patch 8.1.0959: sorting large numbers is not tested
Problem: Sorting large numbers is not tested and does not work properly.
Solution: Add test. Fix comparing lines with and without a number.
(Dominique Pelle, closes #4017)
diff --git a/src/ex_cmds.c b/src/ex_cmds.c
index da10c12..7682931 100644
--- a/src/ex_cmds.c
+++ b/src/ex_cmds.c
@@ -296,16 +296,20 @@
/* Struct to store info to be sorted. */
typedef struct
{
- linenr_T lnum; /* line number */
+ linenr_T lnum; // line number
union {
struct
{
- varnumber_T start_col_nr; /* starting column number */
- varnumber_T end_col_nr; /* ending column number */
+ varnumber_T start_col_nr; // starting column number
+ varnumber_T end_col_nr; // ending column number
} line;
- varnumber_T value; /* value if sorting by integer */
+ struct
+ {
+ varnumber_T value; // value if sorting by integer
+ int is_number; // TRUE when line contains a number
+ } num;
#ifdef FEAT_FLOAT
- float_T value_flt; /* value if sorting by float */
+ float_T value_flt; // value if sorting by float
#endif
} st_u;
} sorti_T;
@@ -335,11 +339,14 @@
if (got_int)
sort_abort = TRUE;
- /* When sorting numbers "start_col_nr" is the number, not the column
- * number. */
if (sort_nr)
- result = l1.st_u.value == l2.st_u.value ? 0
- : l1.st_u.value > l2.st_u.value ? 1 : -1;
+ {
+ if (l1.st_u.num.is_number != l2.st_u.num.is_number)
+ result = l1.st_u.num.is_number - l2.st_u.num.is_number;
+ else
+ result = l1.st_u.num.value == l2.st_u.num.value ? 0
+ : l1.st_u.num.value > l2.st_u.num.value ? 1 : -1;
+ }
#ifdef FEAT_FLOAT
else if (sort_flt)
result = l1.st_u.value_flt == l2.st_u.value_flt ? 0
@@ -553,11 +560,17 @@
if (s > p && s[-1] == '-')
--s; /* include preceding negative sign */
if (*s == NUL)
- /* empty line should sort before any number */
- nrs[lnum - eap->line1].st_u.value = -MAXLNUM;
+ {
+ /* line without number should sort before any number */
+ nrs[lnum - eap->line1].st_u.num.is_number = FALSE;
+ nrs[lnum - eap->line1].st_u.num.value = 0;
+ }
else
+ {
+ nrs[lnum - eap->line1].st_u.num.is_number = TRUE;
vim_str2nr(s, NULL, NULL, sort_what,
- &nrs[lnum - eap->line1].st_u.value, NULL, 0);
+ &nrs[lnum - eap->line1].st_u.num.value, NULL, 0);
+ }
}
#ifdef FEAT_FLOAT
else