feat(test): run checker and interactive Test entirely in the browser

.env.example

@@ -73,16 +73,8 @@ SANDBOX_PIDS_LIMIT=64
 # Worker
 PORT=8080
 WORKER_CONCURRENCY=4
-# WORKER_MODE: "all" (default), "judge" (sandbox only), "platform" (lifecycle/plagiarism only),
-# "test" (checker and interactive Test on the test-judge queue; needs both WASM-OJ dirs below)
+# WORKER_MODE: "all" (default), "judge" (sandbox only), "platform" (lifecycle/plagiarism only)
 WORKER_MODE=all
-# Checker and interactive Test through the WASM-OJ test judge. "all" serves the
-# test-judge queue only when both WASM-OJ dirs are set (docs/runbooks/getting-started.md).
-TEST_JUDGE_ENABLED=false
-# WASM_OJ_RUNTIME_DIR=/home/you/src/wasm-oj-forge/crates/runtime-core/target/release
-# WASM_OJ_TOOLCHAIN_DIR=/home/you/.cache/nojv-wasm-oj-toolchains
-# WASM_OJ_CACHE_DIR=/tmp/wasm-oj
-# TEST_JUDGE_SLOTS=2
 # Minutes a submission may stay pending/running before the platform sweeper marks
 # it system_error and terminates its judge. Range 10–1440; defaults to 10.
 SUBMISSION_PENDING_TIMEOUT_MINUTES=10

.github/renovate.json

@@ -95,12 +95,6 @@
       "matchDepTypes": ["packageManager", "engines"],
       "enabled": false
     },
-    {
-      "description": "WASM-OJ packages are pinned to the forge release in the worker image and checked by the pin test; they move by hand with a forge upgrade",
-      "matchManagers": ["npm"],
-      "matchPackageNames": ["/^@wasm-oj\\//"],
-      "enabled": false
-    },
     {
       "matchManagers": ["github-actions"],
       "schedule": ["* 9-11 1-7 * 1"],
@@ -122,12 +116,6 @@
       "matchPackageNames": ["node"],
       "enabled": false
     },
-    {
-      "description": "The rust image only builds the pinned WASM-OJ runtime; each bump replaces an 82 MB worker layer, so it moves by hand with a forge upgrade",
-      "matchDatasources": ["docker"],
-      "matchPackageNames": ["rust"],
-      "enabled": false
-    },
     {
       "description": "alpine/k8s tracks the cluster's kubectl minor",
       "matchDatasources": ["docker"],

apps/web/messages/en.json

@@ -270,13 +270,8 @@
   "admin_announcementsUnpin": "Unpin",
   "admin_announcementsUnpublish": "Unpublish",
   "admin_cancel": "Cancel",
-  "admin_checkerHelpBody": "Your validator follows the DOMjudge output-validator interface and runs in its own isolated container — the student's program never sees the answer. It is invoked as `validator <input> <judge_answer> <feedback_dir>` with the student output on stdin. It renders an accept/reject verdict only (no partial scoring). Python: `judge_input`, `judge_answer`, `team_output` are pre-bound as strings, and helpers `accept(team_msg=\"\")`, `wrong(team_msg=\"\")`, `judge_log(msg)` write the right feedback files and exit. C++: read `argv[1]` / `argv[2]` and stdin yourself, write `teammessage.txt` into the `argv[3]` feedback dir, and exit 42 (accept) or 43 (wrong) — no judge header required. Feedback is shown to the student. Validator timeout is 30 seconds; a crash or hang becomes a system error and does not count against the student.",
+  "admin_checkerHelpBody": "Your validator follows the DOMjudge output-validator interface. On Submit it runs in the judge sandbox, apart from the student's program, which never sees the answer; on Test it runs in the student's browser and checks the samples. It is invoked as `validator <input> <judge_answer> <feedback_dir>` with the student output on stdin. It renders an accept/reject verdict only (no partial scoring). Python: `judge_input`, `judge_answer`, `team_output` are pre-bound as strings, and `accept(team_msg=\"\")` and `wrong(team_msg=\"\")` write the feedback and exit, and `judge_log(msg)` writes the staff-only judge message. C++: read `argv[1]` / `argv[2]` and stdin yourself, write `teammessage.txt` into the `argv[3]` feedback dir, and exit 42 (accept) or 43 (wrong) — no judge header required. Feedback is shown to the student. The validator gets the larger of 30 seconds and the time limit; a crash or hang becomes a system error and does not count against the student.",
   "admin_checkerHelpTitle": "How to write a checker",
-  "admin_checkSamples": "Check samples with the checker",
-  "admin_checkSamplesRejected": "The checker must accept each sample's output, because students' Test uses it as the answer. Fix the checker or the sample.",
-  "admin_checkSamplesSaveFirst": "Save your changes first to check samples against them.",
-  "admin_checkSamplesUnavailable": "The test judge is unavailable right now. Try again later.",
-  "admin_checkingSamples": "Checking samples…",
   "admin_compareCaseSensitive": "Case-sensitive comparison",
   "admin_compareFloatTolerance": "Float tolerance (ε)",
   "admin_compareFloatToleranceHint": "Leave empty for exact matching. When set, numeric tokens match within absolute OR relative error ε. For anything more, write a Checker.",
@@ -306,7 +301,6 @@
   "admin_envRemove": "Remove env var",
   "admin_envVars": "Environment variables",
   "admin_fileEditable": "Editable",
-  "admin_fileHidden": "Hidden",
   "admin_filePath": "Path",
   "admin_filePathPlaceholder": "main.c or include/header.h",
   "admin_fileReadonly": "Read-only",
@@ -317,16 +311,12 @@
   "admin_inputFormatTooltip": "Describe the input format. Supports Markdown.",
   "admin_interactionFormat": "Interaction notes",
   "admin_interactionFormatTooltip": "Describe what the interactor reads from its input file and how it talks to the student's program. Shown to students. Supports Markdown.",
-  "admin_interactorHelpBody": "Your interactor follows the DOMjudge interactive interface and runs in its own isolated container, wired to the student program by a byte proxy: student stdout flows to interactor stdin, interactor stdout flows to student stdin. It is invoked as `interactor <input> <judge_answer> <feedback_dir>`. Python: `judge_input` and `judge_answer` are pre-bound strings; `read()` reads one line from the student (auto-rejecting if the stream closes) and `write(msg)` sends one line back (already flushed). Helpers `accept(team_msg=\"\")`, `wrong(team_msg=\"\")`, `set_score(x)`, `judge_log(msg)` write the feedback files and exit 42/43. C++: read `argv[1]`/`argv[2]`, talk to the solution over `cin`/`cout` (flush before reading), write `score.txt` / `teammessage.txt` into `argv[3]`, and exit 42 (accept) or 43 (wrong) — no judge header required. The student and interactor share the problem's time limit.",
+  "admin_interactorHelpBody": "Your interactor follows the DOMjudge interactive interface. On Submit it runs in the judge sandbox, and on Test in the student's browser; either way student stdout flows to interactor stdin, and interactor stdout flows to student stdin. It is invoked as `interactor <input> <judge_answer> <feedback_dir>`; Test passes an empty `judge_answer`, so read the secret from `judge_input`. Python: `judge_input` and `judge_answer` are pre-bound strings; `read()` reads one line from the student (auto-rejecting if the stream closes) and `write(msg)` sends one line back (already flushed). `accept(team_msg=\"\")` and `wrong(team_msg=\"\")` write the feedback and exit 42/43, and `judge_log(msg)` writes the staff-only judge message. C++: read `argv[1]`/`argv[2]`, talk to the solution over `cin`/`cout` (flush before reading), write `teammessage.txt` into `argv[3]`, and exit 42 (accept) or 43 (wrong) — no judge header required. The student and interactor share the problem's time limit.",
   "admin_interactorHelpTitle": "How to write an interactor",
   "admin_interactorLanguage": "Interactor language",
   "admin_judgeChecker": "Checker script",
   "admin_judgeInteractive": "Interactive",
-  "admin_judgeProgramDiagnostics": "Build output",
-  "admin_judgeProgramTestFailed": "This judge program can't run in Test. Submissions are still judged normally.",
-  "admin_judgeProgramTestPending": "Preparing this judge program for Test…",
-  "admin_judgeProgramTestReady": "Test can run this judge program.",
-  "admin_judgeProgramTestUnavailable": "Couldn't check this judge program right now.",
+  "admin_judgeProgramReadableNote": "Students can read this program when they press Test.",
   "admin_judgeStandard": "Standard (stdin/stdout diff)",
   "admin_judgeType": "Judge type",
   "admin_judgeTypeHint": "How each testcase is evaluated.",
@@ -403,8 +393,6 @@
   "admin_tabBasicInfo": "Basic Info",
   "admin_tabReports": "Reports",
   "admin_tabJudge": "Judge Settings",
-  "admin_testJudgeDisabled": "Test judging isn't enabled on this server.",
-  "admin_testPythonInteractorUnsupported": "Test doesn't support Python interactors yet.",
   "admin_tabLocked": "Please complete and save basic info first",
   "admin_tabLockedFields": "Fill in and save these in Basic Info first: {fields}",
   "admin_tabOverview": "Overview",
@@ -472,7 +460,7 @@
   "admin_workspaceEmptyFullSource": "No starter template for {language}. Leaving this blank is fine for single-file problems — students will see an empty editor.",
   "admin_workspaceEmptyMultiFile": "No files for {language} yet. Multi-file problems need at least a main file named {filename}.",
   "admin_workspaceFilesHintFullSource": "Optionally provide starter code per language. Students see this template when they open the editor. Leave blank for an empty editor.",
-  "admin_workspaceFilesHintMultiFile": "Provide one editable main file per allowed language, plus any helper files (readonly or hidden).",
+  "admin_workspaceFilesHintMultiFile": "Provide one editable main file per allowed language, plus any readonly helper files.",
   "admin_workspaceFilesTitleFullSource": "Starter templates (optional)",
   "admin_workspaceFilesTitleMultiFile": "Workspace files",
   "admin_workspaceLanguageMissing": "Missing template",
@@ -1145,23 +1133,23 @@
   "editor_invalidSource": "Source code must be nonblank and fit the submission limits: 50,000 characters for a single file, or 500,000 per editable workspace file. Check file paths as well.",
   "editor_invalidRunCases": "Provide 1–10 testcases, with at most 200,000 characters in each input or expected output.",
   "editor_requestTooLarge": "The combined source and testcase request exceeds 2 MiB. Reduce its size and try again.",
-  "admin_workspaceHiddenTestNote": "Browser Test cannot access hidden files. Public code must compile and run without them. If a dependency is safe to share, mark it read-only; otherwise redesign the public test workspace. Hidden files remain server-only.",
   "editor_clientTestCustomImage": "This problem requires a custom container that cannot run in the browser. Ask the author to provide a browser-compatible Test workspace.",
   "editor_clientTestLanguage": "The browser runtime for this language is unavailable. Test runs entirely on your device.",
   "editor_testInteractiveLanguage": "Test can't run interactive problems in JavaScript or TypeScript yet. Switch to another language, or use Submit to have your code judged.",
-  "editor_testNoInteractiveSamples": "This problem has no sample with an interactor input, so there is nothing to test. Use Submit to have your code judged.",
-  "editor_testJudgeBusy": "Test is busy right now. Try again in a moment.",
-  "editor_testJudgeProgramBuildFailed": "This problem's checker or interactor can't run in Test. Submit still judges your code normally.",
-  "editor_testUnavailableForProblem": "Test isn't available for this problem right now. Submit still judges your code normally.",
   "editor_testUnsupportedProblemType": "This problem type doesn't support Test. Use Submit to have your code judged.",
-  "editor_testTooLarge": "This Test request is too large to send to the server. Reduce your program's size or output, or use Submit to have your code judged.",
-  "editor_judgingOnServer": "Judging on the server…",
+  "editor_checkerPreparing": "Preparing checker...",
+  "editor_checkerBuildFailed": "This problem's checker failed to build.",
+  "editor_checkerLoadFailed": "Couldn't load this problem's checker. Reload the page to try again.",
+  "editor_checkerUnavailable": "Couldn't load this problem's checker.",
+  "editor_interactorPreparing": "Preparing interactor...",
+  "editor_interactorBuildFailed": "This problem's interactor failed to build.",
+  "editor_interactorLoadFailed": "Couldn't load this problem's interactor. Reload the page to try again.",
+  "editor_interactorUnavailable": "Couldn't load this problem's interactor.",
   "editor_judgeSystemError": "The judge failed on this case. This is a problem on the judge's side, not with your program.",
-  "editor_interactiveTestNote": "Test runs your program against this problem's interactor on each sample's interactor input; the interactor judges your replies live. You can't add your own cases to an interactive problem.",
+  "editor_interactiveTestNote": "Test runs your program against this problem's interactor on each case's interactor input; the interactor judges your replies live. Cases you add are judged the same way.",
   "editor_checkerCasesNote": "This problem's checker judges the samples only. Cases you add or change just run without a verdict, so check their output yourself.",
   "editor_executed": "Executed",
   "editor_executedNote": "This case ran without being judged. Check its output yourself.",
-  "editor_judgedOnServer": "Judged on the server",
   "editor_judgeFeedback": "Feedback",
   "editor_transcript": "Transcript",
   "editor_transcriptFromInteractor": "From the interactor",
@@ -1596,7 +1584,7 @@
   "problemDetail_loadingSource": "Loading source code...",
   "problemDetail_memoryLimit": "Memory limit",
   "problemDetail_multiFileBadge": "Multi-file",
-  "problemDetail_multiFileHelp": "The problem ships multiple files — one or more you edit plus read-only helpers (and sometimes hidden files). You edit the designated files in the browser and submit them together.",
+  "problemDetail_multiFileHelp": "The problem ships multiple files — one or more you edit plus read-only helpers. You edit the designated files in the browser and submit them together.",
   "problemDetail_noSubmissions": "No submissions yet.",
   "problemDetail_output": "Output",
   "problemDetail_outputFormat": "Output Description",
@@ -1907,12 +1895,9 @@
   "virtualContest_youLabel": "You",
   "workspace_description": "Description",
   "workspace_descriptionPlaceholder": "Describe what this file is for…",
-  "workspace_fileHidden": "This file is hidden",
-  "workspace_fileHiddenNoDescription": "(no description provided)",
   "workspace_filesHeading": "Files",
   "workspace_mustHaveMainFile": "Each enabled language must have an editable file named {filename}",
   "workspace_visibilityEditable": "edit",
-  "workspace_visibilityHidden": "hidden",
   "workspace_visibilityReadonly": "read",
   "contestScoreboard_notStartedHint": "The scoreboard opens once the contest begins.",
   "contestDetail_settingsDeleteConfirmBody": "This will permanently remove the contest draft.",