interpret/src/scan.re.c

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

 Lexer for journal entries.

 "^[edict: expect; there; to; be; args] Something.\n"
 "^[<num>] Something.\n"
 "<num>+.<num>{2}(<num>)$"
 "<bookref> -- \"*\n"

 @std C89/90 */

#include "../src/scan.h"
#include <stdlib.h>
#include <stdio.h>
#include <assert.h>
#include <limits.h>
#include <errno.h>

/* This defines `enum condition`. */
/*!types:re2c*/

int scan(union date32 date, const char *const buffer) {
	const char *YYCURSOR = buffer;
	/*!re2c /**/
	re2c:define:YYCTYPE = char;
	re2c:yyfill:enable = 0;
	*/
}








#if 0

struct scan scan(const char *const buffer) {
	struct scan scan;
	scan.marker = scan.from = scan.cursor = scan.label = scan.buffer = buffer;
	scan.condition = yycline;
	scan.line = 1;
	return scan;
}

int scan_next(struct scan *const s, struct lex *const x) {
	/*!re2c /**/
	re2c:flags:tags = 1;
	re2c:define:YYCTYPE = char;
	re2c:yyfill:enable = 0;
	re2c:define:YYCURSOR = s->cursor;
	re2c:define:YYMARKER = s->marker;
	re2c:define:YYCONDTYPE = 'condition';
	re2c:define:YYGETCONDITION = 's->condition';
	re2c:define:YYGETCONDITION:naked = 1;
	re2c:define:YYSETCONDITION = 's->condition = @@;';
	re2c:define:YYSETCONDITION:naked = 1;
	sentinel = "\x00";
	newline = "\n";
	unix_control = [\x01-\x08\x0a-\x1f\x7f];
	ws = [ \t];
	glyph = [^] \ (sentinel | unix_control | newline | ws);
	keyword = ([a-zA-Z] | [0-9][0-9_\-]*[a-zA-Z]) [a-zA-Z0-9_\-]*;
	decimal = "-"? ([1-9][0-9]* | [0])? "." [0-9]+ | [1-9][0-9]* | [0];
	natural = [1-9][0-9]*;
	date = "-"? natural "-" [0-1][0-9] "-" [0-1][0-9];
	*/
	const char *s0, *s1;
	/*!stags:re2c format = 'const char *@@;\n'; */
	assert(s && x);
	if(!s->buffer) return 0;
	x->line = s->line;
	x->s0 = x->s1 = 0;
scan:
	/*!re2c /**/
	<*> unix_control { return x->symbol = ILLEGAL, 0; }
	<*> * { return x->symbol = SYNTAX, 0; }
	<*> sentinel /* New line always delimits. */
		{ return x->symbol = s->condition == yycline ? END : ILLEGAL, 0; }
	<expect_line> newline => line { x->line = ++s->line; goto scan; }
	/* Symbols that go at the beginning of a line. */
	<line> newline { x->line = ++s->line; goto scan; }
	<line> "[" :=> edict
	<line> "--" :=> source
	<line> "->" :=> location
	<line> "!" => text { return x->symbol = COMPLETE, 1; }
	<line> "^" => text { return x->symbol = CANCELLED, 1; }
	<line> "#" => text { return x->symbol = HEADING, 1; }
	<line> * :=> text

	<text> newline => line { x->line = ++s->line; goto scan; }
	<text, bible> ws+ { goto scan; }
	<text> @s0 glyph+ @s1
		{ x->s0 = s0, x->s1 = s1; return x->symbol = TEXT, 1; }

	bible_ref = natural ":" natural [ab]? ("-" (natural ":")? natural [ab]?)?;
	glyph_minus = glyph \ ['];

	<text> @s0 ("Genesis" | "Exodus" | "Leviticus" | "Numbers" | "Deuteronomy"
	| "Joshua" | "Judges" | "Ruth" | "I"{1,2} " Samuel" | "I"{1,2} " Kings"
	| "I"{1,2} " Chronicles" | "Ezra" | "Nehemiah" | "Esther" | "Job"
	| "Psalms" | "Proverbs" | "Ecclesiastes" | "Song of Solomon" | "Isaiah"
	| "Jeremiah" | "Lamentations" | "Ezekiel" | "Daniel" | "Hosea" | "Joel"
	| "Amos" | "Obadiah" | "Jonah" | "Micah" | "Nahum" | "Habakkuk"
	| "Zephaniah" | "Haggai" | "Zechariah" | "Malachi" | "Matthew" | "Mark"
	| "Luke" | "John" | "Acts" | "Romans" | "I"{1,2} " Corinthians"
	| "Galatians" | "Ephesians" | "Philippians" | "Colossians"
	| "I"{1,2} " Thessalonians" | "I"{1,2} " Timothy" | "Titus" | "Philemon"
	| "Hebrews" | "James" | "I"{1,2} " Peter" | "I"{1,3} " John" | "Jude"
	| "Revelation") @s1 ws* / bible_ref ws+ "--" ws+ "``"
	=> bible { x->s0 = s0, x->s1 = s1; return x->symbol = KJV_BOOK, 1; }
	<bible> @s0 bible_ref @s1 ws+ "--" ws+ "``"
		{ x->s0 = s0, x->s1 = s1; return x->symbol = KJV_CHAPTER_VERSE, 1; }
	<bible> "``" { return x->symbol = KJV_NEXT, 1; }
	<bible> "''" :=> text
	/* fixme: This is a hack that doesn't allow apostrophes at the end of a
	 word, (not sure there are any in the bible.) Is ' terminated by '';
	 otherwise same as glyph+ above. */
	<bible> @s0 ("'"? glyph_minus+ ("'" glyph_minus+)*) @s1
		 { x->s0 = s0, x->s1 = s1; return x->symbol = KJV_TEXT, 1; }
	/* Multiple verses can be present, but they end in ''.
	 Not strictly enforced. */
	<bible> newline / (newline | "``") { x->line = ++s->line; goto scan; }
	<bible> newline { return x->symbol = SYNTAX, 0; }

	<source> @s0 keyword @s1 => expect_line
		 { x->s0 = s0, x->s1 = s1; return x->symbol = SOURCE_RECALL, 1; }

	<location> "" / "(" :=> map
	<location> "[" ws* @s0 keyword @s1 ws* "]"
		{ x->s0 = s0, x->s1 = s1; return x->symbol = LOCATION_SAVE, 1; }
	<location> @s0 keyword @s1 => expect_line
		{ x->s0 = s0, x->s1 = s1; return x->symbol = LOCATION_RECALL, 1; }

	<map> "(" @s0 decimal "," @s1 decimal ")" => expect_caption
		{ x->s0 = s0, x->s1 = s1; return x->symbol = LOCATION, 1; }

	<edict> "source" :=> source
	<edict> "ed" :=> ed
	<edict> "contact" :=> contact
	<edict> "glider" :=> glider
	<edict> "flight" :=> flight
	<edict> "bible" :=> bible
	<edict> "book" :=> book
	<edict> "movie" :=> movie
	<edict> "tv" :=> tv
	<edict> "medication" :=> medication
	<edict> "idea" :=> idea
	<edict> "vaccine" :=> vaccine
	<edict> "in" :=> in
	<edict> "" / natural :=> significant
	<edict> [0-1][0-9] "-" [0-3][0-9]
		", " [0-2][0-9] ":" [0-5][0-9] "] "
		:=> text /* This is likely WhatsApp conversations. Ignore. */
	/* missed, show, 'First, Second', 'Sounds', 'CSS', ..., 'Swanky', 'Shields' */

	/* How did it get into my journal? */
	<edict> "source"
		{ if(s->is_ws_expected || s->edict.size)
			return x->symbol = SYNTAX, 0;
		s->is_ws_expected = 1, s->is_source = 1;
		s->edict.size = 2;
		s->edict.expect[1] = EXPECT_KEYWORD;
		s->edict.expect[0] = EXPECT_END_TEXT;
		return x->symbol = SOURCE, 1; }
	<edict> "default"
		{ if(s->is_ws_expected || !s->is_source)
			return x->symbol = SYNTAX, 0;
		s->is_ws_expected = 1, s->is_source = 0;
		return x->symbol = DEFAULT, 1; }

	/* Editorializing; looking back. */
	<edict> "ed"
		{ if(s->is_ws_expected || s->edict.size)
			return x->symbol = SYNTAX, 0;
		s->is_ws_expected = 1; /* no idea, just copy; probably should do sth */
		s->edict.size = 1;
		s->edict.expect[0] = EXPECT_END_TEXT; /* Pithy comment. */
		return x->symbol = EDITORIALIZING, 1; }

	/* Score. */
	<edict> "significant"
		{ if(s->is_ws_expected || s->edict.size)
			return x->symbol = SYNTAX, 0;
		s->is_ws_expected = 1;
		s->edict.size = 3;
		s->edict.expect[2] = EXPECT_NATURAL; /* Ordinal. */
		s->edict.expect[1] = EXPECT_RESTRICT_TEXT; /* Name. */
		s->edict.expect[0] = EXPECT_DATE; /* Birthday. */
		return x->symbol = SIGNIFICANT, 1; }
	<edict> @s0 natural @s1
		{ if(s->is_ws_expected || s->edict.size)
			return x->symbol = SYNTAX, 0;
		s->is_ws_expected = 1;
		x->s0 = s0, x->s1 = s1;
		return x->symbol = SIGNIFICANT_RECALL, 1; }

	/* General [edict: whatever]. */
	<edict> ws+ { s->is_ws_expected = 0; goto scan; }
	<edict> ":"
		{ if(!s->edict.size) return x->symbol = SYNTAX, 0;
		s->is_ws_expected = 0, s->is_source = 0;
		expect_pop(); goto scan; }
	<edict_keyword> ws* @s0 keyword @s1 ws* ";"?
		{ x->s0 = s0, x->s1 = s1; expect_pop();
		return x->symbol = ARG_KEYWORD, 1; }
	<edict_date> ws* @s0 date @s1 ws* ";"?
		{ x->s0 = s0, x->s1 = s1; expect_pop();
		return x->symbol = ARG_DATE, 1; }
	<edict_natural> ws* @s0 natural @s1 ws* ";"?
		{ x->s0 = s0, x->s1 = s1; expect_pop();
		return x->symbol = ARG_NATURAL, 1; }
	<edict_restrict_text>
		ws* @s0 (glyph \ [;[\]]) ((glyph \ [;[\]]) | ws)* @s1 ws* ";"?
		{ x->s0 = s0, x->s1 = s1; expect_pop();
		return x->symbol = ARG_RESTRICT_TEXT, 1; }
	<edict_end_text>
		ws* @s0 (glyph \ [[\]]) ((glyph \ [[\]]) | ws)* @s1 ws*
		{ x->s0 = s0, x->s1 = s1; expect_pop();
		return x->symbol = ARG_END_TEXT, 1; }
	<edict, edict_end> "]" => expect_line
		{ if(s->edict.size) return 0; goto scan; }
	*/
}

#endif