Hello
I just encoutered a patch(1) limitation when using it on minified json
files from Wordpress. The lines can span more than the maximum of what
patch(1) can cope, which is INT16_MAX. Here is a test for taht:
jot -b A $(( 37268 / 2 )) |tr '\n' 'A' > g.txt
mv g.txt g.txt.orig
sed 's/A/B/' g.txt.orig > g.txt
diff -U4 g.txt.orig g.txt > g.patch
mv g.txt.orig g.txt
patch < g.patch
Currently it dumps a core.
patch(1) tracks line length using a short, hence the INT16_MAX limit.
Changing short for long seems reasonable. Anyone sees something wrong
with this change?
--- patch.c.orig 2024-07-12 03:06:47.101773258 +0200
+++ patch.c 2024-07-12 03:07:58.394948697 +0200
@@ -1015,5 +1015,5 @@
const char *ilineptr;
const char *plineptr;
- short plinelen;
+ long plinelen;
for (iline = base + offset + fuzz; pline <= pat_lines; pline++,
iline++) {
--- pch.c.orig 2024-07-12 03:06:26.498848676 +0200
+++ pch.c 2024-07-12 03:07:54.115023483 +0200
@@ -62,5 +62,5 @@
static LINENUM p_input_line = 0; /* current line # from patch file */
static char **p_line = NULL;/* the text of the hunk */
-static short *p_len = NULL; /* length of each line */
+static long *p_len = NULL; /* length of each line */
static char *p_char = NULL; /* +, -, and ! */
static int hunkmax = INITHUNKMAX; /* size of above arrays to begin with */
@@ -149,5 +149,5 @@
int new_hunkmax;
char **new_p_line;
- short *new_p_len;
+ long *new_p_len;
char *new_p_char;
@@ -1208,5 +1208,5 @@
{
char **tp_line; /* the text of the hunk */
- short *tp_len; /* length of each line */
+ long *tp_len; /* length of each line */
char *tp_char; /* +, -, and ! */
LINENUM i;
@@ -1365,5 +1365,5 @@
* Return the length of a particular patch line.
*/
-short
+long
pch_line_len(LINENUM line)
{
--- pch.h.orig 2024-07-12 03:06:34.298099780 +0200
+++ pch.h 2024-07-12 03:07:19.155488620 +0200
@@ -48,5 +48,5 @@
bool pch_swap(void);
char *pfetch(LINENUM);
-short pch_line_len(LINENUM);
+long pch_line_len(LINENUM);
LINENUM pch_first(void);
LINENUM pch_ptrn_lines(void);
--
Emmanuel Dreyfus
m...@netbsd.org
