File size: 19,437 Bytes
a7caaff
85b7ac8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
c8f3b98
85b7ac8
 
 
 
4de7d31
85b7ac8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4b07aaf
 
 
 
 
 
 
85b7ac8
2794920
85b7ac8
4b07aaf
 
c8f3b98
2794920
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
85b7ac8
 
 
 
c8f3b98
85b7ac8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2794920
 
 
 
 
 
 
85b7ac8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4b07aaf
 
 
 
85b7ac8
 
 
 
 
 
 
 
 
 
 
 
c8f3b98
 
 
 
 
 
 
 
 
 
 
 
 
 
85b7ac8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4b07aaf
85b7ac8
 
 
 
 
2794920
 
 
 
 
 
 
 
 
 
 
 
 
85b7ac8
2794920
85b7ac8
 
 
 
 
4b07aaf
 
c8f3b98
 
 
 
 
85b7ac8
 
 
 
c8f3b98
 
 
 
 
 
 
 
 
 
 
 
 
 
85b7ac8
c8f3b98
85b7ac8
c8f3b98
 
 
 
 
 
 
85b7ac8
c8f3b98
 
 
 
 
85b7ac8
c8f3b98
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
85b7ac8
c8f3b98
85b7ac8
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
"""Core environment logic."""

import copy
import random
from typing import Any, Dict, List, Optional, Tuple

from server.models import (
    Action,
    ActionType,
    ErrorInfo,
    ErrorPhase,
    FileContent,
    FileType,
    Observation,
    TaskDifficulty,
)
from server.simulators.docker_simulator import DockerSimulator
from server.simulators.k8s_simulator import KubernetesSimulator
from server.simulators.workflow_simulator import WorkflowSimulator
from server.tasks.task_registry import TASK_REGISTRY, get_task


class CloudNativeDebugEnvironment:
    MAX_STEPS = 10
    MAX_HINTS = 3

    def _resolve_task_id(self, task_id: Optional[object]) -> Optional[str]:
        """Resolve task identifiers from either canonical string IDs or numeric indices.

        Supports:
        - canonical ids (e.g. "dockerfile_syntax")
        - numeric values (1-based preferred, 0-based tolerated)
        - numeric strings ("1", "0")
        """
        if task_id is None:
            return None

        keys = list(TASK_REGISTRY.keys())

        if isinstance(task_id, int):
            if 1 <= task_id <= len(keys):
                return keys[task_id - 1]
            if 0 <= task_id < len(keys):
                return keys[task_id]
            return str(task_id)

        if isinstance(task_id, str):
            raw = task_id.strip()
            if raw in TASK_REGISTRY:
                return raw
            if raw.isdigit():
                numeric = int(raw)
                if 1 <= numeric <= len(keys):
                    return keys[numeric - 1]
                if 0 <= numeric < len(keys):
                    return keys[numeric]
                return raw
            lowered = raw.lower()
            for key in keys:
                if key.lower() == lowered:
                    return key
            return raw

        return str(task_id)

    def _find_workflow_file(self) -> Optional[FileContent]:
        """Return the first workflow file found in current_files."""
        for path, fc in self.current_files.items():
            if path.startswith(".github/workflows/") and path.endswith(".yml"):
                return fc
        return None

    def _validation_snapshot(self) -> Dict[str, bool]:
        """Return a detailed snapshot of all 7 simulator checks."""
        docker_result = self.docker_sim.validate(self.current_files.get("Dockerfile"), self.current_files)
        workflow_file = self._find_workflow_file()
        workflow_result = self.workflow_sim.validate(workflow_file, self.current_files)
        k8s_result = self.k8s_sim.validate(self.current_files)

        has_docker = "Dockerfile" in self.current_files
        has_workflow = workflow_file is not None
        has_k8s = any(fc.file_type == FileType.KUBERNETES for fc in self.current_files.values())

        snapshot: Dict[str, bool] = {}
        if has_docker:
            snapshot["docker_build_valid"] = bool(docker_result.get("build_success", False))
            snapshot["docker_run_valid"] = bool(docker_result.get("run_success", False))
        if has_workflow:
            snapshot["workflow_parse_valid"] = bool(workflow_result.get("parse_success", False))
            snapshot["workflow_exec_valid"] = bool(workflow_result.get("execution_success", False))
        if has_k8s:
            snapshot["k8s_valid"] = bool(k8s_result.get("valid", True))
            snapshot["k8s_pod_running"] = k8s_result.get("pod_status", "N/A") == "Running"
            svc = k8s_result.get("service_status", "N/A")
            snapshot["k8s_service_active"] = "active" in svc.lower() or svc == "N/A"

        return snapshot

    def __init__(self):
        self.docker_sim = DockerSimulator()
        self.workflow_sim = WorkflowSimulator()
        self.k8s_sim = KubernetesSimulator()

        self.current_task_id: Optional[str] = None
        self.current_scenario_id: Optional[str] = None
        self.current_difficulty: Optional[TaskDifficulty] = None
        self.current_task = None

        self.original_files: Dict[str, FileContent] = {}
        self.current_files: Dict[str, FileContent] = {}
        self.expected_fixes: List[Dict[str, Any]] = []

        self.current_error: Optional[ErrorInfo] = None
        self.issues_total = 0
        self.issues_fixed = 0

        self.step_count = 0
        self.episode_reward = 0.0
        self.done = False
        self.hints_used = 0

        self.trajectory: List[Dict[str, Any]] = []
        self.last_action_success: Optional[bool] = None
        self.last_action_feedback: Optional[str] = None

    def reset(self, task_id: Optional[str] = None, scenario_id: Optional[str] = None, seed: Optional[int] = None):
        if seed is not None:
            random.seed(seed)

        task_id = self._resolve_task_id(task_id)

        if task_id is None:
            task_id = random.choice(list(TASK_REGISTRY.keys()))
        if task_id not in TASK_REGISTRY:
            raise ValueError(f"Unknown task: {task_id}")

        self.current_task_id = task_id
        self.current_task = get_task(task_id)
        self.current_difficulty = self.current_task.DIFFICULTY

        scenario = self.current_task.load_scenario(scenario_id)
        self.current_scenario_id = scenario["id"]

        self.original_files = {
            f["path"]: FileContent(
                path=f["path"],
                content=f["content"],
                file_type=FileType(f["type"]),
                line_count=f["content"].count("\n") + 1,
            )
            for f in scenario["files"]
        }
        self.current_files = copy.deepcopy(self.original_files)

        self.current_error = ErrorInfo(
            phase=ErrorPhase(scenario["error"]["phase"]),
            error_message=scenario["error"]["message"],
            exit_code=scenario["error"].get("exit_code"),
            failed_step=scenario["error"].get("failed_step"),
            line_hint=scenario["error"].get("line_hint"),
        )

        self.expected_fixes = scenario["expected_fixes"]

        # Snapshot the initial broken state from simulators
        self.initial_snapshot = self._validation_snapshot()
        # Count how many checks are initially failing — that's our issues_total
        self.issues_total = sum(1 for v in self.initial_snapshot.values() if not v)
        # Ensure at least 1 issue (the scenario is supposed to be broken)
        self.issues_total = max(1, self.issues_total)
        self.issues_fixed = 0

        self.step_count = 0
        self.episode_reward = 0.0
        self.done = False
        self.hints_used = 0
        self.trajectory = []
        self.last_action_success = None
        self.last_action_feedback = None

        return self.get_observation()

    def step(self, action: Action) -> Tuple[Observation, float, bool, Dict[str, Any]]:
        if self.done:
            return self.get_observation(), 0.0, True, {"error": "Episode already done"}

        self.step_count += 1

        if action.action_type == ActionType.REQUEST_HINT:
            reward, feedback = self._handle_hint_request()
        elif action.action_type == ActionType.SUBMIT:
            reward, feedback = self._handle_submit()
        else:
            reward, feedback = self._handle_edit(action)

        self.last_action_feedback = feedback
        self.episode_reward += reward

        info: Dict[str, Any] = {}
        if self.step_count >= self.MAX_STEPS:
            self.done = True
            info["termination_reason"] = "max_steps"
        elif action.action_type == ActionType.SUBMIT:
            self.done = True
            info["termination_reason"] = "submitted"
        elif self.issues_fixed == self.issues_total:
            self.done = True
            info["termination_reason"] = "all_fixed"

        info["issues_fixed"] = self.issues_fixed
        info["issues_total"] = self.issues_total

        self.trajectory.append(
            {"step": self.step_count, "action": action.model_dump(), "reward": reward, "done": self.done, "info": info}
        )

        return self.get_observation(), reward, self.done, info

    def _handle_edit(self, action: Action) -> Tuple[float, str]:
        if not action.edits:
            self.last_action_success = False
            return 0.0, "No edits provided"

        reward = 0.0
        feedbacks: List[str] = []
        applied_count = 0
        for edit in action.edits:
            if edit.file_path not in self.current_files:
                # Allow creating new files (needed for K8s ConfigMap scenarios etc.)
                if action.action_type == ActionType.EDIT_FILE and edit.new_content:
                    ft = FileType.OTHER
                    if edit.file_path.startswith("k8s/") or edit.file_path.endswith(".yaml") or edit.file_path.endswith(".yml"):
                        ft = FileType.KUBERNETES
                    self.current_files[edit.file_path] = FileContent(
                        path=edit.file_path,
                        content=edit.new_content,
                        file_type=ft,
                        line_count=edit.new_content.count("\n") + 1,
                    )
                    feedbacks.append(f"Created new file: {edit.file_path}")
                    applied_count += 1
                    continue
                feedbacks.append(f"File not found: {edit.file_path}")
                continue

            file_content = self.current_files[edit.file_path]
            lines = file_content.content.split("\n")

            if action.action_type == ActionType.REPLACE_LINE:
                if edit.line_number and 1 <= edit.line_number <= len(lines):
                    lines[edit.line_number - 1] = edit.new_content or ""
                    feedbacks.append(f"Replaced line {edit.line_number} in {edit.file_path}")
                    applied_count += 1
                else:
                    feedbacks.append(f"Invalid line number: {edit.line_number}")
                    continue
            elif action.action_type == ActionType.ADD_LINE:
                insert_at = edit.line_number - 1 if edit.line_number else len(lines)
                insert_at = max(0, min(insert_at, len(lines)))
                lines.insert(insert_at, edit.new_content or "")
                feedbacks.append(f"Added line at {insert_at + 1} in {edit.file_path}")
                applied_count += 1
            elif action.action_type == ActionType.DELETE_LINE:
                if edit.line_number and 1 <= edit.line_number <= len(lines):
                    del lines[edit.line_number - 1]
                    feedbacks.append(f"Deleted line {edit.line_number} in {edit.file_path}")
                    applied_count += 1
                else:
                    feedbacks.append(f"Invalid line number: {edit.line_number}")
                    continue
            elif action.action_type == ActionType.EDIT_FILE:
                if edit.old_content and edit.old_content in file_content.content:
                    new_content = file_content.content.replace(edit.old_content, edit.new_content or "", 1)
                    lines = new_content.split("\n")
                    feedbacks.append(f"Edited content in {edit.file_path}")
                    applied_count += 1
                else:
                    feedbacks.append(f"Content not found in {edit.file_path}")
                    continue
            elif action.action_type == ActionType.ADD_BLOCK:
                insert_at = edit.line_number - 1 if edit.line_number else len(lines)
                insert_at = max(0, min(insert_at, len(lines)))
                block_lines = (edit.new_content or "").split("\n")
                lines[insert_at:insert_at] = block_lines
                feedbacks.append(f"Added block at {insert_at + 1} in {edit.file_path}")
                applied_count += 1
            elif action.action_type == ActionType.DELETE_BLOCK:
                if edit.old_content and edit.old_content in file_content.content:
                    new_content = file_content.content.replace(edit.old_content, "", 1)
                    lines = new_content.split("\n")
                    feedbacks.append(f"Deleted block in {edit.file_path}")
                    applied_count += 1
                else:
                    feedbacks.append(f"Block content not found in {edit.file_path}")
                    continue
            else:
                feedbacks.append(f"Unsupported action type for now: {action.action_type.value}")
                continue

            new_content = "\n".join(lines)
            self.current_files[edit.file_path] = FileContent(
                path=file_content.path,
                content=new_content,
                file_type=file_content.file_type,
                line_count=len(lines),
            )

        reward += self._check_fix_progress()

        if applied_count == 0:
            self.last_action_success = False
            return max(-0.02, reward - 0.02), "; ".join(feedbacks) or "No edit applied"

        self.last_action_success = True
        return max(0.0, reward), "; ".join(feedbacks)

    def _check_fix_progress(self) -> float:
        """Check fix progress by comparing current simulator state against initial broken state.

        Counts how many simulator checks flipped from fail→pass since reset.
        """
        current_snapshot = self._validation_snapshot()

        fixes_now = 0
        for key, initially_broken in self.initial_snapshot.items():
            if not initially_broken and current_snapshot.get(key, False):
                # This check was initially failing and now passes
                fixes_now += 1

        new_fixed = fixes_now - self.issues_fixed
        if new_fixed > 0:
            self.issues_fixed = fixes_now
            return 0.3 * new_fixed
        return 0.0

    def _handle_submit(self) -> Tuple[float, str]:
        docker_result = self.docker_sim.validate(self.current_files.get("Dockerfile"), self.current_files)
        workflow_file = self._find_workflow_file()
        workflow_result = self.workflow_sim.validate(workflow_file, self.current_files)
        k8s_result = self.k8s_sim.validate(self.current_files)

        has_k8s = any(fc.file_type == FileType.KUBERNETES for fc in self.current_files.values())
        has_docker = "Dockerfile" in self.current_files
        has_workflow = workflow_file is not None

        reward = 0.0
        parts: List[str] = []

        # Determine weight distribution based on what file types are present
        if has_docker and has_workflow and has_k8s:
            # Full stack: Docker 20%, Workflow 30%, K8s 30%, fix progress 20%
            docker_w, wf_w, k8s_w = 0.20, 0.30, 0.30
        elif has_docker and has_workflow:
            docker_w, wf_w, k8s_w = 0.50, 0.50, 0.0
        elif has_docker and has_k8s:
            docker_w, wf_w, k8s_w = 0.40, 0.0, 0.40
        elif has_workflow and has_k8s:
            docker_w, wf_w, k8s_w = 0.0, 0.40, 0.40
        elif has_k8s:
            docker_w, wf_w, k8s_w = 0.0, 0.0, 0.80
        elif has_docker:
            docker_w, wf_w, k8s_w = 0.50, 0.0, 0.0
        else:
            docker_w, wf_w, k8s_w = 0.0, 0.50, 0.0

        # Docker validation
        if has_docker:
            if docker_result.get("build_success"):
                reward += docker_w * 0.6
                parts.append("Docker build: PASS")
            else:
                parts.append(f"Docker build: FAIL - {docker_result.get('error', 'unknown')}")

            if docker_result.get("run_success"):
                reward += docker_w * 0.4
                parts.append("Docker run: PASS")
            else:
                parts.append(f"Docker run: FAIL - {docker_result.get('run_error', 'unknown')}")

        # Workflow validation
        if has_workflow:
            if workflow_result["parse_success"]:
                reward += wf_w * 0.4
                parts.append("Workflow parse: PASS")
            else:
                parts.append(f"Workflow parse: FAIL - {workflow_result.get('error', 'unknown')}")

            if workflow_result["execution_success"]:
                reward += wf_w * 0.6
                parts.append("Workflow execution: PASS")
            else:
                parts.append(f"Workflow execution: FAIL - {workflow_result.get('exec_error', 'unknown')}")

        # Kubernetes validation
        if has_k8s:
            if k8s_result["valid"]:
                reward += k8s_w * 0.4
                parts.append("K8s manifests: VALID")
            else:
                k8s_errors = k8s_result.get("errors", [])
                parts.append(f"K8s manifests: INVALID - {'; '.join(k8s_errors[:2])}")

            pod_status = k8s_result.get("pod_status", "N/A")
            if pod_status == "Running":
                reward += k8s_w * 0.3
                parts.append(f"K8s pod status: {pod_status}")
            else:
                parts.append(f"K8s pod status: {pod_status}")

            svc_status = k8s_result.get("service_status", "N/A")
            if "active" in svc_status.lower() or svc_status == "N/A":
                reward += k8s_w * 0.3
                parts.append(f"K8s service: {svc_status}")
            else:
                parts.append(f"K8s service: {svc_status}")

        self.last_action_success = reward >= 0.6
        return reward, "; ".join(parts)

    def _handle_hint_request(self) -> Tuple[float, str]:
        if self.hints_used >= self.MAX_HINTS:
            self.last_action_success = False
            return 0.0, "No hints remaining"

        self.hints_used += 1
        for fix in self.expected_fixes:
            file_path = fix["file"]
            if file_path not in self.current_files:
                continue
            current_content = self.current_files[file_path].content
            is_fixed = False
            if fix["type"] == "contains":
                is_fixed = fix["expected"] in current_content
            elif fix["type"] == "not_contains":
                is_fixed = fix["expected"] not in current_content
            if not is_fixed:
                hint = fix.get("hint", f"Check {file_path}")
                self.last_action_success = True
                return -0.05, f"Hint ({self.hints_used}/{self.MAX_HINTS}): {hint}"

        self.last_action_success = True
        return 0.0, "All known issues appear to be fixed"

    def get_observation(self) -> Observation:
        return Observation(
            task_id=self.current_task_id or "unknown",
            task_description=self.current_task.DESCRIPTION if self.current_task else "",
            difficulty=self.current_difficulty or TaskDifficulty.EASY,
            files=list(self.current_files.values()),
            error=self.current_error or ErrorInfo(
                phase=ErrorPhase.DOCKER_BUILD,
                error_message="No error",
            ),
            available_secrets=self.current_task.AVAILABLE_SECRETS if self.current_task else [],
            runner_os="ubuntu-latest",
            step_number=self.step_count,
            max_steps=self.MAX_STEPS,
            hints_used=self.hints_used,
            hints_available=self.MAX_HINTS - self.hints_used,
            last_action_success=self.last_action_success,
            last_action_feedback=self.last_action_feedback,
            issues_found=self.issues_fixed,
            issues_fixed=self.issues_fixed,
            total_issues=max(1, self.issues_total),
        )