Patch 8.2.3921
Problem: The way xdiff is used is inefficient.
Solution: Use hunk_func instead of the out_line callback. (Lewis Russell,
closes #9344)
Files: src/diff.c
*** ../vim-8.2.3920/src/diff.c 2021-12-27 17:21:38.004449137 +0000
--- src/diff.c 2021-12-28 13:50:11.975517843 +0000
***************
*** 64,69 ****
--- 64,77 ----
garray_T dout_ga; // used for internal diff
} diffout_T;
+ // used for recording hunks from xdiff
+ typedef struct {
+ linenr_T lnum_orig;
+ long count_orig;
+ linenr_T lnum_new;
+ long count_new;
+ } diffhunk_T;
+
// two diff inputs and one result
typedef struct {
diffin_T dio_orig; // original file input
***************
*** 84,95 ****
#ifdef FEAT_FOLDING
static void diff_fold_update(diff_T *dp, int skip_idx);
#endif
! static void diff_read(int idx_orig, int idx_new, diffout_T *fname);
static void diff_copy_entry(diff_T *dprev, diff_T *dp, int idx_orig, int
idx_new);
static diff_T *diff_alloc_new(tabpage_T *tp, diff_T *dprev, diff_T *dp);
! static int parse_diff_ed(char_u *line, linenr_T *lnum_orig, long *count_orig,
linenr_T *lnum_new, long *count_new);
! static int parse_diff_unified(char_u *line, linenr_T *lnum_orig, long
*count_orig, linenr_T *lnum_new, long *count_new);
! static int xdiff_out(void *priv, mmbuffer_t *mb, int nbuf);
#define FOR_ALL_DIFFBLOCKS_IN_TAB(tp, dp) \
for ((dp) = (tp)->tp_first_diff; (dp) != NULL; (dp) = (dp)->df_next)
--- 92,103 ----
#ifdef FEAT_FOLDING
static void diff_fold_update(diff_T *dp, int skip_idx);
#endif
! static void diff_read(int idx_orig, int idx_new, diffio_T *dio);
static void diff_copy_entry(diff_T *dprev, diff_T *dp, int idx_orig, int
idx_new);
static diff_T *diff_alloc_new(tabpage_T *tp, diff_T *dprev, diff_T *dp);
! static int parse_diff_ed(char_u *line, diffhunk_T *hunk);
! static int parse_diff_unified(char_u *line, diffhunk_T *hunk);
! static int xdiff_out(long start_a, long count_a, long start_b, long count_b,
void *priv);
#define FOR_ALL_DIFFBLOCKS_IN_TAB(tp, dp) \
for ((dp) = (tp)->tp_first_diff; (dp) != NULL; (dp) = (dp)->df_next)
***************
*** 880,886 ****
continue;
// Read the diff output and add each entry to the diff list.
! diff_read(idx_orig, idx_new, &dio->dio_diff);
clear_diffin(&dio->dio_new);
clear_diffout(&dio->dio_diff);
--- 888,894 ----
continue;
// Read the diff output and add each entry to the diff list.
! diff_read(idx_orig, idx_new, dio);
clear_diffin(&dio->dio_new);
clear_diffout(&dio->dio_diff);
***************
*** 1118,1124 ****
emit_cfg.ctxlen = 0; // don't need any diff_context here
emit_cb.priv = &diffio->dio_diff;
! emit_cb.out_line = xdiff_out;
if (xdl_diff(&diffio->dio_orig.din_mmfile,
&diffio->dio_new.din_mmfile,
¶m, &emit_cfg, &emit_cb) < 0)
--- 1126,1132 ----
emit_cfg.ctxlen = 0; // don't need any diff_context here
emit_cb.priv = &diffio->dio_diff;
! emit_cfg.hunk_func = xdiff_out;
if (xdl_diff(&diffio->dio_orig.din_mmfile,
&diffio->dio_new.din_mmfile,
¶m, &emit_cfg, &emit_cb) < 0)
***************
*** 1482,1490 ****
wp->w_p_crb = TRUE;
if (!(diff_flags & DIFF_FOLLOWWRAP))
{
! if (!wp->w_p_diff)
wp->w_p_wrap_save = wp->w_p_wrap;
! wp->w_p_wrap = FALSE;
}
# ifdef FEAT_FOLDING
if (!wp->w_p_diff)
--- 1490,1498 ----
wp->w_p_crb = TRUE;
if (!(diff_flags & DIFF_FOLLOWWRAP))
{
! if (!wp->w_p_diff)
wp->w_p_wrap_save = wp->w_p_wrap;
! wp->w_p_wrap = FALSE;
}
# ifdef FEAT_FOLDING
if (!wp->w_p_diff)
***************
*** 1549,1555 ****
if (!(diff_flags & DIFF_FOLLOWWRAP))
{
if (!wp->w_p_wrap)
! wp->w_p_wrap = wp->w_p_wrap_save;
}
#ifdef FEAT_FOLDING
free_string_option(wp->w_p_fdm);
--- 1557,1563 ----
if (!(diff_flags & DIFF_FOLLOWWRAP))
{
if (!wp->w_p_wrap)
! wp->w_p_wrap = wp->w_p_wrap_save;
}
#ifdef FEAT_FOLDING
free_string_option(wp->w_p_fdm);
***************
*** 1607,1626 ****
diff_read(
int idx_orig, // idx of original file
int idx_new, // idx of new file
! diffout_T *dout) // diff output
{
FILE *fd = NULL;
int line_idx = 0;
diff_T *dprev = NULL;
diff_T *dp = curtab->tp_first_diff;
diff_T *dn, *dpl;
char_u linebuf[LBUFLEN]; // only need to hold the diff line
char_u *line;
long off;
int i;
- linenr_T lnum_orig, lnum_new;
- long count_orig, count_new;
int notset = TRUE; // block "*dp" not set yet
enum {
DIFF_ED,
DIFF_UNIFIED,
--- 1615,1635 ----
diff_read(
int idx_orig, // idx of original file
int idx_new, // idx of new file
! diffio_T *dio) // diff output
{
FILE *fd = NULL;
int line_idx = 0;
diff_T *dprev = NULL;
diff_T *dp = curtab->tp_first_diff;
diff_T *dn, *dpl;
+ diffout_T *dout = &dio->dio_diff;
char_u linebuf[LBUFLEN]; // only need to hold the diff line
char_u *line;
long off;
int i;
int notset = TRUE; // block "*dp" not set yet
+ diffhunk_T *hunk;
+
enum {
DIFF_ED,
DIFF_UNIFIED,
***************
*** 1641,1715 ****
}
}
! for (;;)
{
! if (fd == NULL)
{
! if (line_idx >= dout->dout_ga.ga_len)
! break; // did last line
! line = ((char_u **)dout->dout_ga.ga_data)[line_idx++];
! }
! else
! {
! if (vim_fgets(linebuf, LBUFLEN, fd))
! break; // end of file
! line = linebuf;
! }
!
! if (diffstyle == DIFF_NONE)
! {
! // Determine diff style.
! // ed like diff looks like this:
! // {first}[,{last}]c{first}[,{last}]
! // {first}a{first}[,{last}]
! // {first}[,{last}]d{first}
! //
! // unified diff looks like this:
! // --- file1 2018-03-20 13:23:35.783153140 +0100
! // +++ file2 2018-03-20 13:23:41.183156066 +0100
! // @@ -1,3 +1,5 @@
! if (isdigit(*line))
! diffstyle = DIFF_ED;
! else if ((STRNCMP(line, "@@ ", 3) == 0))
! diffstyle = DIFF_UNIFIED;
! else if ((STRNCMP(line, "--- ", 4) == 0)
! && (vim_fgets(linebuf, LBUFLEN, fd) == 0)
! && (STRNCMP(line, "+++ ", 4) == 0)
! && (vim_fgets(linebuf, LBUFLEN, fd) == 0)
! && (STRNCMP(line, "@@ ", 3) == 0))
! diffstyle = DIFF_UNIFIED;
! else
! // Format not recognized yet, skip over this line. Cygwin diff
! // may put a warning at the start of the file.
! continue;
}
! if (diffstyle == DIFF_ED)
! {
! if (!isdigit(*line))
! continue; // not the start of a diff block
! if (parse_diff_ed(line, &lnum_orig, &count_orig,
! &lnum_new, &count_new) == FAIL)
! continue;
! }
! else if (diffstyle == DIFF_UNIFIED)
{
! if (STRNCMP(line, "@@ ", 3) != 0)
! continue; // not the start of a diff block
! if (parse_diff_unified(line, &lnum_orig, &count_orig,
! &lnum_new, &count_new) == FAIL)
! continue;
}
else
{
! emsg(_("E959: Invalid diff format."));
! break;
}
// Go over blocks before the change, for which orig and new are equal.
// Copy blocks from orig to new.
while (dp != NULL
! && lnum_orig > dp->df_lnum[idx_orig] + dp->df_count[idx_orig])
{
if (notset)
diff_copy_entry(dprev, dp, idx_orig, idx_new);
--- 1650,1743 ----
}
}
! if (!dio->dio_internal)
{
! hunk = ALLOC_ONE(diffhunk_T);
! if (hunk == NULL)
{
! emsg(_("E98: Cannot read diff output"));
! return;
}
+ }
! for (;;)
! {
! if (dio->dio_internal)
{
! if (line_idx >= dout->dout_ga.ga_len) {
! break; // did last line
! }
! hunk = ((diffhunk_T **)dout->dout_ga.ga_data)[line_idx++];
}
else
{
! if (fd == NULL)
! {
! if (line_idx >= dout->dout_ga.ga_len)
! break; // did last line
! line = ((char_u **)dout->dout_ga.ga_data)[line_idx++];
! }
! else
! {
! if (vim_fgets(linebuf, LBUFLEN, fd))
! break; // end of file
! line = linebuf;
! }
!
! if (diffstyle == DIFF_NONE)
! {
! // Determine diff style.
! // ed like diff looks like this:
! // {first}[,{last}]c{first}[,{last}]
! // {first}a{first}[,{last}]
! // {first}[,{last}]d{first}
! //
! // unified diff looks like this:
! // --- file1 2018-03-20 13:23:35.783153140 +0100
! // +++ file2 2018-03-20 13:23:41.183156066 +0100
! // @@ -1,3 +1,5 @@
! if (isdigit(*line))
! diffstyle = DIFF_ED;
! else if ((STRNCMP(line, "@@ ", 3) == 0))
! diffstyle = DIFF_UNIFIED;
! else if ((STRNCMP(line, "--- ", 4) == 0)
! && (vim_fgets(linebuf, LBUFLEN, fd) == 0)
! && (STRNCMP(line, "+++ ", 4) == 0)
! && (vim_fgets(linebuf, LBUFLEN, fd) == 0)
! && (STRNCMP(line, "@@ ", 3) == 0))
! diffstyle = DIFF_UNIFIED;
! else
! // Format not recognized yet, skip over this line. Cygwin
! // diff may put a warning at the start of the file.
! continue;
! }
!
! if (diffstyle == DIFF_ED)
! {
! if (!isdigit(*line))
! continue; // not the start of a diff block
! if (parse_diff_ed(line, hunk) == FAIL)
! continue;
! }
! else if (diffstyle == DIFF_UNIFIED)
! {
! if (STRNCMP(line, "@@ ", 3) != 0)
! continue; // not the start of a diff block
! if (parse_diff_unified(line, hunk) == FAIL)
! continue;
! }
! else
! {
! emsg(_("E959: Invalid diff format."));
! break;
! }
}
// Go over blocks before the change, for which orig and new are equal.
// Copy blocks from orig to new.
while (dp != NULL
! && hunk->lnum_orig > dp->df_lnum[idx_orig]
! + dp->df_count[idx_orig])
{
if (notset)
diff_copy_entry(dprev, dp, idx_orig, idx_new);
***************
*** 1719,1759 ****
}
if (dp != NULL
! && lnum_orig <= dp->df_lnum[idx_orig] + dp->df_count[idx_orig]
! && lnum_orig + count_orig >= dp->df_lnum[idx_orig])
{
// New block overlaps with existing block(s).
// First find last block that overlaps.
for (dpl = dp; dpl->df_next != NULL; dpl = dpl->df_next)
! if (lnum_orig + count_orig < dpl->df_next->df_lnum[idx_orig])
break;
// If the newly found block starts before the old one, set the
// start back a number of lines.
! off = dp->df_lnum[idx_orig] - lnum_orig;
if (off > 0)
{
for (i = idx_orig; i < idx_new; ++i)
if (curtab->tp_diffbuf[i] != NULL)
dp->df_lnum[i] -= off;
! dp->df_lnum[idx_new] = lnum_new;
! dp->df_count[idx_new] = count_new;
}
else if (notset)
{
// new block inside existing one, adjust new block
! dp->df_lnum[idx_new] = lnum_new + off;
! dp->df_count[idx_new] = count_new - off;
}
else
// second overlap of new block with existing block
! dp->df_count[idx_new] += count_new - count_orig
+ dpl->df_lnum[idx_orig] + dpl->df_count[idx_orig]
- (dp->df_lnum[idx_orig] + dp->df_count[idx_orig]);
// Adjust the size of the block to include all the lines to the
// end of the existing block or the new diff, whatever ends last.
! off = (lnum_orig + count_orig)
- (dpl->df_lnum[idx_orig] + dpl->df_count[idx_orig]);
if (off < 0)
{
--- 1747,1789 ----
}
if (dp != NULL
! && hunk->lnum_orig <= dp->df_lnum[idx_orig]
! + dp->df_count[idx_orig]
! && hunk->lnum_orig + hunk->count_orig >= dp->df_lnum[idx_orig])
{
// New block overlaps with existing block(s).
// First find last block that overlaps.
for (dpl = dp; dpl->df_next != NULL; dpl = dpl->df_next)
! if (hunk->lnum_orig + hunk->count_orig
! < dpl->df_next->df_lnum[idx_orig])
break;
// If the newly found block starts before the old one, set the
// start back a number of lines.
! off = dp->df_lnum[idx_orig] - hunk->lnum_orig;
if (off > 0)
{
for (i = idx_orig; i < idx_new; ++i)
if (curtab->tp_diffbuf[i] != NULL)
dp->df_lnum[i] -= off;
! dp->df_lnum[idx_new] = hunk->lnum_new;
! dp->df_count[idx_new] = hunk->count_new;
}
else if (notset)
{
// new block inside existing one, adjust new block
! dp->df_lnum[idx_new] = hunk->lnum_new + off;
! dp->df_count[idx_new] = hunk->count_new - off;
}
else
// second overlap of new block with existing block
! dp->df_count[idx_new] += hunk->count_new - hunk->count_orig
+ dpl->df_lnum[idx_orig] + dpl->df_count[idx_orig]
- (dp->df_lnum[idx_orig] + dp->df_count[idx_orig]);
// Adjust the size of the block to include all the lines to the
// end of the existing block or the new diff, whatever ends last.
! off = (hunk->lnum_orig + hunk->count_orig)
- (dpl->df_lnum[idx_orig] + dpl->df_count[idx_orig]);
if (off < 0)
{
***************
*** 1785,1794 ****
if (dp == NULL)
goto done;
! dp->df_lnum[idx_orig] = lnum_orig;
! dp->df_count[idx_orig] = count_orig;
! dp->df_lnum[idx_new] = lnum_new;
! dp->df_count[idx_new] = count_new;
// Set values for other buffers, these must be equal to the
// original buffer, otherwise there would have been a change
--- 1815,1824 ----
if (dp == NULL)
goto done;
! dp->df_lnum[idx_orig] = hunk->lnum_orig;
! dp->df_count[idx_orig] = hunk->count_orig;
! dp->df_lnum[idx_new] = hunk->lnum_new;
! dp->df_count[idx_new] = hunk->count_new;
// Set values for other buffers, these must be equal to the
// original buffer, otherwise there would have been a change
***************
*** 1811,1816 ****
--- 1841,1849 ----
}
done:
+ if (!dio->dio_internal)
+ vim_free(hunk);
+
if (fd != NULL)
fclose(fd);
}
***************
*** 3137,3146 ****
static int
parse_diff_ed(
char_u *line,
! linenr_T *lnum_orig,
! long *count_orig,
! linenr_T *lnum_new,
! long *count_new)
{
char_u *p;
long f1, l1, f2, l2;
--- 3170,3176 ----
static int
parse_diff_ed(
char_u *line,
! diffhunk_T *hunk)
{
char_u *p;
long f1, l1, f2, l2;
***************
*** 3175,3197 ****
if (difftype == 'a')
{
! *lnum_orig = f1 + 1;
! *count_orig = 0;
}
else
{
! *lnum_orig = f1;
! *count_orig = l1 - f1 + 1;
}
if (difftype == 'd')
{
! *lnum_new = f2 + 1;
! *count_new = 0;
}
else
{
! *lnum_new = f2;
! *count_new = l2 - f2 + 1;
}
return OK;
}
--- 3205,3227 ----
if (difftype == 'a')
{
! hunk->lnum_orig = f1 + 1;
! hunk->count_orig = 0;
}
else
{
! hunk->lnum_orig = f1;
! hunk->count_orig = l1 - f1 + 1;
}
if (difftype == 'd')
{
! hunk->lnum_new = f2 + 1;
! hunk->count_new = 0;
}
else
{
! hunk->lnum_new = f2;
! hunk->count_new = l2 - f2 + 1;
}
return OK;
}
***************
*** 3203,3212 ****
static int
parse_diff_unified(
char_u *line,
! linenr_T *lnum_orig,
! long *count_orig,
! linenr_T *lnum_new,
! long *count_new)
{
char_u *p;
long oldline, oldcount, newline, newcount;
--- 3233,3239 ----
static int
parse_diff_unified(
char_u *line,
! diffhunk_T *hunk)
{
char_u *p;
long oldline, oldcount, newline, newcount;
***************
*** 3245,3254 ****
if (newline == 0)
newline = 1;
! *lnum_orig = oldline;
! *count_orig = oldcount;
! *lnum_new = newline;
! *count_new = newcount;
return OK;
}
--- 3272,3281 ----
if (newline == 0)
newline = 1;
! hunk->lnum_orig = oldline;
! hunk->count_orig = oldcount;
! hunk->lnum_new = newline;
! hunk->count_new = newcount;
return OK;
}
***************
*** 3261,3286 ****
* Stores the diff output in a grow array.
*/
static int
! xdiff_out(void *priv, mmbuffer_t *mb, int nbuf)
{
diffout_T *dout = (diffout_T *)priv;
! char_u *p;
!
! // The header line always comes by itself, text lines in at least two
! // parts. We drop the text part.
! if (nbuf > 1)
! return 0;
! // sanity check
! if (STRNCMP(mb[0].ptr, "@@ ", 3) != 0)
! return 0;
if (ga_grow(&dout->dout_ga, 1) == FAIL)
return -1;
! p = vim_strnsave((char_u *)mb[0].ptr, mb[0].size);
! if (p == NULL)
! return -1;
! ((char_u **)dout->dout_ga.ga_data)[dout->dout_ga.ga_len++] = p;
return 0;
}
--- 3288,3314 ----
* Stores the diff output in a grow array.
*/
static int
! xdiff_out(
! long start_a,
! long count_a,
! long start_b,
! long count_b,
! void *priv)
{
diffout_T *dout = (diffout_T *)priv;
! diffhunk_T *p = ALLOC_ONE(diffhunk_T);
! if (p == NULL)
! return -1;
if (ga_grow(&dout->dout_ga, 1) == FAIL)
return -1;
!
! p->lnum_orig = start_a + 1;
! p->count_orig = count_a;
! p->lnum_new = start_b + 1;
! p->count_new = count_b;
! ((diffhunk_T **)dout->dout_ga.ga_data)[dout->dout_ga.ga_len++] = p;
return 0;
}
*** ../vim-8.2.3920/src/version.c 2021-12-28 13:15:02.444896152 +0000
--- src/version.c 2021-12-28 13:51:02.295452062 +0000
***************
*** 751,752 ****
--- 751,754 ----
{ /* Add new patch number below this line */
+ /**/
+ 3921,
/**/
--
You cannot have a baby in one month by getting nine women pregnant.
/// Bram Moolenaar -- [email protected] -- http://www.Moolenaar.net \\\
/// \\\
\\\ sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ ///
\\\ help me help AIDS victims -- http://ICCF-Holland.org ///
--
--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
---
You received this message because you are subscribed to the Google Groups
"vim_dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/vim_dev/20211228135523.544731C0641%40moolenaar.net.
