From 80944488774c20010193dcfa4a2a7295ec090c96 Mon Sep 17 00:00:00 2001
From: jbranchaud <jbranchaud@gmail.com>
Date: Tue, 30 Dec 2025 10:39:03 -0700
Subject: [PATCH] Add Join URI Path Parts as a Ruby TIL

---
 README.md                   |  3 ++-
 ruby/join-uri-path-parts.md | 43 +++++++++++++++++++++++++++++++++++++
 2 files changed, 45 insertions(+), 1 deletion(-)
 create mode 100644 ruby/join-uri-path-parts.md

diff --git a/README.md b/README.md
index 2938f42..0655292 100644
--- a/README.md
+++ b/README.md
@@ -10,7 +10,7 @@ pairing with smart people at Hashrocket.
 
 For a steady stream of TILs, [sign up for my newsletter](https://visualmode.kit.com/newsletter).
 
-_1712 TILs and counting..._
+_1713 TILs and counting..._
 
 See some of the other learning resources I work on:
 
@@ -1412,6 +1412,7 @@ If you've learned something here, support my efforts writing daily TILs by
 - [Install Latest Version Of Ruby With asdf](ruby/install-latest-version-of-ruby-with-asdf.md)
 - [Invoking Rake Tasks Multiple Times](ruby/invoking-rake-tasks-multiple-times.md)
 - [IRB Has Built-In Benchmarking With Ruby 3](ruby/irb-has-built-in-benchmarking-with-ruby-3.md)
+- [Join URI Path Parts](ruby/join-uri-path-parts.md)
 - [Jump Out Of A Nested Context With Throw/Catch](ruby/jump-out-of-a-nested-context-with-throw-catch.md)
 - [Last Raised Exception In The Call Stack](ruby/last-raised-exception-in-the-call-stack.md)
 - [Limit Split](ruby/limit-split.md)
diff --git a/ruby/join-uri-path-parts.md b/ruby/join-uri-path-parts.md
new file mode 100644
index 0000000..9632500
--- /dev/null
+++ b/ruby/join-uri-path-parts.md
@@ -0,0 +1,43 @@
+# Join URI Path Parts
+
+The
+[`URI.join`](https://ruby-doc.org/stdlib-2.5.1/libdoc/uri/rdoc/URI.html#method-c-join)
+method seems like a handy way to combine a base URL with some subpath. However,
+there are some subtle gotchas depending on where forward slashes appear in the
+two arguments.
+
+Let's first look at the, in my opinion, desired behavior:
+
+```ruby
+> URI.join("https://example.com/api/v1/", "users")
+=> #<URI::HTTPS https://example.com/api/v1/users>
+```
+
+The base URL has a trailing slash and the path that I want to join to it has no
+leading slash. The result is a path where `users` is joined to the end of the
+base URL. That's what I'm looking for.
+
+Now, let's see some variations on the above approach that give results that I
+wasn't expecting and don't want.
+
+```ruby
+> URI.join("https://example.com/api/v1", "/users") # 1
+=> #<URI::HTTPS https://example.com/users>
+> URI.join("https://example.com/api/v1", "users") # 2
+=> #<URI::HTTPS https://example.com/api/users>
+> URI.join("https://example.com/api/v1/", "/users") # 3
+=> #<URI::HTTPS https://example.com/users>
+```
+
+1. No trailing slash on the base URL. Leading slash on the path to join. The
+   path portion of the base URL is wiped out and `/users` is joined in.
+2. No trailing slash on the base URL. No leading slash on the path to join. The
+   `users` path replaces the last part of the path in the base URL.
+3. Both a trailing slash in the base URL and a leading slash in the path to
+   join. Same behavior as 1.
+
+I have two takeaways from this:
+- Use with caution. If I'm going to use `URI.join` for this purpose, I need to
+  be careful to only use the form in the first code block.
+- The `URI.join` method is probably meant to be primarily used to join a domain
+  (e.g. `http://example.com`) that has no path with some path segment.