From 2b999e614c48f084f6a14d589d434904aec6d491 Mon Sep 17 00:00:00 2001 From: Rob Landley Date: Tue, 15 Mar 2016 02:23:50 -0500 Subject: Add ls -b and make ls -q work with utf8. --- lib/lib.h | 8 +++- lib/linestack.c | 92 ++++++++++++++++++++++++++++---------------- toys/posix/ls.c | 117 +++++++++++++++++++++++++++++++++----------------------- 3 files changed, 134 insertions(+), 83 deletions(-) diff --git a/lib/lib.h b/lib/lib.h index b12a110b..ef5b5979 100644 --- a/lib/lib.h +++ b/lib/lib.h @@ -223,11 +223,15 @@ void linestack_addstack(struct linestack **lls, struct linestack *throw, void linestack_insert(struct linestack **lls, long pos, char *line, long len); void linestack_append(struct linestack **lls, char *line); struct linestack *linestack_load(char *name); -int crunch_str(char **str, int width, FILE *out, - int (*escout)(FILE *out, int cols, char **buf)); +int crunch_escape(FILE *out, int cols, int wc); +int crunch_rev_escape(FILE *out, int cols, int wc); +int crunch_str(char **str, int width, FILE *out, char *escmore, + int (*escout)(FILE *out, int cols, wchar_t wc)); int draw_str(char *start, int width); int utf8len(char *str); int utf8skip(char *str, int width); +int draw_trim_esc(char *str, int padto, int width, char *escmore, + int (*escout)(FILE *out, int cols,int wc)); int draw_trim(char *str, int padto, int width); // interestingtimes.c diff --git a/lib/linestack.c b/lib/linestack.c index f8ca32a4..eef790b5 100644 --- a/lib/linestack.c +++ b/lib/linestack.c @@ -81,56 +81,75 @@ struct linestack *linestack_load(char *name) // if escout, send it unprintable chars, returns columns output or -1 for // standard escape: ^X if <32, if invliad UTF8, U+XXXX if UTF8 !iswprint() // Returns width in columns, moves *str to end of data consumed. -int crunch_str(char **str, int width, FILE *out, - int (*escout)(FILE *out, int cols, char **buf)) +int crunch_str(char **str, int width, FILE *out, char *escmore, + int (*escout)(FILE *out, int cols, int wc)) { int columns = 0, col, bytes; char *start, *end; - for (end = start = *str; *end;) { - wchar_t wc = *end; + for (end = start = *str; *end; columns += col, end += bytes) { + wchar_t wc; - bytes = 0; - if (*end >= ' ' && (bytes = mbrtowc(&wc, end, 99,0))>0 - && (col = wcwidth(wc))>=0) + if ((bytes = mbrtowc(&wc, end, MB_CUR_MAX, 0))>0 && (col = wcwidth(wc))>=0) { - if (width-columns(col = escout(out, width-columns, &end))) { - char buf[32]; - - tty_esc("7m"); - if (*end < ' ') { - bytes = 1; - sprintf(buf, "^%c", '@'+*end); - } else if (bytes<1) { - bytes = 1; - sprintf(buf, "<%02X>", *end); - } else sprintf(buf, "U+%04X", (unsigned)wc); - col = strlen(buf); - if (width-columns255 || !strchr(escmore, wc)) { + if (width-columns", wc); + else rc = sprintf(buf, "U+%04X", wc); + + if (rc > cols) buf[rc = cols] = 0; + if (out) fputs(buf, out); + + return rc; +} + +int crunch_rev_escape(FILE *out, int cols, int wc) +{ + int rc; + + tty_esc("7m"); + rc = crunch_escape(out, cols, wc); + tty_esc("27m"); + + return rc; +} + // Write width chars at start of string to strdout with standard escapes // Returns length in columns so caller can pad it out with spaces. int draw_str(char *start, int width) { - return crunch_str(&start, width, stdout, 0); + return crunch_str(&start, width, stdout, 0, crunch_rev_escape); } // Return utf8 columns int utf8len(char *str) { - return crunch_str(&str, INT_MAX, 0, 0); + return crunch_str(&str, INT_MAX, 0, 0, crunch_rev_escape); } // Return bytes used by (up to) this many columns @@ -138,14 +157,15 @@ int utf8skip(char *str, int width) { char *s = str; - crunch_str(&s, width, 0, 0); + crunch_str(&s, width, 0, 0, crunch_rev_escape); return s-str; } -// Print utf8 to stdout with standard escapes,trimmed to width and padded +// Print utf8 to stdout with standard escapes, trimmed to width and padded // out to padto. If padto<0 left justify. Returns columns printed -int draw_trim(char *str, int padto, int width) +int draw_trim_esc(char *str, int padto, int width, char *escmore, + int (*escout)(FILE *out, int cols, int wc)) { int apad = abs(padto), len = utf8len(str); @@ -154,8 +174,14 @@ int draw_trim(char *str, int padto, int width) // Left pad if right justified if (padto>0 && apad>len) printf("%*s", apad-len, ""); - crunch_str(&str, len, stdout, 0); + crunch_str(&str, len, stdout, 0, crunch_rev_escape); if (padto<0 && apad>len) printf("%*s", apad-len, ""); return (apad > len) ? apad : len; } + +// draw_trim_esc() with default escape +int draw_trim(char *str, int padto, int width) +{ + return draw_trim_esc(str, padto, width, 0, 0); +} diff --git a/toys/posix/ls.c b/toys/posix/ls.c index 799631b1..bec76e99 100644 --- a/toys/posix/ls.c +++ b/toys/posix/ls.c @@ -5,7 +5,7 @@ * * See http://opengroup.org/onlinepubs/9699919799/utilities/ls.html -USE_LS(NEWTOY(ls, USE_LS_COLOR("(color):;")"ZgoACFHLRSacdfhiklmnpqrstux1[-Cxm1][-Cxml][-Cxmo][-Cxmg][-cu][-ftS][-HL]", TOYFLAG_BIN|TOYFLAG_LOCALE)) +USE_LS(NEWTOY(ls, USE_LS_COLOR("(color):;")"ZgoACFHLRSabcdfhiklmnpqrstux1[-Cxm1][-Cxml][-Cxmo][-Cxmg][-cu][-ftS][-HL][!qb]", TOYFLAG_BIN|TOYFLAG_LOCALE)) config LS bool "ls" @@ -16,24 +16,24 @@ config LS list files what to show: - -a all files including .hidden -c use ctime for timestamps - -d directory, not contents -i inode number - -k block sizes in kilobytes -p put a '/' after dir names - -q unprintable chars as '?' -s size (in blocks) - -u use access time for timestamps -A list all files but . and .. - -H follow command line symlinks -L follow symlinks - -R recursively list files in subdirs -F append /dir *exe @sym |FIFO - -Z security context + -a all files including .hidden -b escape nongraphic chars + -c use ctime for timestamps -d directory, not contents + -i inode number -k block sizes in kilobytes + -p put a '/' after dir names -q unprintable chars as '?' + -s size (in blocks) -u use access time for timestamps + -A list all files but . and .. -H follow command line symlinks + -L follow symlinks -R recursively list files in subdirs + -F append /dir *exe @sym |FIFO -Z security context output formats: - -1 list one file per line -C columns (sorted vertically) - -g like -l but no owner -h human readable sizes - -l long (show full details) -m comma separated - -n like -l but numeric uid/gid -o like -l but no group - -x columns (horizontal sort) + -1 list one file per line -C columns (sorted vertically) + -g like -l but no owner -h human readable sizes + -l long (show full details) -m comma separated + -n like -l but numeric uid/gid -o like -l but no group + -x columns (horizontal sort) sorting (default is alphabetical): - -f unsorted -r reverse -t timestamp -S size + -f unsorted -r reverse -t timestamp -S size config LS_COLOR bool "ls --color" @@ -61,34 +61,53 @@ GLOBALS( unsigned screen_width; int nl_title; - char uid_buf[12], gid_buf[12]; + char uid_buf[12], gid_buf[12], *escmore; ) -// Does two things: 1) Returns wcwidth(utf8) version of strlen, -// 2) replaces unprintable characters input string with '?' wildcard char. -int strwidth(char *s) +// Callback from crunch_str to represent unprintable chars +int crunch_qb(FILE *out, int cols, int wc) { - int total = 0, width, len; - wchar_t c; - - if (!CFG_TOYBOX_I18N) { - total = strlen(s); - if (toys.optflags & FLAG_q) for (; *s; s++) if (!isprint(*s)) *s = '?'; - } else while (*s) { - len = mbrtowc(&c, s, MB_CUR_MAX, 0); - if (len < 1 || (width = wcwidth(c)) < 0) { - total++; - if (toys.optflags & FLAG_q) *s = '?'; - s++; - } else { - s += len; - total += width; + unsigned len = 1; + char buf[32]; + + if (toys.optflags&FLAG_q) *buf = '?'; + else { + if (wc<256) *buf = wc; + // scrute the inscrutable, eff the ineffable, print the unprintable + else len = wcrtomb(buf, wc, 0); + if (toys.optflags&FLAG_b) { + char *to = buf, *from = buf+24; + int i, j; + + memcpy(from, to, 8); + for (i = 0; ist_mode; @@ -292,7 +311,6 @@ static void listfiles(int dirfd, struct dirtree *indir) char tmp[64]; if (-1 == dirfd) { - strwidth(indir->name); perror_msg_raw(indir->name); return; @@ -395,11 +413,12 @@ static void listfiles(int dirfd, struct dirtree *indir) memset(toybuf, ' ', 256); width = 0; for (ul = 0; ulst); mode_t mode = st->st_mode; - char et = endtype(st); + char et = endtype(st), *ss; // Skip directories at the top of the tree when -d isn't set if (S_ISDIR(mode) && !indir->parent && !(flags & FLAG_d)) continue; @@ -428,7 +447,6 @@ static void listfiles(int dirfd, struct dirtree *indir) if (flags & (FLAG_l|FLAG_o|FLAG_n|FLAG_g)) { struct tm *tm; - char *ss; // (long) is to coerce the st types into something we know we can print. mode_to_string(mode, tmp); @@ -436,16 +454,20 @@ static void listfiles(int dirfd, struct dirtree *indir) // print user if (!(flags&FLAG_g)) { - if (flags&FLAG_n) sprintf(ss = tmp, "%u", (unsigned)st->st_uid); - else strwidth(ss = getusername(st->st_uid)); - printf(" %-*s", (int)totals[3], ss); + putchar(' '); + ii = -totals[3]; + if (flags&FLAG_n) printf("%*u", ii, (unsigned)st->st_uid); + else draw_trim_esc(getusername(st->st_uid), ii, abs(ii), TT.escmore, + crunch_qb); } // print group if (!(flags&FLAG_o)) { - if (flags&FLAG_n) sprintf(ss = tmp, "%u", (unsigned)st->st_gid); - else strwidth(ss = getgroupname(st->st_gid)); - printf(" %-*s", (int)totals[4], ss); + putchar(' '); + ii = -totals[4]; + if (flags&FLAG_n) printf("%*u", ii, (unsigned)st->st_gid); + else draw_trim_esc(getgroupname(st->st_gid), ii, abs(ii), TT.escmore, + crunch_qb); } if (flags & FLAG_Z) @@ -472,10 +494,8 @@ static void listfiles(int dirfd, struct dirtree *indir) if (color) printf("\033[%d;%dm", color>>8, color&255); } - if (flags & FLAG_q) { - char *p; - for (p=sort[next]->name; *p; p++) fputc(isprint(*p) ? *p : '?', stdout); - } else xprintf("%s", sort[next]->name); + ss = sort[next]->name; + crunch_str(&ss, INT_MAX, stdout, TT.escmore, crunch_qb); if (color) xprintf("\033[0m"); if ((flags & (FLAG_l|FLAG_o|FLAG_n|FLAG_g)) && S_ISLNK(mode)) { @@ -526,6 +546,7 @@ void ls_main(void) TT.screen_width = 80; terminal_size(&TT.screen_width, NULL); if (TT.screen_width<2) TT.screen_width = 2; + if (toys.optflags&FLAG_b) TT.escmore = " \\"; // Do we have an implied -1 if (!isatty(1)) { -- cgit v1.2.3