specwrk 0.15.8 → 0.15.10

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 0a359bd88ba0482dad8bf78555ce9256e8dfa4f000d48f2e36579df34172a684
-  data.tar.gz: a41d177279abb2622d8778a14344ede5f8eac4cac83bbd99185e6302f956f6e7
+  metadata.gz: 1201e277f5b36c6ac9ff9ac20cf2b591f54d42822b071cf61b57a463164c5c4e
+  data.tar.gz: 0c53e4c86f48062680801f3d0de4ac5f691e55b3afaae98398750b9d8bca865d
 SHA512:
-  metadata.gz: 7faa3a2e27b1338fde5aa02b8ae29482b8203857bfb07a54d5c8ff751e73b051b1bc7c69e05bf1b58a8aa609e9de15b767d78a4002936b83d60f3c2f90f928b7
-  data.tar.gz: d6c167e28a11bcc0798074d4214430430cbdb6e34c8b1f1c5c9bed2c086251736c73aefcc984d5ae10ce47e4204d1d13a4b3119a68ed6e26f262ec657839ab4f
+  metadata.gz: 77142f3689fc39ec758869a42611d40b0f2013ebcba1a3c76a3ed4e2d7b6758f80ac0d9e2a93f58fcf58de9d3ada76ed34380d3a81dfc1bc2d1a49cab98fc67c
+  data.tar.gz: bf4c4f3d3784feaed50f41d84e52dac09d232e6fc9d14e79c00961f79286c25775e4a922c4512a0f17ffd45b4bc863ae61b8fe6a70ab9c233c033303e4c7245f

data/CHANGELOG.md

@@ -1,22 +1,57 @@
-## [Unreleased]
-
 # Changelog
 
-## v0.15.2 — 2025-08-15
-[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.1...v0.15.2)
-- Fix bug where ruby objects were being written to ndjson files isntead of json — [#119](https://github.com/danielwestendorf/specwrk/pull/119) by @danielwestendorf
-
-## v0.15.1 — 2025-08-14
-[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.0...v0.15.1)
-- Add `watch` command to split spec files across processes as they change — [#117](https://github.com/danielwestendorf/specwrk/pull/117) by @danielwestendorf
-- Readme formatting tweaks — @danielwestendorf
-
-## v0.15.0 — 2025-08-08
-[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.14.1...v0.15.0)
-- When output path is specified, write an `.ndjson` file per worker of all examples executed — [#113](https://github.com/danielwestendorf/specwrk/pull/113) (fixes [#10](https://github.com/danielwestendorf/specwrk/issues/10)) by @danielwestendorf
-- Print re-run commands for failures — [#114](https://github.com/danielwestendorf/specwrk/pull/114) (fixes [#7](https://github.com/danielwestendorf/specwrk/issues/7)) by @danielwestendorf
-- Show count of examples that did not execute; make runs resumable by only clearing stores on seed — [#115](https://github.com/danielwestendorf/specwrk/pull/115) by @danielwestendorf
-- Report flake counts and provide flake re-run commands — [#116](https://github.com/danielwestendorf/specwrk/pull/116) (fixes [#112](https://github.com/danielwestendorf/specwrk/issues/112)) by @danielwestendorf
+## v0.15.9 — 2025-08-28  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.8...v0.15.9)  
+- Support lock skipping for read-only endpoints — [#139](https://github.com/danielwestendorf/specwrk/pull/139) by [@danielwestendorf](https://github.com/danielwestendorf)  
+- Fix typo in README — [#137](https://github.com/danielwestendorf/specwrk/pull/137) by [@willnet](https://github.com/willnet)  
+- Support an array of arguments when seeding/starting — [#141](https://github.com/danielwestendorf/specwrk/pull/141) by [@danielwestendorf](https://github.com/danielwestendorf)  
+- Load examples from processing before global lock — [#142](https://github.com/danielwestendorf/specwrk/pull/142) by [@danielwestendorf](https://github.com/danielwestendorf)  
+
+## v0.15.8 — 2025-08-27  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.7...v0.15.8)  
+- Friendly helper message when Redis adapter is not available — [@danielwestendorf](https://github.com/danielwestendorf)  
+
+## v0.15.7 — 2025-08-26  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.6...v0.15.7)  
+- Change request locking from global to per-`run_id` — [#135](https://github.com/danielwestendorf/specwrk/pull/135) by [@danielwestendorf](https://github.com/danielwestendorf)  
+- `rspec-core` is a runtime dependency (not `rspec`) — [#128](https://github.com/danielwestendorf/specwrk/pull/128) by [@bquorning](https://github.com/bquorning)  
+- README typo fix — [#126](https://github.com/danielwestendorf/specwrk/pull/126) by [@brett-anderson](https://github.com/brett-anderson)  
+
+## v0.15.6 — 2025-08-26  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.5...v0.15.6)  
+- Test `Store.adapter_klass` — [@danielwestendorf](https://github.com/danielwestendorf)  
+
+## v0.15.5 — 2025-08-26  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.4...v0.15.5)  
+- Add support for Redis store adapters — [#133](https://github.com/danielwestendorf/specwrk/pull/133) by [@danielwestendorf](https://github.com/danielwestendorf)  
+- Remove support for Ruby 3.1.0 — [@danielwestendorf](https://github.com/danielwestendorf)  
+- Remove `gem-release` from dev dependencies — [@danielwestendorf](https://github.com/danielwestendorf)  
+
+## v0.15.4 — 2025-08-23  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.3...v0.15.4)  
+- Split web endpoints into separate files for easier comprehension — [#129](https://github.com/danielwestendorf/specwrk/pull/129) by [@danielwestendorf](https://github.com/danielwestendorf)  
+- Write the title sequence as watch loops run — [#131](https://github.com/danielwestendorf/specwrk/pull/131) by [@danielwestendorf](https://github.com/danielwestendorf) (fixes [#130](https://github.com/danielwestendorf/specwrk/issues/130))  
+
+## v0.15.3 — 2025-08-22  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.2...v0.15.3)  
+- Clear filter-manager inclusions/exclusions when listing examples — [#122](https://github.com/danielwestendorf/specwrk/pull/122) by [@danielwestendorf](https://github.com/danielwestendorf) (fixes [#121](https://github.com/danielwestendorf/specwrk/issues/121))  
+- Smarter completion-threshold calculation; make runs resumable — [#124](https://github.com/danielwestendorf/specwrk/pull/124) by [@danielwestendorf](https://github.com/danielwestendorf) (fixes [#121](https://github.com/danielwestendorf/specwrk/issues/121))  
+
+## v0.15.2 — 2025-08-15  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.1...v0.15.2)  
+- Fix bug where Ruby objects were being written to NDJSON files instead of JSON — [#119](https://github.com/danielwestendorf/specwrk/pull/119) by [@danielwestendorf](https://github.com/danielwestendorf)  
+
+## v0.15.1 — 2025-08-14  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.15.0...v0.15.1)  
+- Add `watch` command to split spec files across processes as they change — [#117](https://github.com/danielwestendorf/specwrk/pull/117) by [@danielwestendorf](https://github.com/danielwestendorf)  
+- README formatting tweaks — [@danielwestendorf](https://github.com/danielwestendorf)  
+
+## v0.15.0 — 2025-08-08  
+[Compare](https://github.com/danielwestendorf/specwrk/compare/v0.14.1...v0.15.0)  
+- Per-worker NDJSON output when `--output` is specified — [#113](https://github.com/danielwestendorf/specwrk/pull/113) by [@danielwestendorf](https://github.com/danielwestendorf) (fixes [#10](https://github.com/danielwestendorf/specwrk/issues/10))  
+- Print re-run commands for failures — [#114](https://github.com/danielwestendorf/specwrk/pull/114) by [@danielwestendorf](https://github.com/danielwestendorf) (fixes [#7](https://github.com/danielwestendorf/specwrk/issues/7))  
+- Show number of examples that did not execute; make runs resumable — [#115](https://github.com/danielwestendorf/specwrk/pull/115) by [@danielwestendorf](https://github.com/danielwestendorf)  
+- Report flake counts and re-run commands — [#116](https://github.com/danielwestendorf/specwrk/pull/116) by [@danielwestendorf](https://github.com/danielwestendorf) (fixes [#112](https://github.com/danielwestendorf/specwrk/issues/112))  
 
 ## v0.14.1 — 2025-08-07
 [Compare](https://github.com/danielwestendorf/specwrk/compare/v0.14.0...v0.14.1)

data/README.md

@@ -205,7 +205,7 @@ Multi-node, multi-process works best when have many nodes running tests. This di
 To accomplish this, a central queue server is required, examples must be explicitly seeded, and workers explicitly started.
 
 1. Start a centralized queue server (see [Running a persistent Queue Server](#running-a-persistent-queue-server)) 
-2. Seed the server with the specs for the current `SPECWWRK_RUN` pointed at your central server
+2. Seed the server with the specs for the current `SPECWRK_RUN` pointed at your central server
 3. Execute `specwrk work` for the given process count, for the current `SPECWRK_RUN`, pointed at your central server
 
 [GitHub Actions Example](https://github.com/danielwestendorf/specwrk/blob/main/.github/workflows/specwrk-multi-node.yml)
@@ -223,6 +223,8 @@ Start a persistent Queue Server given one of the following methods
 - Secure your server with a key either with the `SPECWRK_SRV_KEY` environment variable or `--key` CLI option
 - Configure the server output to be a persisted volume so your timings survive between system restarts with the `SPECWRK_SRV_STORE_URI` environment variable or `--store-uri` CLI option. By default, `memory:///` will be used for the run's data stores (so run data will no survive server restarts) while `file://#{Dir.tmpdir}` will be used for run timings. Pass `--store-uri file:///whatever/absolute/path` to store all data on disk (required for multiple server processes).
 
+See [specwrk-store-redis_adapter](https://github.com/danielwestendorf/specwrk-store-redis_adapter) for Redis-compatible backed storage.
+
 See [specwrk serve --help](#specwrk-serve) for all possible configuration options.
 
 ### Create a watchfile for the `watch` command
@@ -273,11 +275,16 @@ specwrk is different because it:
 5. Is free.
 
 [parallel_rspec](https://github.com/willbryant/parallel_rspec)
+
 [Knapsack](https://github.com/KnapsackPro/knapsack)
+
 [parallel_tests](https://github.com/grosser/parallel_tests)
+
 [rspecq](https://github.com/skroutz/rspecq)
+
 [RSpec ABQ](https://github.com/rwx-research/rspec-abq)
 
+
 ## Contributing
 
 Bug reports and pull requests are welcome on GitHub at https://github.com/dwestendorf/specwrk.

data/lib/specwrk/cli.rb

@@ -131,9 +131,11 @@ module Specwrk
 
       desc "Seed the server with a list of specs for the run"
       option :max_retries, default: 0, desc: "Number of times an example will be re-run should it fail"
-      argument :dir, required: false, default: "spec", desc: "Relative spec directory to run against"
+      argument :dir, type: :array, required: false, desc: "Relative spec directory to run against, default: spec/"
 
       def call(max_retries:, dir:, **args)
+        dir = ["spec"] if dir.length.zero?
+
         self.class.setup(**args)
 
         require "specwrk/list_examples"
@@ -208,9 +210,11 @@ module Specwrk
 
       desc "Start a server and workers, monitor until complete"
       option :max_retries, default: 0, desc: "Number of times an example will be re-run should it fail"
-      argument :dir, required: false, default: "spec", desc: "Relative spec directory to run against"
+      argument :dir, type: :array, required: false, desc: "Relative spec directory to run against, default: spec/"
 
       def call(max_retries:, dir:, **args)
+        dir = ["spec"] if dir.length.zero?
+
         self.class.setup(**args)
         $stdout.sync = true
 

data/lib/specwrk/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Specwrk
-  VERSION = "0.15.8"
+  VERSION = "0.15.10"
 end

data/lib/specwrk/web/endpoints/base.rb

@@ -46,6 +46,10 @@ module Specwrk
 
         attr_reader :request
 
+        def skip_lock
+          false
+        end
+
         def before_lock
         end
 
@@ -119,7 +123,11 @@ module Specwrk
         end
 
         def with_lock
-          Store.with_lock(URI(ENV.fetch("SPECWRK_SRV_STORE_URI", "memory:///")), run_id) { yield }
+          if skip_lock
+            yield
+          else
+            Store.with_lock(URI(ENV.fetch("SPECWRK_SRV_STORE_URI", "memory:///")), run_id) { yield }
+          end
         end
       end
 

data/lib/specwrk/web/endpoints/complete_and_pop.rb

@@ -19,8 +19,17 @@ module Specwrk
 
         private
 
+        def before_lock
+          processing_examples
+          completed_examples
+        end
+
         def all_examples
-          @all_examples ||= payload.map { |example| [example[:id], example] if processing[example[:id]] }.compact.to_h
+          @all_examples ||= payload.map { |example| [example[:id], example] if processing_examples[example[:id]] }.compact.to_h
+        end
+
+        def processing_examples
+          @processing_examples ||= processing.multi_read(*payload.map { |example| example[:id] })
         end
 
         def completed_examples
@@ -67,7 +76,7 @@ module Specwrk
           # So if we overwrite run times from another process it is nbd
           run_times.merge! run_time_data
 
-          # workers are single proces, single-threaded, so safe to do this work without the lock
+          # workers are single process, single-threaded, so safe to do this work without the lock
           existing_status_counts = worker.multi_read(*EXAMPLE_STATUSES)
           new_status_counts = EXAMPLE_STATUSES.map do |status|
             [status, existing_status_counts.fetch(status, 0) + completed_examples_status_counts.fetch(status, 0)]

data/lib/specwrk/web/endpoints/report.rb

@@ -13,6 +13,12 @@ module Specwrk
 
           [200, {"content-type" => "application/json"}, [JSON.generate(completed_dump)]]
         end
+
+        private
+
+        def skip_lock
+          true
+        end
       end
     end
   end

data/lib/specwrk/web/endpoints/shutdown.rb

@@ -12,6 +12,12 @@ module Specwrk
           [200, {"content-type" => "text/plain"}, ["✌️"]]
         end
 
+        private
+
+        def skip_lock
+          true
+        end
+
         def interupt!
           Thread.new do
             # give the socket a moment to flush the response

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: specwrk
 version: !ruby/object:Gem::Version
-  version: 0.15.8
+  version: 0.15.10
 platform: ruby
 authors:
 - Daniel Westendorf