scdoc(5)

scdoc(1)

# NAME

scdoc - syntax description for scdoc markup language

scdoc - tool for generating *roff*(7) manual pages

# SYNOPSIS

*scdoc* < _input_ > _output_

# DESCRIPTION

scdoc is a tool designed to make the process of writing man pages more

friendly. It converts scdoc files into roff macros, which can then be converted
to man pages or a number of other formats. The syntax is inspired by, but not
directly taken from, markdown. Input files *must* use the UTF-8 encoding.

friendly. It reads scdoc syntax from stdin and writes roff to stdout, suitable
for reading with *man*(1).

# SYNTAX

Input files must use the UTF-8 encoding.

# PREAMBLE

## PREAMBLE

Each scdoc file must begin with the following preamble:

	*name*(_section_)

The *name* is the name of the man page you are writing, and _section_ is the
section you're writing for (see *man*(1) for information on manual sections).

# SECTION HEADERS

## SECTION HEADERS

Each section of your man page should begin with something similar to the
following:

	# HEADER NAME

Subsection headers are also understood - use two hashes. Each header must have
an empty line on either side.

# PARAGRAPHS

## PARAGRAPHS

Begin a new paragraph with an empty line.

# FORMATTING

## FORMATTING

Text can be made *bold* or _underlined_ with asterisks and underscores: \*bold\*
or \_underlined\_.

# INDENTATION

## INDENTATION

You may indent lines with tab characters (*\\t*) to indent them by 4 spaces in
the output. Indented lines may not contain headers.

# LISTS

	The result looks something like this.

	You may use multiple lines and most _formatting_.

Deindent to return to normal.

## LISTS

You may start bulleted lists with dashes, like so:

```
- Item 1
- Item 2

	- Subitem 1
	- Subitem 2

- Item 3
```

You may also use numbered lists like so:

The result looks like this:

```
1. Item 1
2. Item 2
3. Item 3
```

- Item 1
- Item 2
	- Subitem 1
	- Subitem 2
- Item 3

# LITERAL TEXT

## LITERAL TEXT

You may turn off scdoc formatting and output literal text with escape codes and
literal blocks. Inserting a \\ into your source will cause the subsequent symbol
to be treated as a literal and copied directly to the output. You may also make
blocks of literal syntax like so:

```
\`\`\`
_This formatting_ will *not* be interpreted by scdoc.
\`\`\`
```

These blocks will be indented one level. Note that literal text is shown
literally in the man viewer - that is, it's not a means for inserting your own
roff macros into the output.

# AUTHORS

Maintained by Drew DeVault <sir@cmpwn.com>. Up-to-date sources can be found at
https://git.sr.ht/~sircmpwn/scdoc and bugs/patches can be submitted by email to
sir@cmpwn.com.

#include <assert.h>
#include <ctype.h>

#include <stdbool.h>

#include <stdio.h>
#include <stdlib.h>
#include <time.h>
#include <unistd.h>
#include "string.h"
#include "unicode.h"
#include "util.h"

char date[256];

static int parse_section(struct parser *p) {
	str_t *section = str_create();
	uint32_t ch;
	while ((ch = parser_getch(p)) != UTF8_INVALID) {
		if (isdigit(ch)) {
			assert(str_append_ch(section, ch) != -1);
		} else if (ch == ')') {
			if (!section->str) {
				break;
			}
			int sec = strtol(section->str, NULL, 10);
			if (sec < 1 || sec > 9) {
				parser_fatal(p, "Expected section between 1 and 9");
				break;
			}
			str_free(section);
			return sec;
		} else {
			parser_fatal(p, "Expected digit or )");
			break;
		}
	};
	parser_fatal(p, "Expected manual section");
	return -1;
}

static void parse_preamble(struct parser *p) {
	str_t *name = str_create();
	int section = -1;
	uint32_t ch;
	while ((ch = parser_getch(p)) != UTF8_INVALID) {
		if (isalnum(ch)) {
			assert(str_append_ch(name, ch) != -1);
		} else if (ch == '(') {
			section = parse_section(p);
		} else if (ch == '\n') {
			if (name->len == 0) {
				parser_fatal(p, "Expected preamble");
			}
			if (section == -1) {
				parser_fatal(p, "Expected manual section");
			}
			char sec[2] = { '0' + section, 0 };
			roff_macro(p, "TH", name->str, sec, date, NULL);
			break;
		}
	}
	str_free(name);
}

static void parse_format(struct parser *p, enum formatting fmt) {
	char formats[FORMAT_LAST] = {
		[FORMAT_BOLD] = 'B',
		[FORMAT_UNDERLINE] = 'I',
	};
	if (p->flags) {
		if ((p->flags & ~fmt)) {
			parser_fatal(p, "Cannot nest inline formatting.");
		}
		fprintf(p->output, "\\fR");
	} else {
		fprintf(p->output, "\\f%c", formats[fmt]);
	}
	p->flags ^= fmt;
}

static void parse_text(struct parser *p) {
	uint32_t ch;
	int i = 0;
	while ((ch = parser_getch(p)) != UTF8_INVALID) {
		switch (ch) {
		case '\\':
			ch = parser_getch(p);
			if (ch == UTF8_INVALID) {
				parser_fatal(p, "Unexpected EOF");
			} else if (ch == '\\') {
				fprintf(p->output, "\\\\");
			} else {
				utf8_fputch(p->output, ch);
			}
			break;
		case '*':
			parse_format(p, FORMAT_BOLD);
			break;
		case '_':
			parse_format(p, FORMAT_UNDERLINE);
			break;

		case '\n':
			utf8_fputch(p->output, ch);
			return;

		case '.':
			if (!i) {
				// Escape . if it's the first character
				fprintf(p->output, "\\&.");
				break;
			}
			/* fallthrough */
		default:
			utf8_fputch(p->output, ch);

			break;
		}
		if (ch == '\n') {

			break;
		}
		++i;
	}
}

static void parse_heading(struct parser *p) {
	uint32_t ch;
	int level = 1;
	while ((ch = parser_getch(p)) != UTF8_INVALID) {
		if (ch == '#') {
			++level;
		} else if (ch == ' ') {
			break;
		} else {
			parser_fatal(p, "Invalid start of heading (probably needs a space)");
		}
	}
	switch (level) {
	case 1:
		fprintf(p->output, ".SH ");
		break;
	case 2:
		fprintf(p->output, ".SS ");
		break;
	default:
		parser_fatal(p, "Only headings up to two levels deep are permitted");
		break;
	}
	while ((ch = parser_getch(p)) != UTF8_INVALID) {
		utf8_fputch(p->output, ch);
		if (ch == '\n') {
			break;
		}
	}
}

static int parse_indent(struct parser *p) {

static int parse_indent(struct parser *p, int *indent) {

	int i = 0;
	uint32_t ch;
	while ((ch = parser_getch(p)) == '\t') {
		++i;
	}
	parser_pushch(p, ch);

	if (i == *indent - 1) {
		roff_macro(p, "RE", NULL);
	} else if (i == *indent + 1) {
		roff_macro(p, "RS", "4", NULL);
	} else if (i != *indent && ch == '\t') {
		parser_fatal(p, "(De)indented by an amount greater than 1");
	}
	*indent = i;

	return i;
}

static void parse_document(struct parser *p) {

static void parse_list(struct parser *p, int *indent) {

	uint32_t ch;

	int indent = 0;

	if ((ch = parser_getch(p)) != ' ') {
		parser_fatal(p, "Expected space before start of list entry");
	}
	roff_macro(p, "IP", "\\(bu", "4", NULL);
	parse_text(p);

	do {

		int i = parse_indent(p);
		if (i == indent - 1) {
			roff_macro(p, "RE", NULL);
		} else if (i == indent + 1) {
			roff_macro(p, "RS", "4", NULL);
		} else if (i != indent && ch == '\t') {
			parser_fatal(p, "(De)indented by an amount greater than 1");
		}
		indent = i;

		ch = parser_getch(p);
		if (ch == UTF8_INVALID) {

		parse_indent(p, indent);
		if ((ch = parser_getch(p)) == UTF8_INVALID) {

			break;
		}

		if (indent != 0) {
			// Only text is allowed at this point

		switch (ch) {
		case ' ':
			break;
		case '-':
			if ((ch = parser_getch(p)) != ' ') {
				parser_fatal(p, "Expected space before start of list entry");
			}
			fprintf(p->output, ".sp -1\n");
			roff_macro(p, "IP", "\\(bu", "4", NULL);
			parse_text(p);
			break;
		default:
			fprintf(p->output, "\n");

			parser_pushch(p, ch);

			parse_text(p);
			continue;

			return;

		}

	} while (ch != UTF8_INVALID);
}

static void parse_document(struct parser *p) {
	uint32_t ch;
	int indent = 0;
	do {
		parse_indent(p, &indent);
		if ((ch = parser_getch(p)) == UTF8_INVALID) {
			break;
		}

		switch (ch) {
		case '#':

			if (indent != 0) {
				parser_pushch(p, ch);
				parse_text(p);
				break;
			}

			parse_heading(p);

			break;
		case '-':
			parse_list(p, &indent);

			break;
		case ' ':
			parser_fatal(p, "Tabs are required for indentation");
			break;
		case '\n':
			roff_macro(p, "P", NULL);
			break;
		default:
			parser_pushch(p, ch);
			parse_text(p);
			break;
		}
	} while (ch != UTF8_INVALID);
}

static void output_scdoc_preamble(struct parser *p) {
	// TODO: Add version here
	fprintf(p->output, ".\\\" Generated by scdoc\n");
	fprintf(p->output, ".\\\" Fix weird qutation marks:\n");
	fprintf(p->output, ".\\\" http://bugs.debian.org/507673\n");
	fprintf(p->output, ".\\\" http://lists.gnu.org/archive/html/groff/2009-02/msg00013.html\n");
	fprintf(p->output, ".ie \\n(.g .ds Aq \\(aq\n");
	fprintf(p->output, ".el       .ds Aq '\n");
	fprintf(p->output, ".\\\" Disable hyphenation:\n");
	roff_macro(p, "nh", NULL);
	fprintf(p->output, ".\\\" Generated content:\n");
}

int main(int argc, char **argv) {
	if (argc > 1) {
		fprintf(stderr, "Usage: scdoc < input.scd > output.roff");
		return 1;
	}
	time_t now;
	time(&now);
	struct tm *now_tm = localtime(&now);
	strftime(date, sizeof(date), "%F", now_tm);
	struct parser p = {
		.input = stdin,
		.output = stdout,
		.line = 1,
		.col = 1
	};
	output_scdoc_preamble(&p);
	parse_preamble(&p);
	parse_document(&p);
	return 0;
}