
Instead of special-casing creations and deletions let's just generate a diff for them. This logic of not running a diff under -G if we don't have both sides dates back to the original implementation of -S in52e9578985
([PATCH] Introducing software archaeologist's tool "pickaxe"., 2005-05-21). In the case of -S we were not working with the xdiff interface and needed to do this, but when -G was implemented inf506b8e8b5
(git log/diff: add -G<regexp> that greps in the patch text, 2010-08-23) this logic was diligently copied over. But as the performance test added earlier in this series shows, this does not make much of a difference. With: time GIT_TEST_LONG= GIT_PERF_REPEAT_COUNT=10 GIT_PERF_MAKE_OPTS='-j8 CFLAGS=-O3' ./run origin/next HEAD~ HEAD -- p4209-pickaxe.sh With the HEAD~ commit being the preceding "pickaxe -G: terminate early on matching lines" we get these results. Note that it's only the -G codepaths that are relevant to this change: Test origin/next HEAD~ HEAD ----------------------------------------------------------------------------------------------------------------------------------------- 4209.1: git log -S'int main' <limit-rev>.. 0.35(0.32+0.03) 0.35(0.33+0.02) +0.0% 0.35(0.30+0.05) +0.0% 4209.2: git log -S'æ' <limit-rev>.. 0.46(0.42+0.04) 0.46(0.41+0.05) +0.0% 0.46(0.42+0.04) +0.0% 4209.3: git log --pickaxe-regex -S'(int|void|null)' <limit-rev>.. 0.65(0.62+0.02) 0.64(0.61+0.02) -1.5% 0.64(0.60+0.04) -1.5% 4209.4: git log --pickaxe-regex -S'if *\([^ ]+ & ' <limit-rev>.. 0.52(0.45+0.06) 0.52(0.50+0.01) +0.0% 0.54(0.47+0.04) +3.8% 4209.5: git log --pickaxe-regex -S'[àáâãäåæñøùúûüýþ]' <limit-rev>.. 0.39(0.34+0.05) 0.39(0.34+0.04) +0.0% 0.39(0.36+0.03) +0.0% 4209.6: git log -G'(int|void|null)' <limit-rev>.. 0.60(0.55+0.04) 0.58(0.54+0.03) -3.3% 0.58(0.49+0.08) -3.3% 4209.7: git log -G'if *\([^ ]+ & ' <limit-rev>.. 0.61(0.52+0.06) 0.59(0.53+0.05) -3.3% 0.59(0.54+0.05) -3.3% 4209.8: git log -G'[àáâãäåæñøùúûüýþ]' <limit-rev>.. 0.61(0.51+0.07) 0.58(0.54+0.04) -4.9% 0.57(0.51+0.06) -6.6% 4209.9: git log -i -S'int main' <limit-rev>.. 0.36(0.31+0.04) 0.36(0.34+0.02) +0.0% 0.35(0.32+0.03) -2.8% 4209.10: git log -i -S'æ' <limit-rev>.. 0.36(0.33+0.03) 0.39(0.34+0.01) +8.3% 0.36(0.32+0.03) +0.0% 4209.11: git log -i --pickaxe-regex -S'(int|void|null)' <limit-rev>.. 0.83(0.77+0.05) 0.82(0.77+0.05) -1.2% 0.80(0.75+0.04) -3.6% 4209.12: git log -i --pickaxe-regex -S'if *\([^ ]+ & ' <limit-rev>.. 0.67(0.61+0.03) 0.64(0.61+0.03) -4.5% 0.63(0.61+0.02) -6.0% 4209.13: git log -i --pickaxe-regex -S'[àáâãäåæñøùúûüýþ]' <limit-rev>.. 0.40(0.37+0.02) 0.40(0.37+0.03) +0.0% 0.40(0.36+0.04) +0.0% 4209.14: git log -i -G'(int|void|null)' <limit-rev>.. 0.58(0.51+0.07) 0.59(0.52+0.06) +1.7% 0.58(0.52+0.05) +0.0% 4209.15: git log -i -G'if *\([^ ]+ & ' <limit-rev>.. 0.60(0.54+0.05) 0.60(0.54+0.06) +0.0% 0.60(0.56+0.03) +0.0% 4209.16: git log -i -G'[àáâãäåæñøùúûüýþ]' <limit-rev>.. 0.58(0.51+0.06) 0.57(0.52+0.05) -1.7% 0.60(0.48+0.09) +3.4% This small simplification really doesn't buy us much now, but I've got plans to both convert the pickaxe code to using a PCREv2 backend[1] and to implement additional pickaxe modes to do custom searches through the diff[2]. Always having the diff available under -G is going to help to simplify both of those changes. 1. https://lore.kernel.org/git/20210203032811.14979-22-avarab@gmail.com/ 2. https://lore.kernel.org/git/20190424152215.16251-3-avarab@gmail.com/ Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: Junio C Hamano <gitster@pobox.com>
290 lines
7.1 KiB
C
290 lines
7.1 KiB
C
/*
|
|
* Copyright (C) 2005 Junio C Hamano
|
|
* Copyright (C) 2010 Google Inc.
|
|
*/
|
|
#include "cache.h"
|
|
#include "diff.h"
|
|
#include "diffcore.h"
|
|
#include "xdiff-interface.h"
|
|
#include "kwset.h"
|
|
#include "commit.h"
|
|
#include "quote.h"
|
|
|
|
typedef int (*pickaxe_fn)(mmfile_t *one, mmfile_t *two,
|
|
struct diff_options *o,
|
|
regex_t *regexp, kwset_t kws);
|
|
|
|
struct diffgrep_cb {
|
|
regex_t *regexp;
|
|
int hit;
|
|
};
|
|
|
|
static int diffgrep_consume(void *priv, char *line, unsigned long len)
|
|
{
|
|
struct diffgrep_cb *data = priv;
|
|
regmatch_t regmatch;
|
|
|
|
if (line[0] != '+' && line[0] != '-')
|
|
return 0;
|
|
if (data->hit)
|
|
BUG("Already matched in diffgrep_consume! Broken xdiff_emit_line_fn?");
|
|
if (!regexec_buf(data->regexp, line + 1, len - 1, 1,
|
|
®match, 0)) {
|
|
data->hit = 1;
|
|
return 1;
|
|
}
|
|
return 0;
|
|
}
|
|
|
|
static int diff_grep(mmfile_t *one, mmfile_t *two,
|
|
struct diff_options *o,
|
|
regex_t *regexp, kwset_t kws)
|
|
{
|
|
struct diffgrep_cb ecbdata;
|
|
xpparam_t xpp;
|
|
xdemitconf_t xecfg;
|
|
int ret;
|
|
|
|
/*
|
|
* We have both sides; need to run textual diff and see if
|
|
* the pattern appears on added/deleted lines.
|
|
*/
|
|
memset(&xpp, 0, sizeof(xpp));
|
|
memset(&xecfg, 0, sizeof(xecfg));
|
|
ecbdata.regexp = regexp;
|
|
ecbdata.hit = 0;
|
|
xecfg.ctxlen = o->context;
|
|
xecfg.interhunkctxlen = o->interhunkcontext;
|
|
|
|
/*
|
|
* An xdiff error might be our "data->hit" from above. See the
|
|
* comment for xdiff_emit_line_fn in xdiff-interface.h
|
|
*/
|
|
ret = xdi_diff_outf(one, two, discard_hunk_line, diffgrep_consume,
|
|
&ecbdata, &xpp, &xecfg);
|
|
if (ecbdata.hit)
|
|
return 1;
|
|
if (ret)
|
|
return ret;
|
|
return 0;
|
|
}
|
|
|
|
static unsigned int contains(mmfile_t *mf, regex_t *regexp, kwset_t kws,
|
|
unsigned int limit)
|
|
{
|
|
unsigned int cnt = 0;
|
|
unsigned long sz = mf->size;
|
|
const char *data = mf->ptr;
|
|
|
|
if (regexp) {
|
|
regmatch_t regmatch;
|
|
int flags = 0;
|
|
|
|
while (sz &&
|
|
!regexec_buf(regexp, data, sz, 1, ®match, flags)) {
|
|
flags |= REG_NOTBOL;
|
|
data += regmatch.rm_eo;
|
|
sz -= regmatch.rm_eo;
|
|
if (sz && regmatch.rm_so == regmatch.rm_eo) {
|
|
data++;
|
|
sz--;
|
|
}
|
|
cnt++;
|
|
|
|
if (limit && cnt == limit)
|
|
return cnt;
|
|
}
|
|
|
|
} else { /* Classic exact string match */
|
|
while (sz) {
|
|
struct kwsmatch kwsm;
|
|
size_t offset = kwsexec(kws, data, sz, &kwsm);
|
|
if (offset == -1)
|
|
break;
|
|
sz -= offset + kwsm.size[0];
|
|
data += offset + kwsm.size[0];
|
|
cnt++;
|
|
|
|
if (limit && cnt == limit)
|
|
return cnt;
|
|
}
|
|
}
|
|
return cnt;
|
|
}
|
|
|
|
static int has_changes(mmfile_t *one, mmfile_t *two,
|
|
struct diff_options *o,
|
|
regex_t *regexp, kwset_t kws)
|
|
{
|
|
unsigned int c1 = one ? contains(one, regexp, kws, 0) : 0;
|
|
unsigned int c2 = two ? contains(two, regexp, kws, c1 + 1) : 0;
|
|
return c1 != c2;
|
|
}
|
|
|
|
static int pickaxe_match(struct diff_filepair *p, struct diff_options *o,
|
|
regex_t *regexp, kwset_t kws, pickaxe_fn fn)
|
|
{
|
|
struct userdiff_driver *textconv_one = NULL;
|
|
struct userdiff_driver *textconv_two = NULL;
|
|
mmfile_t mf1, mf2;
|
|
int ret;
|
|
|
|
/* ignore unmerged */
|
|
if (!DIFF_FILE_VALID(p->one) && !DIFF_FILE_VALID(p->two))
|
|
return 0;
|
|
|
|
if (o->objfind) {
|
|
return (DIFF_FILE_VALID(p->one) &&
|
|
oidset_contains(o->objfind, &p->one->oid)) ||
|
|
(DIFF_FILE_VALID(p->two) &&
|
|
oidset_contains(o->objfind, &p->two->oid));
|
|
}
|
|
|
|
if (o->flags.allow_textconv) {
|
|
textconv_one = get_textconv(o->repo, p->one);
|
|
textconv_two = get_textconv(o->repo, p->two);
|
|
}
|
|
|
|
/*
|
|
* If we have an unmodified pair, we know that the count will be the
|
|
* same and don't even have to load the blobs. Unless textconv is in
|
|
* play, _and_ we are using two different textconv filters (e.g.,
|
|
* because a pair is an exact rename with different textconv attributes
|
|
* for each side, which might generate different content).
|
|
*/
|
|
if (textconv_one == textconv_two && diff_unmodified_pair(p))
|
|
return 0;
|
|
|
|
if ((o->pickaxe_opts & DIFF_PICKAXE_KIND_G) &&
|
|
!o->flags.text &&
|
|
((!textconv_one && diff_filespec_is_binary(o->repo, p->one)) ||
|
|
(!textconv_two && diff_filespec_is_binary(o->repo, p->two))))
|
|
return 0;
|
|
|
|
mf1.size = fill_textconv(o->repo, textconv_one, p->one, &mf1.ptr);
|
|
mf2.size = fill_textconv(o->repo, textconv_two, p->two, &mf2.ptr);
|
|
|
|
ret = fn(&mf1, &mf2, o, regexp, kws);
|
|
|
|
if (textconv_one)
|
|
free(mf1.ptr);
|
|
if (textconv_two)
|
|
free(mf2.ptr);
|
|
diff_free_filespec_data(p->one);
|
|
diff_free_filespec_data(p->two);
|
|
|
|
return ret;
|
|
}
|
|
|
|
static void pickaxe(struct diff_queue_struct *q, struct diff_options *o,
|
|
regex_t *regexp, kwset_t kws, pickaxe_fn fn)
|
|
{
|
|
int i;
|
|
struct diff_queue_struct outq;
|
|
|
|
DIFF_QUEUE_CLEAR(&outq);
|
|
|
|
if (o->pickaxe_opts & DIFF_PICKAXE_ALL) {
|
|
/* Showing the whole changeset if needle exists */
|
|
for (i = 0; i < q->nr; i++) {
|
|
struct diff_filepair *p = q->queue[i];
|
|
if (pickaxe_match(p, o, regexp, kws, fn))
|
|
return; /* do not munge the queue */
|
|
}
|
|
|
|
/*
|
|
* Otherwise we will clear the whole queue by copying
|
|
* the empty outq at the end of this function, but
|
|
* first clear the current entries in the queue.
|
|
*/
|
|
for (i = 0; i < q->nr; i++)
|
|
diff_free_filepair(q->queue[i]);
|
|
} else {
|
|
/* Showing only the filepairs that has the needle */
|
|
for (i = 0; i < q->nr; i++) {
|
|
struct diff_filepair *p = q->queue[i];
|
|
if (pickaxe_match(p, o, regexp, kws, fn))
|
|
diff_q(&outq, p);
|
|
else
|
|
diff_free_filepair(p);
|
|
}
|
|
}
|
|
|
|
free(q->queue);
|
|
*q = outq;
|
|
}
|
|
|
|
static void regcomp_or_die(regex_t *regex, const char *needle, int cflags)
|
|
{
|
|
int err = regcomp(regex, needle, cflags);
|
|
if (err) {
|
|
/* The POSIX.2 people are surely sick */
|
|
char errbuf[1024];
|
|
regerror(err, regex, errbuf, 1024);
|
|
die("invalid regex: %s", errbuf);
|
|
}
|
|
}
|
|
|
|
void diffcore_pickaxe(struct diff_options *o)
|
|
{
|
|
const char *needle = o->pickaxe;
|
|
int opts = o->pickaxe_opts;
|
|
regex_t regex, *regexp = NULL;
|
|
kwset_t kws = NULL;
|
|
pickaxe_fn fn;
|
|
|
|
if (opts & ~DIFF_PICKAXE_KIND_OBJFIND &&
|
|
(!needle || !*needle))
|
|
BUG("should have needle under -G or -S");
|
|
if (opts & (DIFF_PICKAXE_REGEX | DIFF_PICKAXE_KIND_G)) {
|
|
int cflags = REG_EXTENDED | REG_NEWLINE;
|
|
if (o->pickaxe_opts & DIFF_PICKAXE_IGNORE_CASE)
|
|
cflags |= REG_ICASE;
|
|
regcomp_or_die(®ex, needle, cflags);
|
|
regexp = ®ex;
|
|
|
|
if (opts & DIFF_PICKAXE_KIND_G)
|
|
fn = diff_grep;
|
|
else if (opts & DIFF_PICKAXE_REGEX)
|
|
fn = has_changes;
|
|
else
|
|
/*
|
|
* We don't need to check the combination of
|
|
* -G and --pickaxe-regex, by the time we get
|
|
* here diff.c has already died if they're
|
|
* combined. See the usage tests in
|
|
* t4209-log-pickaxe.sh.
|
|
*/
|
|
BUG("unreachable");
|
|
} else if (opts & DIFF_PICKAXE_KIND_S) {
|
|
if (o->pickaxe_opts & DIFF_PICKAXE_IGNORE_CASE &&
|
|
has_non_ascii(needle)) {
|
|
struct strbuf sb = STRBUF_INIT;
|
|
int cflags = REG_NEWLINE | REG_ICASE;
|
|
|
|
basic_regex_quote_buf(&sb, needle);
|
|
regcomp_or_die(®ex, sb.buf, cflags);
|
|
strbuf_release(&sb);
|
|
regexp = ®ex;
|
|
} else {
|
|
kws = kwsalloc(o->pickaxe_opts & DIFF_PICKAXE_IGNORE_CASE
|
|
? tolower_trans_tbl : NULL);
|
|
kwsincr(kws, needle, strlen(needle));
|
|
kwsprep(kws);
|
|
}
|
|
fn = has_changes;
|
|
} else if (opts & DIFF_PICKAXE_KIND_OBJFIND) {
|
|
fn = NULL;
|
|
} else {
|
|
BUG("unknown pickaxe_opts flag");
|
|
}
|
|
|
|
pickaxe(&diff_queued_diff, o, regexp, kws, fn);
|
|
|
|
if (regexp)
|
|
regfree(regexp);
|
|
if (kws)
|
|
kwsfree(kws);
|
|
return;
|
|
}
|