mdman/format/man.rs
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457
//! Man-page formatter.
use crate::util::{header_text, parse_name_and_section};
use crate::EventIter;
use anyhow::{bail, Error};
use pulldown_cmark::{Alignment, Event, HeadingLevel, LinkType, Tag, TagEnd};
use std::fmt::Write;
use url::Url;
pub struct ManFormatter {
url: Option<Url>,
}
impl ManFormatter {
pub fn new(url: Option<Url>) -> ManFormatter {
ManFormatter { url }
}
}
impl super::Formatter for ManFormatter {
fn render(&self, input: &str) -> Result<String, Error> {
ManRenderer::render(input, self.url.clone())
}
fn render_options_start(&self) -> &'static str {
// Tell pulldown_cmark to ignore this.
// This will be stripped out later.
"<![CDATA["
}
fn render_options_end(&self) -> &'static str {
"]]>"
}
fn render_option(
&self,
params: &[&str],
block: &str,
_man_name: &str,
) -> Result<String, Error> {
let rendered_options = params
.iter()
.map(|param| {
let r = self.render(param)?;
Ok(r.trim().trim_start_matches(".sp").to_string())
})
.collect::<Result<Vec<_>, Error>>()?;
let rendered_block = self.render(block)?;
let rendered_block = rendered_block.trim().trim_start_matches(".sp").trim();
// .RS = move left margin to right 4.
// .RE = move margin back one level.
Ok(format!(
"\n.sp\n{}\n.RS 4\n{}\n.RE\n",
rendered_options.join(", "),
rendered_block
))
}
fn linkify_man_to_md(&self, name: &str, section: u8) -> Result<String, Error> {
Ok(format!("`{}`({})", name, section))
}
}
#[derive(Copy, Clone)]
enum Font {
Bold,
Italic,
}
impl Font {
fn str_from_stack(font_stack: &[Font]) -> &'static str {
let has_bold = font_stack.iter().any(|font| matches!(font, Font::Bold));
let has_italic = font_stack.iter().any(|font| matches!(font, Font::Italic));
match (has_bold, has_italic) {
(false, false) => "\\fR", // roman (normal)
(false, true) => "\\fI", // italic
(true, false) => "\\fB", // bold
(true, true) => "\\f(BI", // bold italic
}
}
}
struct ManRenderer<'e> {
output: String,
parser: EventIter<'e>,
font_stack: Vec<Font>,
}
impl<'e> ManRenderer<'e> {
fn render(input: &str, url: Option<Url>) -> Result<String, Error> {
let parser = crate::md_parser(input, url);
let output = String::with_capacity(input.len() * 3 / 2);
let mut mr = ManRenderer {
parser,
output,
font_stack: Vec::new(),
};
mr.push_man()?;
Ok(mr.output)
}
fn push_man(&mut self) -> Result<(), Error> {
// If this is true, this is inside a cdata block used for hiding
// content from pulldown_cmark.
let mut in_cdata = false;
// The current list stack. None if unordered, Some if ordered with the
// given number as the current index.
let mut list: Vec<Option<u64>> = Vec::new();
// Used in some cases where spacing isn't desired.
let mut suppress_paragraph = false;
let mut table_cell_index = 0;
let mut last_seen_link_data = None;
while let Some((event, range)) = self.parser.next() {
let this_suppress_paragraph = suppress_paragraph;
suppress_paragraph = false;
match event {
Event::Start(tag) => {
match tag {
Tag::Paragraph => {
if !this_suppress_paragraph {
self.flush();
self.output.push_str(".sp\n");
}
}
Tag::Heading { level, .. } => {
if level == HeadingLevel::H1 {
self.push_top_header()?;
} else if level == HeadingLevel::H2 {
// Section header
let text = header_text(&mut self.parser)?;
self.flush();
write!(self.output, ".SH \"{}\"\n", text)?;
suppress_paragraph = true;
} else {
// Subsection header
let text = header_text(&mut self.parser)?;
self.flush();
write!(self.output, ".SS \"{}\"\n", text)?;
suppress_paragraph = true;
}
}
Tag::BlockQuote(..) => {
self.flush();
// .RS = move left margin over 3
// .ll = shrink line length
self.output.push_str(".RS 3\n.ll -5\n.sp\n");
suppress_paragraph = true;
}
Tag::CodeBlock(_kind) => {
// space down, indent 4, no-fill mode
self.flush();
self.output.push_str(".sp\n.RS 4\n.nf\n");
}
Tag::List(start) => list.push(start),
Tag::Item => {
// Note: This uses explicit movement instead of .IP
// because the spacing on .IP looks weird to me.
// space down, indent 4
self.flush();
self.output.push_str(".sp\n.RS 4\n");
match list.last_mut().expect("item must have list start") {
// Ordered list.
Some(n) => {
// move left 4, output the list index number, move right 1.
write!(self.output, "\\h'-04' {}.\\h'+01'", n)?;
*n += 1;
}
// Unordered list.
None => self.output.push_str("\\h'-04'\\(bu\\h'+02'"),
}
suppress_paragraph = true;
}
Tag::FootnoteDefinition(_label) => unimplemented!(),
Tag::Table(alignment) => {
// Table start
// allbox = draw a box around all the cells
// tab(:) = Use `:` to separate cell data (instead of tab)
// ; = end of options
self.output.push_str(
"\n.TS\n\
allbox tab(:);\n",
);
let alignments: Vec<_> = alignment
.iter()
.map(|a| match a {
Alignment::Left | Alignment::None => "lt",
Alignment::Center => "ct",
Alignment::Right => "rt",
})
.collect();
self.output.push_str(&alignments.join(" "));
self.output.push_str(".\n");
table_cell_index = 0;
}
Tag::TableHead => {
table_cell_index = 0;
}
Tag::TableRow => {
table_cell_index = 0;
self.output.push('\n');
}
Tag::TableCell => {
if table_cell_index != 0 {
// Separator between columns.
self.output.push(':');
}
// Start a text block.
self.output.push_str("T{\n");
table_cell_index += 1
}
Tag::Emphasis => self.push_font(Font::Italic),
Tag::Strong => self.push_font(Font::Bold),
// Strikethrough isn't usually supported for TTY.
Tag::Strikethrough => self.output.push_str("~~"),
Tag::Link {
link_type,
dest_url,
..
} => {
last_seen_link_data = Some((link_type.clone(), dest_url.to_owned()));
if dest_url.starts_with('#') {
// In a man page, page-relative anchors don't
// have much meaning.
continue;
}
match link_type {
LinkType::Autolink | LinkType::Email => {
// The text is a copy of the URL, which is not needed.
match self.parser.next() {
Some((Event::Text(_), _range)) => {}
_ => bail!("expected text after autolink"),
}
}
LinkType::Inline
| LinkType::Reference
| LinkType::Collapsed
| LinkType::Shortcut => {
self.push_font(Font::Italic);
}
// This is currently unused. This is only
// emitted with a broken link callback, but I
// felt it is too annoying to escape `[` in
// option descriptions.
LinkType::ReferenceUnknown
| LinkType::CollapsedUnknown
| LinkType::ShortcutUnknown => {
bail!(
"link with missing reference `{}` located at offset {}",
dest_url,
range.start
);
}
}
}
Tag::Image { .. } => {
bail!("images are not currently supported")
}
Tag::HtmlBlock { .. }
| Tag::MetadataBlock { .. }
| Tag::DefinitionList
| Tag::DefinitionListTitle
| Tag::DefinitionListDefinition => {}
}
}
Event::End(tag_end) => {
match &tag_end {
TagEnd::Paragraph => self.flush(),
TagEnd::Heading(..) => {}
TagEnd::BlockQuote(..) => {
self.flush();
// restore left margin, restore line length
self.output.push_str(".br\n.RE\n.ll\n");
}
TagEnd::CodeBlock => {
self.flush();
// Restore fill mode, move margin back one level.
self.output.push_str(".fi\n.RE\n");
}
TagEnd::List(_) => {
list.pop();
}
TagEnd::Item => {
self.flush();
// Move margin back one level.
self.output.push_str(".RE\n");
}
TagEnd::FootnoteDefinition => {}
TagEnd::Table => {
// Table end
// I don't know why, but the .sp is needed to provide
// space with the following content.
self.output.push_str("\n.TE\n.sp\n");
}
TagEnd::TableHead => {}
TagEnd::TableRow => {}
TagEnd::TableCell => {
// End text block.
self.output.push_str("\nT}");
}
TagEnd::Emphasis | TagEnd::Strong => self.pop_font(),
TagEnd::Strikethrough => self.output.push_str("~~"),
TagEnd::Link => {
if let Some((link_type, ref dest_url)) = last_seen_link_data {
if dest_url.starts_with('#') {
continue;
}
match link_type {
LinkType::Autolink | LinkType::Email => {}
LinkType::Inline
| LinkType::Reference
| LinkType::Collapsed
| LinkType::Shortcut => {
self.pop_font();
self.output.push(' ');
}
_ => {
panic!("unexpected tag {:?}", tag_end);
}
}
write!(self.output, "<{}>", escape(&dest_url)?)?;
}
}
TagEnd::Image
| TagEnd::HtmlBlock
| TagEnd::MetadataBlock(..)
| TagEnd::DefinitionListDefinition
| TagEnd::DefinitionListTitle
| TagEnd::DefinitionList => {}
}
}
Event::Text(t) => {
self.output.push_str(&escape(&t)?);
}
Event::Code(t) => {
self.push_font(Font::Bold);
self.output.push_str(&escape(&t)?);
self.pop_font();
}
Event::Html(t) => {
if t.starts_with("<![CDATA[") {
// CDATA is a special marker used for handling options.
in_cdata = true;
} else if in_cdata {
if t.trim().ends_with("]]>") {
in_cdata = false;
} else if !t.trim().is_empty() {
self.output.push_str(&t);
}
} else {
self.output.push_str(&escape(&t)?);
}
}
Event::FootnoteReference(_t) => {}
Event::SoftBreak => self.output.push('\n'),
Event::HardBreak => {
self.flush();
self.output.push_str(".br\n");
}
Event::Rule => {
self.flush();
// \l' **length** ' Draw horizontal line (default underscore).
// \n(.lu Gets value from register "lu" (current line length)
self.output.push_str("\\l'\\n(.lu'\n");
}
Event::TaskListMarker(_b) => unimplemented!(),
Event::InlineHtml(..) => unimplemented!(),
Event::InlineMath(..) => unimplemented!(),
Event::DisplayMath(..) => unimplemented!(),
}
}
Ok(())
}
fn flush(&mut self) {
if !self.output.ends_with('\n') {
self.output.push('\n');
}
}
/// Switch to the given font.
///
/// Because the troff sequence `\fP` for switching to the "previous" font
/// doesn't support nesting, this needs to emulate it here. This is needed
/// for situations like **hi _there_**.
fn push_font(&mut self, font: Font) {
self.font_stack.push(font);
self.output.push_str(Font::str_from_stack(&self.font_stack));
}
fn pop_font(&mut self) {
self.font_stack.pop();
self.output.push_str(Font::str_from_stack(&self.font_stack));
}
/// Parse and render the first top-level header of the document.
fn push_top_header(&mut self) -> Result<(), Error> {
// This enables the tbl preprocessor for tables.
// This seems to be enabled by default on every modern system I could
// find, but it doesn't seem to hurt to enable this.
self.output.push_str("'\\\" t\n");
// Extract the name of the man page.
let text = header_text(&mut self.parser)?;
let (name, section) = parse_name_and_section(&text)?;
// .TH = Table header
// .nh = disable hyphenation
// .ad l = Left-adjust mode (disable justified).
// .ss sets sentence_space_size to 0 (prevents double spaces after .
// if . is last on the line)
write!(
self.output,
".TH \"{}\" \"{}\"\n\
.nh\n\
.ad l\n\
.ss \\n[.ss] 0\n",
escape(&name.to_uppercase())?,
section
)?;
Ok(())
}
}
fn escape(s: &str) -> Result<String, Error> {
// Note: Possible source on output escape sequences: https://man7.org/linux/man-pages/man7/groff_char.7.html.
// Otherwise, use generic escaping in the form `\[u1EE7]` or `\[u1F994]`.
let mut replaced = s
.replace('\\', "\\(rs")
.replace('-', "\\-")
.replace('\u{00A0}', "\\ ") // non-breaking space (non-stretchable)
.replace('–', "\\[en]") // \u{2013} en-dash
.replace('—', "\\[em]") // \u{2014} em-dash
.replace('‘', "\\[oq]") // \u{2018} left single quote
.replace('’', "\\[cq]") // \u{2019} right single quote or apostrophe
.replace('“', "\\[lq]") // \u{201C} left double quote
.replace('”', "\\[rq]") // \u{201D} right double quote
.replace('…', "\\[u2026]") // \u{2026} ellipsis
.replace('│', "|") // \u{2502} box drawing light vertical (could use \[br])
.replace('├', "|") // \u{251C} box drawings light vertical and right
.replace('└', "`") // \u{2514} box drawings light up and right
.replace('─', "\\-") // \u{2500} box drawing light horizontal
;
if replaced.starts_with('.') {
replaced = format!("\\&.{}", &replaced[1..]);
}
if let Some(ch) = replaced.chars().find(|ch| {
!matches!(ch, '\n' | ' ' | '!'..='/' | '0'..='9'
| ':'..='@' | 'A'..='Z' | '['..='`' | 'a'..='z' | '{'..='~')
}) {
bail!(
"character {:?} is not allowed (update the translation table if needed)",
ch
);
}
Ok(replaced)
}