2018-03-09 22:59:12 +02:00
|
|
|
const stringPadding = require('string-padding');
|
2018-05-16 11:56:30 +02:00
|
|
|
const stringToStream = require('string-to-stream')
|
2017-07-17 20:59:40 +02:00
|
|
|
|
|
|
|
const BLOCK_OPEN = "[[BLOCK_OPEN]]";
|
|
|
|
const BLOCK_CLOSE = "[[BLOCK_CLOSE]]";
|
|
|
|
const NEWLINE = "[[NEWLINE]]";
|
|
|
|
const NEWLINE_MERGED = "[[MERGED]]";
|
|
|
|
const SPACE = "[[SPACE]]";
|
2017-06-25 01:19:11 +02:00
|
|
|
|
2018-05-16 11:09:07 +02:00
|
|
|
function processMdArrayNewLines(md) {
|
2017-06-25 01:19:11 +02:00
|
|
|
while (md.length && md[0] == BLOCK_OPEN) {
|
|
|
|
md.shift();
|
|
|
|
}
|
|
|
|
|
|
|
|
while (md.length && md[md.length - 1] == BLOCK_CLOSE) {
|
|
|
|
md.pop();
|
|
|
|
}
|
|
|
|
|
|
|
|
let temp = [];
|
2018-03-09 22:59:12 +02:00
|
|
|
let last = '';
|
|
|
|
for (let i = 0; i < md.length; i++) { let v = md[i];
|
2017-06-25 01:19:11 +02:00
|
|
|
if (isNewLineBlock(last) && isNewLineBlock(v) && last == v) {
|
|
|
|
// Skip it
|
|
|
|
} else {
|
|
|
|
temp.push(v);
|
|
|
|
}
|
|
|
|
last = v;
|
|
|
|
}
|
|
|
|
md = temp;
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
|
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
temp = [];
|
|
|
|
last = "";
|
2018-03-09 22:59:12 +02:00
|
|
|
for (let i = 0; i < md.length; i++) { let v = md[i];
|
2017-06-25 01:19:11 +02:00
|
|
|
if (last == BLOCK_CLOSE && v == BLOCK_OPEN) {
|
|
|
|
temp.pop();
|
|
|
|
temp.push(NEWLINE_MERGED);
|
|
|
|
} else {
|
|
|
|
temp.push(v);
|
|
|
|
}
|
|
|
|
last = v;
|
|
|
|
}
|
|
|
|
md = temp;
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
|
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
temp = [];
|
|
|
|
last = "";
|
2018-03-09 22:59:12 +02:00
|
|
|
for (let i = 0; i < md.length; i++) { let v = md[i];
|
2017-06-25 01:19:11 +02:00
|
|
|
if (last == NEWLINE && (v == NEWLINE_MERGED || v == BLOCK_CLOSE)) {
|
|
|
|
// Skip it
|
|
|
|
} else {
|
|
|
|
temp.push(v);
|
|
|
|
}
|
|
|
|
last = v;
|
|
|
|
}
|
|
|
|
md = temp;
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
|
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
// NEW!!!
|
|
|
|
temp = [];
|
|
|
|
last = "";
|
2018-03-09 22:59:12 +02:00
|
|
|
for (let i = 0; i < md.length; i++) { let v = md[i];
|
2017-06-25 01:19:11 +02:00
|
|
|
if (last == NEWLINE && (v == NEWLINE_MERGED || v == BLOCK_OPEN)) {
|
|
|
|
// Skip it
|
|
|
|
} else {
|
|
|
|
temp.push(v);
|
|
|
|
}
|
|
|
|
last = v;
|
|
|
|
}
|
|
|
|
md = temp;
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
|
|
|
|
|
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
if (md.length > 2) {
|
|
|
|
if (md[md.length - 2] == NEWLINE_MERGED && md[md.length - 1] == NEWLINE) {
|
|
|
|
md.pop();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
let output = '';
|
|
|
|
let previous = '';
|
2017-06-25 01:19:11 +02:00
|
|
|
let start = true;
|
2018-03-09 22:59:12 +02:00
|
|
|
for (let i = 0; i < md.length; i++) { let v = md[i];
|
|
|
|
let add = '';
|
2017-06-25 01:19:11 +02:00
|
|
|
if (v == BLOCK_CLOSE || v == BLOCK_OPEN || v == NEWLINE || v == NEWLINE_MERGED) {
|
|
|
|
add = "\n";
|
|
|
|
} else if (v == SPACE) {
|
|
|
|
if (previous == SPACE || previous == "\n" || start) {
|
|
|
|
continue; // skip
|
|
|
|
} else {
|
|
|
|
add = " ";
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
add = v;
|
|
|
|
}
|
|
|
|
start = false;
|
|
|
|
output += add;
|
|
|
|
previous = add;
|
|
|
|
}
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
if (!output.trim().length) return '';
|
2017-06-25 01:19:11 +02:00
|
|
|
|
2018-05-16 11:56:30 +02:00
|
|
|
// To simplify the result, we only allow up to one empty line between blocks of text
|
|
|
|
const mergeMultipleNewLines = function(lines) {
|
|
|
|
let output = [];
|
|
|
|
let newlineCount = 0;
|
|
|
|
for (let i = 0; i < lines.length; i++) {
|
|
|
|
const line = lines[i];
|
|
|
|
if (!line.trim()) {
|
|
|
|
newlineCount++;
|
|
|
|
} else {
|
|
|
|
newlineCount = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (newlineCount >= 2) continue;
|
|
|
|
|
|
|
|
output.push(line);
|
|
|
|
}
|
|
|
|
return output;
|
|
|
|
}
|
|
|
|
|
|
|
|
let lines = output.replace(/\\r/g, '').split('\n');
|
|
|
|
lines = formatMdLayout(lines)
|
|
|
|
lines = mergeMultipleNewLines(lines);
|
|
|
|
return lines.join('\n');
|
|
|
|
}
|
|
|
|
|
|
|
|
// While the processMdArrayNewLines() function adds newlines in a way that's technically correct, the resulting Markdown can look messy.
|
|
|
|
// This is because while a "block" element should be surrounded by newlines, in practice, some should be surrounded by TWO new lines, while
|
|
|
|
// others by only ONE.
|
|
|
|
//
|
|
|
|
// For instance, this:
|
|
|
|
//
|
|
|
|
// <li>one</li>
|
|
|
|
// <li>two</li>
|
|
|
|
// <li>three</li>
|
|
|
|
//
|
|
|
|
// should result in this:
|
|
|
|
//
|
|
|
|
// - one
|
|
|
|
// - two
|
|
|
|
// - three
|
|
|
|
//
|
|
|
|
// While this:
|
|
|
|
//
|
|
|
|
// <p>Some long paragraph</p><p>And another one</p><p>And the last paragraph</p>
|
|
|
|
//
|
|
|
|
// should result in this:
|
|
|
|
//
|
|
|
|
// Some long paragraph
|
|
|
|
//
|
|
|
|
// And another one
|
|
|
|
//
|
|
|
|
// And the last paragraph
|
|
|
|
//
|
|
|
|
// So in one case, one newline between tags, and in another two newlines. In HTML this would be done via CSS, but in Markdown we need
|
|
|
|
// to add new lines. It's also important to get these newlines right because two blocks of text next to each others might be renderered
|
|
|
|
// differently than if there's a newlines between them. So the function below parses the almost final MD and add new lines depending
|
|
|
|
// on various rules.
|
|
|
|
|
|
|
|
const isHeading = function(line) {
|
|
|
|
return !!line.match(/^#+\s/);
|
|
|
|
}
|
|
|
|
|
|
|
|
const isListItem = function(line) {
|
|
|
|
return line && line.trim().indexOf('- ') === 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
const isCodeLine = function(line) {
|
|
|
|
return line && line.indexOf('\t') === 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
const isTableLine = function(line) {
|
|
|
|
return line.indexOf('| ') === 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
const isPlainParagraph = function(line) {
|
|
|
|
// Note: if a line is no longer than 80 characters, we don't consider it's a paragraph, which
|
|
|
|
// means no newlines will be added before or after. This is to handle text that has been
|
|
|
|
// written with "hard" new lines.
|
|
|
|
if (!line || line.length < 80) return false;
|
|
|
|
|
|
|
|
if (isListItem(line)) return false;
|
|
|
|
if (isHeading(line)) return false;
|
|
|
|
if (isCodeLine(line)) return false;
|
|
|
|
if (isTableLine(line)) return false;
|
|
|
|
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
|
|
|
|
function formatMdLayout(lines) {
|
|
|
|
let previous = '';
|
|
|
|
let newLines = [];
|
|
|
|
for (let i = 0; i < lines.length; i++) {
|
|
|
|
const line = lines[i];
|
|
|
|
|
|
|
|
// Add a new line at the end of a list of items
|
|
|
|
if (isListItem(previous) && line && !isListItem(line)) {
|
|
|
|
newLines.push('');
|
|
|
|
|
|
|
|
// Add a new line at the beginning of a list of items
|
|
|
|
} else if (isListItem(line) && previous && !isListItem(previous)) {
|
|
|
|
newLines.push('');
|
|
|
|
|
|
|
|
// Add a new line before a heading
|
|
|
|
} else if (isHeading(line) && previous) {
|
|
|
|
newLines.push('');
|
|
|
|
|
|
|
|
// Add a new line after a heading
|
|
|
|
} else if (isHeading(previous) && line) {
|
|
|
|
newLines.push('');
|
|
|
|
|
|
|
|
} else if (isCodeLine(line) && !isCodeLine(previous)) {
|
|
|
|
newLines.push('');
|
|
|
|
|
|
|
|
} else if (!isCodeLine(line) && isCodeLine(previous)) {
|
|
|
|
newLines.push('');
|
|
|
|
|
|
|
|
} else if (isTableLine(line) && !isTableLine(previous)) {
|
|
|
|
newLines.push('');
|
|
|
|
|
|
|
|
} else if (!isTableLine(line) && isTableLine(previous)) {
|
|
|
|
newLines.push('');
|
|
|
|
|
|
|
|
// Add a new line at beginning of paragraph
|
|
|
|
} else if (isPlainParagraph(line) && previous) {
|
|
|
|
newLines.push('');
|
|
|
|
|
|
|
|
// Add a new line at end of paragraph
|
|
|
|
} else if (isPlainParagraph(previous) && line) {
|
|
|
|
newLines.push('');
|
|
|
|
}
|
|
|
|
|
|
|
|
newLines.push(line);
|
|
|
|
previous = newLines[newLines.length - 1];
|
|
|
|
}
|
|
|
|
|
|
|
|
return newLines;
|
2018-05-12 12:48:39 +02:00
|
|
|
}
|
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
function isWhiteSpace(c) {
|
2018-03-09 22:59:12 +02:00
|
|
|
return c == '\n' || c == '\r' || c == '\v' || c == '\f' || c == '\t' || c == ' ';
|
2017-06-25 01:19:11 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
// Like QString::simpified(), except that it preserves non-breaking spaces (which
|
|
|
|
// Evernote uses for identation, etc.)
|
|
|
|
function simplifyString(s) {
|
2018-03-09 22:59:12 +02:00
|
|
|
let output = '';
|
2017-06-25 01:19:11 +02:00
|
|
|
let previousWhite = false;
|
|
|
|
for (let i = 0; i < s.length; i++) {
|
|
|
|
let c = s[i];
|
|
|
|
let isWhite = isWhiteSpace(c);
|
|
|
|
if (previousWhite && isWhite) {
|
|
|
|
// skip
|
|
|
|
} else {
|
|
|
|
output += c;
|
|
|
|
}
|
|
|
|
previousWhite = isWhite;
|
|
|
|
}
|
|
|
|
|
|
|
|
while (output.length && isWhiteSpace(output[0])) output = output.substr(1);
|
|
|
|
while (output.length && isWhiteSpace(output[output.length - 1])) output = output.substr(0, output.length - 1);
|
|
|
|
|
|
|
|
return output;
|
|
|
|
}
|
|
|
|
|
|
|
|
function collapseWhiteSpaceAndAppend(lines, state, text) {
|
2018-05-16 11:56:30 +02:00
|
|
|
if (state.inCode.length) {
|
2018-05-15 14:13:55 +02:00
|
|
|
lines.push(text);
|
2017-06-25 01:19:11 +02:00
|
|
|
} else {
|
|
|
|
// Remove all \n and \r from the left and right of the text
|
|
|
|
while (text.length && (text[0] == "\n" || text[0] == "\r")) text = text.substr(1);
|
|
|
|
while (text.length && (text[text.length - 1] == "\n" || text[text.length - 1] == "\r")) text = text.substr(0, text.length - 1);
|
|
|
|
|
|
|
|
// Collapse all white spaces to just one. If there are spaces to the left and right of the string
|
|
|
|
// also collapse them to just one space.
|
2018-03-09 22:59:12 +02:00
|
|
|
let spaceLeft = text.length && text[0] == ' ';
|
|
|
|
let spaceRight = text.length && text[text.length - 1] == ' ';
|
2017-06-25 01:19:11 +02:00
|
|
|
text = simplifyString(text);
|
|
|
|
|
|
|
|
if (!spaceLeft && !spaceRight && text == "") return lines;
|
|
|
|
|
2017-12-02 14:49:42 +02:00
|
|
|
if (state.inQuote) {
|
|
|
|
// Add a ">" at the beginning of the block then at the beginning of each lines. So it turns this:
|
|
|
|
// "my quote\nsecond line" into this => "> my quote\n> second line"
|
2018-03-09 22:59:12 +02:00
|
|
|
lines.push('> ');
|
|
|
|
if (lines.indexOf('\r') >= 0) {
|
|
|
|
text = text.replace(/\n\r/g, '\n\r> ');
|
2017-12-02 14:49:42 +02:00
|
|
|
} else {
|
2018-03-09 22:59:12 +02:00
|
|
|
text = text.replace(/\n/g, '\n> ');
|
2017-12-02 14:49:42 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
if (spaceLeft) lines.push(SPACE);
|
|
|
|
lines.push(text);
|
|
|
|
if (spaceRight) lines.push(SPACE);
|
|
|
|
}
|
|
|
|
|
|
|
|
return lines;
|
|
|
|
}
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
const imageMimeTypes = ["image/cgm", "image/fits", "image/g3fax", "image/gif", "image/ief", "image/jp2", "image/jpeg", "image/jpm", "image/jpx", "image/naplps", "image/png", "image/prs.btif", "image/prs.pti", "image/t38", "image/tiff", "image/tiff-fx", "image/vnd.adobe.photoshop", "image/vnd.cns.inf2", "image/vnd.djvu", "image/vnd.dwg", "image/vnd.dxf", "image/vnd.fastbidsheet", "image/vnd.fpx", "image/vnd.fst", "image/vnd.fujixerox.edmics-mmr", "image/vnd.fujixerox.edmics-rlc", "image/vnd.globalgraphics.pgb", "image/vnd.microsoft.icon", "image/vnd.mix", "image/vnd.ms-modi", "image/vnd.net-fpx", "image/vnd.sealed.png", "image/vnd.sealedmedia.softseal.gif", "image/vnd.sealedmedia.softseal.jpg", "image/vnd.svf", "image/vnd.wap.wbmp", "image/vnd.xiff"];
|
2017-06-25 01:19:11 +02:00
|
|
|
|
|
|
|
function isImageMimeType(m) {
|
|
|
|
return imageMimeTypes.indexOf(m) >= 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
function addResourceTag(lines, resource, alt = "") {
|
2017-08-01 23:40:14 +02:00
|
|
|
// TODO: refactor to use Resource.markdownTag
|
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
let tagAlt = alt == "" ? resource.alt : alt;
|
2018-03-09 22:59:12 +02:00
|
|
|
if (!tagAlt) tagAlt = '';
|
2017-06-25 01:19:11 +02:00
|
|
|
if (isImageMimeType(resource.mime)) {
|
|
|
|
lines.push("![");
|
|
|
|
lines.push(tagAlt);
|
|
|
|
lines.push("](:/" + resource.id + ")");
|
|
|
|
} else {
|
|
|
|
lines.push("[");
|
|
|
|
lines.push(tagAlt);
|
|
|
|
lines.push("](:/" + resource.id + ")");
|
|
|
|
}
|
|
|
|
|
|
|
|
return lines;
|
|
|
|
}
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
function isBlockTag(n) {
|
2018-05-16 11:09:07 +02:00
|
|
|
return ["div", "p", "dl", "dd", 'dt', "center", 'address'].indexOf(n) >= 0;
|
2017-06-25 01:19:11 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
function isStrongTag(n) {
|
2018-03-09 22:59:12 +02:00
|
|
|
return n == "strong" || n == "b" || n == 'big';
|
2017-12-06 21:29:58 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
function isStrikeTag(n) {
|
2018-03-09 22:59:12 +02:00
|
|
|
return n == "strike" || n == "s" || n == 'del';
|
2017-06-25 01:19:11 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
function isEmTag(n) {
|
|
|
|
return n == "em" || n == "i" || n == "u";
|
|
|
|
}
|
|
|
|
|
|
|
|
function isAnchor(n) {
|
|
|
|
return n == "a";
|
|
|
|
}
|
|
|
|
|
|
|
|
function isIgnoredEndTag(n) {
|
2018-05-16 11:09:07 +02:00
|
|
|
return ["en-note", "en-todo", "span", "body", "html", "font", "br", 'hr', 'tbody', 'sup', 'img', 'abbr', 'cite', 'thead', 'small', 'tt', 'sub', 'colgroup', 'col', 'ins', 'caption', 'var', 'map', 'area'].indexOf(n) >= 0;
|
2017-06-25 01:19:11 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
function isListTag(n) {
|
|
|
|
return n == "ol" || n == "ul";
|
|
|
|
}
|
|
|
|
|
|
|
|
// Elements that don't require any special treatment beside adding a newline character
|
|
|
|
function isNewLineOnlyEndTag(n) {
|
2018-05-16 11:09:07 +02:00
|
|
|
return ["div", "p", "li", "h1", "h2", "h3", "h4", "h5", 'h6', "dl", "dd", 'dt', "center", 'address'].indexOf(n) >= 0;
|
|
|
|
}
|
|
|
|
|
2017-12-08 19:41:32 +02:00
|
|
|
function isInlineCodeTag(n) {
|
2018-03-09 22:59:12 +02:00
|
|
|
return ['samp', 'kbd'].indexOf(n) >= 0;
|
2017-12-08 19:41:32 +02:00
|
|
|
}
|
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
function isNewLineBlock(s) {
|
|
|
|
return s == BLOCK_OPEN || s == BLOCK_CLOSE;
|
|
|
|
}
|
|
|
|
|
|
|
|
function xmlNodeText(xmlNode) {
|
2018-03-09 22:59:12 +02:00
|
|
|
if (!xmlNode || !xmlNode.length) return '';
|
2017-06-25 01:19:11 +02:00
|
|
|
return xmlNode[0];
|
|
|
|
}
|
|
|
|
|
2017-12-06 21:29:58 +02:00
|
|
|
function attributeToLowerCase(node) {
|
|
|
|
if (!node.attributes) return {};
|
|
|
|
let output = {};
|
|
|
|
for (let n in node.attributes) {
|
|
|
|
if (!node.attributes.hasOwnProperty(n)) continue;
|
|
|
|
output[n.toLowerCase()] = node.attributes[n];
|
|
|
|
}
|
|
|
|
return output;
|
|
|
|
}
|
|
|
|
|
2018-05-16 11:09:07 +02:00
|
|
|
function enexXmlToMdArray(stream, resources) {
|
2017-12-06 21:29:58 +02:00
|
|
|
let remainingResources = resources.slice();
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
const removeRemainingResource = (id) => {
|
2017-12-06 21:29:58 +02:00
|
|
|
for (let i = 0; i < remainingResources.length; i++) {
|
|
|
|
const r = remainingResources[i];
|
|
|
|
if (r.id === id) {
|
|
|
|
remainingResources.splice(i, 1);
|
|
|
|
}
|
|
|
|
}
|
2018-03-09 22:59:12 +02:00
|
|
|
}
|
2017-06-25 01:19:11 +02:00
|
|
|
|
|
|
|
return new Promise((resolve, reject) => {
|
|
|
|
let state = {
|
2018-05-16 11:56:30 +02:00
|
|
|
inCode: [],
|
|
|
|
inPre: false,
|
2017-12-02 14:49:42 +02:00
|
|
|
inQuote: false,
|
2017-06-25 01:19:11 +02:00
|
|
|
lists: [],
|
|
|
|
anchorAttributes: [],
|
|
|
|
};
|
|
|
|
|
2018-05-16 11:09:07 +02:00
|
|
|
let options = {};
|
2017-12-06 21:29:58 +02:00
|
|
|
let strict = false;
|
2018-05-16 11:09:07 +02:00
|
|
|
var saxStream = require('sax').createStream(strict, options)
|
2017-06-25 01:19:11 +02:00
|
|
|
|
2017-07-13 21:29:10 +02:00
|
|
|
let section = {
|
2018-03-09 22:59:12 +02:00
|
|
|
type: 'text',
|
2017-07-13 21:29:10 +02:00
|
|
|
lines: [],
|
|
|
|
parent: null,
|
|
|
|
};
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
saxStream.on('error', function(e) {
|
2017-12-06 21:29:58 +02:00
|
|
|
console.warn(e);
|
2018-03-09 22:59:12 +02:00
|
|
|
//reject(e);
|
|
|
|
})
|
2017-06-25 01:19:11 +02:00
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
saxStream.on('text', function(text) {
|
|
|
|
if (['table', 'tr', 'tbody'].indexOf(section.type) >= 0) return;
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines = collapseWhiteSpaceAndAppend(section.lines, state, text);
|
2018-03-09 22:59:12 +02:00
|
|
|
})
|
2017-06-25 01:19:11 +02:00
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
saxStream.on('opentag', function(node) {
|
2017-12-06 21:29:58 +02:00
|
|
|
const nodeAttributes = attributeToLowerCase(node);
|
2017-06-25 01:19:11 +02:00
|
|
|
let n = node.name.toLowerCase();
|
2018-10-01 21:55:24 +02:00
|
|
|
|
|
|
|
const currentList = state.lists && state.lists.length ? state.lists[state.lists.length - 1] : null;
|
|
|
|
|
|
|
|
// Kind of a hack: If we are inside a list, at the beginning of an item (when a "- " or "1. " has been added
|
|
|
|
// but no other text yet), if the current tag is eg. a <div> or any other block tag, we skip it, so that a new line
|
|
|
|
// does not get created. It is to handle list4.html test case.
|
|
|
|
// https://github.com/laurent22/joplin/issues/832
|
|
|
|
if (currentList) {
|
|
|
|
if (!currentList.startedText && isBlockTag(n)) return;
|
|
|
|
currentList.startedText = true;
|
|
|
|
}
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
if (n == 'en-note') {
|
2017-06-25 01:19:11 +02:00
|
|
|
// Start of note
|
|
|
|
} else if (isBlockTag(n)) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push(BLOCK_OPEN);
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'table') {
|
2017-07-13 21:29:10 +02:00
|
|
|
let newSection = {
|
2018-03-09 22:59:12 +02:00
|
|
|
type: 'table',
|
2017-07-13 21:29:10 +02:00
|
|
|
lines: [],
|
|
|
|
parent: section,
|
|
|
|
};
|
|
|
|
section.lines.push(newSection);
|
|
|
|
section = newSection;
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'tbody' || n == 'thead') {
|
2017-07-13 21:29:10 +02:00
|
|
|
// Ignore it
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'tr') {
|
|
|
|
if (section.type != 'table') {
|
|
|
|
console.warn('Found a <tr> tag outside of a table');
|
2017-12-06 21:29:58 +02:00
|
|
|
return;
|
|
|
|
}
|
2017-07-13 21:29:10 +02:00
|
|
|
|
|
|
|
let newSection = {
|
2018-03-09 22:59:12 +02:00
|
|
|
type: 'tr',
|
2017-07-13 21:29:10 +02:00
|
|
|
lines: [],
|
|
|
|
parent: section,
|
2017-07-17 20:59:40 +02:00
|
|
|
isHeader: false,
|
2018-03-09 22:59:12 +02:00
|
|
|
}
|
2017-07-13 21:29:10 +02:00
|
|
|
|
|
|
|
section.lines.push(newSection);
|
|
|
|
section = newSection;
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'td' || n == 'th') {
|
|
|
|
if (section.type != 'tr') {
|
|
|
|
console.warn('Found a <td> tag outside of a <tr>');
|
2017-12-06 21:29:58 +02:00
|
|
|
return;
|
|
|
|
}
|
2017-07-13 21:29:10 +02:00
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
if (n == 'th') section.isHeader = true;
|
2017-07-17 20:59:40 +02:00
|
|
|
|
2017-07-13 21:29:10 +02:00
|
|
|
let newSection = {
|
2018-03-09 22:59:12 +02:00
|
|
|
type: 'td',
|
2017-07-13 21:29:10 +02:00
|
|
|
lines: [],
|
|
|
|
parent: section,
|
|
|
|
};
|
|
|
|
|
|
|
|
section.lines.push(newSection);
|
|
|
|
section = newSection;
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (isListTag(n)) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push(BLOCK_OPEN);
|
2018-10-01 21:55:24 +02:00
|
|
|
state.lists.push({ tag: n, counter: 1, startedText: false });
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'li') {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push(BLOCK_OPEN);
|
2017-06-25 01:19:11 +02:00
|
|
|
if (!state.lists.length) {
|
2018-05-16 11:56:30 +02:00
|
|
|
console.warn("Found <li> tag without being inside a list");
|
2017-06-25 01:19:11 +02:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
let container = state.lists[state.lists.length - 1];
|
2018-10-01 21:55:24 +02:00
|
|
|
container.startedText = false;
|
2017-06-25 01:19:11 +02:00
|
|
|
if (container.tag == "ul") {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push("- ");
|
2017-06-25 01:19:11 +02:00
|
|
|
} else {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push(container.counter + '. ');
|
2017-06-25 01:19:11 +02:00
|
|
|
container.counter++;
|
|
|
|
}
|
|
|
|
} else if (isStrongTag(n)) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push("**");
|
2017-12-06 21:29:58 +02:00
|
|
|
} else if (isStrikeTag(n)) {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push('(');
|
2017-12-08 19:41:32 +02:00
|
|
|
} else if (isInlineCodeTag(n)) {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push('`');
|
|
|
|
} else if (n == 'q') {
|
2017-12-02 14:49:42 +02:00
|
|
|
section.lines.push('"');
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'img') {
|
2018-05-16 11:09:07 +02:00
|
|
|
// TODO: TEST IMAGE
|
2018-03-09 22:59:12 +02:00
|
|
|
if (nodeAttributes.src) { // Many (most?) img tags don't have no source associated, especially when they were imported from HTML
|
|
|
|
let s = '![';
|
2017-12-06 21:29:58 +02:00
|
|
|
if (nodeAttributes.alt) s += nodeAttributes.alt;
|
2018-05-16 11:09:07 +02:00
|
|
|
s += '](' + nodeAttributes.src + ')';
|
2017-12-06 21:29:58 +02:00
|
|
|
section.lines.push(s);
|
|
|
|
}
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (isAnchor(n)) {
|
2017-12-06 21:29:58 +02:00
|
|
|
state.anchorAttributes.push(nodeAttributes);
|
2018-05-16 11:56:30 +02:00
|
|
|
// Need to add the '[' via this function to make sure that links within code blocks
|
|
|
|
// are handled correctly.
|
|
|
|
collapseWhiteSpaceAndAppend(section.lines, state, '[');
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (isEmTag(n)) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push("*");
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "en-todo") {
|
2018-03-09 22:59:12 +02:00
|
|
|
let x = nodeAttributes && nodeAttributes.checked && nodeAttributes.checked.toLowerCase() == 'true' ? 'X' : ' ';
|
|
|
|
section.lines.push('- [' + x + '] ');
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "hr") {
|
2017-07-15 18:14:15 +02:00
|
|
|
// Needs to be surrounded by new lines so that it's properly rendered as a line when converting to HTML
|
|
|
|
section.lines.push(NEWLINE);
|
2018-05-16 11:56:30 +02:00
|
|
|
section.lines.push('* * *');
|
2017-07-15 18:14:15 +02:00
|
|
|
section.lines.push(NEWLINE);
|
|
|
|
section.lines.push(NEWLINE);
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "h1") {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push(BLOCK_OPEN); section.lines.push("# ");
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "h2") {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push(BLOCK_OPEN); section.lines.push("## ");
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "h3") {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push(BLOCK_OPEN); section.lines.push("### ");
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "h4") {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push(BLOCK_OPEN); section.lines.push("#### ");
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "h5") {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push(BLOCK_OPEN); section.lines.push("##### ");
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "h6") {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push(BLOCK_OPEN); section.lines.push("###### ");
|
|
|
|
} else if (n == 'blockquote') {
|
2017-12-02 14:49:42 +02:00
|
|
|
section.lines.push(BLOCK_OPEN);
|
|
|
|
state.inQuote = true;
|
2018-05-16 11:56:30 +02:00
|
|
|
} else if (n === 'code') {
|
|
|
|
state.inCode.push(true);
|
|
|
|
state.currentCode = '';
|
|
|
|
|
|
|
|
let newSection = {
|
|
|
|
type: 'code',
|
|
|
|
lines: [],
|
|
|
|
parent: section,
|
|
|
|
}
|
|
|
|
|
|
|
|
section.lines.push(newSection);
|
|
|
|
section = newSection;
|
|
|
|
} else if (n === 'pre') {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push(BLOCK_OPEN);
|
2018-05-16 11:56:30 +02:00
|
|
|
state.inPre = true;
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "br") {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push(NEWLINE);
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (n == "en-media") {
|
2017-12-06 21:29:58 +02:00
|
|
|
const hash = nodeAttributes.hash;
|
2017-06-25 01:19:11 +02:00
|
|
|
|
|
|
|
let resource = null;
|
|
|
|
for (let i = 0; i < resources.length; i++) {
|
|
|
|
let r = resources[i];
|
|
|
|
if (r.id == hash) {
|
|
|
|
resource = r;
|
2017-12-06 21:29:58 +02:00
|
|
|
removeRemainingResource(r.id);
|
2017-06-25 01:19:11 +02:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!resource) {
|
|
|
|
// This is a bit of a hack. Notes sometime have resources attached to it, but those <resource> tags don't contain
|
|
|
|
// an "objID" tag, making it impossible to reference the resource. However, in this case the content of the note
|
|
|
|
// will contain a corresponding <en-media/> tag, which has the ID in the "hash" attribute. All this information
|
|
|
|
// has been collected above so we now set the resource ID to the hash attribute of the en-media tags. Here's an
|
|
|
|
// example of note that shows this problem:
|
|
|
|
|
|
|
|
// <?xml version="1.0" encoding="UTF-8"?>
|
|
|
|
// <!DOCTYPE en-export SYSTEM "http://xml.evernote.com/pub/evernote-export2.dtd">
|
|
|
|
// <en-export export-date="20161221T203133Z" application="Evernote/Windows" version="6.x">
|
|
|
|
// <note>
|
|
|
|
// <title>Commande</title>
|
|
|
|
// <content>
|
|
|
|
// <![CDATA[
|
|
|
|
// <?xml version="1.0" encoding="UTF-8"?>
|
|
|
|
// <!DOCTYPE en-note SYSTEM "http://xml.evernote.com/pub/enml2.dtd">
|
|
|
|
// <en-note>
|
|
|
|
// <en-media alt="your QR code" hash="216a16a1bbe007fba4ccf60b118b4ccc" type="image/png"></en-media>
|
|
|
|
// </en-note>
|
|
|
|
// ]]>
|
|
|
|
// </content>
|
|
|
|
// <created>20160921T203424Z</created>
|
|
|
|
// <updated>20160921T203438Z</updated>
|
|
|
|
// <note-attributes>
|
|
|
|
// <reminder-order>20160902T140445Z</reminder-order>
|
|
|
|
// <reminder-done-time>20160924T101120Z</reminder-done-time>
|
|
|
|
// </note-attributes>
|
|
|
|
// <resource>
|
|
|
|
// <data encoding="base64">........</data>
|
|
|
|
// <mime>image/png</mime>
|
|
|
|
// <width>150</width>
|
|
|
|
// <height>150</height>
|
|
|
|
// </resource>
|
|
|
|
// </note>
|
|
|
|
// </en-export>
|
|
|
|
|
2018-03-23 19:59:18 +02:00
|
|
|
// Note that there's also the case of resources with no ID where the ID is actually the MD5 of the content.
|
|
|
|
// This is handled in import-enex.js
|
|
|
|
|
2017-06-25 01:19:11 +02:00
|
|
|
let found = false;
|
2017-12-06 21:29:58 +02:00
|
|
|
for (let i = 0; i < remainingResources.length; i++) {
|
|
|
|
let r = remainingResources[i];
|
2017-06-25 01:19:11 +02:00
|
|
|
if (!r.id) {
|
2018-03-23 19:59:18 +02:00
|
|
|
resource = Object.assign({}, r);
|
|
|
|
resource.id = hash;
|
|
|
|
remainingResources.splice(i, 1);
|
2017-06-25 01:19:11 +02:00
|
|
|
found = true;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!found) {
|
2018-03-09 22:59:12 +02:00
|
|
|
console.warn('Hash with no associated resource: ' + hash);
|
2017-06-25 01:19:11 +02:00
|
|
|
}
|
2018-03-23 19:59:18 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
// If the resource does not appear among the note's resources, it
|
|
|
|
// means it's an attachement. It will be appended along with the
|
|
|
|
// other remaining resources at the bottom of the markdown text.
|
|
|
|
if (resource && !!resource.id) {
|
|
|
|
section.lines = addResourceTag(section.lines, resource, nodeAttributes.alt);
|
2017-06-25 01:19:11 +02:00
|
|
|
}
|
2018-05-16 11:09:07 +02:00
|
|
|
} else if (["span", "font", 'sup', 'cite', 'abbr', 'small', 'tt', 'sub', 'colgroup', 'col', 'ins', 'caption', 'var', 'map', 'area'].indexOf(n) >= 0) {
|
2017-12-06 21:29:58 +02:00
|
|
|
// Inline tags that can be ignored in Markdown
|
2017-06-25 01:19:11 +02:00
|
|
|
} else {
|
2018-05-16 11:09:07 +02:00
|
|
|
console.warn("Unsupported start tag: " + n);
|
2017-06-25 01:19:11 +02:00
|
|
|
}
|
2018-03-09 22:59:12 +02:00
|
|
|
})
|
2017-06-25 01:19:11 +02:00
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
saxStream.on('closetag', function(n) {
|
2017-12-06 21:29:58 +02:00
|
|
|
n = n ? n.toLowerCase() : n;
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
if (n == 'en-note') {
|
2017-06-25 01:19:11 +02:00
|
|
|
// End of note
|
|
|
|
} else if (isNewLineOnlyEndTag(n)) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push(BLOCK_CLOSE);
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'td' || n == 'th') {
|
2017-12-06 21:29:58 +02:00
|
|
|
if (section && section.parent) section = section.parent;
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'tr') {
|
2017-12-06 21:29:58 +02:00
|
|
|
if (section && section.parent) section = section.parent;
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'table') {
|
2017-12-06 21:29:58 +02:00
|
|
|
if (section && section.parent) section = section.parent;
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (isIgnoredEndTag(n)) {
|
|
|
|
// Skip
|
|
|
|
} else if (isListTag(n)) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push(BLOCK_CLOSE);
|
2017-06-25 01:19:11 +02:00
|
|
|
state.lists.pop();
|
|
|
|
} else if (isStrongTag(n)) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push("**");
|
2017-12-06 21:29:58 +02:00
|
|
|
} else if (isStrikeTag(n)) {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push(')');
|
2017-12-08 19:41:32 +02:00
|
|
|
} else if (isInlineCodeTag(n)) {
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push('`');
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (isEmTag(n)) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push("*");
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'q') {
|
2017-12-02 14:49:42 +02:00
|
|
|
section.lines.push('"');
|
2018-03-09 22:59:12 +02:00
|
|
|
} else if (n == 'blockquote') {
|
2017-12-02 14:49:42 +02:00
|
|
|
section.lines.push(BLOCK_OPEN);
|
|
|
|
state.inQuote = false;
|
2018-05-16 11:56:30 +02:00
|
|
|
} else if (n === 'code') {
|
|
|
|
state.inCode.pop();
|
|
|
|
|
|
|
|
if (!state.inCode.length) {
|
|
|
|
const codeLines = section.lines.join('').split('\n');
|
|
|
|
section.lines = [];
|
|
|
|
if (codeLines.length > 1) {
|
|
|
|
for (let i = 0; i < codeLines.length; i++) {
|
|
|
|
if (i > 0) section.lines.push('\n');
|
|
|
|
section.lines.push('\t' + codeLines[i]);
|
|
|
|
}
|
|
|
|
} else {
|
|
|
|
section.lines.push('`' + codeLines.join('') + '`');
|
|
|
|
}
|
|
|
|
|
|
|
|
if (section && section.parent) section = section.parent;
|
|
|
|
}
|
|
|
|
} else if (n === 'pre') {
|
|
|
|
state.inPre = false;
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push(BLOCK_CLOSE);
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (isAnchor(n)) {
|
|
|
|
let attributes = state.anchorAttributes.pop();
|
2018-03-09 22:59:12 +02:00
|
|
|
let url = attributes && attributes.href ? attributes.href : '';
|
2017-07-13 21:16:01 +02:00
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
if (section.lines.length < 1) throw new Error('Invalid anchor tag closing'); // Sanity check, but normally not possible
|
2017-07-13 21:16:01 +02:00
|
|
|
|
|
|
|
// When closing the anchor tag, check if there's is any text content. If not
|
|
|
|
// put the URL as is (don't wrap it in [](url)). The markdown parser, using
|
|
|
|
// GitHub flavour, will turn this URL into a link. This is to generate slightly
|
|
|
|
// cleaner markdown.
|
2018-05-16 11:09:46 +02:00
|
|
|
|
|
|
|
// Need to loop on the previous tags so as to skip the special ones, which are not relevant for the below algorithm.
|
|
|
|
let previous = null;
|
|
|
|
for (let i = section.lines.length - 1; i >= 0; i--) {
|
|
|
|
previous = section.lines[i];
|
|
|
|
if ([BLOCK_OPEN, BLOCK_CLOSE, NEWLINE, NEWLINE_MERGED, SPACE].indexOf(previous) >= 0 || !previous) {
|
|
|
|
continue;
|
|
|
|
} else {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
if (previous == '[') {
|
2018-05-16 11:09:46 +02:00
|
|
|
// We have a link that had some content but, after parsing, nothing is left. The content was most likely
|
|
|
|
// something that shows up via CSS and which we cannot support. For example:
|
|
|
|
//
|
|
|
|
// <a onclick="return vote()" href="vote?id=17045576">
|
|
|
|
// <div class="votearrow" title="upvote"></div>
|
|
|
|
// </a>
|
|
|
|
//
|
|
|
|
// In the case above the arrow is displayed via CSS.
|
|
|
|
// It is useless to display the full URL since often it is not relevant for a note (for example
|
|
|
|
// it's interactive bits) and it's not user-generated content such as a URL that would appear in a comment.
|
|
|
|
// So in this case, we still want to preserve the information but display it in a discreet way as a simple [L].
|
|
|
|
|
|
|
|
// Need to pop everything inside the current [] because it can only be special chars that we don't want (they would create uncessary newlines)
|
|
|
|
for (let i = section.lines.length - 1; i >= 0; i--) {
|
|
|
|
if (section.lines[i] !== '[') {
|
|
|
|
section.lines.pop();
|
|
|
|
} else {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!url) {
|
|
|
|
// If there's no URL and no content, pop the [ and don't save any content.
|
|
|
|
section.lines.pop();
|
|
|
|
} else {
|
|
|
|
section.lines.push('(L)');
|
|
|
|
section.lines.push('](' + url + ')');
|
|
|
|
}
|
2017-07-13 21:16:01 +02:00
|
|
|
} else if (!previous || previous == url) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.pop();
|
|
|
|
section.lines.pop();
|
2018-03-23 19:59:18 +02:00
|
|
|
section.lines.push(url);
|
2017-07-13 21:16:01 +02:00
|
|
|
} else {
|
2017-12-07 02:35:15 +02:00
|
|
|
// Need to remove any new line character between the current ']' and the previous '['
|
|
|
|
// otherwise it won't render properly.
|
2017-12-07 02:57:36 +02:00
|
|
|
let allSpaces = true;
|
2017-12-07 02:35:15 +02:00
|
|
|
for (let i = section.lines.length - 1; i >= 0; i--) {
|
|
|
|
const c = section.lines[i];
|
2018-03-09 22:59:12 +02:00
|
|
|
if (c === '[') {
|
2017-12-07 02:35:15 +02:00
|
|
|
break;
|
|
|
|
} else {
|
|
|
|
if (c === BLOCK_CLOSE || c === BLOCK_OPEN || c === NEWLINE) {
|
|
|
|
section.lines[i] = SPACE;
|
2017-12-07 02:57:36 +02:00
|
|
|
} else {
|
|
|
|
if (!isWhiteSpace(c)) allSpaces = false;
|
2017-12-07 02:35:15 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2017-12-07 02:57:36 +02:00
|
|
|
|
|
|
|
if (allSpaces) {
|
|
|
|
for (let i = section.lines.length - 1; i >= 0; i--) {
|
|
|
|
const c = section.lines.pop();
|
2018-03-09 22:59:12 +02:00
|
|
|
if (c === '[') break;
|
|
|
|
}
|
2018-03-23 19:59:18 +02:00
|
|
|
section.lines.push(url);
|
2017-12-07 02:57:36 +02:00
|
|
|
} else {
|
2018-05-16 11:09:46 +02:00
|
|
|
|
|
|
|
// Eg. converts:
|
|
|
|
// [ Sign in ](https://example.com)
|
|
|
|
// to:
|
|
|
|
// [Sign in](https://example.com)
|
|
|
|
const trimTextStartAndEndSpaces = function(lines) {
|
|
|
|
let firstBracketIndex = 0;
|
|
|
|
let foundFirstNonWhite = false;
|
|
|
|
for (let i = lines.length - 1; i >= 0; i--) {
|
|
|
|
const l = lines[i];
|
|
|
|
if (!foundFirstNonWhite && (l === SPACE || l === ' ' || !l)) {
|
|
|
|
lines.pop();
|
|
|
|
} else {
|
|
|
|
foundFirstNonWhite = true;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (l === '[') {
|
|
|
|
firstBracketIndex = i;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
for (let i = firstBracketIndex + 1; i < lines.length; i++) {
|
|
|
|
const l = lines[i];
|
|
|
|
if (l === SPACE || l === ' ' ||!l) {
|
|
|
|
lines.splice(i, 1);
|
|
|
|
} else {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return lines;
|
|
|
|
}
|
|
|
|
|
|
|
|
section.lines = trimTextStartAndEndSpaces(section.lines);
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
section.lines.push('](' + url + ')');
|
2017-12-07 02:57:36 +02:00
|
|
|
}
|
2017-07-13 21:16:01 +02:00
|
|
|
}
|
2017-06-25 01:19:11 +02:00
|
|
|
} else if (isListTag(n)) {
|
2017-07-13 21:29:10 +02:00
|
|
|
section.lines.push(BLOCK_CLOSE);
|
2017-06-25 01:19:11 +02:00
|
|
|
state.lists.pop();
|
|
|
|
} else if (n == "en-media") {
|
|
|
|
// Skip
|
|
|
|
} else if (isIgnoredEndTag(n)) {
|
|
|
|
// Skip
|
|
|
|
} else {
|
2018-05-16 11:09:07 +02:00
|
|
|
console.warn("Unsupported end tag: " + n);
|
2017-06-25 01:19:11 +02:00
|
|
|
}
|
2018-05-16 11:09:07 +02:00
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
})
|
2017-06-25 01:19:11 +02:00
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
saxStream.on('attribute', function(attr) {
|
|
|
|
|
|
|
|
})
|
|
|
|
|
|
|
|
saxStream.on('end', function() {
|
2017-06-25 01:19:11 +02:00
|
|
|
resolve({
|
2017-07-13 21:29:10 +02:00
|
|
|
content: section,
|
2017-12-06 21:29:58 +02:00
|
|
|
resources: remainingResources,
|
2017-06-25 01:19:11 +02:00
|
|
|
});
|
2018-03-09 22:59:12 +02:00
|
|
|
})
|
2017-06-25 01:19:11 +02:00
|
|
|
|
|
|
|
stream.pipe(saxStream);
|
|
|
|
});
|
|
|
|
}
|
|
|
|
|
2017-12-07 02:35:15 +02:00
|
|
|
function tableHasSubTables(table) {
|
2017-07-17 20:59:40 +02:00
|
|
|
for (let trIndex = 0; trIndex < table.lines.length; trIndex++) {
|
|
|
|
const tr = table.lines[trIndex];
|
2018-05-16 11:09:46 +02:00
|
|
|
if (!tr || !tr.lines) continue;
|
|
|
|
|
2017-07-17 20:59:40 +02:00
|
|
|
for (let tdIndex = 0; tdIndex < tr.lines.length; tdIndex++) {
|
|
|
|
const td = tr.lines[tdIndex];
|
2017-12-07 02:35:15 +02:00
|
|
|
for (let i = 0; i < td.lines.length; i++) {
|
2018-03-09 22:59:12 +02:00
|
|
|
if (typeof td.lines[i] === 'object') return true;
|
2017-12-07 02:35:15 +02:00
|
|
|
}
|
2017-07-17 20:59:40 +02:00
|
|
|
}
|
|
|
|
}
|
2017-12-07 02:35:15 +02:00
|
|
|
return false;
|
2017-07-17 20:59:40 +02:00
|
|
|
}
|
2017-07-13 21:29:10 +02:00
|
|
|
|
2017-12-07 02:35:15 +02:00
|
|
|
// Markdown tables don't support tables within tables, which is common in notes that are complete web pages, for example when imported
|
|
|
|
// via Web Clipper. So to handle this, we render all the outer tables as regular text (as if replacing all the <table>, <tr> and <td>
|
|
|
|
// elements by <div>) and only the inner ones, those that don't contain any other tables, are rendered as actual tables. This is generally
|
|
|
|
// the required behaviour since the outer tables are usually for layout and the inner ones are the content.
|
|
|
|
function drawTable(table) {
|
2017-07-17 20:59:40 +02:00
|
|
|
// | First Header | Second Header |
|
|
|
|
// | ------------- | ------------- |
|
|
|
|
// | Content Cell | Content Cell |
|
|
|
|
// | Content Cell | Content Cell |
|
|
|
|
|
2017-07-20 20:58:06 +02:00
|
|
|
// There must be at least 3 dashes separating each header cell.
|
|
|
|
// https://gist.github.com/IanWang/28965e13cdafdef4e11dc91f578d160d#tables
|
2017-12-07 02:35:15 +02:00
|
|
|
|
|
|
|
const flatRender = tableHasSubTables(table); // Render the table has regular text
|
2017-07-17 20:59:40 +02:00
|
|
|
let lines = [];
|
2017-12-07 02:35:15 +02:00
|
|
|
lines.push(BLOCK_OPEN);
|
2017-07-17 20:59:40 +02:00
|
|
|
let headerDone = false;
|
|
|
|
for (let trIndex = 0; trIndex < table.lines.length; trIndex++) {
|
|
|
|
const tr = table.lines[trIndex];
|
|
|
|
const isHeader = tr.isHeader;
|
|
|
|
let line = [];
|
|
|
|
let headerLine = [];
|
|
|
|
let emptyHeader = null;
|
2017-12-07 02:35:15 +02:00
|
|
|
for (let tdIndex = 0; tdIndex < tr.lines.length; tdIndex++) {
|
|
|
|
const td = tr.lines[tdIndex];
|
2017-07-17 20:59:40 +02:00
|
|
|
|
2017-12-07 02:35:15 +02:00
|
|
|
if (flatRender) {
|
|
|
|
line.push(BLOCK_OPEN);
|
|
|
|
|
|
|
|
let currentCells = [];
|
|
|
|
|
|
|
|
const renderCurrentCells = () => {
|
|
|
|
if (!currentCells.length) return;
|
2018-05-16 11:09:07 +02:00
|
|
|
const cellText = processMdArrayNewLines(currentCells);
|
2017-12-07 02:35:15 +02:00
|
|
|
line.push(cellText);
|
|
|
|
currentCells = [];
|
2018-03-09 22:59:12 +02:00
|
|
|
}
|
2017-12-07 02:35:15 +02:00
|
|
|
|
|
|
|
// In here, recursively render the tables
|
|
|
|
for (let i = 0; i < td.lines.length; i++) {
|
|
|
|
const c = td.lines[i];
|
2018-05-16 11:56:30 +02:00
|
|
|
if (typeof c === 'object' && ['table', 'td', 'tr', 'th'].indexOf(c.type) >= 0) { // This is a table
|
2017-12-07 02:35:15 +02:00
|
|
|
renderCurrentCells();
|
|
|
|
currentCells = currentCells.concat(drawTable(c));
|
2018-03-09 22:59:12 +02:00
|
|
|
} else { // This is plain text
|
2017-12-07 02:35:15 +02:00
|
|
|
currentCells.push(c);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
renderCurrentCells();
|
|
|
|
|
|
|
|
line.push(BLOCK_CLOSE);
|
2018-03-09 22:59:12 +02:00
|
|
|
} else { // Regular table rendering
|
2017-12-07 02:35:15 +02:00
|
|
|
|
2018-02-17 21:28:24 +02:00
|
|
|
// A cell in a Markdown table cannot have actual new lines so replace
|
|
|
|
// them with <br>, which are supported by the markdown renderers.
|
2018-05-16 11:56:30 +02:00
|
|
|
let cellText = processMdArrayNewLines(td.lines, true)
|
|
|
|
let lines = cellText.split('\n');
|
|
|
|
lines = postProcessMarkdown(lines);
|
|
|
|
cellText = lines.join('\n').replace(/\n+/g, "<br>");
|
2018-03-23 19:59:18 +02:00
|
|
|
|
|
|
|
// Inside tables cells, "|" needs to be escaped
|
|
|
|
cellText = cellText.replace(/\|/g, "\\|");
|
2017-12-07 02:35:15 +02:00
|
|
|
|
2018-03-23 19:59:18 +02:00
|
|
|
// Previously the width of the cell was as big as the content since it looks nicer, however that often doesn't work
|
|
|
|
// since the content can be very long, resulting in unreadable markdown. So no solution is perfect but making it a
|
|
|
|
// width of 3 is a bit better. Note that 3 is the minimum width of a cell - below this, it won't be rendered by
|
|
|
|
// markdown parsers.
|
|
|
|
const width = 3;
|
2018-03-09 22:59:12 +02:00
|
|
|
line.push(stringPadding(cellText, width, ' ', stringPadding.RIGHT));
|
2017-12-07 02:35:15 +02:00
|
|
|
|
|
|
|
if (!headerDone) {
|
|
|
|
if (!isHeader) {
|
|
|
|
if (!emptyHeader) emptyHeader = [];
|
2018-03-09 22:59:12 +02:00
|
|
|
let h = stringPadding(' ', width, ' ', stringPadding.RIGHT);
|
2017-12-07 02:35:15 +02:00
|
|
|
emptyHeader.push(h);
|
|
|
|
}
|
2018-03-09 22:59:12 +02:00
|
|
|
headerLine.push('-'.repeat(width));
|
2017-07-13 21:29:10 +02:00
|
|
|
}
|
2018-03-09 22:59:12 +02:00
|
|
|
|
2017-07-13 21:29:10 +02:00
|
|
|
}
|
2017-07-17 20:59:40 +02:00
|
|
|
}
|
|
|
|
|
2017-12-07 02:35:15 +02:00
|
|
|
if (flatRender) {
|
2017-07-17 20:59:40 +02:00
|
|
|
headerDone = true;
|
2017-12-07 02:35:15 +02:00
|
|
|
lines.push(BLOCK_OPEN);
|
|
|
|
lines = lines.concat(line);
|
|
|
|
lines.push(BLOCK_CLOSE);
|
|
|
|
} else {
|
|
|
|
if (emptyHeader) {
|
2018-03-09 22:59:12 +02:00
|
|
|
lines.push('| ' + emptyHeader.join(' | ') + ' |');
|
|
|
|
lines.push('| ' + headerLine.join(' | ') + ' |');
|
2017-12-07 02:35:15 +02:00
|
|
|
headerDone = true;
|
|
|
|
}
|
2017-07-17 20:59:40 +02:00
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
lines.push('| ' + line.join(' | ') + ' |');
|
2017-07-17 20:59:40 +02:00
|
|
|
|
2017-12-07 02:35:15 +02:00
|
|
|
if (!headerDone) {
|
2018-03-09 22:59:12 +02:00
|
|
|
lines.push('| ' + headerLine.join(' | ') + ' |');
|
2017-12-07 02:35:15 +02:00
|
|
|
headerDone = true;
|
|
|
|
}
|
2017-07-13 21:29:10 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2017-12-07 02:35:15 +02:00
|
|
|
lines.push(BLOCK_CLOSE);
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
return flatRender ? lines : lines.join('<<<<:D>>>>' + NEWLINE + '<<<<:D>>>>').split('<<<<:D>>>>');
|
2017-07-13 21:29:10 +02:00
|
|
|
}
|
|
|
|
|
2018-05-16 11:56:30 +02:00
|
|
|
function postProcessMarkdown(lines) {
|
|
|
|
// After importing HTML, the resulting Markdown often has empty lines at the beginning and end due to
|
|
|
|
// block start/end or elements that were ignored, etc. If these white spaces were intended it's not really
|
|
|
|
// possible to detect it, so simply trim them all so that the result is more deterministic and can be
|
|
|
|
// easily unit tested.
|
|
|
|
const trimEmptyLines = function(lines) {
|
|
|
|
while (lines.length) {
|
|
|
|
if (!lines[0].trim()) {
|
|
|
|
lines.splice(0, 1);
|
|
|
|
} else {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
while (lines.length) {
|
|
|
|
if (!lines[lines.length - 1].trim()) {
|
|
|
|
lines.pop();
|
|
|
|
} else {
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return lines;
|
|
|
|
}
|
|
|
|
|
|
|
|
function cleanUpSpaces(lines) {
|
|
|
|
const output = [];
|
|
|
|
|
|
|
|
for (let i = 0; i < lines.length; i++) {
|
|
|
|
let line = lines[i];
|
|
|
|
|
|
|
|
if (line.length) {
|
|
|
|
// eg. " - Some list item" => " - Some list item"
|
|
|
|
// Note that spaces before the "-" are preserved
|
|
|
|
line = line.replace(/^(\s+|)-\s+/, '$1- ')
|
|
|
|
|
|
|
|
// eg "Some text " => "Some text"
|
|
|
|
line = line.replace(/^(.*?)\s+$/, '$1')
|
|
|
|
}
|
|
|
|
|
|
|
|
output.push(line);
|
|
|
|
}
|
|
|
|
|
|
|
|
return output;
|
|
|
|
}
|
|
|
|
|
|
|
|
lines = trimEmptyLines(lines)
|
|
|
|
lines = cleanUpSpaces(lines)
|
|
|
|
|
|
|
|
return lines;
|
|
|
|
}
|
|
|
|
|
|
|
|
async function enexXmlToMd(xmlString, resources, options = {}) {
|
|
|
|
const stream = stringToStream(xmlString);
|
|
|
|
let result = await enexXmlToMdArray(stream, resources, options);
|
2018-05-15 14:26:53 +02:00
|
|
|
|
|
|
|
let mdLines = [];
|
|
|
|
|
|
|
|
for (let i = 0; i < result.content.lines.length; i++) {
|
|
|
|
let line = result.content.lines[i];
|
2018-05-16 11:56:30 +02:00
|
|
|
if (typeof line === 'object' && line.type === 'table') { // A table
|
2018-05-15 14:26:53 +02:00
|
|
|
const table = line;
|
|
|
|
const tableLines = drawTable(table);
|
|
|
|
mdLines = mdLines.concat(tableLines);
|
2018-05-16 11:56:30 +02:00
|
|
|
} else if (typeof line === 'object' && line.type === 'code') {
|
|
|
|
mdLines = mdLines.concat(line.lines);
|
|
|
|
} else if (typeof line === 'object') {
|
|
|
|
console.warn('Unhandled object type:', line);
|
|
|
|
mdLines = mdLines.concat(line.lines);
|
2018-05-15 14:26:53 +02:00
|
|
|
} else { // an actual line
|
|
|
|
mdLines.push(line);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
let firstAttachment = true;
|
|
|
|
for (let i = 0; i < result.resources.length; i++) {
|
|
|
|
let r = result.resources[i];
|
|
|
|
if (firstAttachment) mdLines.push(NEWLINE);
|
|
|
|
mdLines.push(NEWLINE);
|
|
|
|
mdLines = addResourceTag(mdLines, r, r.filename);
|
|
|
|
firstAttachment = false;
|
|
|
|
}
|
|
|
|
|
2018-05-16 11:56:30 +02:00
|
|
|
let output = processMdArrayNewLines(mdLines).split('\n')
|
|
|
|
|
|
|
|
output = postProcessMarkdown(output);
|
|
|
|
|
|
|
|
return output.join('\n');
|
2017-06-25 01:19:11 +02:00
|
|
|
}
|
|
|
|
|
2018-03-09 22:59:12 +02:00
|
|
|
module.exports = { enexXmlToMd, processMdArrayNewLines, NEWLINE, addResourceTag };
|