Move performance readme to benchmarking folder

README.md

@@ -63,31 +63,3 @@ The IO format will write to anything that responds to the `#write` method.
 # config/environments/production.rb (or staging, etc)
 config.logger = Timber::Logger.new(Timber::LogDevices::IO.new(STDOUT)))
 ```
-
-## Performance
-
-Timber was designed with an obsessive focus on performance and resource usage. In the `/benchmark` folder you'll find benchmarking tests.
-
-The following results were run on a bare metal server in order to achieve consistent and repeatable results.
-
-```console
-+---------------------------+------------+-----------------+------------------+
-|    Timber benchmarking. 10 requests per test. Times are "real" CPU time.    |
-+---------------------------+------------+-----------------+------------------+
-|                           | Total      | Per request avg | Per request diff |
-+---------------------------+------------+-----------------+------------------+
-| Control                   | 0.00568414 | 0.00056841      |                  |
-| Timber probes only        | 0.00673819 | 0.00067382      | 0.0001054        |
-| Timber probes and logging | 0.00912786 | 0.00091279      | 0.00023897       |
-+---------------------------+------------+-----------------+------------------+
-```
-
-1. `Control` - This is vanilla rails app without Timber installed.
-2. `Timber probels only` - The same rails app but with the `Timber::Probes` installed, isolating probes performance.
-3. `Timber probes and logging` - Testing the full Timber library, probes and adding context to each log line.
-
-The benchmark can be run yourself via:
-
-```console
-$ appraisal ruby benchmark/rails_request.rb
-```

benchmark/README.md

@@ -0,0 +1,27 @@
+## Benchmarking & Performance
+
+Timber was designed with an obsessive focus on performance and resource usage. In the `/benchmark` folder you'll find benchmarking tests.
+
+The following results were run on a bare metal server in order to achieve consistent and repeatable results.
+
+```console
++---------------------------+------------+-----------------+------------------+
+|    Timber benchmarking. 10 requests per test. Times are "real" CPU time.    |
++---------------------------+------------+-----------------+------------------+
+|                           | Total      | Per request avg | Per request diff |
++---------------------------+------------+-----------------+------------------+
+| Control                   | 0.00568414 | 0.00056841      |                  |
+| Timber probes only        | 0.00673819 | 0.00067382      | 0.0001054        |
+| Timber probes and logging | 0.00912786 | 0.00091279      | 0.00023897       |
++---------------------------+------------+-----------------+------------------+
+```
+
+1. `Control` - This is vanilla rails app without Timber installed.
+2. `Timber probels only` - The same rails app but with the `Timber::Probes` installed, isolating probes performance.
+3. `Timber probes and logging` - Testing the full Timber library, probes and adding context to each log line.
+
+The benchmark can be run yourself via:
+
+```console
+$ appraisal ruby benchmark/rails_request.rb
+```

lib/timber/macros/deep_set.rb

@@ -0,0 +1,13 @@
+module Timber
+  module Macros
+    module DeepSet
+      def set(hash, path, value)
+        keys = path.split(".")
+        target_hash = keys[0..-2].inject(hash) do |acc, key|
+          acc[key.to_sym] || raise("could not find key #{value.inspect} for #{hash}")
+        end
+        target_hash[keys.last] = value
+      end
+    end
+  end
+end