You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

185 lines
5.1 KiB

// Copyright Joyent, Inc. and other Node contributors.
//
// Permission is hereby granted, free of charge, to any person obtaining a
// copy of this software and associated documentation files (the
// "Software"), to deal in the Software without restriction, including
// without limitation the rights to use, copy, modify, merge, publish,
// distribute, sublicense, and/or sell copies of the Software, and to permit
// persons to whom the Software is furnished to do so, subject to the
// following conditions:
//
// The above copyright notice and this permission notice shall be included
// in all copies or substantial portions of the Software.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
// OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN
// NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
// DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
// OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
// USE OR OTHER DEALINGS IN THE SOFTWARE.
var fs = require('fs');
var marked = require('marked');
var path = require('path');
module.exports = toHTML;
function toHTML(input, filename, template, cb) {
var lexed = marked.lexer(input);
fs.readFile(template, 'utf8', function(er, template) {
if (er) return cb(er);
render(lexed, filename, template, cb);
});
}
function render(lexed, filename, template, cb) {
// get the section
var section = getSection(lexed);
filename = path.basename(filename, '.markdown');
lexed = parseLists(lexed);
// generate the table of contents.
// this mutates the lexed contents in-place.
buildToc(lexed, filename, function(er, toc) {
if (er) return cb(er);
template = template.replace(/__FILENAME__/g, filename);
template = template.replace(/__SECTION__/g, section);
template = template.replace(/__VERSION__/g, process.version);
template = template.replace(/__TOC__/g, toc);
// content has to be the last thing we do with
// the lexed tokens, because it's destructive.
content = marked.parser(lexed);
template = template.replace(/__CONTENT__/g, content);
cb(null, template);
});
}
// just update the list item text in-place.
// lists that come right after a heading are what we're after.
function parseLists(input) {
var state = null;
var depth = 0;
var output = [];
output.links = input.links;
input.forEach(function(tok) {
if (tok.type === 'code' && tok.text.match(/Stability:.*/g)) {
tok.text = parseAPIHeader(tok.text);
output.push({ type: 'html', text: tok.text });
return;
}
if (state === null) {
if (tok.type === 'heading') {
state = 'AFTERHEADING';
}
output.push(tok);
return;
}
if (state === 'AFTERHEADING') {
if (tok.type === 'list_start') {
state = 'LIST';
if (depth === 0) {
output.push({ type:'html', text: '<div class="signature">' });
}
depth++;
output.push(tok);
return;
}
state = null;
output.push(tok);
return;
}
if (state === 'LIST') {
if (tok.type === 'list_start') {
depth++;
output.push(tok);
return;
}
if (tok.type === 'list_end') {
depth--;
if (depth === 0) {
state = null;
output.push({ type:'html', text: '</div>' });
}
output.push(tok);
return;
}
if (tok.text) {
tok.text = parseListItem(tok.text);
}
}
output.push(tok);
});
return output;
}
function parseListItem(text) {
text = text.replace(/\{([^\}]+)\}/, '<span class="type">$1</span>');
//XXX maybe put more stuff here?
return text;
}
function parseAPIHeader(text) {
text = text.replace(/(.*:)\s(\d)([\s\S]*)/,
'<pre class="api_stability_$2">$1 $2$3</pre>');
return text;
}
// section is just the first heading
function getSection(lexed) {
var section = '';
for (var i = 0, l = lexed.length; i < l; i++) {
var tok = lexed[i];
if (tok.type === 'heading') return tok.text;
}
return '';
}
function buildToc(lexed, filename, cb) {
var indent = 0;
var toc = [];
var depth = 0;
lexed.forEach(function(tok) {
if (tok.type !== 'heading') return;
if (tok.depth - depth > 1) {
return cb(new Error('Inappropriate heading level\n' +
JSON.stringify(tok)));
}
depth = tok.depth;
var id = getId(filename + '_' + tok.text.trim());
toc.push(new Array((depth - 1) * 2 + 1).join(' ') +
'* <a href="#' + id + '">' +
tok.text + '</a>');
tok.text += '<span><a class="mark" href="#' + id + '" ' +
'id="' + id + '">#</a></span>';
});
toc = marked.parse(toc.join('\n'));
cb(null, toc);
}
var idCounters = {};
function getId(text) {
text = text.toLowerCase();
text = text.replace(/[^a-z0-9]+/g, '_');
text = text.replace(/^_+|_+$/, '');
text = text.replace(/^([^a-z])/, '_$1');
if (idCounters.hasOwnProperty(text)) {
text += '_' + (++idCounters[text]);
} else {
idCounters[text] = 0;
}
return text;
}