feat: suggest correct paths based on levenshtein distance

Author: tlouisse

.changeset/gentle-crabs-provide.md

@@ -0,0 +1,5 @@
+---
+'check-html-links': minor
+---
+
+suggest correct paths based on levenshtein distance

packages/check-html-links/src/CheckHtmlLinksCli.js

@@ -75,7 +75,7 @@ export class CheckHtmlLinksCli {
         )} missing reference targets (used by ${referenceCount} links) while checking ${
           files.length
         } files:`,
-        ...formatErrors(errors)
+        ...formatErrors(errors, { files })
           .split('\n')
           .map(line => `  ${line}`),
         `Checking links duration: ${performance[0]}s ${performance[1] / 1000000}ms`,

packages/check-html-links/src/formatErrors.js

@@ -1,13 +1,15 @@
 import path from 'path';
 import chalk from 'chalk';
+import levenshtein from './levenshtein.js';
 
 /** @typedef {import('../types/main').Error} Error */
 
 /**
  * @param {Error[]} errors
- * @param {*} relativeFrom
+ * @param {{ relativeFrom?: string; files: string[] }} opts
  */
-export function formatErrors(errors, relativeFrom = process.cwd()) {
+// @ts-expect-error we need empty obj to destructure from
+export function formatErrors(errors, { relativeFrom = process.cwd(), files } = {}) {
   let output = [];
   let number = 0;
   for (const error of errors) {
@@ -41,6 +43,27 @@ export function formatErrors(errors, relativeFrom = process.cwd()) {
       output.push(`  ... ${more} more references to this target`);
     }
     output.push('');
+
+    /**
+     * Also consider finding the updated path. This can be useful when documentation is restructured
+     * For instance, the folder name was changed, but the file name was not.
+     */
+    let suggestion;
+    let lowestScore = -1;
+    files.forEach(file => {
+      const filePathToCompare = file.replace(relativeFrom + '/', '');
+      const score = levenshtein(filePathToCompare, filePath);
+      if (score && (lowestScore === -1 || score < lowestScore)) {
+        lowestScore = score;
+        suggestion = filePathToCompare;
+      }
+    });
+
+    if (suggestion) {
+      output.push(
+        chalk.italic(`Suggestion: did you mean ${chalk.magenta(suggestion)} instead?\n\n`),
+      );
+    }
   }
   return output.join('\n');
 }

packages/check-html-links/src/levenshtein.js

@@ -0,0 +1,107 @@
+/* eslint-disable */
+// https://github.com/gustf/js-levenshtein/blob/master/index.js
+
+/**
+ * @param {number} d0
+ * @param {number} d1
+ * @param {number} d2
+ * @param {any} bx
+ * @param {any} ay
+ */
+function _min(d0, d1, d2, bx, ay) {
+  return d0 < d1 || d2 < d1 ? (d0 > d2 ? d2 + 1 : d0 + 1) : bx === ay ? d1 : d1 + 1;
+}
+
+/**
+ * @param {string} a
+ * @param {string} b
+ * @returns {number|undefined}
+ */
+export default function (a, b) {
+  if (a === b) {
+    return 0;
+  }
+
+  if (a.length > b.length) {
+    var tmp = a;
+    a = b;
+    b = tmp;
+  }
+
+  var la = a.length;
+  var lb = b.length;
+
+  while (la > 0 && a.charCodeAt(la - 1) === b.charCodeAt(lb - 1)) {
+    la--;
+    lb--;
+  }
+
+  var offset = 0;
+
+  while (offset < la && a.charCodeAt(offset) === b.charCodeAt(offset)) {
+    offset++;
+  }
+
+  la -= offset;
+  lb -= offset;
+
+  if (la === 0 || lb < 3) {
+    return lb;
+  }
+
+  var x = 0;
+  var y;
+  var d0;
+  var d1;
+  var d2;
+  var d3;
+  var dd;
+  var dy;
+  var ay;
+  var bx0;
+  var bx1;
+  var bx2;
+  var bx3;
+
+  var vector = [];
+
+  for (y = 0; y < la; y++) {
+    vector.push(y + 1);
+    vector.push(a.charCodeAt(offset + y));
+  }
+
+  var len = vector.length - 1;
+
+  for (; x < lb - 3; ) {
+    bx0 = b.charCodeAt(offset + (d0 = x));
+    bx1 = b.charCodeAt(offset + (d1 = x + 1));
+    bx2 = b.charCodeAt(offset + (d2 = x + 2));
+    bx3 = b.charCodeAt(offset + (d3 = x + 3));
+    dd = x += 4;
+    for (y = 0; y < len; y += 2) {
+      dy = vector[y];
+      ay = vector[y + 1];
+      d0 = _min(dy, d0, d1, bx0, ay);
+      d1 = _min(d0, d1, d2, bx1, ay);
+      d2 = _min(d1, d2, d3, bx2, ay);
+      dd = _min(d2, d3, dd, bx3, ay);
+      vector[y] = dd;
+      d3 = d2;
+      d2 = d1;
+      d1 = d0;
+      d0 = dy;
+    }
+  }
+
+  for (; x < lb; ) {
+    bx0 = b.charCodeAt(offset + (d0 = x));
+    dd = ++x;
+    for (y = 0; y < len; y += 2) {
+      dy = vector[y];
+      vector[y] = dd = _min(dy, d0, dd, bx0, vector[y + 1]);
+      d0 = dy;
+    }
+  }
+
+  return dd;
+}

packages/check-html-links/test-node/formatErrors.test.js

@@ -1,13 +1,20 @@
 import chai from 'chai';
 import chalk from 'chalk';
-import { execute } from './test-helpers.js';
 import { formatErrors } from 'check-html-links';
+import path from 'path';
+import { fileURLToPath } from 'url';
+import { execute } from './test-helpers.js';
+import { listFiles } from '../src/listFiles.js';
 
 const { expect } = chai;
+const __dirname = path.dirname(fileURLToPath(import.meta.url));
 
 async function executeAndFormat(inPath) {
   const { errors, cleanup } = await execute(inPath);
-  return formatErrors(cleanup(errors));
+  const testDir = path.join(__dirname, inPath.split('/').join(path.sep));
+  const rootDir = path.resolve(testDir);
+  const files = await listFiles('**/*.html', rootDir);
+  return formatErrors(cleanup(errors), { files });
 }
 
 describe('formatErrors', () => {
@@ -16,23 +23,34 @@ describe('formatErrors', () => {
     chalk.level = 0;
   });
 
-  it('prints a nice summery', async () => {
+  it('prints a nice summary', async () => {
     const result = await executeAndFormat('fixtures/test-case');
     expect(result.trim().split('\n')).to.deep.equal([
       '1. missing id="my-teams" in fixtures/test-case/price/index.html',
       '  from fixtures/test-case/history/index.html:1:9 via href="/price/#my-teams"',
       '',
+      'Suggestion: did you mean test-node/fixtures/test-case/price/index.html instead?',
+      '',
+      '',
       '2. missing file fixtures/test-case/about/images/team.png',
       '  from fixtures/test-case/about/index.html:3:10 via src="./images/team.png"',
       '',
+      'Suggestion: did you mean test-node/fixtures/test-case/about/index.html instead?',
+      '',
+      '',
       '3. missing reference target fixtures/test-case/aboot',
       '  from fixtures/test-case/about/index.html:6:11 via href="/aboot"',
       '  from fixtures/test-case/history/index.html:4:11 via href="/aboot"',
       '  from fixtures/test-case/index.html:4:11 via href="/aboot"',
       '  ... 2 more references to this target',
       '',
+      'Suggestion: did you mean test-node/fixtures/test-case/index.html instead?',
+      '',
+      '',
       '4. missing reference target fixtures/test-case/prce',
       '  from fixtures/test-case/index.html:1:9 via href="./prce"',
+      '',
+      'Suggestion: did you mean test-node/fixtures/test-case/index.html instead?',
     ]);
   });
 });