client: handle wide chars for table rows
PATCH v2 1/1 client: handle wide chars for table rows
Find out printing width of wide chars via wcwidth().
--- client/display.c | 18 ++++++++++++++---- client/main.c | 3 +++ 2 files changed, 17 insertions(+), 4 deletions(-)
client/display.c +14 -4
@@ -30,6 +30,7 @@#include <sys/stat.h> #include <sys/ioctl.h> #include <unistd.h> +#include <wchar.h> #include <readline/history.h> #include <readline/readline.h>
@@ -401,18 +402,27 @@ static char* next_line(char *s, unsigned int *max, char **color_out)unsigned int i;int last_space =-1 ;int last_color =-1 ; +int s_len =strlen (s);/* Find the last space before 'max', as well as any color */ -for (i =0 ; i <= *max && s[i] !='\0' ; i++) { +for (i =0 ; i <= *max && i != s_len; i++) {if (s[i] ==' ' ) last_space = i;else if (s[i] ==0x1b ) {/* color escape won't count for column width */ *max +=color_end (s + i); last_color = i; -/* Add width for non-codepoint UTF-8 bytes */ - }else if (((uint8_t )s[i] >>6 ) ==2 ) - *max +=1 ; +/* Non-ASCII */ + }else if (s[i] &0x80 ) { +wchar_t w; +int w_mblen =l_utf8_get_codepoint (&s[i], s_len - i, &w); +if (w_mblen >0 ) { +/* Compensate max bytes */ + *max += w_mblen -wcwidth (w); +/* Skip other bytes of this codepoint */ + i += w_mblen -1 ; + } + } }/* Reached the end of the string within the column bounds */
client/main.c +3 -0
@@ -25,6 +25,7 @@#endif #include <errno.h> +#include <locale.h> #include <signal.h> #include <ell/ell.h>
@@ -50,6 +51,8 @@ int main(int argc, char *argv[]) {bool all_done; +setlocale (LC_CTYPE ,"" ); +if (!l_main_init ())return EXIT_FAILURE ;
Signature · 1 line
2.38.0
Hi Pinghao,
Quoted 24 lines · Pinghao Wu
Find out printing width of wide chars via
wcwidth().
---
client/display.c | 18 ++++++++++++++---- client/main.c | 3 +++2 files changed, 17 insertions(+), 4 deletions(-)client/display.c +1
-0@@ -30,6 +30,7 @@ #include <sys/stat.h> #include <sys/ioctl.h> #include <unistd.h> +#include <wchar.h>
#include <readline/history.h>
#include <readline/readline.h>client/display.c +1
-0@@ -401,18 +402,27 @@ static char* next_line(char *s, unsigned int *max, char **color_out) unsigned int i; int last_space = -1; int last_color = -1; + int s_len = strlen(s);
Quoted 4 lines
/* Find the last space before 'max', as well as any color */ - for (i = 0; i <= *max && s[i] != '\0'; i++) {
+ for (i = 0; i <= *max && i !=s_len; i++) {
I know you didn't write the for-loop originally, but now since you're incrementing i by some value larger than 1 it makes more sense to use a while loop and break out if i > *max. In addition I think you could blow past the i != s_len check.
I think this entire thing could be simplified now with you're changes by putting l_utf8_get_codepoint into a while loop and always calling wcwidth() for each character. Then we aren't distinguishing between ascii and utf-8, just always increment i/max by the needed values (also taking into account the space/color checks obviously).
Quoted 47 lines
if (s[i] == ' ')
last_space = i;
else if (s[i] ==0x1b) {
/* color escape won't count for column width */
*max +=color_end(s + i);last_color= i;
- /* Add width for non-codepoint UTF-8 bytes */ - } else if (((uint8_t)s[i] >> 6) == 2)
- *max += 1;
+ /* Non-ASCII */
+ } else if (s[i] &0x80) {
+ wchar_t w;
+ intw_mblen=l_utf8_get_codepoint(&s[i], s_len - i, &w);
+ if (w_mblen> 0) {
+ /* Compensate max bytes */ + *max +=w_mblen- wcwidth(w); + /* Skip other bytes of this codepoint */
+ i +=w_mblen- 1;
+ }
+ }
}
/* Reached the end of the string within the column bounds */client/main.c +0
-0@@ -25,6 +25,7 @@ #endif
#include <errno.h>
+#include <locale.h>
#include <signal.h>
#include <ell/ell.h>client/main.c +0
-0@@ -50,6 +51,8 @@ int main(int argc, char *argv[]) { bool all_done;
+ setlocale(LC_CTYPE, "");
+
if (!l_main_init())
returnEXIT_FAILURE;