iframe-proxy

vdye · 2023-03-10T18:53:24Z

Closes #23

Summary

This pull request creates a Cucumber/TypeScript end-to-end testing framework for the Git Bundle Server (see #23 for background on the framework & reasoning for why TypeScript was selected). Two classes of tests are created to start with: "basic" tests (demonstrating expected typical workflows) and "performance" tests (comparing performance with & without use of the bundle server on large repositories).

The commits are broken down as follows:

Commit 1 sets up a TypeScript Cucumber project with no tests
Commits 2-4 create the basic scenario tests, including the corresponding test step definitions & helper classes.
Commit 5 adds the performance tests, as well as the filters to prevent running those tests by default
Commit 6 updates all of the workflows to use the 'stable' version of Go (in preparation for a new workflow that does the same)
Commit 7 adds an e2e-test target to the Makefile
Commit 8 adds a new test.yml workflow

Performance results

The performance tests are quite slow, so they aren't run by default in the Actions workflow. For reference, some example results on my machine were:

Clone execution time for https://github.com/git/git.git: 20.70720636224747s (bundle URI) vs. 27.499098003387452s (no bundle URI)
Clone execution time for https://github.com/torvalds/linux.git: 474.9542240447998s (bundle URI) vs. 623.6930130381584s (no bundle URI)

Note that this only compares the clone time; it also took torvalds/linux ~10 minutes to clone & generate its 4.1Gb(!) base bundle. Since that scale is more reflective of the monorepos this bundle server would be used with than the small repos used in the "basic" tests, it's probably worth doing some more detailed profiling on the bundle server. Hopefully, we'd find some places where we can improve performance.

Set up a Cucumber (TypeScript) test framework in the 'test/e2e' directory. This commit contains only the infrastructure needed to write and run the tests; the tests themselves will be added in future commits. Signed-off-by: Victoria Dye <vdye@github.com>

Add 'basic.feature' to test common workflows using the bundle server. Initially, add a common "Background" step starting the bundle web server at the specified port, and a common "After" step run when a test ends (successful or not) stopping that web server process. Signed-off-by: Victoria Dye <vdye@github.com>

derrickstolee

I was attempting to run these locally, but I'm having issues with node.js version at least 14 working with an npm install on my Ubuntu machine. Don't let my inability to install software block this effort.

derrickstolee · 2023-03-17T13:25:22Z

+  Scenario: A user can clone with a bundle URI pointing to the bundle server
+    Given a remote repository 'https://github.com/vdye/asset-hash.git'
+    Given the bundle server has been initialized with the remote repo
+    When I clone from the remote repo with a bundle URI
+    Then bundles are downloaded and used


This is an interesting formal mechanism for describing tests. Each of these lines are repeatable steps that can be mixed and matched. For instance, I could imagine (not looking ahead) a test like:

Scenario: User fails to clone from a non-existing bundle URI Given a remote repository 'https://github.com/vdye/asset-hash.git' When I clone from the remote repo with a bundle URI Then bundles are not downloaded and clone succeeds with a warning

The only new step to build is the Then statement. Neat!

derrickstolee · 2023-03-17T13:27:36Z

+  // Ensure warning wasn't thrown
+  clonedRepo.cloneResult.stderr.toString().split("\n").forEach(function (line) {
+    if (line.startsWith("warning: failed to download bundle from URI")) {
+      assert.fail(line)
+    }


I wonder if it would be worth asking git for-each-ref --format="%(refname)" "refs/bundles/*" if it has at least one ref?

derrickstolee · 2023-03-17T13:28:35Z

+  result = clonedRepo.runGit("for-each-ref", "--format=%(refname)")
+  utils.assertStatus(0, result, "git for-each-ref failed")
+
+  const bundleRefRegex = new RegExp('^refs/bundles/.*$')
+  const bundleRefs = result.stdout.toString().split("\n").filter(function (line) {
+    return bundleRefRegex.test(line)


Ah, you do it here! We could add the refs/bundles/* filter to the command, then we don't need to regex on the result, just look for at least one line.

derrickstolee · 2023-03-17T13:38:53Z

+  Scenario: A user can fetch with a bundle server that's behind and get all updates
+    Given a new remote repository with main branch 'main'
+    Given another user pushed 10 commits to 'main'
+    Given the bundle server has been initialized with the remote repo
+    Given I cloned from the remote repo with a bundle URI
+    Given another user pushed 2 commits to 'main'
+    When I fetch from the remote
+    Then I am up-to-date with 'main'
+    Then my repo's bundles are not up-to-date with 'main'
+
+  Scenario: A user will fetch incremental bundles to stay up-to-date
+    Given a new remote repository with main branch 'main'
+    Given another user pushed 10 commits to 'main'
+    Given the bundle server has been initialized with the remote repo
+    Given I cloned from the remote repo with a bundle URI
+    Given another user pushed 2 commits to 'main'
+    Given the bundle server was updated for the remote repo
+    When I fetch from the remote
+    Then I am up-to-date with 'main'
+    Then my repo's bundles are up-to-date with 'main'
+
+  Scenario: A user can fetch force-pushed refs from the bundle server
+    Given a new remote repository with main branch 'main'
+    Given another user pushed 10 commits to 'main'
+    Given the bundle server has been initialized with the remote repo
+    Given I cloned from the remote repo with a bundle URI
+    Given another user removed 2 commits and added 4 commits to 'main'
+    Given the bundle server was updated for the remote repo
+    When I fetch from the remote
+    Then I am up-to-date with 'main'
+    Then my repo's bundles are up-to-date with 'main'


😍 this is amazing to read and clearly understand the scenarios.

derrickstolee · 2023-03-17T13:40:48Z

+
+    const clonedRepo = this.getRepo(user)
+
+    // TODO: figure out a better way to check whether the repo is empty


git rev-list --all -n 1 has no output?

derrickstolee · 2023-03-17T13:53:08Z

+e2e-test: build
+	@echo
+	@echo "======== Running end-to-end tests ========"
+	$(RM) -r $(TESTDIR)
+	@if [ -n "$(TEST_E2E_PERF)" ]; then \


Looks like make e2e-test and make e2e-test TEST_E2E_PERF=1 will run the tests and performance tests, respectively.

Could you update the README with these test instructions?

I also had to update my version of node.js to work with cucumber. It was really old, but maybe that should be listed as a dependency for developers?

derrickstolee · 2023-03-17T13:55:03Z

+on:
+  pull_request:
+  workflow_dispatch:


As you mentioned, "only on PRs" is quite nice because that way it's not on arbitrary pushes and not re-testing after a PR merged (and was subject to required builds).

The manual trigger of the workflow (with optional perf tests!) is excellent.

derrickstolee · 2023-03-17T14:46:40Z

+    Examples:
+      | repo                                  |
+      | https://github.com/git/git.git        | # takes ~2 minutes
+      | https://github.com/torvalds/linux.git | # takes ~30(!) minutes


It took me a while to realize that this literally supplies the input to the tests. Awesome.

Here are my test results for the data you have so far:

Clone execution time for https://github.com/git/git.git: 11.64416905105114s (bundle URI) vs. 22.594083791017532s (no bundle URI) Clone execution time for https://github.com/torvalds/linux.git: 168.95288394904136s (bundle URI) vs. 302.48135003399847s (no bundle URI)

Here are some others I tested:

Clone execution time for https://github.com/git-for-windows/git.git: 14.297109727025033s (bundle URI) vs. 29.787424183011055s (no bundle URI) Clone execution time for https://github.com/kubernetes/kubernetes.git: 37.84783263194561s (bundle URI) vs. 41.304983952999116s (no bundle URI) Clone execution time for https://github.com/homebrew/homebrew-core: 31.01036406993866s (bundle URI) vs. 357.1608746279478s (no bundle URI)

If you want to include any of them, here is my example list (minus git-for-windows/git since it's not materially different from git/git):

Examples: | repo | | https://github.com/git/git.git | # takes ~2 minutes | https://github.com/kubernetes/kubernetes.git | # takes ~3 minutes | https://github.com/homebrew/homebrew-core | # takes ~15(!) minutes | https://github.com/torvalds/linux.git | # takes ~30(!) minutes

Thanks, I'll add those cases!

Seeing your results, my timing comments could be misleading - there will be a ton of variability in execution time depending on internet speed (primarily), system resources, etc. I'll remove them and add a disclaimer in README.md along the lines of "The performance tests clone very large repos and can take anywhere from ~30 minutes to multiple hours to run, depending on internet connectivity and other system resources".

On that note, I'll also add an @online tag and corresponding "offline" cucumber.js profile to allow people to run tests without internet.

ldennington

Great stuff! ✨ Thanks for laying this great foundation.

ldennington · 2023-03-17T15:36:49Z

@@ -0,0 +1,7 @@
+import { defineParameterType } from '@cucumber/cucumber'
+
+defineParameterType({


You've found a lot of cool ways to integrate useful Cucumber features - it's been really helpful for ramping!

derrickstolee · 2023-03-17T19:11:51Z

I played around with extending your tests, so I could get a feel of how it works. The result is de09244, which let me repeat the performance tests, but used a second machine with existing bundles instead of creating new ones from scratch. (Perhaps these should be the same, with the environment variable switching modes at runtime instead of failing without it.)

Add a test and its corresponding step definitions verifying that a remote repository initialized in a bundle server can be cloned with '--bundle-uri' to download bundles. Helped-by: Derrick Stolee <derrickstolee@github.com> Signed-off-by: Victoria Dye <vdye@github.com>

Add scenarios to the end-to-end tests involving updates to the remote and/or bundle server. As part of this update, add a custom Cucumber expression for matching "are/are not" to a boolean. Helped-by: Derrick Stolee <derrickstolee@github.com> Signed-off-by: Victoria Dye <vdye@github.com>

Add a test that captures timing measurements (in 'console.log()' printouts) for cloning various large repositories. The tests are divided into two example sets: one for tests that run on modern hardware with a good internet connection for a *total* of less than 15 minutes (currently take ~7 minutes), and the other for tests that take longer than that. The latter set is tagged with '@slow' and the 'default' profile is updated to skip them. Add an 'all' profile that runs these slow performance tests alongside the "regular" tests. Helped-by: Derrick Stolee <derrickstolee@github.com> Signed-off-by: Victoria Dye <vdye@github.com>

@online

Tag tests that require internet access with '@online' and add a Cucumber profile that skips those tests. This allows developers to easily run tests on the bundle server in an offline or bandwidth-limited environment. Signed-off-by: Victoria Dye <vdye@github.com>

Update the workflows that build the project to use the fixed version of Go tied to the project in 'go.mod'. The 'go.mod' can be used as the sole source of truth on the supported version of Go in the repository, making it easier to upgrade and use new features when needed. Signed-off-by: Victoria Dye <vdye@github.com>

Add a target to the Makefile that installs the end-to-end test NPM dependencies then runs the tests. By default, this will only run tests matching the "default" profile of the tests (excluding performance tests). The 'E2E_FLAGS' variable can be used to specify the profile(s) used by the tests (e.g., '--all' to include the tests excluded by default). Add documentation on how to run the end-to-end tests to 'README.md'. Signed-off-by: Victoria Dye <vdye@github.com>

Add a dedicated 'test.yml' workflow that executes the end-to-end tests. Unlike CI, these tests are only run automatically for pull requests (although they can be manually triggered via workflow dispatch). By default, the tests will run with the "default" profile; this can be changed via an input variable in the workflow dispatch. Because the end-to-end tests deal with external dependencies (namely, Git), add an 'install-dependencies.sh' script to install and configure the dependencies on a GitHub Actions runner. The goal of this workflow is for it to eventually be used to run multiple types of long-running tests, such as integration tests or installer validation tests. Signed-off-by: Victoria Dye <vdye@github.com>

vdye · 2023-03-17T20:35:11Z

I ended up making more substantial changes than I initially intended. 😅 Beyond the suggestions in @derrickstolee's review:

changed setup-go to use version in go.mod rather than the stable.
split the performance test into @slow and not-@slow, adding the non-@slow tests to the default end-to-end execution.
rather than create a new npm script per profile, pass the profiles as command line args in run-e2e-tests.sh.
related to ^, change TEST_E2E_PERF to the more generic E2E_FLAGS, with documentation for the valid options.

If anyone's interested, here's the range-diff!

Range diff

1:  a817a95 ! 1:  e8a38b0 e2e: add initial bundle server clone test
    @@ Commit message
         repository initialized in a bundle server can be cloned with '--bundle-uri'
         to download bundles.
     
    +    Helped-by: Derrick Stolee <derrickstolee@github.com>
         Signed-off-by: Victoria Dye <vdye@github.com>
     
      ## .gitignore ##
    @@ test/e2e/features/step_definitions/repository.ts (new)
     +  const actualURI = result.stdout.toString().trim()
     +  assert.strictEqual(actualURI, this.bundleServer.bundleUri())
     +
    -+  result = clonedRepo.runGit("for-each-ref", "--format=%(refname)")
    ++  result = clonedRepo.runGit("for-each-ref", "--format=%(refname)", "refs/bundles/*")
     +  utils.assertStatus(0, result, "git for-each-ref failed")
     +
    -+  const bundleRefRegex = new RegExp('^refs/bundles/.*$')
    -+  const bundleRefs = result.stdout.toString().split("\n").filter(function (line) {
    -+    return bundleRefRegex.test(line)
    ++  const bundleRefs = result.stdout.toString().split("\n").filter(function(line) {
    ++    return line.trim() != ""
     +  })
     +  assert.strict(bundleRefs.length > 0, "No bundle refs found in the repo")
     +})
2:  99d8db1 ! 2:  368fa56 e2e: add tests updating the remote & bundle server
    @@ Commit message
         bundle server. As part of this update, add a custom Cucumber expression for
         matching "are/are not" to a boolean.
     
    +    Helped-by: Derrick Stolee <derrickstolee@github.com>
         Signed-off-by: Victoria Dye <vdye@github.com>
     
      ## test/e2e/cucumber.js ##
    @@ test/e2e/features/support/world.ts: export class BundleServerWorld extends World
     +
     +    const clonedRepo = this.getRepo(user)
     +
    -+    // TODO: figure out a better way to check whether the repo is empty
    -+    if (clonedRepo.runGit("show").status != 0) {
    ++    const result = clonedRepo.runGit("rev-list", "--all", "-n", "1")
    ++    if (result.stdout.toString().trim() == "") {
     +      // Repo is empty, so make sure we're on the right branch
     +      utils.assertStatus(0, clonedRepo.runGit("branch", "-m", branch))
     +    } else {
3:  40dd79c ! 3:  c9ee08f e2e: add performance test
    @@ Commit message
         e2e: add performance test
     
         Add a test that captures timing measurements (in 'console.log()' printouts)
    -    for cloning various large repositories. Because these tests can be
    -    exceptionally slow, tag them with '@slow' and make sure they are not run in
    -    the 'default' profile. Add an 'all' profile that runs the performance tests
    -    alongside the "regular" tests.
    +    for cloning various large repositories. The tests are divided into two
    +    example sets: one for tests that run on modern hardware with a good internet
    +    connection for a *total* of less than 15 minutes (currently take ~7
    +    minutes), and the other for tests that take longer than that. The latter set
    +    is tagged with '@slow' and the 'default' profile is updated to skip them.
    +    Add an 'all' profile that runs these slow performance tests alongside the
    +    "regular" tests.
     
    +    Helped-by: Derrick Stolee <derrickstolee@github.com>
         Signed-off-by: Victoria Dye <vdye@github.com>
     
      ## test/e2e/cucumber.js ##
    @@ test/e2e/features/performance.feature (new)
     +  Background: The bundle web server is running
     +    Given the bundle web server was started at port 8080
     +
    -+  @slow
     +  Scenario Outline: Comparing clone performance
     +    Given a remote repository '<repo>'
     +    Given the bundle server has been initialized with the remote repo
    @@ test/e2e/features/performance.feature (new)
     +    Then I compare the clone execution times
     +
     +    Examples:
    -+      | repo                                  |
    -+      | https://github.com/git/git.git        | # takes ~2 minutes
    -+      | https://github.com/torvalds/linux.git | # takes ~30(!) minutes
    ++      | repo                                         |
    ++      | https://github.com/git/git.git               |
    ++      | https://github.com/kubernetes/kubernetes.git |
    ++
    ++    @slow
    ++    Examples:
    ++      | repo                                      |
    ++      | https://github.com/homebrew/homebrew-core |
    ++      | https://github.com/torvalds/linux.git     |
     
      ## test/e2e/features/step_definitions/repository.ts ##
     @@ test/e2e/features/step_definitions/repository.ts: When('I clone from the remote repo with a bundle URI', async function (this: Bun
    @@ test/e2e/features/step_definitions/repository.ts: Then('my repo\'s bundles {bool
     +  utils.assertStatus(0, myClone.cloneResult)
     +  utils.assertStatus(0, otherClone.cloneResult)
     +
    -+  console.log(`\nClone execution time for ${this.remote!.remoteUri}: ${myClone.cloneTimeMs / 1000}s (bundle URI) vs. ${otherClone.cloneTimeMs / 1000}s (no bundle URI)`)
    ++  console.log(`\nClone execution time for ${this.remote!.remoteUri}: ${(myClone.cloneTimeMs / 1000).toFixed(2)}s (bundle URI) vs. ${(otherClone.cloneTimeMs / 1000).toFixed(2)}s (no bundle URI)`)
     +})
    -
    - ## test/e2e/package.json ##
    -@@
    -   "description": "",
    -   "main": "index.js",
    -   "scripts": {
    --    "test": "cucumber-js"
    -+    "test": "cucumber-js",
    -+    "test-all": "cucumber-js -p all"
    -   },
    -   "license": "MIT",
    -   "devDependencies": {
4:  3d477c9 < -:  ------- workflows: update 'go-version' to 'stable'
-:  ------- > 4:  9c7860f e2e: add profile for running offline tests
-:  ------- > 5:  704f46f workflows: use go version specified in 'go.mod'
5:  cfcf0b0 ! 6:  a6ff664 Makefile: add 'e2e-test' target
    @@ Commit message
         Add a target to the Makefile that installs the end-to-end test NPM
         dependencies then runs the tests. By default, this will only run tests
         matching the "default" profile of the tests (excluding performance tests).
    -    The 'TEST_E2E_PERF' flag, if it has a value, will instead run the tests with
    -    the 'all' profile (including the performance tests).
    +    The 'E2E_FLAGS' variable can be used to specify the profile(s) used by the
    +    tests (e.g., '--all' to include the tests excluded by default).
    +
    +    Add documentation on how to run the end-to-end tests to 'README.md'.
     
         Signed-off-by: Victoria Dye <vdye@github.com>
     
    @@ Makefile: PACKAGE_ARCH := $(GOARCH)
      APPLE_APP_IDENTITY =
      APPLE_INST_IDENTITY =
      APPLE_KEYCHAIN_PROFILE =
    -+TEST_E2E_PERF=
    ++E2E_FLAGS=
      
      # Build targets
      .PHONY: build
    @@ Makefile: doc:
     +	@echo
     +	@echo "======== Running end-to-end tests ========"
     +	$(RM) -r $(TESTDIR)
    -+	@if [ -n "$(TEST_E2E_PERF)" ]; then \
    -+		./scripts/run-e2e-tests.sh --all; \
    -+	else \
    -+		./scripts/run-e2e-tests.sh; \
    -+	fi
    ++	@scripts/run-e2e-tests.sh $(E2E_FLAGS)
     +
      # Installation targets
      .PHONY: install
    @@ Makefile: clean:
      	$(RM) -r $(DOCDIR)
     +	$(RM) -r $(TESTDIR)
     
    + ## README.md ##
    +@@ README.md: $ go build -o bin/ ./...
    + 
    + ### Testing and Linting
    + 
    +-To run the project's unit tests, navigate to the repository root directory and
    +-run `go test -v ./...`.
    ++Unless otherwise specified, run commands from the repository root.
    + 
    +-To run the project's linter, navigate to the repository root directory and run
    +-`go vet ./...`.
    ++#### Unit tests
    ++
    ++```
    ++go test -v ./...
    ++```
    ++
    ++#### Linter
    ++
    ++```
    ++go vet ./...
    ++```
    ++
    ++#### End-to-end tests
    ++
    ++In order to run these tests, you need to have a recent version of
    ++[Node.js](https://nodejs.org) (current LTS version is a pretty safe bet) and NPM
    ++installed.
    ++
    ++For the standard set of tests (i.e., excluding exceptionally slow tests), run:
    ++
    ++```
    ++make e2e-test
    ++```
    ++
    ++To configure the test execution and filtering, set the `E2E_FLAGS` build
    ++variable. The available options are:
    ++
    ++* `--offline`: run all tests except those that require internet access.
    ++* `--all`: run all tests, including slow performance tests.
    ++
    ++The above modes are mutually exclusive; if multiple are specified, only the last
    ++will be used. For example, `E2E_FLAGS="--offline --all"` is equivalent to
    ++`E2E_FLAGS="--all"`.
    ++
    ++:warning: The performance tests that are excluded by default clone very large
    ++repos from the internet and can take anywhere from ~30 minutes to multiple hours
    ++to run, depending on internet connectivity and other system resources.
    + 
    + ## License
    + 
    +
      ## scripts/run-e2e-tests.sh (new) ##
     @@
     +#!/bin/bash
    @@ scripts/run-e2e-tests.sh (new)
     +TESTDIR="$THISDIR/../test/e2e"
     +
     +# Defaults
    -+NPM_TARGET="test"
    ++ARGS=()
     +
     +# Parse script arguments
     +for i in "$@"
     +do
     +case "$i" in
    ++	--offline)
    ++	ARGS+=("-p" "offline")
    ++	shift # past argument
    ++	;;
     +	--all)
    -+	NPM_TARGET="test-all"
    ++	ARGS+=("-p" "all")
     +	shift # past argument
     +	;;
     +	*)
    @@ scripts/run-e2e-tests.sh (new)
     +cd "$TESTDIR"
     +
     +npm install
    -+npm run $NPM_TARGET
    ++npm run test -- ${ARGS:+${ARGS[*]}}
6:  617da16 ! 7:  f0b17de workflows: add a workflow to run end-to-end tests
    @@ Commit message
         Add a dedicated 'test.yml' workflow that executes the end-to-end tests.
         Unlike CI, these tests are only run automatically for pull requests
         (although they can be manually triggered via workflow dispatch). By default,
    -    the tests will run with the "default" (no perf test) profile; this can be
    -    changed via an input variable in the workflow dispatch.
    +    the tests will run with the "default" profile; this can be changed via an
    +    input variable in the workflow dispatch.
     
         Because the end-to-end tests deal with external dependencies (namely, Git),
         add an 'install-dependencies.sh' script to install and configure the
    @@ .github/workflows/test.yml (new)
     +  pull_request:
     +  workflow_dispatch:
     +    inputs:
    -+      test-perf:
    ++      run-all:
     +        type: boolean
    -+        description: 'Whether to run the (slow) end-to-end perf tests'
    ++        description: 'Include tests that are excluded by default due to slowness'
     +        default: false
     +
     +jobs:
    @@ .github/workflows/test.yml (new)
     +    runs-on: ubuntu-latest
     +
     +    steps:
    ++      - name: Clone repository
    ++        uses: actions/checkout@v3
    ++
     +      - name: Setup Go
     +        uses: actions/setup-go@v3
     +        with:
    -+          go-version: 'stable'
    ++          go-version-file: 'go.mod'
     +
     +      - name: Setup Node.js
     +        uses: actions/setup-node@v3
     +        with:
     +          node-version: 18
     +
    -+      - name: Clone repository
    -+        uses: actions/checkout@v3
    -+
     +      - name: Install system dependencies for end-to-end tests
     +        run: build/ci/install-dependencies.sh
     +        shell: bash
     +
     +      - name: Enable perf tests
    -+        if: ${{ github.event.inputs.test-perf }}
    -+        run: echo "TEST_E2E_PERF=1" >> $GITHUB_ENV
    ++        if: ${{ github.event.inputs.run-all }}
    ++        run: echo "E2E_FLAGS='--all'" >> $GITHUB_ENV
     +
     +      - name: Run end-to-end tests
     +        env:
    -+          TEST_E2E_PERF: ${{matrix.jobs.goarch}}
    -+        run: make e2e-test TEST_E2E_PERF="$TEST_E2E_PERF"
    ++          E2E_FLAGS: ${{env.E2E_FLAGS}}
    ++        run: make e2e-test E2E_FLAGS="$E2E_FLAGS"
     
      ## build/ci/install-dependencies.sh (new) ##
     @@

derrickstolee

I like the new E2E_FLAGS interface and doc updates, in particular.

vdye force-pushed the vdye/e2e-test-ts branch from 975dc90 to c77f762 Compare March 15, 2023 18:12

vdye changed the base branch from main to vdye/misc-bugfixes March 15, 2023 18:13

vdye force-pushed the vdye/e2e-test-ts branch from 862cb96 to bbd993d Compare March 15, 2023 19:41

Base automatically changed from vdye/misc-bugfixes to main March 15, 2023 20:19

vdye added 2 commits March 15, 2023 15:35

vdye force-pushed the vdye/e2e-test-ts branch from bbd993d to 617da16 Compare March 16, 2023 16:30

vdye self-assigned this Mar 16, 2023

vdye added this to the v1.0 milestone Mar 16, 2023

vdye marked this pull request as ready for review March 16, 2023 17:27

vdye requested review from derrickstolee and ldennington as code owners March 16, 2023 17:27

vdye changed the title ~~WIP: end-to-end tests~~ Add end-to-end tests Mar 16, 2023

derrickstolee approved these changes Mar 17, 2023

View reviewed changes

derrickstolee reviewed Mar 17, 2023

View reviewed changes

Comment thread test/e2e/features/step_definitions/repository.ts Outdated

ldennington approved these changes Mar 17, 2023

View reviewed changes

vdye added 7 commits March 17, 2023 13:25

vdye force-pushed the vdye/e2e-test-ts branch from 617da16 to f0b17de Compare March 17, 2023 20:27

derrickstolee approved these changes Mar 20, 2023

View reviewed changes

vdye merged commit 7ea9882 into main Mar 20, 2023

vdye deleted the vdye/e2e-test-ts branch March 20, 2023 15:51

ldennington mentioned this pull request Mar 30, 2023

Add integration tests #39

Merged


		const clonedRepo = this.getRepo(user)

		// TODO: figure out a better way to check whether the repo is empty

		@@ -0,0 +1,7 @@
		import { defineParameterType } from '@cucumber/cucumber'

		defineParameterType({

Sunbelt Computer Software

PL/B Language Development and Support

Conversation

vdye commented Mar 10, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Performance results

Uh oh!

derrickstolee left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

ldennington left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

derrickstolee commented Mar 17, 2023

Uh oh!

vdye commented Mar 17, 2023

Uh oh!

derrickstolee left a comment

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

vdye commented Mar 10, 2023 •

edited

Loading