/** * CommonMark scanning utilities built on `StringReader`: link label * normalization, HTML character reference decoding, backslash unescaping, and * link destination/title/label scanners. */ import {StringBuilder, StringReader, asciiLowerByte, isSpaceOrTab, isAsciiWhitespace, isAsciiPunctuation, isAsciiAlpha, isAsciiDigit, isAsciiAlphanumeric, isAsciiHexDigit, hexDigitValue, isUnicodeWhitespace, isUnicodePunctuation} from 'zena:string'; /** * Strips leading and trailing spaces (`0x20`) and tabs (`0x09`). */ export function trimSpaces(s: String): String { let r = new StringReader(s); r.skipSpaces(); let start = r.mark(); var lastNonSpace = start; while (!r.isAtEnd) { let cp = r.advance(); if (!isSpaceOrTab(cp)) { lastNonSpace = r.mark(); } } return r.sliceRange(start, lastNonSpace); } /** * Advances `r` past spaces, tabs, and at most one line ending (`\r\n` or `\n`) * followed by spaces and tabs. */ export function skipSpacesAndOneLineEnd(r: StringReader): void { r.skipSpaces(); if (r.matchCodePoint(0x0D)) { r.matchCodePoint(0x0A); r.skipSpaces(); } else if (r.matchCodePoint(0x0A)) { r.skipSpaces(); } } /** * Normalizes a link reference label: trims surrounding whitespace, collapses * internal whitespace runs to a single space (`0x20`), and folds case. * Returns an empty string if the label contains no non-whitespace characters. */ export function normalizeLinkLabel(raw: String): String { let trimmed = raw.trim(); if (trimmed.isEmpty) { return ''; } let r = new StringReader(trimmed); let sb = new StringBuilder(); var inSpace = false; while (!r.isAtEnd) { let cp = r.advance(); if (isAsciiWhitespace(cp)) { if (!inSpace) { sb.appendByte(0x20); inSpace = true; } } else { sb.appendCodePoint(asciiLowerByte(cp)); inSpace = false; } } return sb.toString(); } /** * Resolves common named HTML entities (`&name;`) to their string value, * or returns `null` when `name` is unrecognized. */ function lookupNamedEntity(name: String): String | null { if (name == 'amp' || name == 'AMP') { return '&'; } if (name == 'lt' || name == 'LT') { return '<'; } if (name == 'gt' || name == 'GT') { return '>'; } if (name == 'quot' || name == 'QUOT') { return '"'; } if (name == 'apos') { return '\''; } if (name == 'nbsp') { return String.fromCodePoint(0x00A0); } if (name == 'copy' || name == 'COPY') { return '©'; } if (name == 'reg' || name == 'REG') { return '®'; } if (name == 'trade' || name == 'TRADE') { return '™'; } if (name == 'mdash') { return '—'; } if (name == 'ndash') { return '–'; } if (name == 'hellip') { return '…'; } if (name == 'laquo') { return '«'; } if (name == 'raquo') { return '»'; } if (name == 'lsquo') { return '‘'; } if (name == 'rsquo') { return '’'; } if (name == 'ldquo') { return '“'; } if (name == 'rdquo') { return '”'; } if (name == 'bull') { return '•'; } if (name == 'middot') { return '·'; } if (name == 'sect') { return '§'; } if (name == 'para') { return '¶'; } if (name == 'deg') { return '°'; } if (name == 'plusmn') { return '±'; } if (name == 'times') { return '×'; } if (name == 'divide') { return '÷'; } if (name == 'euro') { return '€'; } if (name == 'pound') { return '£'; } if (name == 'yen') { return '¥'; } if (name == 'cent') { return '¢'; } if (name == 'larr') { return '←'; } if (name == 'rarr') { return '→'; } if (name == 'uarr') { return '↑'; } if (name == 'darr') { return '↓'; } if (name == 'harr') { return '↔'; } if (name == 'ZeroWidthSpace') { return String.fromCodePoint(0x200B); } if (name == 'ouml') { return 'ö'; } if (name == 'Ouml') { return 'Ö'; } if (name == 'auml') { return 'ä'; } if (name == 'Auml') { return 'Ä'; } if (name == 'uuml') { return 'ü'; } if (name == 'Uuml') { return 'Ü'; } if (name == 'eacute') { return 'é'; } if (name == 'Eacute') { return 'É'; } if (name == 'frac12') { return '½'; } if (name == 'frac14') { return '¼'; } if (name == 'frac34') { return '¾'; } return null; } /** * Attempts to decode an HTML character reference starting with `&` at the * current position of `r`. If valid, appends the decoded character(s) to `sb`, * advances `r` past the closing `;`, and returns `true`; otherwise resets `r` * and returns `false`. */ export function tryDecodeEntity(r: StringReader, sb: StringBuilder): boolean { let saved = r.mark(); if (!r.matchCodePoint(0x26)) { // & return false; } if (r.matchCodePoint(0x23)) { // # let next = r.peek(); if (next == 0x78 || next == 0x58) { // x or X r.advance(); var count = 0; var value = 0; while (count < 6 && isAsciiHexDigit(r.peek())) { value = value << 4 | hexDigitValue(r.advance()); count += 1; } if (count > 0 && r.matchCodePoint(0x3B)) { // ; sb.appendCodePoint(if (value <= 0) 0xFFFD else value); return true; } r.reset(saved); return false; } var count = 0; var value = 0; while (count < 7 && isAsciiDigit(r.peek())) { value = value * 10 + (r.advance() - 0x30); count += 1; } if (count > 0 && r.matchCodePoint(0x3B)) { // ; sb.appendCodePoint(if (value <= 0) 0xFFFD else value); return true; } r.reset(saved); return false; } if (isAsciiAlpha(r.peek())) { let nameStart = r.mark(); var count = 0; while (count < 32 && isAsciiAlphanumeric(r.peek())) { r.advance(); count += 1; } let nameEnd = r.mark(); if (r.matchCodePoint(0x3B)) { // ; let name = r.sliceRange(nameStart, nameEnd); let decoded = lookupNamedEntity(name); if (decoded != null) { sb.append(decoded); return true; } } } r.reset(saved); return false; } /** * Unescapes CommonMark backslash escapes (`\!`, `\*`, etc.) and HTML character * references (`&`, `#`, `#`) in `s`. */ export function unescapeString(s: String): String { if (!s.contains('\\') && !s.contains('&')) { return s; } let r = new StringReader(s); let sb = new StringBuilder(); var last = r.mark(); while (!r.isAtEnd) { let pos = r.mark(); let cp = r.peek(); if (cp == 0x5C) { // \ r.advance(); if (!r.isAtEnd && isAsciiPunctuation(r.peek())) { if (pos > last) { sb.append(r.sliceRange(last, pos)); } sb.appendCodePoint(r.advance()); last = r.mark(); } } else if (cp == 0x26) { // & if (pos > last) { sb.append(r.sliceRange(last, pos)); last = pos; } if (tryDecodeEntity(r, sb)) { last = r.mark(); } else { r.advance(); } } else { r.advance(); } } if (last < r.mark()) { sb.append(r.sliceFrom(last)); } return sb.toString(); } /** * Scans a CommonMark link destination at the current position of `r`. * Returns `(matched, unescapedDestination)`. On failure, `r` is reset to its * initial position. */ export function scanLinkDestination(r: StringReader): inline (boolean, String) { if (r.isAtEnd) { return (false, ''); } let saved = r.mark(); if (r.matchCodePoint(0x3C)) { // <...> let contentStart = r.mark(); while (!r.isAtEnd) { let pos = r.mark(); let cp = r.advance(); if (cp == 0x0A || cp == 0x0D || cp == 0x3C) { r.reset(saved); return (false, ''); } if (cp == 0x3E) { // > let raw = r.sliceRange(contentStart, pos); return (true, unescapeString(raw)); } if (cp == 0x5C && !r.isAtEnd && isAsciiPunctuation(r.peek())) { r.advance(); } } r.reset(saved); return (false, ''); } var openParens = 0; while (!r.isAtEnd) { let cp = r.peek(); if (cp <= 0x20 || cp == 0x7F) { break; } if (cp == 0x5C) { r.advance(); if (!r.isAtEnd && isAsciiPunctuation(r.peek())) { r.advance(); } } else if (cp == 0x28) { // ( openParens += 1; if (openParens > 32) { r.reset(saved); return (false, ''); } r.advance(); } else if (cp == 0x29) { // ) if (openParens == 0) { break; } openParens -= 1; r.advance(); } else { r.advance(); } } if (openParens != 0 || r.mark() == saved) { r.reset(saved); return (false, ''); } return (true, unescapeString(r.sliceFrom(saved))); } /** * Scans a CommonMark link title (`"..."`, `'...'`, or `(...)`) at the current * position of `r`. Returns `(matched, unescapedTitle)`. On failure, `r` is * reset to its initial position. */ export function scanLinkTitle(r: StringReader): inline (boolean, String) { if (r.isAtEnd) { return (false, ''); } let saved = r.mark(); let opener = r.peek(); var closer = 0; if (opener == 0x22) { // " closer = 0x22; } else if (opener == 0x27) { // ' closer = 0x27; } else if (opener == 0x28) { // ( closer = 0x29; } else { return (false, ''); } r.advance(); let contentStart = r.mark(); while (!r.isAtEnd) { let pos = r.mark(); let cp = r.advance(); if (cp == closer) { let raw = r.sliceRange(contentStart, pos); return (true, unescapeString(raw)); } if (opener == 0x28 && cp == 0x28) { // Unescaped '(' inside (...) title is disallowed r.reset(saved); return (false, ''); } if (cp == 0x5C && !r.isAtEnd && isAsciiPunctuation(r.peek())) { r.advance(); } } r.reset(saved); return (false, ''); } /** * Scans a link label `[...]` starting at the current position of `r`. * Returns `(matched, rawInterior)`. On failure, `r` is reset to its initial * position. */ export function scanLinkLabel(r: StringReader): inline (boolean, String) { let saved = r.mark(); if (!r.matchCodePoint(0x5B)) { // [ return (false, ''); } let contentStart = r.mark(); var charCount = 0; while (!r.isAtEnd && charCount <= 999) { let pos = r.mark(); let cp = r.advance(); if (cp == 0x5D) { // ] return (true, r.sliceRange(contentStart, pos)); } if (cp == 0x5B) { // unescaped [ r.reset(saved); return (false, ''); } if (cp == 0x5C && !r.isAtEnd && isAsciiPunctuation(r.peek())) { r.advance(); charCount += 2; } else { charCount += 1; } } r.reset(saved); return (false, ''); }