feat(paste): CSV/TSV → markdown table on tabular paste

- src/utils/csv-to-table.js — pure module. Auto-detects the delimiter
  (tab wins over comma; comma wins over semicolon). RFC 4180-style
  parsing handles quoted fields, escaped quotes (""), and CRLF. Cells
  are escaped for markdown tables (\\ for backslashes, \| for pipes,
  newlines stripped). Alignment: a column is right-aligned when every
  non-empty cell matches a numeric pattern AND there are ≥2 rows or
  ≥1 multi-character cell (single-char numbers like "1" alone are
  too ambiguous to call as numeric — could be labels). Single-row input
  produces a header-only table (no fabricated "Column N" labels, no
  body).
- src/editor/smart-paste.js — paste handler now tries CSV first (it
  needs no async), falls back to the existing URL → title flow.

Tests (30 new, tests/csv-to-table.test.js):
  - detectDelimiter: tab > comma > semicolon, empty/non-string safe
  - parseRows: simple, RFC 4180 quoted, escaped quotes, CRLF, multi-line
  - escapeCell: pipes, backslash-first escaping, newline strip, null/number
  - csvToTable: simple CSV, single-row (header-only), TSV, empty input
  - alignment: numeric detection (≥2 rows OR ≥1 multi-char cell), currency,
    percentages, text columns stay left
  - escaping inside cells (pipes, newlines)
  - ragged-row padding
  - looksLikeCsv: true for tab/comma multi-row, false for single row,
    column-count mismatch, prose, very short input

Full suite: 72 suites, 861 tests, lint+format clean.

Amit Haridas

src/editor/smart-paste.js

@@ -15,6 +15,7 @@
  * @returns {import('@codemirror/view').Extension}
  */
 const { EditorView } = require('@codemirror/view');
+const { csvToTable, looksLikeCsv } = require('../utils/csv-to-table');
 
 const URL_ONLY_RE = /^\s*(https?:\/\/[^\s]+)\s*$/i;
 const TIMEOUT_MS = 4000;
@@ -28,14 +29,27 @@ function smartPaste(deps) {
   return EditorView.domEventHandlers({
     paste(event, view) {
       const text = event.clipboardData && event.clipboardData.getData('text/plain');
-      const match = text && URL_ONLY_RE.exec(text);
-      if (!match) return; // not a URL-only paste — let the default handler run
+      if (!text) return;
 
-      const url = match[1];
+      // CSV/TSV flavor first — it's instant and self-contained (no async).
+      if (looksLikeCsv(text)) {
         event.preventDefault();
+        const table = csvToTable(text);
+        if (!table) return;
+        const head = view.state.selection.main.head;
+        view.dispatch({
+          changes: { from: head, insert: table },
+          selection: { anchor: head + table.length },
+        });
+        return;
+      }
 
-      // Insert the URL immediately so the paste isn't lost on slow networks,
-      // then async-fetch the title and rewrite the just-pasted range.
+      // URL flavor — paste the URL immediately, then rewrite with the title.
+      const urlMatch = URL_ONLY_RE.exec(text);
+      if (!urlMatch) return; // not our department — let the default handler run
+
+      const url = urlMatch[1];
+      event.preventDefault();
       const head = view.state.selection.main.head;
       const from = head;
 
@@ -44,16 +58,12 @@ function smartPaste(deps) {
         selection: { anchor: from + url.length },
       });
 
-      // Best-effort fetch; if it fails, leave the URL as-is.
       Promise.race([
         fetchTitle({ url, timeoutMs: TIMEOUT_MS }),
         new Promise((resolve) => setTimeout(() => resolve(null), TIMEOUT_MS)),
       ])
         .then((result) => {
           if (!result || !result.title) return;
-          // The user may have continued typing in the meantime. Cap the
-          // rewrite at the original insertion length so we don't clobber
-          // anything else.
           const currentLen = view.state.doc.length;
           const rewriteTo = Math.min(from + url.length, currentLen);
           if (rewriteTo <= from) return;

src/utils/csv-to-table.js

@@ -0,0 +1,190 @@
+/**
+ * CSV → markdown table converter.
+ *
+ * Smart-paste flavor: when the user pastes tab-separated text (Excel,
+ * Sheets, Numbers copies) or CSV, convert to a markdown pipe-table.
+ * Pure module — no IO, no DOM — so the renderer can require it directly
+ * and main never has to touch the clipboard.
+ *
+ * Grammar (deliberately minimal):
+ *   - Rows split on newlines (LF or CRLF)
+ *   - Cells split on the configured delimiter (tab, comma, or semicolon)
+ *   - Quoted fields per RFC 4180: "…" may contain the delimiter, "" is an
+ *     escaped quote, multi-line quoted fields are supported
+ *   - First row becomes the header (only when there are ≥2 rows; otherwise
+ *     the single row is treated as data with no header)
+ *
+ * Alignment column is added per the longest cell in each column.
+ *
+ * @module csv-to-table
+ */
+
+/**
+ * Detect the delimiter by sampling the first non-empty line.
+ * Tabs win over commas when both are present (Excel/Sheets default).
+ */
+function detectDelimiter(text) {
+  if (typeof text !== 'string' || text.length === 0) return ',';
+  const firstLine = text.split(/\r?\n/, 1)[0] || '';
+  const counts = {
+    '\t': (firstLine.match(/\t/g) || []).length,
+    ',': (firstLine.match(/,/g) || []).length,
+    ';': (firstLine.match(/;/g) || []).length,
+  };
+  let best = ',';
+  let bestCount = counts[','];
+  for (const d of ['\t', ',', ';']) {
+    if (counts[d] > bestCount) {
+      best = d;
+      bestCount = counts[d];
+    }
+  }
+  return bestCount === 0 ? ',' : best;
+}
+
+/** Parse a CSV string into rows of cells. Handles RFC 4180 quoting. */
+function parseRows(text, delimiter) {
+  const rows = [];
+  let row = [];
+  let cell = '';
+  let inQuotes = false;
+  for (let i = 0; i < text.length; i++) {
+    const ch = text[i];
+    if (inQuotes) {
+      if (ch === '"') {
+        if (text[i + 1] === '"') {
+          cell += '"';
+          i++;
+        } else {
+          inQuotes = false;
+        }
+      } else {
+        cell += ch;
+      }
+      continue;
+    }
+    if (ch === '"') {
+      inQuotes = true;
+    } else if (ch === delimiter) {
+      row.push(cell);
+      cell = '';
+    } else if (ch === '\n' || ch === '\r') {
+      row.push(cell);
+      if (row.length > 1 || row[0] !== '') rows.push(row);
+      row = [];
+      cell = '';
+      // Swallow the \n after \r
+      if (ch === '\r' && text[i + 1] === '\n') i++;
+    } else {
+      cell += ch;
+    }
+  }
+  if (cell !== '' || row.length > 0) {
+    row.push(cell);
+    rows.push(row);
+  }
+  return rows;
+}
+
+/** Escape a cell so it can sit inside a markdown table cell. */
+function escapeCell(value) {
+  if (value === undefined || value === null) return '';
+  return String(value)
+    .replace(/\\/g, '\\\\')
+    .replace(/\|/g, '\\|')
+    .replace(/[\r\n]+/g, '');
+}
+
+/** Detect alignment per column from the data. */
+function alignmentFor(rows, colIdx) {
+  const nonEmpty = rows.filter((r) => (r[colIdx] || '').trim() !== '');
+  if (nonEmpty.length === 0) return '---';
+  const numericish = nonEmpty.every((r) => {
+    const v = (r[colIdx] || '').trim().replace(/^[-$€£¥+]/, '');
+    return /^-?[\d,.]+%?$/.test(v);
+  });
+  // Numeric when:
+  //  - every cell matches the numeric pattern AND
+  //  - we have ≥ 2 non-empty rows OR ≥ 1 multi-character cell (a single
+  //    "1" or "5" is too ambiguous — the column might really be labels)
+  if (!numericish) return '---';
+  const hasMultiChar = nonEmpty.some((r) => (r[colIdx] || '').length >= 2);
+  if (nonEmpty.length >= 2 || hasMultiChar) return '---:';
+  return '---';
+}
+
+/**
+ * Convert CSV / TSV text to a markdown table.
+ *
+ * @param {string} text
+ * @param {object} [opts]
+ * @param {string} [opts.delimiter]   auto-detected when omitted
+ * @param {boolean} [opts.firstRowIsHeader=true]
+ * @returns {string} markdown table, or empty string when input has no rows
+ */
+function csvToTable(text, opts = {}) {
+  if (typeof text !== 'string' || text.trim() === '') return '';
+  const delimiter = opts.delimiter || detectDelimiter(text);
+  const rows = parseRows(text, delimiter).filter((r) => !(r.length === 1 && r[0] === ''));
+  if (rows.length === 0) return '';
+  // Normalize column count
+  const colCount = rows.reduce((m, r) => Math.max(m, r.length), 0);
+  for (const r of rows) {
+    while (r.length < colCount) r.push('');
+  }
+  const firstRowIsHeader = opts.firstRowIsHeader !== false && rows.length >= 2;
+
+  let header, body;
+  if (firstRowIsHeader) {
+    header = rows[0];
+    body = rows.slice(1);
+  } else if (rows.length === 1) {
+    // Single-row input: treat the row as the header and emit no body — a
+    // table with one row of data and fabricated "Column N" labels would be
+    // useless noise.
+    header = rows[0];
+    body = [];
+  } else {
+    header = Array.from({ length: colCount }, (_, i) => `Column ${i + 1}`);
+    body = rows;
+  }
+
+  // Alignment comes from the DATA rows only; headers are text by convention.
+  const dataForAlignment = firstRowIsHeader ? rows.slice(1) : rows;
+  const alignments = [];
+  for (let i = 0; i < colCount; i++) alignments.push(alignmentFor(dataForAlignment, i));
+
+  const lines = [];
+  lines.push('| ' + header.map(escapeCell).join(' | ') + ' |');
+  lines.push('| ' + alignments.join(' | ') + ' |');
+  for (let i = 0; i < body.length; i++) {
+    lines.push('| ' + body[i].map(escapeCell).join(' | ') + ' |');
+  }
+  return lines.join('\n');
+}
+
+/** Heuristic: is this paste candidate a CSV/TSV? */
+function looksLikeCsv(text) {
+  if (typeof text !== 'string' || text.length < 4) return false;
+  const lines = text.split(/\r?\n/).filter((l) => l.length > 0);
+  if (lines.length < 2) return false;
+  // Every non-empty line must have the same number of delimiters as the
+  // first, allowing a ±1 slack for trailing empties.
+  const delim = detectDelimiter(text);
+  const counts = lines.slice(0, 5).map((l) => (l.match(new RegExp(escapeRegex(delim), 'g')) || []).length);
+  const first = counts[0];
+  if (first < 1) return false;
+  return counts.every((c) => Math.abs(c - first) <= 1);
+}
+
+function escapeRegex(s) {
+  return String(s).replace(/[.*+?^${}()|[\]\\]/g, '\\$&');
+}
+
+module.exports = {
+  detectDelimiter,
+  parseRows,
+  escapeCell,
+  csvToTable,
+  looksLikeCsv,
+};

tests/csv-to-table.test.js

@@ -0,0 +1,171 @@
+/**
+ * @jest-environment node
+ *
+ * csv-to-table tests — pure module.
+ */
+const { detectDelimiter, csvToTable, looksLikeCsv, parseRows, escapeCell } = require('../src/utils/csv-to-table');
+
+describe('detectDelimiter', () => {
+  test('prefers tabs over commas when both are present', () => {
+    expect(detectDelimiter('a\tb\tc\n1\t2\t3')).toBe('\t');
+  });
+
+  test('falls back to commas', () => {
+    expect(detectDelimiter('a,b,c\n1,2,3')).toBe(',');
+  });
+
+  test('uses semicolons when commas are absent', () => {
+    expect(detectDelimiter('a;b;c\n1;2;3')).toBe(';');
+  });
+
+  test('returns comma for empty / non-string input', () => {
+    expect(detectDelimiter('')).toBe(',');
+    expect(detectDelimiter(null)).toBe(',');
+  });
+});
+
+describe('parseRows (RFC 4180 quoting)', () => {
+  test('parses simple comma-separated rows', () => {
+    expect(parseRows('a,b,c\n1,2,3', ',')).toEqual([
+      ['a', 'b', 'c'],
+      ['1', '2', '3'],
+    ]);
+  });
+
+  test('handles quoted fields with embedded delimiter', () => {
+    expect(parseRows('"a, b",c', ',')).toEqual([['a, b', 'c']]);
+  });
+
+  test('handles escaped double-quotes inside quoted fields', () => {
+    expect(parseRows('"he said ""hi""",ok', ',')).toEqual([['he said "hi"', 'ok']]);
+  });
+
+  test('handles CRLF line endings', () => {
+    expect(parseRows('a,b\r\n1,2', ',')).toEqual([
+      ['a', 'b'],
+      ['1', '2'],
+    ]);
+  });
+
+  test('handles multi-line quoted fields', () => {
+    expect(parseRows('"line1\nline2",x', ',')).toEqual([['line1\nline2', 'x']]);
+  });
+});
+
+describe('escapeCell', () => {
+  test('escapes pipes (table syntax)', () => {
+    expect(escapeCell('a | b')).toBe('a \\| b');
+  });
+
+  test('escapes backslashes first', () => {
+    expect(escapeCell('a\\b')).toBe('a\\\\b');
+    expect(escapeCell('a\\|b')).toBe('a\\\\\\|b');
+  });
+
+  test('strips newlines (cells must be one logical line)', () => {
+    expect(escapeCell('a\nb')).toBe('ab');
+    expect(escapeCell('a\r\nb')).toBe('ab');
+    // Tabs survive — they don't break markdown tables and may carry intent.
+    expect(escapeCell('a\tb')).toBe('a\tb');
+  });
+
+  test('handles null / undefined / numbers', () => {
+    expect(escapeCell(null)).toBe('');
+    expect(escapeCell(undefined)).toBe('');
+    expect(escapeCell(42)).toBe('42');
+  });
+});
+
+describe('csvToTable — basic', () => {
+  test('converts simple CSV with first row as header', () => {
+    const md = csvToTable('name,age\nAlice,30\nBob,25');
+    // age column has 2 numeric values (30, 25) → right-aligned
+    expect(md).toBe('| name | age |\n| --- | ---: |\n| Alice | 30 |\n| Bob | 25 |');
+  });
+
+  test('handles a single row (no header)', () => {
+    const md = csvToTable('just,one,row');
+    expect(md).toBe('| just | one | row |\n| --- | --- | --- |');
+  });
+
+  test('TSV (Excel/Sheets paste)', () => {
+    const md = csvToTable('a\tb\tc\n1\t2\t3');
+    expect(md).toBe('| a | b | c |\n| --- | --- | --- |\n| 1 | 2 | 3 |');
+  });
+
+  test('returns empty string for empty / whitespace-only input', () => {
+    expect(csvToTable('')).toBe('');
+    expect(csvToTable('   \n   ')).toBe('');
+  });
+});
+
+describe('csvToTable — alignment', () => {
+  test('right-aligns numeric columns', () => {
+    const md = csvToTable('name,count\nalpha,5\nbeta,12');
+    expect(md).toContain('| name | count |');
+    expect(md).toContain('| --- | ---: |');
+  });
+
+  test('keeps left alignment for text columns', () => {
+    const md = csvToTable('name,city\nAlice,Paris');
+    expect(md).toContain('| --- | --- |');
+  });
+
+  test('recognizes currency-formatted numbers as numeric', () => {
+    const md = csvToTable('item,price\napple,$1.50\npear,2.30');
+    expect(md).toContain('| --- | ---: |');
+  });
+
+  test('recognizes percentages', () => {
+    const md = csvToTable('metric,rate\nuptime,99.5%');
+    expect(md).toContain('| --- | ---: |');
+  });
+});
+
+describe('csvToTable — escaping', () => {
+  test('escapes pipes inside cells', () => {
+    const md = csvToTable('a,b\nfoo|bar,baz');
+    expect(md).toContain('| foo\\|bar | baz |');
+  });
+
+  test('escapes newlines inside cells', () => {
+    const md = csvToTable('a,b\n"line1\nline2",x');
+    expect(md).toContain('| line1line2 | x |');
+  });
+});
+
+describe('csvToTable — ragged rows', () => {
+  test('pads short rows with empty cells so the table stays rectangular', () => {
+    const md = csvToTable('a,b,c\n1,2');
+    const dataLines = md.split('\n').slice(2); // skip header + sep
+    expect(dataLines[0]).toBe('| 1 | 2 |  |');
+  });
+});
+
+describe('looksLikeCsv', () => {
+  test('returns true for a tab-separated multi-row paste', () => {
+    expect(looksLikeCsv('a\tb\tc\n1\t2\t3')).toBe(true);
+  });
+
+  test('returns true for a comma-separated multi-row paste', () => {
+    expect(looksLikeCsv('a,b,c\n1,2,3')).toBe(true);
+  });
+
+  test('returns false for a single row', () => {
+    expect(looksLikeCsv('a,b,c')).toBe(false);
+  });
+
+  test('returns false when the column counts don\'t line up at all', () => {
+    expect(looksLikeCsv('a,b,c\n1\n2,3,4,5')).toBe(false);
+  });
+
+  test('returns false for very short / empty input', () => {
+    expect(looksLikeCsv('')).toBe(false);
+    expect(looksLikeCsv('ab')).toBe(false);
+    expect(looksLikeCsv(null)).toBe(false);
+  });
+
+  test('returns false for prose (no delimiter)', () => {
+    expect(looksLikeCsv('The quick brown fox jumps over the lazy dog.\nThe cat sat on the mat.')).toBe(false);
+  });
+});