/** @license 2022 Neil Edelman, distributed under the terms of the
 [MIT License](https://opensource.org/licenses/MIT).

 Scan journal entries for kjv references. */

#include "../src/source.h"
#include "../src/kjv.h"
#include "../src/pair.h"
#include <inttypes.h> /* C99 */
#include <stdlib.h>
#include <stdio.h>
#include <errno.h>
#include <assert.h>
#include <limits.h>


static void kjvline_to_string(const union line64 line, const struct kjvrange *u,
	char (*const a)[12]) { (void)u; date32_to_string(line.date, a); }
static int kjvline_compare(const union line64 a, const union line64 b)
	{ return a.u64 > b.u64; }
#define TREE_NAME kjvline
#define TREE_KEY union line64
#define TREE_VALUE struct kjvrange
#define TREE_COMPARE
#define TREE_TO_STRING
#define TREE_BODY
#include "../src/tree.h"


/*!conditions:re2c*/

static int scan(union date32 date, const char *const buffer,
	struct kjvline_tree *const lines) {
	const char *YYCURSOR, *YYMARKER, *yyt1, *yyt2, *yyt3, *s0, *s1, *t0, *t1;
	enum kjv_book book = Revelation;
	uint32_t chapter = 0, verse = 0, verse_end = 0;
	enum YYCONDTYPE condition = yycline;
	size_t line = 1;
	const char *why = "unexpected";
	assert(buffer && lines);
	YYCURSOR = YYMARKER = yyt1 = buffer;
	/*!re2c /**/
	re2c:define:YYCTYPE = char;
	re2c:yyfill:enable = 0;
	re2c:define:YYGETCONDITION = "condition";
	re2c:define:YYSETCONDITION = "condition = @@;";
	re2c:define:YYGETCONDITION:naked = 1;
	re2c:define:YYSETCONDITION:naked = 1;

	unix_control = [\x01-\x08\x0b-\x1f\x7f];
	ws = [ \t];
	glyph = [^] \ ("\x00" | "\n" | unix_control | ws);
	natural = [1-9][0-9]*;
	engage = ws+ "--" ws+;
	/* (natural ":")? Don't use for memorizing and use for reading, I think? */
	/*("``"|"\"") This is not in the next book. */
	lookat = ws* natural ":" natural [ab]? ("-" natural [ab]?)? engage;
	first = ("I" | "1") " "?;
	second = ("II" | "2") " "?;
	third = ("III" | "3") " "?;
	*/
	for( ; ; ) { /*!re2c /**/
		/* Default ignore. */
		<skip> [^\n\x00] { continue; }
		<skip> "\x00" { why = "no newline at end of file"; goto catch; }
		<line> "\x00" { return 1; }
		<line, skip> "\n" => line { line++; continue; }
		<line> * :=> skip

		/* Books. */
		<line> "Genesis" / lookat => book { book = Genesis; continue; }
		<line> "Exodus" / lookat => book { book = Exodus; continue; }
		<line> "Leviticus" / lookat => book { book = Leviticus; continue; }
		<line> "Numbers" / lookat => book { book = Numbers; continue; }
		<line> "Deuteronomy" / lookat => book { book = Deuteronomy; continue; }
		<line> "Joshua" / lookat => book { book = Joshua; continue; }
		<line> "Judges" / lookat => book { book = Judges; continue; }
		<line> "Ruth" / lookat => book { book = Ruth; continue; }
		<line> first "Samuel" / lookat => book { book = ISamuel; continue; }
		<line> second "Samuel" / lookat => book { book = IISamuel; continue; }
		<line> first "Kings" / lookat => book { book = IKings; continue; }
		<line> second "Kings" / lookat => book { book = IIKings; continue; }
		<line> first "Chronicles" / lookat
			=> book { book = IChronicles; continue; }
		<line> second "Chronicles" / lookat
			=> book { book = IIChronicles; continue; }
		<line> "Ezra" / lookat => book { book = Ezra; continue; }
		<line> "Nehemiah" / lookat => book { book = Nehemiah; continue; }
		<line> "Esther" / lookat => book { book = Esther; continue; }
		<line> "Job" / lookat => book { book = Job; continue; }
		<line> "Psalms" / lookat => book { book = Psalms; continue; }
		<line> "Proverbs" / lookat => book { book = Proverbs; continue; }
		<line> "Ecclesiastes" / lookat
			=> book { book = Ecclesiastes; continue; }
		<line> "Song of Solomon" / lookat
			=> book { book = Song_of_Solomon; continue; }
		<line> "Isaiah" / lookat => book { book = Isaiah; continue; }
		<line> "Jeremiah" / lookat => book { book = Jeremiah; continue; }
		<line> "Lamentations" / lookat
			=> book { book = Lamentations; continue; }
		<line> "Ezekiel" / lookat => book { book = Ezekiel; continue; }
		<line> "Daniel" / lookat => book { book = Daniel; continue; }
		<line> "Hosea" / lookat => book { book = Hosea; continue; }
		<line> "Joel" / lookat => book { book = Joel; continue; }
		<line> "Amos" / lookat => book { book = Amos; continue; }
		<line> "Obadiah" / lookat => book { book = Obadiah; continue; }
		<line> "Jonah" / lookat => book { book = Jonah; continue; }
		<line> "Micah" / lookat => book { book = Micah; continue; }
		<line> "Nahum" / lookat => book { book = Nahum; continue; }
		<line> "Habakkuk" / lookat => book { book = Habakkuk; continue; }
		<line> "Zephaniah" / lookat => book { book = Zephaniah; continue; }
		<line> "Haggai" / lookat => book { book = Haggai; continue; }
		<line> "Zechariah" / lookat => book { book = Zechariah; continue; }
		<line> "Malachi" / lookat => book { book = Malachi; continue; }
		<line> "Matthew" / lookat => book { book = Matthew; continue; }
		<line> "Mark" / lookat => book { book = Mark; continue; }
		<line> "Luke" / lookat => book { book = Luke; continue; }
		<line> "John" / lookat => book { book = John; continue; }
		<line> "Acts" / lookat => book { book = Acts; continue; }
		<line> "Romans" / lookat => book { book = Romans; continue; }
		<line> first "Corinthians" / lookat
			=> book { book = ICorinthians; continue; }
		<line> second "Corinthians" / lookat
			=> book { book = IICorinthians; continue; }
		<line> "Galatians" / lookat => book { book = Galatians; continue; }
		<line> "Ephesians" / lookat => book { book = Ephesians; continue; }
		<line> "Philippians" / lookat => book { book = Philippians; continue; }
		<line> "Colossians" / lookat => book { book = Colossians; continue; }
		<line> first "Thessalonians" / lookat
			=> book { book = IThessalonians; continue; }
		<line> second "Thessalonians" / lookat
			=> book { book = IIThessalonians; continue; }
		<line> first "Timothy" / lookat => book { book = ITimothy; continue; }
		<line> second "Timothy" / lookat => book { book = IITimothy; continue; }
		<line> "Titus" / lookat => book { book = Titus; continue; }
		<line> "Philemon" / lookat => book { book = Philemon; continue; }
		<line> "Hebrews" / lookat => book { book = Hebrews; continue; }
		<line> "James" / lookat => book { book = James; continue; }
		<line> first "Peter" / lookat => book { book = IPeter; continue; }
		<line> second "Peter" / lookat => book { book = IIPeter; continue; }
		<line> first "John" / lookat => book { book = IJohn; continue; }
		<line> second "John" / lookat => book { book = IIJohn; continue; }
		<line> third "John" / lookat => book { book = IIIJohn; continue; }
		<line> "Jude" / lookat => book { book = Jude; continue; }
		<line> "Revelation" / lookat => book { book = Revelation; continue; }

		/* Extract further information. */
		<book> * { why = "default unrecognized"; goto catch; }
		/* 19:15a, just ignore the a. */
		<book> ws+ @s0 natural @s1 ":" @t0 natural @t1 [ab]? {
			if(chapter || verse || verse_end)
				{ why = "reference unrecognized"; goto catch; }
			if(!pair_to_natural(s0, s1, &chapter)
				|| !pair_to_natural(t0, t1, &verse))
				{ why = "reference numerical error"; goto catch; }
			continue;
		}
		<book> "-" @s0 natural @s1 [ab]? { /* Verse range. */
			if(!chapter || !verse || verse_end)
				{ why = "range unrecognized"; goto catch; }
			if(!pair_to_natural(s0, s1, &verse_end))
				{ why = "range numerical error"; goto catch; }
			continue;
		}
		<book> engage => skip {
			if(!chapter || !verse) { why = "missing information"; goto catch; }
			if(verse_end && verse_end <= verse)
				{ why = "interval error"; goto catch; } /* 0 or valid. */
			const union line64 key
				= {{ (uint32_t)line, {{ date.day, date.month, date.year }} }};
			struct kjvrange *value;
			switch(kjvline_tree_assign(lines, key, &value)) { /* fixme */
			case TREE_PRESENT: why = "duplicate key";
			case TREE_ERROR: goto catch;
			case TREE_ABSENT:
				value->start.book = book;
				value->start.chapter = chapter;
				value->start.verse = verse;
				value->verse_end = verse_end;
				break;
			}
			book = Revelation, chapter = 0, verse = 0, verse_end = 0;
			continue;
		}
	*/ }
	assert(0); /* Never gets here. */
catch:
	if(!errno) errno = EILSEQ;
	{
		char datestr[12];
		date32_to_string(date, &datestr);
		fprintf(stderr, "KJV lines: %s.\n"
			"%s line %zu: %s.\n",
			kjvline_tree_to_string(lines), datestr, line, why);
	}
	return 0;
}

void kjv_line_(struct kjvline_tree *const f) { kjvline_tree_(f); }

struct kjvline_tree kjv_line(struct journal *const j) {
	struct kjvline_tree lines;
	struct journal_iterator it;
	union date32 date;
	const char *text;
	assert(j);
	lines = kjvline_tree();
	it = journal_iterator(j);
	while(journal_next(&it, &date, &text))
		if(!scan(date, text, &lines)) goto catch;
	goto finally;
catch:
	kjv_line_(&lines);
finally:
	return lines;
}

int kjv_line_is_empty(const struct kjvline_tree *const lines)
	{ return !lines || !lines->root.node; }

const char *kjv_line_to_string(const struct kjvline_tree *const lines)
	{ return kjvline_tree_to_string(lines); }

struct kjvline_tree_iterator kjv_line_iterator(struct kjvline_tree *const lines)
	{ return kjvline_tree_iterator(lines); }

int kjv_line_next(struct kjvline_tree_iterator *const it, union line64 *const k,
	const struct kjvrange **const v) {
	assert(it && k && v);
	if(!kjvline_tree_next(it)) return 0;
	*k = kjvline_tree_key(it);
	*v = kjvline_tree_value(it);
	return 1;
}

/** @return Success. */
int kjv_graphs(struct journal *const jrnl, struct sources *const srcs) {
	struct kjvline_tree lines = {0};
	struct kjvcount_table count = {0};
	struct kjvset_table set = kjv_set();
	size_t no_total;
	const char *reason = 0;

	count = kjv_count(&no_total);
	fprintf(stderr, "KJV count: %s.\n", kjv_count_to_string(&count));
	if(!no_total) { reason = "kjv failed to load"; goto catch; }

	lines = kjv_line(jrnl);
	fprintf(stderr, "KJV lines: %s.\n", kjv_line_to_string(&lines));
	if(kjv_line_is_empty(&lines))
		{ reason = "KJV lines parsing failed"; goto catch; }

	struct kjvline_tree_iterator it = kjv_line_iterator(&lines);
	union line64 line;
	const struct kjvrange *range;

	/*printf("set term postscript eps enhanced\n"
		"set output \"kjv.eps\"\n");*/
	/*printf("set terminal cairolatex standalone pdf size "
		"16cm,10.5cm dashed transparent\n"
		"set output \"kjv.tex\"\n");*/
	/* https://stackoverflow.com/a/12601553 */
	printf("set terminal pngcairo dashed transparent truecolor"
		" size 840, 480 fontscale 1\n"
		"set output \"kjv.png\"\n");
	printf("$Data <<EOD\n"
		"# date, verse, words, set / %zu, source\n", no_total);
	while(kjv_line_next(&it, &line, &range)) {
		char citestr[12], datestr[12];
		size_t words = 0, newwords = 0;
		const struct source *src = source_lookup(srcs, line); /* Source. */
		if(!src || !src->name.a) { errno = EDOM; goto catch; }
		date32_to_string(line.date, &datestr); /* Date. */
		kjvcite_to_string(range->start, &citestr); /* KJV cite. */
		for(union kjvcite c = range->start; ; c.verse++) {
			size_t w = kjv_count_get(&count, c);
			words += w;
			switch (kjv_set_add(&set, c)) {
			case TABLE_ERROR: goto catch;
			case TABLE_ABSENT: newwords += w;
			case TABLE_PRESENT: break;
			} /* while(); */
			if(!range->verse_end || range->verse_end <= c.verse) break;
		}
		printf("%s, %s", datestr, citestr);
		if(range->verse_end) printf("-%" PRIu32, range->verse_end);
		printf(", %zu, %zu, %.*s\n",
			words, newwords, (int)(src->name.b - src->name.a), src->name.a);
	}
	printf("EOD\n"
		"# theozh https://stackoverflow.com/a/75466214/2472827\n"
		"# get a unique list from datablock\n"
		"addToList(list,col) = list.( strstrt(list,'\"'.strcol(col).'\"') \\\n"
		" > 0 ? '' : ' \"'.strcol(col).'\"')\n"
		"Uniqs = ''\n"
		"stats $Data u (Uniqs=addToList(Uniqs,5)) nooutput\n"
		"Uniq(i) = word(Uniqs,i)\n"
		"getIndex(s) = sum [_i=1:words(Uniqs)] s eq word(Uniqs,_i) ? _i : 0\n"
		"\n"
		"stats $Data u 3 nooutput\n"
		"cumsum = STATS_sum\n"
		"stats $Data u 4 nooutput\n"
		"setsum = STATS_sum\n"
		"\n"
		"myTimeFmt = \"%%Y-%%m-%%d\"\n"
		"set format x myTimeFmt timedate\n"
		"set xtics format myTimeFmt rotate by -30\n"
		"set format y \"%%g%%%%\"\n"
		"set grid\n"
		"set key out reverse Left noautotitle\n"
		"set style fill solid 0.5\n"
		"unset border\n"
		"set autoscale xfix # max? hack: can't get x to extend further\n"
		"\n"
		"set label sprintf(\"%%u cumulative words (duplicate verses counted)\","
		" cumsum) center at graph 0.5, first cumsum*100/%zu offset 0,0.5\n"
		"set label sprintf(\"%%u unique KJV verse words memorized\", setsum) "
		"center at graph 0.5, first setsum*100/%zu offset 0,0.5\n"
		"\n"
		"plot \\\n"
		" cumsum*100/%zu w l lc \"grey\" dt 2 lw 1, \\\n"
		" total=0 $Data u"
		" (timecolumn(1,myTimeFmt)):(dy=$3*100/%zu,total=total+dy) \\\n"
		" w steps lc \"grey\" dt 2 lw 1, \\\n" /* `pngcairo` has trouble :[. */
		" total=0 $Data u"
		" (timecolumn(1,myTimeFmt)):(dy=$4*100/%zu,total=total+dy) \\\n"
		" w steps lc \"black\" dt 1 lw 1, \\\n"
		" setsum*100/%zu w l lc \"black\" dt 1 lw 1, \\\n"
		" total=0 '' u \\\n"
		" (timecolumn(1,myTimeFmt)): \\\n"
		" (dy=$4*100/%zu,total=total+dy,total/2.): \\\n"
		" (43200): \\\n"
		" (total/2.): \\\n"
		" (getIndex(strcol(5))) w boxxy lc var lw 1, \\\n"
		" for [i=1:words(Uniqs)] keyentry w boxxy lc i ti Uniq(i)\n",
		no_total, no_total, no_total, no_total, no_total, no_total, no_total);
	goto finally;
catch:
	perror(reason);
finally:
	kjv_line_(&lines);
	kjv_count_(&count);
	return !reason;
}

int main(void) {
	int success = EXIT_SUCCESS;
	const char *reason = 0;
	errno = 0;
	struct journal jrnl = {0};
	struct sources srcs = {0};

	jrnl = journal("../journal");
	fprintf(stderr, "Journal: %s.\n", journal_to_string(&jrnl));
	if(journal_is_empty(&jrnl))
		{ reason = "journal failed to load"; goto catch; }

	srcs = sources(&jrnl);
	//fprintf(stderr, "Sources: %s.\n", sources_to_string(&srcs));
	if(sources_is_empty(&srcs))
		{ reason = "sources failed to parse"; goto catch; }

	if(!kjv_graphs(&jrnl, &srcs))
		{ reason = "graphs failed"; goto catch; }

	goto finally;
catch:
	success = EXIT_FAILURE;
	perror("journal");
	if(reason) fprintf(stderr, "Details: %s.\n", reason);
finally:
	sources_(&srcs);
	journal_(&jrnl);
	return success;
}