Fix URL parser edge cases from PR review

- Strip .git suffix from repo/project names (clone URLs now work)
- Conservative self-hosted detection (hostname.startsWith instead of includes)
- CLI: Reorder args so options before URL are handled correctly

Fixes:
1. URLs like https://github.com/owner/repo.git now parse correctly
2. notgitlab.com no longer incorrectly matches as GitLab
3.  now works (options can be anywhere)

Added 8 new tests for edge cases.

Agent-Id: agent-ce81a04d-72f2-4289-8eb7-c3074d7d8030

Author: richhankins

src/bin/index.ts

@@ -30,16 +30,23 @@ program.addCommand(mcpCommand);
 program.addCommand(agentCommand);
 
 // Auto-detect URL mode: ctxc index <url> -> ctxc index url <url>
-// This allows users to skip the 'url' subcommand when providing a URL directly
+// Scan for URL anywhere after 'index' to support: ctxc index -i name https://...
 const indexIdx = process.argv.indexOf("index");
-if (indexIdx !== -1 && indexIdx + 1 < process.argv.length) {
-  const nextArg = process.argv[indexIdx + 1];
+if (indexIdx !== -1) {
   const subcommands = ["url", "github", "gitlab", "bitbucket", "website"];
-  if (
-    nextArg.match(/^https?:\/\//) &&
-    !subcommands.includes(nextArg)
-  ) {
-    process.argv.splice(indexIdx + 1, 0, "url");
+  // Find first URL-like argument after 'index'
+  for (let i = indexIdx + 1; i < process.argv.length; i++) {
+    const arg = process.argv[i];
+    // Stop if we hit a known subcommand
+    if (subcommands.includes(arg)) break;
+    // Found a URL - reorder args to put 'url' and the URL right after 'index'
+    if (arg.match(/^https?:\/\//)) {
+      // Remove the URL from its current position
+      process.argv.splice(i, 1);
+      // Insert 'url' <url> right after 'index'
+      process.argv.splice(indexIdx + 1, 0, "url", arg);
+      break;
+    }
   }
 }
 

src/core/url-parser.test.ts

@@ -161,3 +161,60 @@ describe("parseSourceUrl", () => {
   });
 });
 
+
+describe("Edge cases", () => {
+  describe(".git suffix handling", () => {
+    it("strips .git suffix from GitHub URLs", () => {
+      const result = parseSourceUrl("https://github.com/owner/repo.git");
+      expect(result.type).toBe("github");
+      expect(result.config).toEqual({ owner: "owner", repo: "repo", ref: "HEAD" });
+      expect(result.defaultIndexName).toBe("repo");
+    });
+
+    it("strips .git suffix from GitLab URLs", () => {
+      const result = parseSourceUrl("https://gitlab.com/group/project.git");
+      expect(result.type).toBe("gitlab");
+      expect(result.config).toEqual({ projectId: "group/project", ref: "HEAD", baseUrl: undefined });
+      expect(result.defaultIndexName).toBe("project");
+    });
+
+    it("strips .git suffix from Bitbucket URLs", () => {
+      const result = parseSourceUrl("https://bitbucket.org/workspace/repo.git");
+      expect(result.type).toBe("bitbucket");
+      expect(result.config).toEqual({
+        workspace: "workspace",
+        repo: "repo",
+        ref: "HEAD",
+        baseUrl: undefined,
+      });
+      expect(result.defaultIndexName).toBe("repo");
+    });
+  });
+
+  describe("Conservative self-hosted detection", () => {
+    it("detects gitlab.company.com as GitLab", () => {
+      const result = parseSourceUrl("https://gitlab.company.com/team/project");
+      expect(result.type).toBe("gitlab");
+    });
+
+    it("does NOT match notgitlab.com as GitLab", () => {
+      const result = parseSourceUrl("https://notgitlab.com/some/path");
+      expect(result.type).toBe("website");
+    });
+
+    it("does NOT match mygitlabserver.com as GitLab", () => {
+      const result = parseSourceUrl("https://mygitlabserver.com/some/path");
+      expect(result.type).toBe("website");
+    });
+
+    it("detects bitbucket.company.com as Bitbucket", () => {
+      const result = parseSourceUrl("https://bitbucket.company.com/workspace/repo");
+      expect(result.type).toBe("bitbucket");
+    });
+
+    it("does NOT match notbitbucket.org as Bitbucket", () => {
+      const result = parseSourceUrl("https://notbitbucket.org/some/path");
+      expect(result.type).toBe("website");
+    });
+  });
+});

src/core/url-parser.ts

@@ -9,6 +9,14 @@ import type { GitLabSourceConfig } from "../sources/gitlab.js";
 import type { BitBucketSourceConfig } from "../sources/bitbucket.js";
 import type { WebsiteSourceConfig } from "../sources/website.js";
 
+/**
+ * Strip .git suffix from repo/project names
+ */
+function stripGitSuffix(name: string): string {
+  return name.endsWith(".git") ? name.slice(0, -4) : name;
+}
+
+
 /**
  * Result of parsing a source URL
  */
@@ -43,12 +51,12 @@ export function parseSourceUrl(urlString: string): ParsedUrl {
   }
 
   // GitLab (gitlab.com or hostname contains "gitlab")
-  if (hostname === "gitlab.com" || hostname.includes("gitlab")) {
+  if (hostname === "gitlab.com" || hostname.startsWith("gitlab.")) {
     return parseGitLabUrl(url);
   }
 
   // Bitbucket (bitbucket.org or hostname contains "bitbucket")
-  if (hostname === "bitbucket.org" || hostname.includes("bitbucket")) {
+  if (hostname === "bitbucket.org" || hostname.startsWith("bitbucket.")) {
     return parseBitBucketUrl(url);
   }
 
@@ -76,7 +84,7 @@ function parseGitHubUrl(url: URL): ParsedUrl {
   }
 
   const owner = pathParts[0];
-  const repo = pathParts[1];
+  const repo = stripGitSuffix(pathParts[1]);
   let ref = "HEAD";
 
   // Check for tree/branch or commit/sha patterns
@@ -121,6 +129,11 @@ function parseGitLabUrl(url: URL): ParsedUrl {
     }
   }
 
+  // Strip .git suffix from project name if present
+  const lastPart = projectParts[projectParts.length - 1];
+  if (lastPart.endsWith(".git")) {
+    projectParts[projectParts.length - 1] = stripGitSuffix(lastPart);
+  }
   const projectId = projectParts.join("/");
   const projectName = projectParts[projectParts.length - 1];
 
@@ -149,7 +162,7 @@ function parseBitBucketUrl(url: URL): ParsedUrl {
   }
 
   const workspace = pathParts[0];
-  const repo = pathParts[1];
+  const repo = stripGitSuffix(pathParts[1]);
   let ref = "HEAD";
 
   // Check for /src/branch or /branch/name patterns