fix: remove external depedency for string similarity calculation (#9944)

* fix: remove depedency for edit distance * test: ut
OfficeDev · Sep 14, 2023 · 60d5b8a · 60d5b8a
1 parent 3ea9bb2
commit 60d5b8a
Show file tree

Hide file tree

Showing 5 changed files with 341 additions and 78 deletions.
diff --git a/packages/cli/package.json b/packages/cli/package.json
@@ -109,7 +109,6 @@
     "figures": "^3.2.0",
     "fs-extra": "^9.1.0",
     "lodash": "^4.17.21",
-    "natural": "^6.7.2",
     "node-machine-id": "^1.1.12",
     "open": "^8.2.1",
     "semver": "^7.5.4",

diff --git a/packages/cli/src/commands/engine.ts b/packages/cli/src/commands/engine.ts
@@ -43,10 +43,9 @@ import {
 import CliTelemetry from "../telemetry/cliTelemetry";
 import { TelemetryComponentType, TelemetryProperty } from "../telemetry/cliTelemetryEvents";
 import UI from "../userInteraction";
-import { CliConfigOptions, UserSettings } from "../userSetttings";
-import { getSystemInputs } from "../utils";
+import { CliConfigOptions } from "../userSetttings";
+import { editDistance, getSystemInputs } from "../utils";
 import { helper } from "./helper";
-import natural from "natural";
 
 class CLIEngine {
   /**
@@ -280,7 +279,7 @@ class CLIEngine {
     let mini = token.length;
     let mostSimilarCommand: CLICommand | undefined = undefined;
     for (const cmd of context.command.commands || []) {
-      const d = natural.LevenshteinDistance(token, cmd.name);
+      const d = editDistance(token, cmd.name);
       if (d < mini && d <= 2) {
         mini = d;
         mostSimilarCommand = cmd;

diff --git a/packages/cli/src/utils.ts b/packages/cli/src/utils.ts
@@ -235,3 +235,33 @@ export async function getTemplates(): Promise<Sample[]> {
   });
   return samples;
 }
+
+export function editDistance(s1: string, s2: string): number {
+  const len1 = s1.length;
+  const len2 = s2.length;
+
+  // Create a 2D array to store the edit distances
+  const dp: number[][] = new Array(len1 + 1).fill(0).map(() => new Array(len2 + 1).fill(0));
+
+  // Initialize the first row and column
+  for (let i = 0; i <= len1; i++) {
+    dp[i][0] = i;
+  }
+  for (let j = 0; j <= len2; j++) {
+    dp[0][j] = j;
+  }
+
+  // Calculate the edit distance using dynamic programming
+  for (let i = 1; i <= len1; i++) {
+    for (let j = 1; j <= len2; j++) {
+      const cost = s1[i - 1] === s2[j - 1] ? 0 : 1;
+      dp[i][j] = Math.min(
+        dp[i - 1][j] + 1, // Deletion
+        dp[i][j - 1] + 1, // Insertion
+        dp[i - 1][j - 1] + cost // Substitution
+      );
+    }
+  }
+
+  return dp[len1][len2];
+}
diff --git a/packages/cli/tests/unit/utils.tests.ts b/packages/cli/tests/unit/utils.tests.ts
@@ -13,6 +13,7 @@ import activate from "../../src/activate";
 import AzureAccountManager from "../../src/commonlib/azureLogin";
 import { UserSettings } from "../../src/userSetttings";
 import {
+  editDistance,
   flattenNodes,
   getColorizedString,
   getSettingsVersion,
@@ -416,3 +417,28 @@ describe("activate", async () => {
     }
   });
 });
+
+describe("editDistance", async () => {
+  it("happy", async () => {
+    {
+      const d = editDistance("a", "b");
+      expect(d).equals(1);
+    }
+    {
+      const d = editDistance("abc", "abd");
+      expect(d).equals(1);
+    }
+    {
+      const d = editDistance("abc", "aabbc");
+      expect(d).equals(2);
+    }
+    {
+      const d = editDistance("", "abc");
+      expect(d).equals(3);
+    }
+    {
+      const d = editDistance("abc", "");
+      expect(d).equals(3);
+    }
+  });
+});