Merge main into feat/io/view-togggle

Resolved conflicts:
- scrapers/atcoder.py: kept defensive if tests else '' checks
- scrapers/codechef.py: kept defensive if tests else '' checks
- tests/test_scrapers.py: kept comprehensive validation from main
- lua/cp/ui/views.lua: removed misplaced navigation code from loop

.pre-commit-config.yaml

@@ -2,7 +2,7 @@ minimum_pre_commit_version: '3.5.0'
 
 repos:
   - repo: https://github.com/JohnnyMorganz/StyLua
-    rev: v2.1.0
+    rev: v2.3.1
     hooks:
       - id: stylua-github
         name: stylua (Lua formatter)
@@ -10,7 +10,7 @@ repos:
         pass_filenames: true
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.6.9
+    rev: v0.14.3
     hooks:
       - id: ruff-format
         name: ruff (format)
@@ -30,7 +30,7 @@ repos:
         pass_filenames: false
 
   - repo: https://github.com/pre-commit/mirrors-prettier
-    rev: v3.1.0
+    rev: v4.0.0-alpha.8
     hooks:
       - id: prettier
         name: prettier (format markdown)

doc/cp.nvim.txt

@@ -34,15 +34,30 @@ COMMANDS                                                         *cp-commands*
                                     :CP codeforces 1933 --lang python
 <
         View Commands ~
-            :CP run [--debug] [n]
+            :CP run [all|n|n,m,...] [--debug]
                                 Run tests in I/O view (see |cp-io-view|).
                                 Lightweight split showing test verdicts.
-                                Without [n]: runs all tests, shows verdict summary
+
-                                With [n]: runs test n, shows detailed output
+                                Execution modes:
+                                • :CP run              Combined: single execution with all tests
+                                                       (auto-switches to individual when multiple samples)
+                                • :CP run all          Individual: N separate executions
+                                • :CP run n            Individual: run test n only
+                                • :CP run n,m,...      Individual: run specific tests (e.g. nth and mth)
+
                                 --debug: Use debug build (builds to build/<name>.dbg)
+
+                                Combined mode runs all test inputs in one execution (matching
+                                platform behavior for multi-test problems). When a problem has
+                                multiple independent sample test cases, :CP run auto-switches to
+                                individual mode to run each sample separately.
+
                                 Examples: >
-                                    :CP run              " All tests
+                                    :CP run              " Combined: all tests, one execution
-                                    :CP run --debug 2    " Test 2, debug build
+                                    :CP run all          " Individual: all tests, N executions
+                                    :CP run 2            " Individual: test 2 only
+                                    :CP run 1,3,5        " Individual: tests 1, 3, and 5
+                                    :CP run all --debug  " Individual with debug build
 <
             :CP panel [--debug] [n]
                                 Open full-screen test panel (see |cp-panel|).
@@ -536,10 +551,27 @@ Example: Setting up and solving AtCoder contest ABC324
 I/O VIEW                                                          *cp-io-view*
 
 The I/O view provides lightweight test feedback in persistent side splits.
-All test outputs are concatenated with verdict summaries at the bottom.
+Test outputs are concatenated with verdict summaries at the bottom.
 The |cp-panel| offers more fine-grained analysis with diff modes.
 
-Access the I/O view with :CP run [n]
+Execution Modes ~
+
+The I/O view supports two execution modes:
+
+Combined Mode (:CP run with single sample)
+    • Single execution with all test inputs concatenated
+    • Matches platform behavior (e.g. Codeforces multi-test format)
+    • Shows one verdict for the entire execution
+    • Input split: All test inputs concatenated
+    • Output split: Single program output + verdict
+    • Used when problem has one sample containing multiple test cases
+
+Individual Mode (:CP run all / :CP run n / :CP run n,m,...)
+    • Separate execution for each test case
+    • Per-test verdicts for debugging
+    • Input split: Selected test inputs concatenated
+    • Output split: All test outputs concatenated + per-test verdicts
+    • Auto-selected when problem has multiple independent samples
 
 Layout ~
 
@@ -561,7 +593,7 @@ The I/O view appears as 30% width splits on the right side: >
     └──────────────────────────┴─────────────────────────────────────────────┘
 <
 The output split shows:
-1. Concatenated test outputs (separated by blank lines)
+1. Program output (raw, preserving all formatting)
 2. Space-aligned verdict summary with:
    - Test number and status (AC/WA/TLE/MLE/RTE with color highlighting)
    - Runtime: actual/limit in milliseconds
@@ -570,8 +602,10 @@ The output split shows:
 
 Usage ~
 
-    :CP run             Run all tests
+    :CP run             Combined mode: all tests in one execution
-    :CP run 3           Run test 3 only
+    :CP run all         Individual mode: all tests separately
+    :CP run 3           Individual mode: test 3 only
+    :CP run 1,3,5       Individual mode: specific tests (1, 3, and 5)
 
 Navigation ~
 

lua/cp/commands/init.lua

@@ -17,8 +17,11 @@ local actions = constants.ACTIONS
 ---@field problem_id? string
 ---@field interactor_cmd? string
 ---@field test_index? integer
+---@field test_indices? integer[]
+---@field mode? string
 ---@field debug? boolean
 ---@field language? string
+---@field subcommand? string
 
 --- Turn raw args into normalized structure to later dispatch
 ---@param args string[] The raw command-line mode args
@@ -75,7 +78,7 @@ local function parse_command(args)
       return { type = 'action', action = 'edit', test_index = test_index }
     elseif first == 'run' or first == 'panel' then
       local debug = false
-      local test_index = nil
+      local test_indices = nil
       local mode = 'combined'
 
       if #args == 2 then
@@ -84,20 +87,39 @@ local function parse_command(args)
         elseif args[2] == 'all' then
           mode = 'individual'
         else
-          local idx = tonumber(args[2])
+          if args[2]:find(',') then
-          if not idx then
+            local indices = {}
-            return {
+            for num in args[2]:gmatch('[^,]+') do
-              type = 'error',
+              local idx = tonumber(num)
-              message = ("Invalid argument '%s': expected test number, 'all', or --debug"):format(
+              if not idx or idx < 1 or idx ~= math.floor(idx) then
-                args[2]
+                return {
-              ),
+                  type = 'error',
-            }
+                  message = ("Invalid test index '%s' in list"):format(num),
+                }
+              end
+              table.insert(indices, idx)
+            end
+            if #indices == 0 then
+              return { type = 'error', message = 'No valid test indices provided' }
+            end
+            test_indices = indices
+            mode = 'individual'
+          else
+            local idx = tonumber(args[2])
+            if not idx then
+              return {
+                type = 'error',
+                message = ("Invalid argument '%s': expected test number(s), 'all', or --debug"):format(
+                  args[2]
+                ),
+              }
+            end
+            if idx < 1 or idx ~= math.floor(idx) then
+              return { type = 'error', message = ("'%s' is not a valid test index"):format(idx) }
+            end
+            test_indices = { idx }
+            mode = 'individual'
           end
-          if idx < 1 or idx ~= math.floor(idx) then
-            return { type = 'error', message = ("'%s' is not a valid test index"):format(idx) }
-          end
-          test_index = idx
-          mode = 'individual'
         end
       elseif #args == 3 then
         if args[2] == 'all' then
@@ -109,6 +131,30 @@ local function parse_command(args)
             }
           end
           debug = true
+        elseif args[2]:find(',') then
+          local indices = {}
+          for num in args[2]:gmatch('[^,]+') do
+            local idx = tonumber(num)
+            if not idx or idx < 1 or idx ~= math.floor(idx) then
+              return {
+                type = 'error',
+                message = ("Invalid test index '%s' in list"):format(num),
+              }
+            end
+            table.insert(indices, idx)
+          end
+          if #indices == 0 then
+            return { type = 'error', message = 'No valid test indices provided' }
+          end
+          if args[3] ~= '--debug' then
+            return {
+              type = 'error',
+              message = ("Invalid argument '%s': expected --debug"):format(args[3]),
+            }
+          end
+          test_indices = indices
+          mode = 'individual'
+          debug = true
         else
           local idx = tonumber(args[2])
           if not idx then
@@ -126,21 +172,23 @@ local function parse_command(args)
               message = ("Invalid argument '%s': expected --debug"):format(args[3]),
             }
           end
-          test_index = idx
+          test_indices = { idx }
           mode = 'individual'
           debug = true
         end
       elseif #args > 3 then
         return {
           type = 'error',
-          message = 'Too many arguments. Usage: :CP ' .. first .. ' [all|test_num] [--debug]',
+          message = 'Too many arguments. Usage: :CP '
+            .. first
+            .. ' [all|test_num[,test_num...]] [--debug]',
         }
       end
 
       return {
         type = 'action',
         action = first,
-        test_index = test_index,
+        test_indices = test_indices,
         debug = debug,
         mode = mode,
       }
@@ -221,9 +269,12 @@ function M.handle_command(opts)
     if cmd.action == 'interact' then
       ui.toggle_interactive(cmd.interactor_cmd)
     elseif cmd.action == 'run' then
-      ui.run_io_view(cmd.test_index, cmd.debug, cmd.mode)
+      ui.run_io_view(cmd.test_indices, cmd.debug, cmd.mode)
     elseif cmd.action == 'panel' then
-      ui.toggle_panel({ debug = cmd.debug, test_index = cmd.test_index })
+      ui.toggle_panel({
+        debug = cmd.debug,
+        test_index = cmd.test_indices and cmd.test_indices[1] or nil,
+      })
     elseif cmd.action == 'next' then
       setup.navigate_problem(1, cmd.language)
     elseif cmd.action == 'prev' then

lua/cp/ui/edit.lua

@@ -274,10 +274,25 @@ local function save_all_tests()
   local is_multi_test = contest_data.problems[contest_data.index_map[problem_id]].multi_test
     or false
 
+  -- Generate combined test from individual test cases
+  local combined_input = table.concat(
+    vim.tbl_map(function(tc)
+      return tc.input
+    end, edit_state.test_cases),
+    '\n'
+  )
+  local combined_expected = table.concat(
+    vim.tbl_map(function(tc)
+      return tc.expected
+    end, edit_state.test_cases),
+    '\n'
+  )
+
   cache.set_test_cases(
     platform,
     contest_id,
     problem_id,
+    { input = combined_input, expected = combined_expected },
     edit_state.test_cases,
     edit_state.constraints and edit_state.constraints.timeout_ms or 0,
     edit_state.constraints and edit_state.constraints.memory_mb or 0,

lua/cp/ui/views.lua

@@ -405,7 +405,7 @@ function M.ensure_io_view()
   end
 end
 
-function M.run_io_view(test_index, debug, mode)
+function M.run_io_view(test_indices_arg, debug, mode)
   mode = mode or 'combined'
 
   local platform, contest_id, problem_id =
@@ -425,6 +425,13 @@ function M.run_io_view(test_index, debug, mode)
     return
   end
 
+  if mode == 'combined' then
+    local test_cases = cache.get_test_cases(platform, contest_id, problem_id)
+    if test_cases and #test_cases > 1 then
+      mode = 'individual'
+    end
+  end
+
   M.ensure_io_view()
 
   local run = require('cp.runner.run')
@@ -447,19 +454,21 @@ function M.run_io_view(test_index, debug, mode)
   if mode == 'individual' then
     local test_state = run.get_panel_state()
 
-    if test_index then
+    if test_indices_arg then
-      if test_index < 1 or test_index > #test_state.test_cases then
+      for _, idx in ipairs(test_indices_arg) do
-        logger.log(
+        if idx < 1 or idx > #test_state.test_cases then
-          string.format(
+          logger.log(
-            'Test %d does not exist (only %d tests available)',
+            string.format(
-            test_index,
+              'Test %d does not exist (only %d tests available)',
-            #test_state.test_cases
+              idx,
-          ),
+              #test_state.test_cases
-          vim.log.levels.WARN
+            ),
-        )
+            vim.log.levels.WARN
-        return
+          )
+          return
+        end
       end
-      test_indices = { test_index }
+      test_indices = test_indices_arg
     else
       for i = 1, #test_state.test_cases do
         test_indices[i] = i
@@ -512,6 +521,11 @@ function M.run_io_view(test_index, debug, mode)
   if mode == 'combined' then
     local combined = cache.get_combined_test(platform, contest_id, problem_id)
 
+    if not combined then
+      logger.log('No combined test found', vim.log.levels.ERROR)
+      return
+    end
+
     run.load_test_cases()
 
     local result = run.run_combined_test(debug)

scrapers/atcoder.py

@@ -71,7 +71,7 @@ def _retry_after_requests(details):
     on_backoff=_retry_after_requests,
 )
 def _fetch(url: str) -> str:
-    r = _session.get(url, headers=HEADERS, timeout=TIMEOUT_SECONDS)
+    r = _session.get(url, headers=HEADERS, timeout=TIMEOUT_SECONDS, verify=False)
     if r.status_code in RETRY_STATUS:
         raise requests.HTTPError(response=r)
     r.raise_for_status()
@@ -243,7 +243,8 @@ def _to_problem_summaries(rows: list[dict[str, str]]) -> list[ProblemSummary]:
 
 async def _fetch_all_contests_async() -> list[ContestSummary]:
     async with httpx.AsyncClient(
-        limits=httpx.Limits(max_connections=100, max_keepalive_connections=100)
+        limits=httpx.Limits(max_connections=100, max_keepalive_connections=100),
+        verify=False,
     ) as client:
         first_html = await _get_async(client, ARCHIVE_URL)
         last = _parse_last_page(first_html)

tests/test_scrapers.py

@@ -58,18 +58,38 @@ def test_scraper_offline_fixture_matrix(run_scraper_offline, scraper, mode):
         assert len(objs) >= 1, "No test objects returned"
         validated_any = False
         for obj in objs:
-            assert "problem_id" in obj, "Missing problem_id"
+            if "success" in obj and "tests" in obj and "problem_id" in obj:
-            assert obj["problem_id"] != "", "Empty problem_id"
+                tr = TestsResult.model_validate(obj)
-            assert "combined" in obj, "Missing combined field"
+                assert tr.problem_id != ""
-            assert isinstance(obj["combined"], dict), "combined must be a dict"
+                assert isinstance(tr.tests, list)
-            assert "input" in obj["combined"], "Missing combined.input"
+                assert hasattr(tr, "combined"), "Missing combined field"
-            assert "expected" in obj["combined"], "Missing combined.expected"
+                assert tr.combined is not None, "combined field is None"
-            assert "tests" in obj, "Missing tests field"
+                assert hasattr(tr.combined, "input"), "combined missing input"
-            assert isinstance(obj["tests"], list), "tests must be a list"
+                assert hasattr(tr.combined, "expected"), "combined missing expected"
-            assert "timeout_ms" in obj, "Missing timeout_ms"
+                assert isinstance(tr.combined.input, str), "combined.input not string"
-            assert "memory_mb" in obj, "Missing memory_mb"
+                assert isinstance(tr.combined.expected, str), (
-            assert "interactive" in obj, "Missing interactive"
+                    "combined.expected not string"
-            assert "multi_test" in obj, "Missing multi_test field"
+                )
-            assert isinstance(obj["multi_test"], bool), "multi_test must be bool"
+                assert hasattr(tr, "multi_test"), "Missing multi_test field"
-            validated_any = True
+                assert isinstance(tr.multi_test, bool), "multi_test not boolean"
+                validated_any = True
+            else:
+                assert "problem_id" in obj
+                assert "tests" in obj and isinstance(obj["tests"], list)
+                assert (
+                    "timeout_ms" in obj and "memory_mb" in obj and "interactive" in obj
+                )
+                assert "combined" in obj, "Missing combined field in raw JSON"
+                assert isinstance(obj["combined"], dict), "combined not a dict"
+                assert "input" in obj["combined"], "combined missing input key"
+                assert "expected" in obj["combined"], "combined missing expected key"
+                assert isinstance(obj["combined"]["input"], str), (
+                    "combined.input not string"
+                )
+                assert isinstance(obj["combined"]["expected"], str), (
+                    "combined.expected not string"
+                )
+                assert "multi_test" in obj, "Missing multi_test field in raw JSON"
+                assert isinstance(obj["multi_test"], bool), "multi_test not boolean"
+                validated_any = True
         assert validated_any, "No valid tests payloads validated"