add: dev workflow shenanigans
some shit
This commit is contained in:
+153
-87
@@ -19,24 +19,26 @@
|
||||
#include <wchar.h>
|
||||
#include <wctype.h>
|
||||
|
||||
enum {
|
||||
F_LINES = 1 << 0, /* -l: count '\n' */
|
||||
F_WORDS = 1 << 1, /* -w: whitespace-separated tokens */
|
||||
F_CHARS = 1 << 2, /* -m: multibyte characters */
|
||||
F_BYTES = 1 << 3, /* -c: bytes */
|
||||
enum
|
||||
{
|
||||
F_LINES = 1 << 0, /* -l: count '\n' */
|
||||
F_WORDS = 1 << 1, /* -w: whitespace-separated tokens */
|
||||
F_CHARS = 1 << 2, /* -m: multibyte characters */
|
||||
F_BYTES = 1 << 3, /* -c: bytes */
|
||||
};
|
||||
|
||||
static int flags = 0;
|
||||
|
||||
typedef struct {
|
||||
typedef struct
|
||||
{
|
||||
long long lines;
|
||||
long long words;
|
||||
long long chars;
|
||||
long long bytes;
|
||||
int ok; /* read succeeded */
|
||||
int ok; /* read succeeded */
|
||||
} counts_t;
|
||||
|
||||
static unsigned char ws_tab[256]; /* ws_tab[c] = 1 if c is whitespace */
|
||||
static unsigned char ws_tab[256]; /* ws_tab[c] = 1 if c is whitespace */
|
||||
|
||||
static void init_ws_tab(void)
|
||||
{
|
||||
@@ -46,18 +48,17 @@ static void init_ws_tab(void)
|
||||
|
||||
static void usage(FILE *out)
|
||||
{
|
||||
fprintf(out,
|
||||
"usage: fastwc [-lwc] [-m] [file...]\n"
|
||||
"\n"
|
||||
"Count lines, words, and bytes (default) or selected counts.\n"
|
||||
"With no file, or when file is -, read standard input.\n"
|
||||
"\n"
|
||||
" -l count lines\n"
|
||||
" -w count words\n"
|
||||
" -c count bytes\n"
|
||||
" -m count characters\n"
|
||||
" --help display this help and exit\n"
|
||||
" --version output version information and exit\n");
|
||||
fprintf(out, "usage: fastwc [-lwc] [-m] [file...]\n"
|
||||
"\n"
|
||||
"Count lines, words, and bytes (default) or selected counts.\n"
|
||||
"With no file, or when file is -, read standard input.\n"
|
||||
"\n"
|
||||
" -l count lines\n"
|
||||
" -w count words\n"
|
||||
" -c count bytes\n"
|
||||
" -m count characters\n"
|
||||
" --help display this help and exit\n"
|
||||
" --version output version information and exit\n");
|
||||
}
|
||||
|
||||
/*
|
||||
@@ -75,7 +76,8 @@ static long long count_newlines(const unsigned char *s, size_t n)
|
||||
long long k = 0;
|
||||
size_t i = 0;
|
||||
|
||||
for (; i + 8 <= n; i += 8) {
|
||||
for (; i + 8 <= n; i += 8)
|
||||
{
|
||||
uint64_t x;
|
||||
memcpy(&x, s + i, 8);
|
||||
x = (x ^ nl) & cl;
|
||||
@@ -99,7 +101,8 @@ static long long count_words(const unsigned char *s, size_t n, int *prev_ws)
|
||||
size_t i = 0;
|
||||
int prev = *prev_ws;
|
||||
|
||||
for (; i + 8 <= n; i += 8) {
|
||||
for (; i + 8 <= n; i += 8)
|
||||
{
|
||||
uint8_t m = 0;
|
||||
m |= (uint8_t)ws_tab[s[i + 0]] << 0;
|
||||
m |= (uint8_t)ws_tab[s[i + 1]] << 1;
|
||||
@@ -117,7 +120,8 @@ static long long count_words(const unsigned char *s, size_t n, int *prev_ws)
|
||||
prev = (m >> 7) & 1;
|
||||
}
|
||||
|
||||
for (; i < n; i++) {
|
||||
for (; i < n; i++)
|
||||
{
|
||||
int ws = ws_tab[s[i]];
|
||||
if (prev && !ws)
|
||||
w++;
|
||||
@@ -142,27 +146,34 @@ static void count_stream_mb(FILE *fp, counts_t *c)
|
||||
|
||||
memset(&st, 0, sizeof st);
|
||||
|
||||
while ((nread = fread(buf + pend, 1, sizeof buf - pend, fp)) > 0) {
|
||||
while ((nread = fread(buf + pend, 1, sizeof buf - pend, fp)) > 0)
|
||||
{
|
||||
size_t n = nread + pend;
|
||||
size_t i = 0;
|
||||
|
||||
c->bytes += (long long)nread;
|
||||
|
||||
while (i < n) {
|
||||
while (i < n)
|
||||
{
|
||||
wchar_t wc;
|
||||
size_t r;
|
||||
|
||||
if (buf[i] < 0x80) {
|
||||
if (buf[i] < 0x80)
|
||||
{
|
||||
wc = buf[i];
|
||||
r = 1;
|
||||
} else {
|
||||
}
|
||||
else
|
||||
{
|
||||
r = mbrtowc(&wc, (const char *)buf + i, n - i, &st);
|
||||
if (r == (size_t)-2) { /* incomplete: carry over */
|
||||
if (r == (size_t)-2)
|
||||
{ /* incomplete: carry over */
|
||||
pend = n - i;
|
||||
memmove(buf, buf + i, pend);
|
||||
break;
|
||||
}
|
||||
if (r == (size_t)-1) { /* invalid sequence */
|
||||
if (r == (size_t)-1)
|
||||
{ /* invalid sequence */
|
||||
memset(&st, 0, sizeof st);
|
||||
wc = L'\xfffd';
|
||||
r = 1;
|
||||
@@ -171,9 +182,12 @@ static void count_stream_mb(FILE *fp, counts_t *c)
|
||||
|
||||
if (wc == L'\n')
|
||||
c->lines++;
|
||||
if (iswspace(wc)) {
|
||||
if (iswspace(wc))
|
||||
{
|
||||
prev_ws = 1;
|
||||
} else if (prev_ws) {
|
||||
}
|
||||
else if (prev_ws)
|
||||
{
|
||||
c->words++;
|
||||
prev_ws = 0;
|
||||
}
|
||||
@@ -192,16 +206,21 @@ static void count_stream_mb(FILE *fp, counts_t *c)
|
||||
/* Fast path (-l/-w/-c): one pass, per-chunk memchr + table counting. */
|
||||
static void count_stream(FILE *fp, counts_t *c)
|
||||
{
|
||||
static unsigned char buf[1 << 17]; /* 128 KiB */
|
||||
static unsigned char buf[1 << 17]; /* 128 KiB */
|
||||
size_t nread;
|
||||
int prev_ws = 1; /* start of file: as if preceded by whitespace */
|
||||
int prev_ws = 1; /* start of file: as if preceded by whitespace */
|
||||
|
||||
if (flags & F_CHARS) {
|
||||
if (flags & F_CHARS)
|
||||
{
|
||||
count_stream_mb(fp, c);
|
||||
return;
|
||||
}
|
||||
|
||||
while ((nread = fread(buf, 1, sizeof buf, fp)) > 0) {
|
||||
for (;;)
|
||||
{
|
||||
nread = fread(buf, 1, sizeof buf, fp); /* NOLINT: EOF-state FP */
|
||||
if (nread == 0)
|
||||
break;
|
||||
c->bytes += (long long)nread;
|
||||
if (flags & F_LINES)
|
||||
c->lines += count_newlines(buf, nread);
|
||||
@@ -217,38 +236,53 @@ static void count_file(const char *path, counts_t *c)
|
||||
{
|
||||
FILE *fp;
|
||||
|
||||
if (strcmp(path, "-") == 0) {
|
||||
fp = stdin;
|
||||
} else {
|
||||
fp = fopen(path, "rb");
|
||||
if (fp == NULL) {
|
||||
fprintf(stderr, "fastwc: %s: %s\n", path, strerror(errno));
|
||||
c->ok = 0;
|
||||
return;
|
||||
}
|
||||
if (strcmp(path, "-") == 0)
|
||||
{
|
||||
count_stream(stdin, c);
|
||||
if (ferror(stdin))
|
||||
fprintf(stderr, "fastwc: standard input: read error: %s\n",
|
||||
strerror(errno));
|
||||
return;
|
||||
}
|
||||
|
||||
fp = fopen(path, "rb");
|
||||
if (fp == NULL)
|
||||
{
|
||||
fprintf(stderr, "fastwc: %s: %s\n", path, strerror(errno));
|
||||
c->ok = 0;
|
||||
return;
|
||||
}
|
||||
|
||||
count_stream(fp, c);
|
||||
|
||||
if (ferror(fp))
|
||||
fprintf(stderr, "fastwc: %s: read error: %s\n",
|
||||
strcmp(path, "-") == 0 ? "standard input" : path,
|
||||
strerror(errno));
|
||||
fprintf(stderr, "fastwc: %s: read error: %s\n", path, strerror(errno));
|
||||
|
||||
if (fp != stdin)
|
||||
fclose(fp);
|
||||
fclose(fp);
|
||||
}
|
||||
|
||||
static int col_width(long long v)
|
||||
{
|
||||
int w = 1;
|
||||
while (v >= 10) {
|
||||
while (v >= 10)
|
||||
{
|
||||
v /= 10;
|
||||
w++;
|
||||
}
|
||||
return w;
|
||||
}
|
||||
|
||||
static void widen(int *width, long long v, int enabled)
|
||||
{
|
||||
int w;
|
||||
|
||||
if (!enabled)
|
||||
return;
|
||||
w = col_width(v);
|
||||
if (w > *width)
|
||||
*width = w;
|
||||
}
|
||||
|
||||
int main(int argc, char **argv)
|
||||
{
|
||||
counts_t *rows;
|
||||
@@ -258,29 +292,43 @@ int main(int argc, char **argv)
|
||||
|
||||
init_ws_tab();
|
||||
|
||||
for (a = 1; a < argc; a++) {
|
||||
for (a = 1; a < argc; a++)
|
||||
{
|
||||
const char *arg = argv[a];
|
||||
|
||||
if (arg[0] != '-' || arg[1] == '\0')
|
||||
break; /* first file argument */
|
||||
if (strcmp(arg, "--") == 0) {
|
||||
break; /* first file argument */
|
||||
if (strcmp(arg, "--") == 0)
|
||||
{
|
||||
a++;
|
||||
break;
|
||||
}
|
||||
if (strcmp(arg, "--help") == 0) {
|
||||
if (strcmp(arg, "--help") == 0)
|
||||
{
|
||||
usage(stdout);
|
||||
return 0;
|
||||
}
|
||||
if (strcmp(arg, "--version") == 0) {
|
||||
if (strcmp(arg, "--version") == 0)
|
||||
{
|
||||
printf("fastwc 0.1.0\n");
|
||||
return 0;
|
||||
}
|
||||
for (const char *p = arg + 1; *p; p++) {
|
||||
switch (*p) {
|
||||
case 'l': flags |= F_LINES; break;
|
||||
case 'w': flags |= F_WORDS; break;
|
||||
case 'c': flags |= F_BYTES; break;
|
||||
case 'm': flags |= F_CHARS; break;
|
||||
for (const char *p = arg + 1; *p; p++)
|
||||
{
|
||||
switch (*p)
|
||||
{
|
||||
case 'l':
|
||||
flags |= F_LINES;
|
||||
break;
|
||||
case 'w':
|
||||
flags |= F_WORDS;
|
||||
break;
|
||||
case 'c':
|
||||
flags |= F_BYTES;
|
||||
break;
|
||||
case 'm':
|
||||
flags |= F_CHARS;
|
||||
break;
|
||||
default:
|
||||
fprintf(stderr, "fastwc: invalid option -- '%c'\n", *p);
|
||||
usage(stderr);
|
||||
@@ -290,25 +338,30 @@ int main(int argc, char **argv)
|
||||
}
|
||||
|
||||
if (flags == 0)
|
||||
flags = F_LINES | F_WORDS | F_BYTES; /* wc default: -l -w -c */
|
||||
flags = F_LINES | F_WORDS | F_BYTES; /* wc default: -l -w -c */
|
||||
|
||||
if (flags & F_CHARS)
|
||||
setlocale(LC_CTYPE, "");
|
||||
|
||||
nfiles = argc - a;
|
||||
if (nfiles == 0) {
|
||||
if (nfiles == 0)
|
||||
{
|
||||
rows = calloc(1, sizeof *rows);
|
||||
rows[0].ok = 1;
|
||||
count_stream(stdin, &rows[0]);
|
||||
if (!rows[0].ok) {
|
||||
if (!rows[0].ok)
|
||||
{
|
||||
fprintf(stderr, "fastwc: standard input: read error: %s\n",
|
||||
strerror(errno));
|
||||
failed = 1;
|
||||
}
|
||||
nfiles = 1;
|
||||
} else {
|
||||
}
|
||||
else
|
||||
{
|
||||
rows = calloc((size_t)nfiles, sizeof *rows);
|
||||
for (i = 0; i < nfiles; i++) {
|
||||
for (i = 0; i < nfiles; i++)
|
||||
{
|
||||
rows[i].ok = 1;
|
||||
count_file(argv[a + i], &rows[i]);
|
||||
if (!rows[i].ok)
|
||||
@@ -320,36 +373,49 @@ int main(int argc, char **argv)
|
||||
int wl = 1, ww = 1, wm = 1, wb = 1;
|
||||
long long tl = 0, tw = 0, tm = 0, tb = 0;
|
||||
|
||||
for (i = 0; i < nfiles; i++) {
|
||||
for (i = 0; i < nfiles; i++)
|
||||
{
|
||||
counts_t *r = &rows[i];
|
||||
int x;
|
||||
tl += r->lines; tw += r->words; tm += r->chars; tb += r->bytes;
|
||||
if ((flags & F_LINES) && (x = col_width(r->lines)) > wl) wl = x;
|
||||
if ((flags & F_WORDS) && (x = col_width(r->words)) > ww) ww = x;
|
||||
if ((flags & F_CHARS) && (x = col_width(r->chars)) > wm) wm = x;
|
||||
if ((flags & F_BYTES) && (x = col_width(r->bytes)) > wb) wb = x;
|
||||
tl += r->lines;
|
||||
tw += r->words;
|
||||
tm += r->chars;
|
||||
tb += r->bytes;
|
||||
widen(&wl, r->lines, flags & F_LINES);
|
||||
widen(&ww, r->words, flags & F_WORDS);
|
||||
widen(&wm, r->chars, flags & F_CHARS);
|
||||
widen(&wb, r->bytes, flags & F_BYTES);
|
||||
}
|
||||
if ((flags & F_LINES) && col_width(tl) > wl) wl = col_width(tl);
|
||||
if ((flags & F_WORDS) && col_width(tw) > ww) ww = col_width(tw);
|
||||
if ((flags & F_CHARS) && col_width(tm) > wm) wm = col_width(tm);
|
||||
if ((flags & F_BYTES) && col_width(tb) > wb) wb = col_width(tb);
|
||||
widen(&wl, tl, flags & F_LINES);
|
||||
widen(&ww, tw, flags & F_WORDS);
|
||||
widen(&wm, tm, flags & F_CHARS);
|
||||
widen(&wb, tb, flags & F_BYTES);
|
||||
|
||||
for (i = 0; i < nfiles; i++) {
|
||||
for (i = 0; i < nfiles; i++)
|
||||
{
|
||||
counts_t *r = &rows[i];
|
||||
if (flags & F_LINES) printf("%*lld ", wl, r->lines);
|
||||
if (flags & F_WORDS) printf("%*lld ", ww, r->words);
|
||||
if (flags & F_CHARS) printf("%*lld ", wm, r->chars);
|
||||
if (flags & F_BYTES) printf("%*lld ", wb, r->bytes);
|
||||
if (flags & F_LINES)
|
||||
printf("%*lld ", wl, r->lines);
|
||||
if (flags & F_WORDS)
|
||||
printf("%*lld ", ww, r->words);
|
||||
if (flags & F_CHARS)
|
||||
printf("%*lld ", wm, r->chars);
|
||||
if (flags & F_BYTES)
|
||||
printf("%*lld ", wb, r->bytes);
|
||||
if (argc - a > 0)
|
||||
printf("%s", argv[a + i]);
|
||||
printf("\n");
|
||||
}
|
||||
|
||||
if (argc - a > 1) {
|
||||
if (flags & F_LINES) printf("%*lld ", wl, tl);
|
||||
if (flags & F_WORDS) printf("%*lld ", ww, tw);
|
||||
if (flags & F_CHARS) printf("%*lld ", wm, tm);
|
||||
if (flags & F_BYTES) printf("%*lld ", wb, tb);
|
||||
if (argc - a > 1)
|
||||
{
|
||||
if (flags & F_LINES)
|
||||
printf("%*lld ", wl, tl);
|
||||
if (flags & F_WORDS)
|
||||
printf("%*lld ", ww, tw);
|
||||
if (flags & F_CHARS)
|
||||
printf("%*lld ", wm, tm);
|
||||
if (flags & F_BYTES)
|
||||
printf("%*lld ", wb, tb);
|
||||
printf("total\n");
|
||||
}
|
||||
|
||||
|
||||
Reference in New Issue
Block a user