cirno/plan9port
1
0
Fork 0
mirror of https://github.com/9fans/plan9port.git synced 2025-01-12 11:10:07 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

ed: handle Unicode beyond the BMP correctly in list mode.

Browse source 
List mode was constrained to the BMP. This change introduces
the following new list mode convention, using Go string literal syntax:

Non-printing ASCII characters display as \xhh.
Non-ASCII characters in the BMP display as \uhhhh.
Characters beyond the BMP display as \Uhhhhhhhh.
This commit is contained in:
sean 2020-05-21 16:10:30 +01:00 committed by Russ Cox
parent 3850e6e177
commit 95220bf887
2 changed files with 42 additions and 11 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
man/man1/ed.1
View file

@ -441,10 +441,18 @@ a backspace as
.LR \eb , .LR \eb ,
backslashes as backslashes as
.LR \e\e , .LR \e\e ,
and non-printing characters as and non-printing ASCII characters as
a backslash, an a backslash, an
.LR x , .LR x ,
and four hexadecimal digits. and two hexadecimal digits.
non-ASCII characters in the Basic Multilingual Plane
are printed as a backslash, a small
.LR u ,
and four hexadecimal digits; and characters above the
Basic Multilingual Plane are printed as a backslash,
a big
.LR U ,
and six hexadecimal digits.
Long lines are folded, Long lines are folded,
with the second and subsequent sub-lines indented one tab stop. with the second and subsequent sub-lines indented one tab stop.
If the last character in the line is a blank, If the last character in the line is a blank,

41
src/cmd/ed.c
View file

@ -21,6 +21,12 @@ enum
EOF = -1 EOF = -1
}; };
enum
{
LINELEN = 70, /* max number of glyphs in a display line */
BELL = 6 /* A char could require up to BELL glyphs to display */
};
void (*oldhup)(int); void (*oldhup)(int);
void (*oldquit)(int); void (*oldquit)(int);
int* addr1; int* addr1;
@ -40,7 +46,7 @@ int ichanged;
int io; int io;
Biobuf iobuf; Biobuf iobuf;
int lastc; int lastc;
char line[70]; char line[LINELEN];
Rune* linebp; Rune* linebp;
Rune linebuf[LBSIZE]; Rune linebuf[LBSIZE];
int listf; int listf;
@ -1543,7 +1549,7 @@ putchr(int ac)
*lp++ = 'n'; *lp++ = 'n';
} }
} else { } else {
if(col > (72-6-2)) { if(col > (LINELEN-BELL)) {
col = 8; col = 8;
*lp++ = '\\'; *lp++ = '\\';
*lp++ = '\n'; *lp++ = '\n';
@ -1558,15 +1564,32 @@ putchr(int ac)
if(c == '\t') if(c == '\t')
c = 't'; c = 't';
col++; col++;
} else } else if (c<' ' || c=='\177') {
if(c<' ' || c>='\177') {
*lp++ = '\\'; *lp++ = '\\';
*lp++ = 'x'; *lp++ = 'x';
*lp++ = hex[c>>12]; *lp++ = hex[(c>>4)&0xF];
*lp++ = hex[c>>8&0xF]; c = hex[c&0xF];
*lp++ = hex[c>>4&0xF]; col += 3;
c = hex[c&0xF]; } else if (c>'\177' && c<=0xFFFF) {
*lp++ = '\\';
*lp++ = 'u';
*lp++ = hex[(c>>12)&0xF];
*lp++ = hex[(c>>8)&0xF];
*lp++ = hex[(c>>4)&0xF];
c = hex[c&0xF];
col += 5; col += 5;
} else if (c>0xFFFF) {
*lp++ = '\\';
*lp++ = 'U';
*lp++ = hex[(c>>28)&0xF];
*lp++ = hex[(c>>24)&0xF];
*lp++ = hex[(c>>20)&0xF];
*lp++ = hex[(c>>16)&0xF];
*lp++ = hex[(c>>12)&0xF];
*lp++ = hex[(c>>8)&0xF];
*lp++ = hex[(c>>4)&0xF];
c = hex[c&0xF];
col += 9;
} }
} }
} }
@ -1574,7 +1597,7 @@ putchr(int ac)
rune = c; rune = c;
lp += runetochar(lp, &rune); lp += runetochar(lp, &rune);
if(c == '\n' || lp >= &line[sizeof(line)-5]) { if(c == '\n' || lp >= &line[LINELEN-BELL]) {
linp = line; linp = line;
write(oflag? 2: 1, line, lp-line); write(oflag? 2: 1, line, lp-line);
return; return;