diff options
author | altaf-creator <dev@altafcreator.com> | 2024-05-12 12:14:02 +0700 |
---|---|---|
committer | altaf-creator <dev@altafcreator.com> | 2024-05-12 12:14:02 +0700 |
commit | d607ac12097afb5cb6f398a4e7b5cf4316efedc6 (patch) | |
tree | 6f4bc5b98a6ff3a1c3189f7ef9b570c0481e100d /packages/markdown-it-14.1.0/lib/rules_block | |
parent | 7441f212967256ac4c9a93ba0b1f026308a8bfb6 (diff) |
self host
Diffstat (limited to 'packages/markdown-it-14.1.0/lib/rules_block')
12 files changed, 1612 insertions, 0 deletions
diff --git a/packages/markdown-it-14.1.0/lib/rules_block/blockquote.mjs b/packages/markdown-it-14.1.0/lib/rules_block/blockquote.mjs new file mode 100644 index 0000000..b61da02 --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/blockquote.mjs @@ -0,0 +1,209 @@ +// Block quotes + +import { isSpace } from '../common/utils.mjs' + +export default function blockquote (state, startLine, endLine, silent) { + let pos = state.bMarks[startLine] + state.tShift[startLine] + let max = state.eMarks[startLine] + + const oldLineMax = state.lineMax + + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[startLine] - state.blkIndent >= 4) { return false } + + // check the block quote marker + if (state.src.charCodeAt(pos) !== 0x3E/* > */) { return false } + + // we know that it's going to be a valid blockquote, + // so no point trying to find the end of it in silent mode + if (silent) { return true } + + const oldBMarks = [] + const oldBSCount = [] + const oldSCount = [] + const oldTShift = [] + + const terminatorRules = state.md.block.ruler.getRules('blockquote') + + const oldParentType = state.parentType + state.parentType = 'blockquote' + let lastLineEmpty = false + let nextLine + + // Search the end of the block + // + // Block ends with either: + // 1. an empty line outside: + // ``` + // > test + // + // ``` + // 2. an empty line inside: + // ``` + // > + // test + // ``` + // 3. another tag: + // ``` + // > test + // - - - + // ``` + for (nextLine = startLine; nextLine < endLine; nextLine++) { + // check if it's outdented, i.e. it's inside list item and indented + // less than said list item: + // + // ``` + // 1. anything + // > current blockquote + // 2. checking this line + // ``` + const isOutdented = state.sCount[nextLine] < state.blkIndent + + pos = state.bMarks[nextLine] + state.tShift[nextLine] + max = state.eMarks[nextLine] + + if (pos >= max) { + // Case 1: line is not inside the blockquote, and this line is empty. + break + } + + if (state.src.charCodeAt(pos++) === 0x3E/* > */ && !isOutdented) { + // This line is inside the blockquote. + + // set offset past spaces and ">" + let initial = state.sCount[nextLine] + 1 + let spaceAfterMarker + let adjustTab + + // skip one optional space after '>' + if (state.src.charCodeAt(pos) === 0x20 /* space */) { + // ' > test ' + // ^ -- position start of line here: + pos++ + initial++ + adjustTab = false + spaceAfterMarker = true + } else if (state.src.charCodeAt(pos) === 0x09 /* tab */) { + spaceAfterMarker = true + + if ((state.bsCount[nextLine] + initial) % 4 === 3) { + // ' >\t test ' + // ^ -- position start of line here (tab has width===1) + pos++ + initial++ + adjustTab = false + } else { + // ' >\t test ' + // ^ -- position start of line here + shift bsCount slightly + // to make extra space appear + adjustTab = true + } + } else { + spaceAfterMarker = false + } + + let offset = initial + oldBMarks.push(state.bMarks[nextLine]) + state.bMarks[nextLine] = pos + + while (pos < max) { + const ch = state.src.charCodeAt(pos) + + if (isSpace(ch)) { + if (ch === 0x09) { + offset += 4 - (offset + state.bsCount[nextLine] + (adjustTab ? 1 : 0)) % 4 + } else { + offset++ + } + } else { + break + } + + pos++ + } + + lastLineEmpty = pos >= max + + oldBSCount.push(state.bsCount[nextLine]) + state.bsCount[nextLine] = state.sCount[nextLine] + 1 + (spaceAfterMarker ? 1 : 0) + + oldSCount.push(state.sCount[nextLine]) + state.sCount[nextLine] = offset - initial + + oldTShift.push(state.tShift[nextLine]) + state.tShift[nextLine] = pos - state.bMarks[nextLine] + continue + } + + // Case 2: line is not inside the blockquote, and the last line was empty. + if (lastLineEmpty) { break } + + // Case 3: another tag found. + let terminate = false + for (let i = 0, l = terminatorRules.length; i < l; i++) { + if (terminatorRules[i](state, nextLine, endLine, true)) { + terminate = true + break + } + } + + if (terminate) { + // Quirk to enforce "hard termination mode" for paragraphs; + // normally if you call `tokenize(state, startLine, nextLine)`, + // paragraphs will look below nextLine for paragraph continuation, + // but if blockquote is terminated by another tag, they shouldn't + state.lineMax = nextLine + + if (state.blkIndent !== 0) { + // state.blkIndent was non-zero, we now set it to zero, + // so we need to re-calculate all offsets to appear as + // if indent wasn't changed + oldBMarks.push(state.bMarks[nextLine]) + oldBSCount.push(state.bsCount[nextLine]) + oldTShift.push(state.tShift[nextLine]) + oldSCount.push(state.sCount[nextLine]) + state.sCount[nextLine] -= state.blkIndent + } + + break + } + + oldBMarks.push(state.bMarks[nextLine]) + oldBSCount.push(state.bsCount[nextLine]) + oldTShift.push(state.tShift[nextLine]) + oldSCount.push(state.sCount[nextLine]) + + // A negative indentation means that this is a paragraph continuation + // + state.sCount[nextLine] = -1 + } + + const oldIndent = state.blkIndent + state.blkIndent = 0 + + const token_o = state.push('blockquote_open', 'blockquote', 1) + token_o.markup = '>' + const lines = [startLine, 0] + token_o.map = lines + + state.md.block.tokenize(state, startLine, nextLine) + + const token_c = state.push('blockquote_close', 'blockquote', -1) + token_c.markup = '>' + + state.lineMax = oldLineMax + state.parentType = oldParentType + lines[1] = state.line + + // Restore original tShift; this might not be necessary since the parser + // has already been here, but just to make sure we can do that. + for (let i = 0; i < oldTShift.length; i++) { + state.bMarks[i + startLine] = oldBMarks[i] + state.tShift[i + startLine] = oldTShift[i] + state.sCount[i + startLine] = oldSCount[i] + state.bsCount[i + startLine] = oldBSCount[i] + } + state.blkIndent = oldIndent + + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/code.mjs b/packages/markdown-it-14.1.0/lib/rules_block/code.mjs new file mode 100644 index 0000000..e45e6f9 --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/code.mjs @@ -0,0 +1,30 @@ +// Code block (4 spaces padded) + +export default function code (state, startLine, endLine/*, silent */) { + if (state.sCount[startLine] - state.blkIndent < 4) { return false } + + let nextLine = startLine + 1 + let last = nextLine + + while (nextLine < endLine) { + if (state.isEmpty(nextLine)) { + nextLine++ + continue + } + + if (state.sCount[nextLine] - state.blkIndent >= 4) { + nextLine++ + last = nextLine + continue + } + break + } + + state.line = last + + const token = state.push('code_block', 'code', 0) + token.content = state.getLines(startLine, last, 4 + state.blkIndent, false) + '\n' + token.map = [startLine, state.line] + + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/fence.mjs b/packages/markdown-it-14.1.0/lib/rules_block/fence.mjs new file mode 100644 index 0000000..930f7b3 --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/fence.mjs @@ -0,0 +1,94 @@ +// fences (``` lang, ~~~ lang) + +export default function fence (state, startLine, endLine, silent) { + let pos = state.bMarks[startLine] + state.tShift[startLine] + let max = state.eMarks[startLine] + + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[startLine] - state.blkIndent >= 4) { return false } + + if (pos + 3 > max) { return false } + + const marker = state.src.charCodeAt(pos) + + if (marker !== 0x7E/* ~ */ && marker !== 0x60 /* ` */) { + return false + } + + // scan marker length + let mem = pos + pos = state.skipChars(pos, marker) + + let len = pos - mem + + if (len < 3) { return false } + + const markup = state.src.slice(mem, pos) + const params = state.src.slice(pos, max) + + if (marker === 0x60 /* ` */) { + if (params.indexOf(String.fromCharCode(marker)) >= 0) { + return false + } + } + + // Since start is found, we can report success here in validation mode + if (silent) { return true } + + // search end of block + let nextLine = startLine + let haveEndMarker = false + + for (;;) { + nextLine++ + if (nextLine >= endLine) { + // unclosed block should be autoclosed by end of document. + // also block seems to be autoclosed by end of parent + break + } + + pos = mem = state.bMarks[nextLine] + state.tShift[nextLine] + max = state.eMarks[nextLine] + + if (pos < max && state.sCount[nextLine] < state.blkIndent) { + // non-empty line with negative indent should stop the list: + // - ``` + // test + break + } + + if (state.src.charCodeAt(pos) !== marker) { continue } + + if (state.sCount[nextLine] - state.blkIndent >= 4) { + // closing fence should be indented less than 4 spaces + continue + } + + pos = state.skipChars(pos, marker) + + // closing code fence must be at least as long as the opening one + if (pos - mem < len) { continue } + + // make sure tail has spaces only + pos = state.skipSpaces(pos) + + if (pos < max) { continue } + + haveEndMarker = true + // found! + break + } + + // If a fence has heading spaces, they should be removed from its inner block + len = state.sCount[startLine] + + state.line = nextLine + (haveEndMarker ? 1 : 0) + + const token = state.push('fence', 'code', 0) + token.info = params + token.content = state.getLines(startLine + 1, nextLine, len, true) + token.markup = markup + token.map = [startLine, state.line] + + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/heading.mjs b/packages/markdown-it-14.1.0/lib/rules_block/heading.mjs new file mode 100644 index 0000000..d2f7b79 --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/heading.mjs @@ -0,0 +1,51 @@ +// heading (#, ##, ...) + +import { isSpace } from '../common/utils.mjs' + +export default function heading (state, startLine, endLine, silent) { + let pos = state.bMarks[startLine] + state.tShift[startLine] + let max = state.eMarks[startLine] + + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[startLine] - state.blkIndent >= 4) { return false } + + let ch = state.src.charCodeAt(pos) + + if (ch !== 0x23/* # */ || pos >= max) { return false } + + // count heading level + let level = 1 + ch = state.src.charCodeAt(++pos) + while (ch === 0x23/* # */ && pos < max && level <= 6) { + level++ + ch = state.src.charCodeAt(++pos) + } + + if (level > 6 || (pos < max && !isSpace(ch))) { return false } + + if (silent) { return true } + + // Let's cut tails like ' ### ' from the end of string + + max = state.skipSpacesBack(max, pos) + const tmp = state.skipCharsBack(max, 0x23, pos) // # + if (tmp > pos && isSpace(state.src.charCodeAt(tmp - 1))) { + max = tmp + } + + state.line = startLine + 1 + + const token_o = state.push('heading_open', 'h' + String(level), 1) + token_o.markup = '########'.slice(0, level) + token_o.map = [startLine, state.line] + + const token_i = state.push('inline', '', 0) + token_i.content = state.src.slice(pos, max).trim() + token_i.map = [startLine, state.line] + token_i.children = [] + + const token_c = state.push('heading_close', 'h' + String(level), -1) + token_c.markup = '########'.slice(0, level) + + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/hr.mjs b/packages/markdown-it-14.1.0/lib/rules_block/hr.mjs new file mode 100644 index 0000000..d467b21 --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/hr.mjs @@ -0,0 +1,40 @@ +// Horizontal rule + +import { isSpace } from '../common/utils.mjs' + +export default function hr (state, startLine, endLine, silent) { + const max = state.eMarks[startLine] + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[startLine] - state.blkIndent >= 4) { return false } + + let pos = state.bMarks[startLine] + state.tShift[startLine] + const marker = state.src.charCodeAt(pos++) + + // Check hr marker + if (marker !== 0x2A/* * */ && + marker !== 0x2D/* - */ && + marker !== 0x5F/* _ */) { + return false + } + + // markers can be mixed with spaces, but there should be at least 3 of them + + let cnt = 1 + while (pos < max) { + const ch = state.src.charCodeAt(pos++) + if (ch !== marker && !isSpace(ch)) { return false } + if (ch === marker) { cnt++ } + } + + if (cnt < 3) { return false } + + if (silent) { return true } + + state.line = startLine + 1 + + const token = state.push('hr', 'hr', 0) + token.map = [startLine, state.line] + token.markup = Array(cnt + 1).join(String.fromCharCode(marker)) + + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/html_block.mjs b/packages/markdown-it-14.1.0/lib/rules_block/html_block.mjs new file mode 100644 index 0000000..197520f --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/html_block.mjs @@ -0,0 +1,69 @@ +// HTML block + +import block_names from '../common/html_blocks.mjs' +import { HTML_OPEN_CLOSE_TAG_RE } from '../common/html_re.mjs' + +// An array of opening and corresponding closing sequences for html tags, +// last argument defines whether it can terminate a paragraph or not +// +const HTML_SEQUENCES = [ + [/^<(script|pre|style|textarea)(?=(\s|>|$))/i, /<\/(script|pre|style|textarea)>/i, true], + [/^<!--/, /-->/, true], + [/^<\?/, /\?>/, true], + [/^<![A-Z]/, />/, true], + [/^<!\[CDATA\[/, /\]\]>/, true], + [new RegExp('^</?(' + block_names.join('|') + ')(?=(\\s|/?>|$))', 'i'), /^$/, true], + [new RegExp(HTML_OPEN_CLOSE_TAG_RE.source + '\\s*$'), /^$/, false] +] + +export default function html_block (state, startLine, endLine, silent) { + let pos = state.bMarks[startLine] + state.tShift[startLine] + let max = state.eMarks[startLine] + + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[startLine] - state.blkIndent >= 4) { return false } + + if (!state.md.options.html) { return false } + + if (state.src.charCodeAt(pos) !== 0x3C/* < */) { return false } + + let lineText = state.src.slice(pos, max) + + let i = 0 + for (; i < HTML_SEQUENCES.length; i++) { + if (HTML_SEQUENCES[i][0].test(lineText)) { break } + } + if (i === HTML_SEQUENCES.length) { return false } + + if (silent) { + // true if this sequence can be a terminator, false otherwise + return HTML_SEQUENCES[i][2] + } + + let nextLine = startLine + 1 + + // If we are here - we detected HTML block. + // Let's roll down till block end. + if (!HTML_SEQUENCES[i][1].test(lineText)) { + for (; nextLine < endLine; nextLine++) { + if (state.sCount[nextLine] < state.blkIndent) { break } + + pos = state.bMarks[nextLine] + state.tShift[nextLine] + max = state.eMarks[nextLine] + lineText = state.src.slice(pos, max) + + if (HTML_SEQUENCES[i][1].test(lineText)) { + if (lineText.length !== 0) { nextLine++ } + break + } + } + } + + state.line = nextLine + + const token = state.push('html_block', '', 0) + token.map = [startLine, nextLine] + token.content = state.getLines(startLine, nextLine, state.blkIndent, true) + + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/lheading.mjs b/packages/markdown-it-14.1.0/lib/rules_block/lheading.mjs new file mode 100644 index 0000000..ee3b9a3 --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/lheading.mjs @@ -0,0 +1,82 @@ +// lheading (---, ===) + +export default function lheading (state, startLine, endLine/*, silent */) { + const terminatorRules = state.md.block.ruler.getRules('paragraph') + + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[startLine] - state.blkIndent >= 4) { return false } + + const oldParentType = state.parentType + state.parentType = 'paragraph' // use paragraph to match terminatorRules + + // jump line-by-line until empty one or EOF + let level = 0 + let marker + let nextLine = startLine + 1 + + for (; nextLine < endLine && !state.isEmpty(nextLine); nextLine++) { + // this would be a code block normally, but after paragraph + // it's considered a lazy continuation regardless of what's there + if (state.sCount[nextLine] - state.blkIndent > 3) { continue } + + // + // Check for underline in setext header + // + if (state.sCount[nextLine] >= state.blkIndent) { + let pos = state.bMarks[nextLine] + state.tShift[nextLine] + const max = state.eMarks[nextLine] + + if (pos < max) { + marker = state.src.charCodeAt(pos) + + if (marker === 0x2D/* - */ || marker === 0x3D/* = */) { + pos = state.skipChars(pos, marker) + pos = state.skipSpaces(pos) + + if (pos >= max) { + level = (marker === 0x3D/* = */ ? 1 : 2) + break + } + } + } + } + + // quirk for blockquotes, this line should already be checked by that rule + if (state.sCount[nextLine] < 0) { continue } + + // Some tags can terminate paragraph without empty line. + let terminate = false + for (let i = 0, l = terminatorRules.length; i < l; i++) { + if (terminatorRules[i](state, nextLine, endLine, true)) { + terminate = true + break + } + } + if (terminate) { break } + } + + if (!level) { + // Didn't find valid underline + return false + } + + const content = state.getLines(startLine, nextLine, state.blkIndent, false).trim() + + state.line = nextLine + 1 + + const token_o = state.push('heading_open', 'h' + String(level), 1) + token_o.markup = String.fromCharCode(marker) + token_o.map = [startLine, state.line] + + const token_i = state.push('inline', '', 0) + token_i.content = content + token_i.map = [startLine, state.line - 1] + token_i.children = [] + + const token_c = state.push('heading_close', 'h' + String(level), -1) + token_c.markup = String.fromCharCode(marker) + + state.parentType = oldParentType + + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/list.mjs b/packages/markdown-it-14.1.0/lib/rules_block/list.mjs new file mode 100644 index 0000000..fb53abd --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/list.mjs @@ -0,0 +1,331 @@ +// Lists + +import { isSpace } from '../common/utils.mjs' + +// Search `[-+*][\n ]`, returns next pos after marker on success +// or -1 on fail. +function skipBulletListMarker (state, startLine) { + const max = state.eMarks[startLine] + let pos = state.bMarks[startLine] + state.tShift[startLine] + + const marker = state.src.charCodeAt(pos++) + // Check bullet + if (marker !== 0x2A/* * */ && + marker !== 0x2D/* - */ && + marker !== 0x2B/* + */) { + return -1 + } + + if (pos < max) { + const ch = state.src.charCodeAt(pos) + + if (!isSpace(ch)) { + // " -test " - is not a list item + return -1 + } + } + + return pos +} + +// Search `\d+[.)][\n ]`, returns next pos after marker on success +// or -1 on fail. +function skipOrderedListMarker (state, startLine) { + const start = state.bMarks[startLine] + state.tShift[startLine] + const max = state.eMarks[startLine] + let pos = start + + // List marker should have at least 2 chars (digit + dot) + if (pos + 1 >= max) { return -1 } + + let ch = state.src.charCodeAt(pos++) + + if (ch < 0x30/* 0 */ || ch > 0x39/* 9 */) { return -1 } + + for (;;) { + // EOL -> fail + if (pos >= max) { return -1 } + + ch = state.src.charCodeAt(pos++) + + if (ch >= 0x30/* 0 */ && ch <= 0x39/* 9 */) { + // List marker should have no more than 9 digits + // (prevents integer overflow in browsers) + if (pos - start >= 10) { return -1 } + + continue + } + + // found valid marker + if (ch === 0x29/* ) */ || ch === 0x2e/* . */) { + break + } + + return -1 + } + + if (pos < max) { + ch = state.src.charCodeAt(pos) + + if (!isSpace(ch)) { + // " 1.test " - is not a list item + return -1 + } + } + return pos +} + +function markTightParagraphs (state, idx) { + const level = state.level + 2 + + for (let i = idx + 2, l = state.tokens.length - 2; i < l; i++) { + if (state.tokens[i].level === level && state.tokens[i].type === 'paragraph_open') { + state.tokens[i + 2].hidden = true + state.tokens[i].hidden = true + i += 2 + } + } +} + +export default function list (state, startLine, endLine, silent) { + let max, pos, start, token + let nextLine = startLine + let tight = true + + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[nextLine] - state.blkIndent >= 4) { return false } + + // Special case: + // - item 1 + // - item 2 + // - item 3 + // - item 4 + // - this one is a paragraph continuation + if (state.listIndent >= 0 && + state.sCount[nextLine] - state.listIndent >= 4 && + state.sCount[nextLine] < state.blkIndent) { + return false + } + + let isTerminatingParagraph = false + + // limit conditions when list can interrupt + // a paragraph (validation mode only) + if (silent && state.parentType === 'paragraph') { + // Next list item should still terminate previous list item; + // + // This code can fail if plugins use blkIndent as well as lists, + // but I hope the spec gets fixed long before that happens. + // + if (state.sCount[nextLine] >= state.blkIndent) { + isTerminatingParagraph = true + } + } + + // Detect list type and position after marker + let isOrdered + let markerValue + let posAfterMarker + if ((posAfterMarker = skipOrderedListMarker(state, nextLine)) >= 0) { + isOrdered = true + start = state.bMarks[nextLine] + state.tShift[nextLine] + markerValue = Number(state.src.slice(start, posAfterMarker - 1)) + + // If we're starting a new ordered list right after + // a paragraph, it should start with 1. + if (isTerminatingParagraph && markerValue !== 1) return false + } else if ((posAfterMarker = skipBulletListMarker(state, nextLine)) >= 0) { + isOrdered = false + } else { + return false + } + + // If we're starting a new unordered list right after + // a paragraph, first line should not be empty. + if (isTerminatingParagraph) { + if (state.skipSpaces(posAfterMarker) >= state.eMarks[nextLine]) return false + } + + // For validation mode we can terminate immediately + if (silent) { return true } + + // We should terminate list on style change. Remember first one to compare. + const markerCharCode = state.src.charCodeAt(posAfterMarker - 1) + + // Start list + const listTokIdx = state.tokens.length + + if (isOrdered) { + token = state.push('ordered_list_open', 'ol', 1) + if (markerValue !== 1) { + token.attrs = [['start', markerValue]] + } + } else { + token = state.push('bullet_list_open', 'ul', 1) + } + + const listLines = [nextLine, 0] + token.map = listLines + token.markup = String.fromCharCode(markerCharCode) + + // + // Iterate list items + // + + let prevEmptyEnd = false + const terminatorRules = state.md.block.ruler.getRules('list') + + const oldParentType = state.parentType + state.parentType = 'list' + + while (nextLine < endLine) { + pos = posAfterMarker + max = state.eMarks[nextLine] + + const initial = state.sCount[nextLine] + posAfterMarker - (state.bMarks[nextLine] + state.tShift[nextLine]) + let offset = initial + + while (pos < max) { + const ch = state.src.charCodeAt(pos) + + if (ch === 0x09) { + offset += 4 - (offset + state.bsCount[nextLine]) % 4 + } else if (ch === 0x20) { + offset++ + } else { + break + } + + pos++ + } + + const contentStart = pos + let indentAfterMarker + + if (contentStart >= max) { + // trimming space in "- \n 3" case, indent is 1 here + indentAfterMarker = 1 + } else { + indentAfterMarker = offset - initial + } + + // If we have more than 4 spaces, the indent is 1 + // (the rest is just indented code block) + if (indentAfterMarker > 4) { indentAfterMarker = 1 } + + // " - test" + // ^^^^^ - calculating total length of this thing + const indent = initial + indentAfterMarker + + // Run subparser & write tokens + token = state.push('list_item_open', 'li', 1) + token.markup = String.fromCharCode(markerCharCode) + const itemLines = [nextLine, 0] + token.map = itemLines + if (isOrdered) { + token.info = state.src.slice(start, posAfterMarker - 1) + } + + // change current state, then restore it after parser subcall + const oldTight = state.tight + const oldTShift = state.tShift[nextLine] + const oldSCount = state.sCount[nextLine] + + // - example list + // ^ listIndent position will be here + // ^ blkIndent position will be here + // + const oldListIndent = state.listIndent + state.listIndent = state.blkIndent + state.blkIndent = indent + + state.tight = true + state.tShift[nextLine] = contentStart - state.bMarks[nextLine] + state.sCount[nextLine] = offset + + if (contentStart >= max && state.isEmpty(nextLine + 1)) { + // workaround for this case + // (list item is empty, list terminates before "foo"): + // ~~~~~~~~ + // - + // + // foo + // ~~~~~~~~ + state.line = Math.min(state.line + 2, endLine) + } else { + state.md.block.tokenize(state, nextLine, endLine, true) + } + + // If any of list item is tight, mark list as tight + if (!state.tight || prevEmptyEnd) { + tight = false + } + // Item become loose if finish with empty line, + // but we should filter last element, because it means list finish + prevEmptyEnd = (state.line - nextLine) > 1 && state.isEmpty(state.line - 1) + + state.blkIndent = state.listIndent + state.listIndent = oldListIndent + state.tShift[nextLine] = oldTShift + state.sCount[nextLine] = oldSCount + state.tight = oldTight + + token = state.push('list_item_close', 'li', -1) + token.markup = String.fromCharCode(markerCharCode) + + nextLine = state.line + itemLines[1] = nextLine + + if (nextLine >= endLine) { break } + + // + // Try to check if list is terminated or continued. + // + if (state.sCount[nextLine] < state.blkIndent) { break } + + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[nextLine] - state.blkIndent >= 4) { break } + + // fail if terminating block found + let terminate = false + for (let i = 0, l = terminatorRules.length; i < l; i++) { + if (terminatorRules[i](state, nextLine, endLine, true)) { + terminate = true + break + } + } + if (terminate) { break } + + // fail if list has another type + if (isOrdered) { + posAfterMarker = skipOrderedListMarker(state, nextLine) + if (posAfterMarker < 0) { break } + start = state.bMarks[nextLine] + state.tShift[nextLine] + } else { + posAfterMarker = skipBulletListMarker(state, nextLine) + if (posAfterMarker < 0) { break } + } + + if (markerCharCode !== state.src.charCodeAt(posAfterMarker - 1)) { break } + } + + // Finalize list + if (isOrdered) { + token = state.push('ordered_list_close', 'ol', -1) + } else { + token = state.push('bullet_list_close', 'ul', -1) + } + token.markup = String.fromCharCode(markerCharCode) + + listLines[1] = nextLine + state.line = nextLine + + state.parentType = oldParentType + + // mark paragraphs tight if needed + if (tight) { + markTightParagraphs(state, listTokIdx) + } + + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/paragraph.mjs b/packages/markdown-it-14.1.0/lib/rules_block/paragraph.mjs new file mode 100644 index 0000000..6ecdcef --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/paragraph.mjs @@ -0,0 +1,46 @@ +// Paragraph + +export default function paragraph (state, startLine, endLine) { + const terminatorRules = state.md.block.ruler.getRules('paragraph') + const oldParentType = state.parentType + let nextLine = startLine + 1 + state.parentType = 'paragraph' + + // jump line-by-line until empty one or EOF + for (; nextLine < endLine && !state.isEmpty(nextLine); nextLine++) { + // this would be a code block normally, but after paragraph + // it's considered a lazy continuation regardless of what's there + if (state.sCount[nextLine] - state.blkIndent > 3) { continue } + + // quirk for blockquotes, this line should already be checked by that rule + if (state.sCount[nextLine] < 0) { continue } + + // Some tags can terminate paragraph without empty line. + let terminate = false + for (let i = 0, l = terminatorRules.length; i < l; i++) { + if (terminatorRules[i](state, nextLine, endLine, true)) { + terminate = true + break + } + } + if (terminate) { break } + } + + const content = state.getLines(startLine, nextLine, state.blkIndent, false).trim() + + state.line = nextLine + + const token_o = state.push('paragraph_open', 'p', 1) + token_o.map = [startLine, state.line] + + const token_i = state.push('inline', '', 0) + token_i.content = content + token_i.map = [startLine, state.line] + token_i.children = [] + + state.push('paragraph_close', 'p', -1) + + state.parentType = oldParentType + + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/reference.mjs b/packages/markdown-it-14.1.0/lib/rules_block/reference.mjs new file mode 100644 index 0000000..4166286 --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/reference.mjs @@ -0,0 +1,212 @@ +import { isSpace, normalizeReference } from '../common/utils.mjs' + +export default function reference (state, startLine, _endLine, silent) { + let pos = state.bMarks[startLine] + state.tShift[startLine] + let max = state.eMarks[startLine] + let nextLine = startLine + 1 + + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[startLine] - state.blkIndent >= 4) { return false } + + if (state.src.charCodeAt(pos) !== 0x5B/* [ */) { return false } + + function getNextLine (nextLine) { + const endLine = state.lineMax + + if (nextLine >= endLine || state.isEmpty(nextLine)) { + // empty line or end of input + return null + } + + let isContinuation = false + + // this would be a code block normally, but after paragraph + // it's considered a lazy continuation regardless of what's there + if (state.sCount[nextLine] - state.blkIndent > 3) { isContinuation = true } + + // quirk for blockquotes, this line should already be checked by that rule + if (state.sCount[nextLine] < 0) { isContinuation = true } + + if (!isContinuation) { + const terminatorRules = state.md.block.ruler.getRules('reference') + const oldParentType = state.parentType + state.parentType = 'reference' + + // Some tags can terminate paragraph without empty line. + let terminate = false + for (let i = 0, l = terminatorRules.length; i < l; i++) { + if (terminatorRules[i](state, nextLine, endLine, true)) { + terminate = true + break + } + } + + state.parentType = oldParentType + if (terminate) { + // terminated by another block + return null + } + } + + const pos = state.bMarks[nextLine] + state.tShift[nextLine] + const max = state.eMarks[nextLine] + + // max + 1 explicitly includes the newline + return state.src.slice(pos, max + 1) + } + + let str = state.src.slice(pos, max + 1) + + max = str.length + let labelEnd = -1 + + for (pos = 1; pos < max; pos++) { + const ch = str.charCodeAt(pos) + if (ch === 0x5B /* [ */) { + return false + } else if (ch === 0x5D /* ] */) { + labelEnd = pos + break + } else if (ch === 0x0A /* \n */) { + const lineContent = getNextLine(nextLine) + if (lineContent !== null) { + str += lineContent + max = str.length + nextLine++ + } + } else if (ch === 0x5C /* \ */) { + pos++ + if (pos < max && str.charCodeAt(pos) === 0x0A) { + const lineContent = getNextLine(nextLine) + if (lineContent !== null) { + str += lineContent + max = str.length + nextLine++ + } + } + } + } + + if (labelEnd < 0 || str.charCodeAt(labelEnd + 1) !== 0x3A/* : */) { return false } + + // [label]: destination 'title' + // ^^^ skip optional whitespace here + for (pos = labelEnd + 2; pos < max; pos++) { + const ch = str.charCodeAt(pos) + if (ch === 0x0A) { + const lineContent = getNextLine(nextLine) + if (lineContent !== null) { + str += lineContent + max = str.length + nextLine++ + } + } else if (isSpace(ch)) { + /* eslint no-empty:0 */ + } else { + break + } + } + + // [label]: destination 'title' + // ^^^^^^^^^^^ parse this + const destRes = state.md.helpers.parseLinkDestination(str, pos, max) + if (!destRes.ok) { return false } + + const href = state.md.normalizeLink(destRes.str) + if (!state.md.validateLink(href)) { return false } + + pos = destRes.pos + + // save cursor state, we could require to rollback later + const destEndPos = pos + const destEndLineNo = nextLine + + // [label]: destination 'title' + // ^^^ skipping those spaces + const start = pos + for (; pos < max; pos++) { + const ch = str.charCodeAt(pos) + if (ch === 0x0A) { + const lineContent = getNextLine(nextLine) + if (lineContent !== null) { + str += lineContent + max = str.length + nextLine++ + } + } else if (isSpace(ch)) { + /* eslint no-empty:0 */ + } else { + break + } + } + + // [label]: destination 'title' + // ^^^^^^^ parse this + let titleRes = state.md.helpers.parseLinkTitle(str, pos, max) + while (titleRes.can_continue) { + const lineContent = getNextLine(nextLine) + if (lineContent === null) break + str += lineContent + pos = max + max = str.length + nextLine++ + titleRes = state.md.helpers.parseLinkTitle(str, pos, max, titleRes) + } + let title + + if (pos < max && start !== pos && titleRes.ok) { + title = titleRes.str + pos = titleRes.pos + } else { + title = '' + pos = destEndPos + nextLine = destEndLineNo + } + + // skip trailing spaces until the rest of the line + while (pos < max) { + const ch = str.charCodeAt(pos) + if (!isSpace(ch)) { break } + pos++ + } + + if (pos < max && str.charCodeAt(pos) !== 0x0A) { + if (title) { + // garbage at the end of the line after title, + // but it could still be a valid reference if we roll back + title = '' + pos = destEndPos + nextLine = destEndLineNo + while (pos < max) { + const ch = str.charCodeAt(pos) + if (!isSpace(ch)) { break } + pos++ + } + } + } + + if (pos < max && str.charCodeAt(pos) !== 0x0A) { + // garbage at the end of the line + return false + } + + const label = normalizeReference(str.slice(1, labelEnd)) + if (!label) { + // CommonMark 0.20 disallows empty labels + return false + } + + // Reference can not terminate anything. This check is for safety only. + /* istanbul ignore if */ + if (silent) { return true } + + if (typeof state.env.references === 'undefined') { + state.env.references = {} + } + if (typeof state.env.references[label] === 'undefined') { + state.env.references[label] = { title, href } + } + + state.line = nextLine + return true +} diff --git a/packages/markdown-it-14.1.0/lib/rules_block/state_block.mjs b/packages/markdown-it-14.1.0/lib/rules_block/state_block.mjs new file mode 100644 index 0000000..3c2a876 --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/state_block.mjs @@ -0,0 +1,220 @@ +// Parser state class + +import Token from '../token.mjs' +import { isSpace } from '../common/utils.mjs' + +function StateBlock (src, md, env, tokens) { + this.src = src + + // link to parser instance + this.md = md + + this.env = env + + // + // Internal state vartiables + // + + this.tokens = tokens + + this.bMarks = [] // line begin offsets for fast jumps + this.eMarks = [] // line end offsets for fast jumps + this.tShift = [] // offsets of the first non-space characters (tabs not expanded) + this.sCount = [] // indents for each line (tabs expanded) + + // An amount of virtual spaces (tabs expanded) between beginning + // of each line (bMarks) and real beginning of that line. + // + // It exists only as a hack because blockquotes override bMarks + // losing information in the process. + // + // It's used only when expanding tabs, you can think about it as + // an initial tab length, e.g. bsCount=21 applied to string `\t123` + // means first tab should be expanded to 4-21%4 === 3 spaces. + // + this.bsCount = [] + + // block parser variables + + // required block content indent (for example, if we are + // inside a list, it would be positioned after list marker) + this.blkIndent = 0 + this.line = 0 // line index in src + this.lineMax = 0 // lines count + this.tight = false // loose/tight mode for lists + this.ddIndent = -1 // indent of the current dd block (-1 if there isn't any) + this.listIndent = -1 // indent of the current list block (-1 if there isn't any) + + // can be 'blockquote', 'list', 'root', 'paragraph' or 'reference' + // used in lists to determine if they interrupt a paragraph + this.parentType = 'root' + + this.level = 0 + + // Create caches + // Generate markers. + const s = this.src + + for (let start = 0, pos = 0, indent = 0, offset = 0, len = s.length, indent_found = false; pos < len; pos++) { + const ch = s.charCodeAt(pos) + + if (!indent_found) { + if (isSpace(ch)) { + indent++ + + if (ch === 0x09) { + offset += 4 - offset % 4 + } else { + offset++ + } + continue + } else { + indent_found = true + } + } + + if (ch === 0x0A || pos === len - 1) { + if (ch !== 0x0A) { pos++ } + this.bMarks.push(start) + this.eMarks.push(pos) + this.tShift.push(indent) + this.sCount.push(offset) + this.bsCount.push(0) + + indent_found = false + indent = 0 + offset = 0 + start = pos + 1 + } + } + + // Push fake entry to simplify cache bounds checks + this.bMarks.push(s.length) + this.eMarks.push(s.length) + this.tShift.push(0) + this.sCount.push(0) + this.bsCount.push(0) + + this.lineMax = this.bMarks.length - 1 // don't count last fake line +} + +// Push new token to "stream". +// +StateBlock.prototype.push = function (type, tag, nesting) { + const token = new Token(type, tag, nesting) + token.block = true + + if (nesting < 0) this.level-- // closing tag + token.level = this.level + if (nesting > 0) this.level++ // opening tag + + this.tokens.push(token) + return token +} + +StateBlock.prototype.isEmpty = function isEmpty (line) { + return this.bMarks[line] + this.tShift[line] >= this.eMarks[line] +} + +StateBlock.prototype.skipEmptyLines = function skipEmptyLines (from) { + for (let max = this.lineMax; from < max; from++) { + if (this.bMarks[from] + this.tShift[from] < this.eMarks[from]) { + break + } + } + return from +} + +// Skip spaces from given position. +StateBlock.prototype.skipSpaces = function skipSpaces (pos) { + for (let max = this.src.length; pos < max; pos++) { + const ch = this.src.charCodeAt(pos) + if (!isSpace(ch)) { break } + } + return pos +} + +// Skip spaces from given position in reverse. +StateBlock.prototype.skipSpacesBack = function skipSpacesBack (pos, min) { + if (pos <= min) { return pos } + + while (pos > min) { + if (!isSpace(this.src.charCodeAt(--pos))) { return pos + 1 } + } + return pos +} + +// Skip char codes from given position +StateBlock.prototype.skipChars = function skipChars (pos, code) { + for (let max = this.src.length; pos < max; pos++) { + if (this.src.charCodeAt(pos) !== code) { break } + } + return pos +} + +// Skip char codes reverse from given position - 1 +StateBlock.prototype.skipCharsBack = function skipCharsBack (pos, code, min) { + if (pos <= min) { return pos } + + while (pos > min) { + if (code !== this.src.charCodeAt(--pos)) { return pos + 1 } + } + return pos +} + +// cut lines range from source. +StateBlock.prototype.getLines = function getLines (begin, end, indent, keepLastLF) { + if (begin >= end) { + return '' + } + + const queue = new Array(end - begin) + + for (let i = 0, line = begin; line < end; line++, i++) { + let lineIndent = 0 + const lineStart = this.bMarks[line] + let first = lineStart + let last + + if (line + 1 < end || keepLastLF) { + // No need for bounds check because we have fake entry on tail. + last = this.eMarks[line] + 1 + } else { + last = this.eMarks[line] + } + + while (first < last && lineIndent < indent) { + const ch = this.src.charCodeAt(first) + + if (isSpace(ch)) { + if (ch === 0x09) { + lineIndent += 4 - (lineIndent + this.bsCount[line]) % 4 + } else { + lineIndent++ + } + } else if (first - lineStart < this.tShift[line]) { + // patched tShift masked characters to look like spaces (blockquotes, list markers) + lineIndent++ + } else { + break + } + + first++ + } + + if (lineIndent > indent) { + // partially expanding tabs in code blocks, e.g '\t\tfoobar' + // with indent=2 becomes ' \tfoobar' + queue[i] = new Array(lineIndent - indent + 1).join(' ') + this.src.slice(first, last) + } else { + queue[i] = this.src.slice(first, last) + } + } + + return queue.join('') +} + +// re-export Token class to use in block rules +StateBlock.prototype.Token = Token + +export default StateBlock diff --git a/packages/markdown-it-14.1.0/lib/rules_block/table.mjs b/packages/markdown-it-14.1.0/lib/rules_block/table.mjs new file mode 100644 index 0000000..be0ba0a --- /dev/null +++ b/packages/markdown-it-14.1.0/lib/rules_block/table.mjs @@ -0,0 +1,228 @@ +// GFM table, https://github.github.com/gfm/#tables-extension- + +import { isSpace } from '../common/utils.mjs' + +// Limit the amount of empty autocompleted cells in a table, +// see https://github.com/markdown-it/markdown-it/issues/1000, +// +// Both pulldown-cmark and commonmark-hs limit the number of cells this way to ~200k. +// We set it to 65k, which can expand user input by a factor of x370 +// (256x256 square is 1.8kB expanded into 650kB). +const MAX_AUTOCOMPLETED_CELLS = 0x10000 + +function getLine (state, line) { + const pos = state.bMarks[line] + state.tShift[line] + const max = state.eMarks[line] + + return state.src.slice(pos, max) +} + +function escapedSplit (str) { + const result = [] + const max = str.length + + let pos = 0 + let ch = str.charCodeAt(pos) + let isEscaped = false + let lastPos = 0 + let current = '' + + while (pos < max) { + if (ch === 0x7c/* | */) { + if (!isEscaped) { + // pipe separating cells, '|' + result.push(current + str.substring(lastPos, pos)) + current = '' + lastPos = pos + 1 + } else { + // escaped pipe, '\|' + current += str.substring(lastPos, pos - 1) + lastPos = pos + } + } + + isEscaped = (ch === 0x5c/* \ */) + pos++ + + ch = str.charCodeAt(pos) + } + + result.push(current + str.substring(lastPos)) + + return result +} + +export default function table (state, startLine, endLine, silent) { + // should have at least two lines + if (startLine + 2 > endLine) { return false } + + let nextLine = startLine + 1 + + if (state.sCount[nextLine] < state.blkIndent) { return false } + + // if it's indented more than 3 spaces, it should be a code block + if (state.sCount[nextLine] - state.blkIndent >= 4) { return false } + + // first character of the second line should be '|', '-', ':', + // and no other characters are allowed but spaces; + // basically, this is the equivalent of /^[-:|][-:|\s]*$/ regexp + + let pos = state.bMarks[nextLine] + state.tShift[nextLine] + if (pos >= state.eMarks[nextLine]) { return false } + + const firstCh = state.src.charCodeAt(pos++) + if (firstCh !== 0x7C/* | */ && firstCh !== 0x2D/* - */ && firstCh !== 0x3A/* : */) { return false } + + if (pos >= state.eMarks[nextLine]) { return false } + + const secondCh = state.src.charCodeAt(pos++) + if (secondCh !== 0x7C/* | */ && secondCh !== 0x2D/* - */ && secondCh !== 0x3A/* : */ && !isSpace(secondCh)) { + return false + } + + // if first character is '-', then second character must not be a space + // (due to parsing ambiguity with list) + if (firstCh === 0x2D/* - */ && isSpace(secondCh)) { return false } + + while (pos < state.eMarks[nextLine]) { + const ch = state.src.charCodeAt(pos) + + if (ch !== 0x7C/* | */ && ch !== 0x2D/* - */ && ch !== 0x3A/* : */ && !isSpace(ch)) { return false } + + pos++ + } + + let lineText = getLine(state, startLine + 1) + let columns = lineText.split('|') + const aligns = [] + for (let i = 0; i < columns.length; i++) { + const t = columns[i].trim() + if (!t) { + // allow empty columns before and after table, but not in between columns; + // e.g. allow ` |---| `, disallow ` ---||--- ` + if (i === 0 || i === columns.length - 1) { + continue + } else { + return false + } + } + + if (!/^:?-+:?$/.test(t)) { return false } + if (t.charCodeAt(t.length - 1) === 0x3A/* : */) { + aligns.push(t.charCodeAt(0) === 0x3A/* : */ ? 'center' : 'right') + } else if (t.charCodeAt(0) === 0x3A/* : */) { + aligns.push('left') + } else { + aligns.push('') + } + } + + lineText = getLine(state, startLine).trim() + if (lineText.indexOf('|') === -1) { return false } + if (state.sCount[startLine] - state.blkIndent >= 4) { return false } + columns = escapedSplit(lineText) + if (columns.length && columns[0] === '') columns.shift() + if (columns.length && columns[columns.length - 1] === '') columns.pop() + + // header row will define an amount of columns in the entire table, + // and align row should be exactly the same (the rest of the rows can differ) + const columnCount = columns.length + if (columnCount === 0 || columnCount !== aligns.length) { return false } + + if (silent) { return true } + + const oldParentType = state.parentType + state.parentType = 'table' + + // use 'blockquote' lists for termination because it's + // the most similar to tables + const terminatorRules = state.md.block.ruler.getRules('blockquote') + + const token_to = state.push('table_open', 'table', 1) + const tableLines = [startLine, 0] + token_to.map = tableLines + + const token_tho = state.push('thead_open', 'thead', 1) + token_tho.map = [startLine, startLine + 1] + + const token_htro = state.push('tr_open', 'tr', 1) + token_htro.map = [startLine, startLine + 1] + + for (let i = 0; i < columns.length; i++) { + const token_ho = state.push('th_open', 'th', 1) + if (aligns[i]) { + token_ho.attrs = [['style', 'text-align:' + aligns[i]]] + } + + const token_il = state.push('inline', '', 0) + token_il.content = columns[i].trim() + token_il.children = [] + + state.push('th_close', 'th', -1) + } + + state.push('tr_close', 'tr', -1) + state.push('thead_close', 'thead', -1) + + let tbodyLines + let autocompletedCells = 0 + + for (nextLine = startLine + 2; nextLine < endLine; nextLine++) { + if (state.sCount[nextLine] < state.blkIndent) { break } + + let terminate = false + for (let i = 0, l = terminatorRules.length; i < l; i++) { + if (terminatorRules[i](state, nextLine, endLine, true)) { + terminate = true + break + } + } + + if (terminate) { break } + lineText = getLine(state, nextLine).trim() + if (!lineText) { break } + if (state.sCount[nextLine] - state.blkIndent >= 4) { break } + columns = escapedSplit(lineText) + if (columns.length && columns[0] === '') columns.shift() + if (columns.length && columns[columns.length - 1] === '') columns.pop() + + // note: autocomplete count can be negative if user specifies more columns than header, + // but that does not affect intended use (which is limiting expansion) + autocompletedCells += columnCount - columns.length + if (autocompletedCells > MAX_AUTOCOMPLETED_CELLS) { break } + + if (nextLine === startLine + 2) { + const token_tbo = state.push('tbody_open', 'tbody', 1) + token_tbo.map = tbodyLines = [startLine + 2, 0] + } + + const token_tro = state.push('tr_open', 'tr', 1) + token_tro.map = [nextLine, nextLine + 1] + + for (let i = 0; i < columnCount; i++) { + const token_tdo = state.push('td_open', 'td', 1) + if (aligns[i]) { + token_tdo.attrs = [['style', 'text-align:' + aligns[i]]] + } + + const token_il = state.push('inline', '', 0) + token_il.content = columns[i] ? columns[i].trim() : '' + token_il.children = [] + + state.push('td_close', 'td', -1) + } + state.push('tr_close', 'tr', -1) + } + + if (tbodyLines) { + state.push('tbody_close', 'tbody', -1) + tbodyLines[1] = nextLine + } + + state.push('table_close', 'table', -1) + tableLines[1] = nextLine + + state.parentType = oldParentType + state.line = nextLine + return true +} |