXIST || errno == EACCES) {
/* We may be trying to create a file where a directory
* used to be.
*/
struct stat st;
if (!lstat(path, &st) && (!S_ISDIR(st.st_mode) || !rmdir(path)))
errno = EEXIST;
}
if (errno == EEXIST) {
unsigned int nr = getpid();
for (;;) {
const char *newpath;
newpath = mkpath("%s~%u", path, nr);
if (!try_create_file(newpath, mode, buf, size)) {
if (!rename(newpath, path))
return;
unlink(newpath);
break;
}
if (errno != EEXIST)
break;
++nr;
}
}
die("unable to write file %s mode %o", path, mode);
}
static void create_file(struct patch *patch)
{
char *path = patch->new_name;
unsigned mode = patch->new_mode;
unsigned long size = patch->resultsize;
char *buf = patch->result;
if (!mode)
mode = S_IFREG | 0644;
create_one_file(path, mode, buf, size);
add_index_file(path, mode, buf, size);
}
/* phase zero is to remove, phase one is to create */
static void write_out_one_result(struct patch *patch, int phase)
{
if (patch->is_delete > 0) {
if (phase == 0)
remove_file(patch, 1);
return;
}
if (patch->is_new > 0 || patch->is_copy) {
if (phase == 1)
create_file(patch);
return;
}
/*
* Rename or modification boils down to the same
* thing: remove the old, write the new
*/
if (phase == 0)
remove_file(patch, patch->is_rename);
if (phase == 1)
create_file(patch);
}
static int write_out_one_reject(struct patch *patch)
{
FILE *rej;
char namebuf[PATH_MAX];
struct fragment *frag;
int cnt = 0;
for (cnt = 0, frag = patch->fragments; frag; frag = frag->next) {
if (!frag->rejected)
continue;
cnt++;
}
if (!cnt) {
if (apply_verbosely)
say_patch_name(stderr,
"Applied patch ", patch, " cleanly.\n");
return 0;
}
/* This should not happen, because a removal patch that leaves
* contents are marked "rejected" at the patch level.
*/
if (!patch->new_name)
die("internal error");
/* Say this even without --verbose */
say_patch_name(stderr, "Applying patch ", patch, " with");
fprintf(stderr, " %d rejects...\n", cnt);
cnt = strlen(patch->new_name);
if (ARRAY_SIZE(namebuf) <= cnt + 5) {
cnt = ARRAY_SIZE(namebuf) - 5;
fprintf(stderr,
"warning: truncating .rej filename to %.*s.rej",
cnt - 1, patch->new_name);
}
memcpy(namebuf, patch->new_name, cnt);
memcpy(namebuf + cnt, ".rej", 5);
rej = fopen(namebuf, "w");
if (!rej)
return error("cannot open %s: %s", namebuf, strerror(errno));
/* Normal git tools never deal with .rej, so do not pretend
* this is a git patch by saying --git nor give extended
* headers. While at it, maybe please "kompare" that wants
* the trailing TAB and some garbage at the end of line ;-).
*/
fprintf(rej, "diff a/%s b/%s\t(rejected hunks)\n",
patch->new_name, patch->new_name);
for (cnt = 1, frag = patch->fragments;
frag;
cnt++, frag = frag->next) {
if (!frag->rejected) {
fprintf(stderr, "Hunk #%d applied cleanly.\n", cnt);
continue;
}
fprintf(stderr, "Rejected hunk #%d.\n", cnt);
fprintf(rej, "%.*s", frag->size, frag->patch);
if (frag->patch[frag->size-1] != '\n')
fputc('\n', rej);
}
fclose(rej);
return -1;
}
static int write_out_results(struct patch *list, int skipped_patch)
{
int phase;
int errs = 0;
struct patch *l;
if (!list && !skipped_patch)
return error("No changes");
for (phase = 0; phase < 2; phase++) {
l = list;
while (l) {
if (l->rejected)
errs = 1;
else {
write_out_one_result(l, phase);
if (phase == 1 && write_out_one_reject(l))
errs = 1;
}
l = l->next;
}
}
return errs;
}
static struct lock_file lock_file;
static struct excludes {
struct excludes *next;
const char *path;
} *excludes;
static int use_patch(struct patch *p)
{
const char *pathname = p->new_name ? p->new_name : p->old_name;
struct excludes *x = excludes;
while (x) {
if (fnmatch(x->path, pathname, 0) == 0)
return 0;
x = x->next;
}
if (0 < prefix_length) {
int pathlen = strlen(pathname);
if (pathlen <= prefix_length ||
memcmp(prefix, pathname, prefix_length))
return 0;
}
return 1;
}
static void prefix_one(char **name)
{
char *old_name = *name;
if (!old_name)
return;
*name = xstrdup(prefix_filename(prefix, prefix_length, *name));
free(old_name);
}
static void prefix_patches(struct patch *p)
{
if (!prefix || p->is_toplevel_relative)
return;
for ( ; p; p = p->next) {
if (p->new_name == p->old_name) {
char *prefixed = p->new_name;
prefix_one(&prefixed);
p->new_name = p->old_name = prefixed;
}
else {
prefix_one(&p->new_name);
prefix_one(&p->old_name);
}
}
}
static int apply_patch(int fd, const char *filename, int inaccurate_eof)
{
size_t offset;
struct strbuf buf;
struct patch *list = NULL, **listp = &list;
int skipped_patch = 0;
strbuf_init(&buf, 0);
patch_input_file = filename;
read_patch_file(&buf, fd);
offset = 0;
while (offset < buf.len) {
struct patch *patch;
int nr;
patch = xcalloc(1, sizeof(*patch));
patch->inaccurate_eof = inaccurate_eof;
nr = parse_chunk(buf.buf + offset, buf.len - offset, patch);
if (nr < 0)
break;
if (apply_in_reverse)
reverse_patches(patch);
if (prefix)
prefix_patches(patch);
if (use_patch(patch)) {
patch_stats(patch);
*listp = patch;
listp = &patch->next;
}
else {
/* perhaps free it a bit better? */
free(patch);
skipped_patch++;
}
offset += nr;
}
if (whitespace_error && (new_whitespace == error_on_whitespace))
apply = 0;
update_index = check_index && apply;
if (update_index && newfd < 0)
newfd = hold_locked_index(&lock_file, 1);
if (check_index) {
if (read_cache() < 0)
die("unable to read index file");
}
if ((check || apply) &&
check_patch_list(list) < 0 &&
!apply_with_reject)
exit(1);
if (apply && write_out_results(list, skipped_patch))
exit(1);
if (fake_ancestor)
build_fake_ancestor(list, fake_ancestor);
if (diffstat)
stat_patch_list(list);
if (numstat)
numstat_patch_list(list);
if (summary)
summary_patch_list(list);
strbuf_release(&buf);
return 0;
}
static int git_apply_config(const char *var, const char *value)
{
if (!strcmp(var, "apply.whitespace")) {
apply_default_whitespace = xstrdup(value);
return 0;
}
return git_default_config(var, value);
}
int cmd_apply(int argc, const char **argv, const char *unused_prefix)
{
int i;
int read_stdin = 1;
int inaccurate_eof = 0;
int errs = 0;
int is_not_gitdir = 0;
const char *whitespace_option = NULL;
prefix = setup_git_directory_gently(&is_not_gitdir);
prefix_length = prefix ? strlen(prefix) : 0;
git_config(git_apply_config);
if (apply_default_whitespace)
parse_whitespace_option(apply_default_whitespace);
for (i = 1; i < argc; i++) {
const char *arg = argv[i];
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
char *end;
int fd;
if (!strcmp(arg, "-")) {
errs |= apply_patch(0, "<stdin>", inaccurate_eof);
read_stdin = 0;
continue;
}
if (!prefixcmp(arg, "--exclude=")) {
struct excludes *x = xmalloc(sizeof(*x));
x->path = arg + 10;
x->next = excludes;
excludes = x;
continue;
}
if (!prefixcmp(arg, "-p")) {
p_value = atoi(arg + 2);
p_value_known = 1;
continue;
}
if (!strcmp(arg, "--no-add")) {
no_add = 1;
continue;
}
if (!strcmp(arg, "--stat")) {
apply = 0;
diffstat = 1;
continue;
}
if (!strcmp(arg, "--allow-binary-replacement") ||
!strcmp(arg, "--binary")) {
continue; /* now no-op */
}
if (!strcmp(arg, "--numstat")) {
apply = 0;
numstat = 1;
continue;
}
if (!strcmp(arg, "--summary")) {
apply = 0;
summary = 1;
continue;
}
if (!strcmp(arg, "--check")) {
apply = 0;
check = 1;
continue;
}
if (!strcmp(arg, "--index")) {
if (is_not_gitdir)
die("--index outside a repository");
check_index = 1;
continue;
}
if (!strcmp(arg, "--cached")) {
if (is_not_gitdir)
die("--cached outside a repository");
check_index = 1;
cached = 1;
continue;
}
if (!strcmp(arg, "--apply")) {
apply = 1;
continue;
}
if (!strcmp(arg, "--build-fake-ancestor")) {
apply = 0;
if (++i >= argc)
die ("need a filename");
fake_ancestor = argv[i];
continue;
}
if (!strcmp(arg, "-z")) {
line_termination = 0;
continue;
}
if (!prefixcmp(arg, "-C")) {
Implement limited context matching in git-apply. Ok this really should be the good version. The option handling has been reworked to be automation safe. Currently to import the -mm tree I have to work around git-apply by using patch. Because some of Andrews patches in quilt will only apply with fuzz. I started out implementing a --fuzz option and then I realized fuzz is not a very safe concept for an automated system. What you really want is a minimum number of context lines that must match. This allows policy to be set without knowing how many lines of context a patch actually provides. By default the policy remains to match all provided lines of context. Allowng git-apply to match a restricted set of context makes it much easier to import the -mm tree into git. I am still only processing 1.5 to 1.6 patches a second for the 692 patches in 2.6.17-rc1-mm2 is still painful but it does help. If I just loop through all of Andrews patches in order and run git-apply --index -C1 I process the entire patchset in 1m53s or about 6 patches per second. So running git-mailinfo, git-write-tree, git-commit-tree, and git-update-ref everytime has a measurable impact, and shows things can be speeded up even more. All of these timings were taking on my poor 700Mhz Athlon with 512MB of ram. So people with fast machiens should see much better performance. When a match is found after the number of context are reduced a warning is generated. Since this is a rare event and possibly dangerous this seems to make sense. Unless you are patching a single file the error message is a little bit terse at the moment, but it should be easy to go back and fix. I have also updated the documentation for git-apply to reflect the new -C option that sets the minimum number of context lines that must match. Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: Junio C Hamano <junkio@cox.net>
2006-04-10 17:33:06 +08:00
p_context = strtoul(arg + 2, &end, 0);
if (*end != '\0')
die("unrecognized context count '%s'", arg + 2);
continue;
}
if (!prefixcmp(arg, "--whitespace=")) {
whitespace_option = arg + 13;
parse_whitespace_option(arg + 13);
continue;
}
if (!strcmp(arg, "-R") || !strcmp(arg, "--reverse")) {
apply_in_reverse = 1;
continue;
}
if (!strcmp(arg, "--unidiff-zero")) {
unidiff_zero = 1;
continue;
}
if (!strcmp(arg, "--reject")) {
apply = apply_with_reject = apply_verbosely = 1;
continue;
}
if (!strcmp(arg, "-v") || !strcmp(arg, "--verbose")) {
apply_verbosely = 1;
continue;
}
if (!strcmp(arg, "--inaccurate-eof")) {
inaccurate_eof = 1;
continue;
}
if (0 < prefix_length)
arg = prefix_filename(prefix, prefix_length, arg);
fd = open(arg, O_RDONLY);
if (fd < 0)
usage(apply_usage);
read_stdin = 0;
set_default_whitespace_mode(whitespace_option);
errs |= apply_patch(fd, arg, inaccurate_eof);
close(fd);
}
set_default_whitespace_mode(whitespace_option);
if (read_stdin)
errs |= apply_patch(0, "<stdin>", inaccurate_eof);
if (whitespace_error) {
if (squelch_whitespace_errors &&
squelch_whitespace_errors < whitespace_error) {
int squelched =
whitespace_error - squelch_whitespace_errors;
fprintf(stderr, "warning: squelched %d "
"whitespace error%s\n",
squelched,
squelched == 1 ? "" : "s");
}
if (new_whitespace == error_on_whitespace)
die("%d line%s add%s whitespace errors.",
whitespace_error,
whitespace_error == 1 ? "" : "s",
whitespace_error == 1 ? "s" : "");
if (applied_after_fixing_ws)
fprintf(stderr, "warning: %d line%s applied after"
" fixing whitespace errors.\n",
applied_after_fixing_ws,
applied_after_fixing_ws == 1 ? "" : "s");
else if (whitespace_error)
fprintf(stderr, "warning: %d line%s add%s whitespace errors.\n",
whitespace_error,
whitespace_error == 1 ? "" : "s",
whitespace_error == 1 ? "s" : "");
}
if (update_index) {
if (write_cache(newfd, active_cache, active_nr) ||
close(newfd) || commit_locked_index(&lock_file))
die("Unable to write new index file");
}
return !!errs;
}