426 lines
13 KiB
C
426 lines
13 KiB
C
/** @std GNU C11 */
|
|
|
|
#include "lex.h"
|
|
#include <unistd.h> /* chdir (POSIX) */
|
|
#include <sys/types.h> /* mode_t (POSIX) */
|
|
#include <sys/stat.h> /* umask (POSIX) */
|
|
#include <dirent.h> /* opendir readdir closedir */
|
|
#include <limits.h>
|
|
#include <stdio.h>
|
|
#include <stdlib.h>
|
|
#include <stdint.h>
|
|
|
|
#if INT_MAX >= 100000000000
|
|
#error int_to_string requires truncation on this compiler.
|
|
#endif
|
|
static void int_to_string(const int *const n, char (*const a)[12])
|
|
{ sprintf(*a, "%d", *n); }
|
|
#define ARRAY_NAME int
|
|
#define ARRAY_TYPE int
|
|
#define ARRAY_EXPECT_TRAIT
|
|
#include "array.h"
|
|
#define ARRAY_TO_STRING &int_to_string
|
|
#include "array.h"
|
|
static int int_cmp(const int *const a, const int *const b)
|
|
{ return (*b < *a) - (*a < *b); }
|
|
static int void_int_cmp(const void *const a, const void *const b)
|
|
{ return int_cmp(a, b); }
|
|
|
|
#define ARRAY_NAME char
|
|
#define ARRAY_TYPE char
|
|
#include "array.h"
|
|
/** Append a text file, `fn`, to `c`, and add a '\0'.
|
|
@return Success. A partial read is failure. @throws[fopen, fread, malloc]
|
|
@throws[EISEQ] The text file has embedded nulls.
|
|
@throws[ERANGE] If the standard library does not follow POSIX. */
|
|
static int append_file(struct char_array *c, const char *const fn) {
|
|
FILE *fp = 0;
|
|
const size_t granularity = 1024;
|
|
size_t nread;
|
|
char *cursor;
|
|
int success = 0;
|
|
assert(c && fn);
|
|
if(!(fp = fopen(fn, "r"))) goto catch;
|
|
/* Read entire file in chunks. */
|
|
do if(!(cursor = char_array_buffer(c, granularity))
|
|
|| (nread = fread(cursor, 1, granularity, fp), ferror(fp))
|
|
|| !char_array_append(c, nread)) goto catch;
|
|
while(nread == granularity);
|
|
/* File to `C` string. */
|
|
if(!(cursor = char_array_new(c))) goto catch;
|
|
*cursor = '\0';
|
|
/* Binary files with embedded '\0' are not allowed. */
|
|
if(strchr(c->data, '\0') != cursor) { errno = EILSEQ; goto catch; }
|
|
{ success = 1; goto finally; }
|
|
catch:
|
|
if(!errno) errno = EILSEQ; /* Will never be true on POSIX. */
|
|
finally:
|
|
if(fp && fclose(fp)) success = 0;
|
|
return success;
|
|
}
|
|
|
|
/** Is `y` a leap-year? */
|
|
static int leap(int y) {
|
|
assert(y >= 1582);
|
|
if(!(y % 400)) return 1;
|
|
if(!(y % 100)) return 0;
|
|
if(!(y % 4)) return 1;
|
|
return 0;
|
|
}
|
|
/** Assumes: reverse ordering of byte-fields; unsigned is defined; C11 and GNU
|
|
anonymous unions. */
|
|
union date32 {
|
|
uint32_t u32;
|
|
struct { unsigned day : 5, month : 4, year : 23; };
|
|
};
|
|
static int date_mixup(union date32 a, union date32 b) { return a.u32 > b.u32; }
|
|
static void date32_to_string(const union date32 d, char (*const z)[12]) {
|
|
assert(d.year < 10000 && d.month && d.month <= 31 && d.day && d.day <= 31);
|
|
sprintf(*z, "%u-%2.2u-%2.2u", d.year % 10000, d.month, d.day);
|
|
}
|
|
|
|
/** Convert or narrower type or return zero. */
|
|
static union date32 date_to_32(const int y, const int m, const int d) {
|
|
union date32 d32 = { 0 };
|
|
/* Leap year calculations only work at y>=1 and Gregorian Calendar and max
|
|
23 bits. */
|
|
if(y < 1582 || y > 8388607 || m < 1 || m > 12 || d < 1 || d > 31) goto no;
|
|
switch(m) {
|
|
case 1: case 3: case 5: case 7: case 8: case 10: case 12: break;
|
|
case 4: case 6: case 9: case 11: if(d > 30) goto no; break;
|
|
case 2: if(d > 28 + leap(y)) goto no; break;
|
|
default: assert(0); break;
|
|
}
|
|
d32.year = (unsigned)y, d32.month = (unsigned)m, d32.day = (unsigned)d;
|
|
no:
|
|
return d32;
|
|
}
|
|
|
|
/** Tomohiko Sakamoto comp.lang.c 1993-04-10. */
|
|
static unsigned weekday(union date32 d) {
|
|
d.year -= d.month < 3;
|
|
return (d.year + d.year / 4 - d.year / 100 + d.year / 400
|
|
+ "-bed=pen+mad."[d.month] + d.day) % 7;
|
|
}
|
|
|
|
|
|
/* Contained in <lex.h> to share with <lex.re_c.c>. */
|
|
#define ARRAY_NAME lex
|
|
#define ARRAY_TYPE struct lex
|
|
#include "array.h"
|
|
struct page_tree_entry_c;
|
|
static void entry_to_string(struct page_tree_entry_c, char (*)[12]);
|
|
struct page {
|
|
struct char_array entry;
|
|
struct lex_array meaning;
|
|
};
|
|
#define TREE_NAME page
|
|
#define TREE_KEY union date32
|
|
#define TREE_VALUE struct page
|
|
#define TREE_COMPARE &date_mixup
|
|
#define TREE_EXPECT_TRAIT
|
|
#include "tree.h"
|
|
#define TREE_TO_STRING &entry_to_string
|
|
#include "tree.h"
|
|
static void entry_to_string(const struct page_tree_entry_c entry,
|
|
char (*const z)[12]) { date32_to_string(*entry.key, z); }
|
|
|
|
struct source { char *key, *desc; };
|
|
|
|
/*
|
|
### plot with steps
|
|
reset session
|
|
|
|
$Data <<EOD
|
|
1,1,0
|
|
1,2,0
|
|
1,3,0
|
|
1,4,2
|
|
1,5,1
|
|
1,6,3
|
|
1,7,3
|
|
1,8,1
|
|
1,9,3
|
|
1,10,8
|
|
1,11,1
|
|
1,12,0
|
|
1,13,3
|
|
EOD
|
|
|
|
set title "Cumulative count" font ",16"
|
|
set xlabel "episode"
|
|
set ylabel "cumulative count"
|
|
set xtics 1
|
|
set key bottom right
|
|
set grid
|
|
unset border
|
|
|
|
set datafile separator comma
|
|
plot $Data u 2:($3) smooth cumulative with steps lw 2 lc "red" ti "cumulative count"
|
|
### end of code
|
|
*/
|
|
|
|
static int bible_graph(/*const*/ struct page_tree *const journal) {
|
|
enum { CHILL, BOOK, CHAPTER, WORD, NEXT } state = CHILL;
|
|
struct page_tree_entry entry = { 0, 0 };
|
|
struct lex *lex = 0;
|
|
size_t count = 0;
|
|
for(struct page_tree_iterator p_it = page_tree_iterator(journal);
|
|
(entry = page_tree_next(&p_it)).key; ) {
|
|
struct page *const page = entry.value;
|
|
for(struct lex_array_iterator l_it = lex_array_iterator(&page->meaning);
|
|
(lex = lex_array_next(&l_it)); ) {
|
|
switch(lex->symbol) {
|
|
case KJV_BOOK:
|
|
if(state != CHILL && state != WORD) goto catch;
|
|
if(state == WORD) printf("\n");
|
|
fprintf(stderr, "%d-%.2d-%.2d: \"%.*s\", ",
|
|
entry.key->year, entry.key->month, entry.key->day,
|
|
(int)(lex->s1 - lex->s0), lex->s0);
|
|
state = BOOK;
|
|
break;
|
|
case KJV_CHAPTER_VERSE:
|
|
if(state != BOOK) goto catch;
|
|
printf("\"%.*s\", \"", (int)(lex->s1 - lex->s0), lex->s0);
|
|
state = CHAPTER;
|
|
break;
|
|
case KJV_NEXT:
|
|
if(state != WORD) goto catch;
|
|
printf("\"\n");
|
|
break;
|
|
case KJV_TEXT:
|
|
if(state != WORD && state != CHAPTER && state != NEXT)
|
|
goto catch;
|
|
printf("%s%.*s<%d>", state == WORD ? "*" : "",
|
|
(int)(lex->s1 - lex->s0 < 0 ? 10 : lex->s1 - lex->s0), lex->s0, (int)(lex->s1 - lex->s0));
|
|
count++;
|
|
state = WORD;
|
|
break;
|
|
default:
|
|
if(state != CHILL && state != WORD) goto catch;
|
|
if(state == WORD) printf("\"\n"), state = CHILL;
|
|
break;
|
|
}
|
|
}
|
|
if(state != CHILL && state != WORD) goto catch;
|
|
if(state == WORD) printf("\n"), state = CHILL;
|
|
}
|
|
printf("Count: %lu.\n", (unsigned long)count);
|
|
return 1;
|
|
catch:
|
|
fprintf(stderr, "Bible error.\n");
|
|
if(entry.key) {
|
|
fprintf(stderr, "On date: %d-%.2d-%.2d.\n",
|
|
entry.key->year, entry.key->month, entry.key->day);
|
|
if(lex) fprintf(stderr, "At line %lu.\n", (unsigned long)lex->line);
|
|
}
|
|
errno = EILSEQ;
|
|
return 0;
|
|
}
|
|
|
|
#define C_BLACK "\033[0;30m"
|
|
#define C_RED "\033[0;31m"
|
|
#define C_GREEN "\033[0;32m"
|
|
#define C_YELLOW "\033[0;33m"
|
|
#define C_BLUE "\033[0;34m"
|
|
#define C_PURPLE "\033[0;35m"
|
|
#define C_CYAN "\033[0;36m"
|
|
#define C_WHITE "\033[0;37m"
|
|
|
|
#define CB_BLACK "\033[1;30m"
|
|
#define CB_RED "\033[1;31m"
|
|
#define CB_GREEN "\033[1;32m"
|
|
#define CB_YELLOW "\033[1;33m"
|
|
#define CB_BLUE "\033[1;34m"
|
|
#define CB_PURPLE "\033[1;35m"
|
|
#define CB_CYAN "\033[1;36m"
|
|
#define CB_WHITE "\033[1;37m"
|
|
|
|
#define C_RESET "\033[0m"
|
|
|
|
int main(int argc, char **argv) {
|
|
int success = EXIT_SUCCESS;
|
|
char *intent = 0;
|
|
|
|
/* For reading in files, overwritten. */
|
|
DIR *dir = 0;
|
|
struct dirent *de;
|
|
struct int_array years = int_array(), months = int_array(),
|
|
days = int_array();
|
|
int *y, *y_end;
|
|
|
|
struct page_tree journal = page_tree();
|
|
|
|
errno = 0;
|
|
if(argc != 2) { intent = "needs journal location, which should"
|
|
" contain <year>/<month>/<day>.txt"; goto catch; }
|
|
|
|
/* Get the years list as directories matching a year. */
|
|
if(chdir(argv[1]) == -1 || !(dir = opendir("."))) goto catch;
|
|
while((de = readdir(dir))) {
|
|
struct stat st;
|
|
int year, *p;
|
|
if(!lex_looks_like_year(de->d_name, &year)) continue;
|
|
if(stat(de->d_name, &st)) goto catch;
|
|
if(!S_ISDIR(st.st_mode)) continue;
|
|
if(!(p = int_array_new(&years))) goto catch;
|
|
*p = year;
|
|
}
|
|
closedir(dir), dir = 0;
|
|
/* Sort the years for sensible ordering of parsing. */
|
|
qsort(years.data, years.size, sizeof *years.data, &void_int_cmp);
|
|
fprintf(stderr, "Years in <<%s>>: %s.\n",
|
|
argv[1], int_array_to_string(&years));
|
|
|
|
/* Go though each year. */
|
|
for(y = years.data, y_end = y + years.size; y < y_end; y++) {
|
|
char fn[64];
|
|
int *m, *m_end;
|
|
sprintf(fn, "%d", *y);
|
|
|
|
/* Get the months as directories. */
|
|
if(chdir(fn) == -1 || !(dir = opendir("."))) goto catch;
|
|
while((de = readdir(dir))) {
|
|
struct stat st;
|
|
int month, *p;
|
|
if(!(month = lex_looks_like_month(de->d_name))) continue;
|
|
if(stat(de->d_name, &st)) goto catch;
|
|
if(!S_ISDIR(st.st_mode)) continue;
|
|
if(!(p = int_array_new(&months))) goto catch;
|
|
*p = month;
|
|
}
|
|
closedir(dir), dir = 0;
|
|
qsort(months.data, months.size, sizeof *months.data, &void_int_cmp);
|
|
fprintf(stderr, "Months in <<%s>>: %s.)\n",
|
|
fn, int_array_to_string(&months));
|
|
|
|
/* Go though each month. */
|
|
for(m = months.data, m_end = m + months.size; m < m_end; m++) {
|
|
int *d, *d_end;
|
|
sprintf(fn, "%.2d", *m);
|
|
|
|
/* Get the days as files. */
|
|
if(chdir(fn) == -1 || !(dir = opendir("."))) goto catch;
|
|
while((de = readdir(dir))) {
|
|
struct stat st;
|
|
int day, *p;
|
|
/* fixme: Have yyyy-mm-dd to figure out how many days. */
|
|
if(!(day = lex_looks_like_day(de->d_name))) continue;
|
|
if(stat(de->d_name, &st)) goto catch;
|
|
if(S_ISDIR(st.st_mode)) continue;
|
|
if(!(p = int_array_new(&days))) goto catch;
|
|
*p = day;
|
|
}
|
|
closedir(dir), dir = 0;
|
|
qsort(days.data, days.size, sizeof *days.data, &void_int_cmp);
|
|
fprintf(stderr, "Days in <<%s>>: %s.\n",
|
|
fn, int_array_to_string(&days));
|
|
|
|
for(d = days.data, d_end = d + days.size; d < d_end; d++) {
|
|
struct lex *lex = 0;
|
|
struct page *page = 0;
|
|
union date32 d32;
|
|
if(!(d32 = date_to_32(*y, *m, *d)).year) { errno = EILSEQ;
|
|
intent = "date parse error"; goto syntax; }
|
|
sprintf(fn, "%.2d.txt", *d);
|
|
if(page_tree_bulk_add(&journal, d32, &page) != TREE_UNIQUE) {
|
|
if(!errno) intent = "not unique", errno = EDOM;
|
|
goto syntax;
|
|
}
|
|
page->entry = char_array();
|
|
page->meaning = lex_array();
|
|
if(!append_file(&page->entry, fn)) goto syntax;
|
|
int first = 1;
|
|
for(lex_reset(page->entry.data); ; ) {
|
|
if(!(lex = lex_array_new(&page->meaning))) goto syntax;
|
|
if(!lex_next(lex)) {
|
|
if(lex->symbol != END) { errno = EILSEQ; goto syntax; }
|
|
break; /* Terminated successfully. */
|
|
}
|
|
switch(lex->symbol) {
|
|
case TEXT: printf("%s%.*s",
|
|
first ? "" : " ", (int)(lex->s1 - lex->s0), lex->s0);
|
|
first = 0; break;
|
|
case PARAGRAPH: printf("\n" C_RESET); break;
|
|
case KJV_BOOK: printf(C_YELLOW "%.*s",
|
|
(int)(lex->s1 - lex->s0), lex->s0); break;
|
|
case KJV_CHAPTER_VERSE: printf(" ch. %.*s",
|
|
(int)(lex->s1 - lex->s0), lex->s0); break;
|
|
case KJV_TEXT: printf("%s%.*s",
|
|
(int)(lex->s1 - lex->s0), lex->s0); break;
|
|
case KJV_NEXT: printf("(next)\n"); break;
|
|
default:
|
|
fprintf(stderr, "%lu: %s",
|
|
(unsigned long)lex->line, lex_symbols[lex->symbol]);
|
|
if(lex->s0 && lex->s1) {
|
|
if(lex->s0 + INT_MAX < lex->s1)
|
|
intent = "line too long", errno = EILSEQ;
|
|
else
|
|
fprintf(stderr, " <<%.*s>>",
|
|
(int)(lex->s1 - lex->s0), lex->s0);
|
|
}
|
|
fprintf(stderr, ".\n");
|
|
break;
|
|
}
|
|
}
|
|
continue;
|
|
syntax:
|
|
fprintf(stderr, "On date: %d-%.2d-%.2d.\n", *y, *m, *d);
|
|
if(!page) goto catch;
|
|
if(!lex) { fprintf(stderr, "While parsing <<<\n%s>>>.\n",
|
|
page->entry.data); goto catch; }
|
|
for(struct lex_array_iterator it
|
|
= lex_array_iterator(&page->meaning);
|
|
lex = lex_array_next(&it); ) {
|
|
fprintf(stderr, "%lu: %s",
|
|
(unsigned long)lex->line, lex_symbols[lex->symbol]);
|
|
if(lex->s0 && lex->s1) {
|
|
if(lex->s0 + INT_MAX < lex->s1)
|
|
intent = "line too long", errno = EILSEQ;
|
|
else
|
|
fprintf(stderr, " <<%.*s>>",
|
|
(int)(lex->s1 - lex->s0), lex->s0);
|
|
}
|
|
fprintf(stderr, ".\n");
|
|
}
|
|
goto catch;
|
|
}
|
|
|
|
int_array_clear(&days);
|
|
if(chdir("..") == -1) goto catch;
|
|
}
|
|
|
|
int_array_clear(&months);
|
|
if(chdir("..") == -1) goto catch;
|
|
/* fixme: Expand, contact is the next thing that it doesn't get. */
|
|
if(*y == 1993/*1996*/) break;
|
|
}
|
|
page_tree_bulk_finish(&journal);
|
|
int_array_(&years), int_array_(&months), int_array_(&days);
|
|
fprintf(stderr, "Journal has entries: %s\n", page_tree_to_string(&journal));
|
|
|
|
/* Do something interesting? */
|
|
if(!bible_graph(&journal)) goto catch;
|
|
|
|
goto finally;
|
|
catch:
|
|
success = EXIT_FAILURE;
|
|
perror("interpret");
|
|
if(intent) fprintf(stderr, "Further explanation: %s.\n", intent);
|
|
finally:
|
|
if(dir && closedir(dir)) success = EXIT_FAILURE, perror("dir");
|
|
int_array_(&years), int_array_(&months), int_array_(&days);
|
|
struct page_tree_entry entry;
|
|
for(struct page_tree_iterator it = page_tree_iterator(&journal);
|
|
(entry = page_tree_next(&it)).key; ) {
|
|
struct page *const page = entry.value;
|
|
char z[12];
|
|
date32_to_string(*entry.key, &z);
|
|
lex_array_(&page->meaning);
|
|
char_array_(&page->entry);
|
|
}
|
|
return success;
|
|
}
|
|
|
|
/* Popular KJV 738137 words. */
|