File size: 7,570 Bytes
fc6ff5a
d581a4f
adea8c3
548872c
f8670cd
548872c
adea8c3
 
 
 
548872c
adea8c3
d581a4f
548872c
d581a4f
 
548872c
 
d581a4f
 
 
 
 
 
 
548872c
 
d581a4f
 
 
 
 
 
 
 
 
 
548872c
 
d581a4f
 
 
 
 
 
 
 
 
 
 
548872c
 
 
d581a4f
 
548872c
 
 
d581a4f
 
 
 
 
f8670cd
d581a4f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f8670cd
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d581a4f
0320a8d
548872c
 
 
 
 
d581a4f
fc6ff5a
d581a4f
fc6ff5a
0320a8d
d581a4f
548872c
 
d581a4f
 
f8670cd
 
 
 
 
d581a4f
 
548872c
 
d581a4f
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
548872c
d581a4f
 
 
 
 
 
 
 
 
 
 
548872c
 
74df718
 
 
548872c
d581a4f
 
 
 
 
 
 
548872c
d581a4f
 
 
 
548872c
d581a4f
 
 
 
 
 
548872c
d581a4f
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
from datetime import datetime, timezone
from typing import List, Optional, Set
from codelens_env.models import (
    TaskId, Action, Observation, StepResult, ResetResult,
    ActionType, ActionRecord, EpisodeResult, Severity, GroundTruthIssue, Reward
)
from codelens_env.scenarios import get_scenario
from codelens_env.graders.bug_grader import grade_bug_detection
from codelens_env.graders.security_grader import grade_security_audit
from codelens_env.graders.arch_grader import grade_architectural_review

class CodeLensEnv:
    MAX_NOISE_BUDGET = 5
    TASK_MAX_STEPS = {
        TaskId.BUG_DETECTION: 10,
        TaskId.SECURITY_AUDIT: 15,
        TaskId.ARCHITECTURAL_REVIEW: 20,
    }
    SEVERITY_WEIGHTS = {
        Severity.CRITICAL: 1.0,
        Severity.HIGH: 0.8,
        Severity.MEDIUM: 0.5,
        Severity.LOW: 0.2,
        Severity.INFO: 0.0,
    }

    def __init__(self):
        self.task_id: Optional[TaskId] = None
        self.seed: int = 42
        self.scenario = None
        self.step_count: int = 0
        self.noise_budget: int = self.MAX_NOISE_BUDGET
        self.history: List[ActionRecord] = []
        self.matched_issue_ids: Set[str] = set()
        self.done: bool = False
        self.terminated_reason: str = ""
        self.episode_id: str = ""

    def reset(self, task_id: TaskId, seed: int = 42) -> ResetResult:
        self.scenario = get_scenario(task_id, seed)
        self.task_id = task_id
        self.seed = seed
        self.step_count = 0
        self.noise_budget = self.MAX_NOISE_BUDGET
        self.history = []
        self.matched_issue_ids = set()
        self.done = False
        self.terminated_reason = ""
        
        obs = self._build_observation()
        return ResetResult(
            task_id=task_id,
            seed=seed,
            scenario_hash=self.scenario.hash,
            observation=obs
        )

    def step(self, action: Action) -> StepResult:
        if self.done:
            raise ValueError("Episode is already finished")

        self.step_count += 1
        reward = 0.0
        match = None  # Track matched ground truth issue (if any)
        
        # Determine terminal state and reward
        if action.action_type in (ActionType.APPROVE, ActionType.REQUEST_CHANGES):
            self.done = True
            self.terminated_reason = "terminal_action"
            reward = 0.0 # Grader handles final score
        
        elif action.action_type == ActionType.FLAG_ISSUE:
            match = None
            for issue in self.scenario.ground_truth_issues:
                if self._is_match(action, issue):
                    match = issue
                    break
            
            if match:
                if match.id not in self.matched_issue_ids:
                    reward = self.SEVERITY_WEIGHTS.get(match.severity, 0.0)
                    self.matched_issue_ids.add(match.id)
                else:
                    # Already matched: penalty
                    reward = -0.05
                    self.noise_budget -= 1
                    if self.noise_budget <= 0:
                        self.done = True
                        self.terminated_reason = "noise_exhausted"
            else:
                # False positive: penalty
                reward = -0.05
                self.noise_budget -= 1
                if self.noise_budget <= 0:
                    self.done = True
                    self.terminated_reason = "noise_exhausted"
        
        # Max steps check
        max_steps = self.TASK_MAX_STEPS.get(self.task_id, 10)
        if not self.done and self.step_count >= max_steps:
            self.done = True
            self.terminated_reason = "max_steps"

        # Build reward reason
        if action.action_type in (ActionType.APPROVE, ActionType.REQUEST_CHANGES):
            reward_reason = "Terminal action submitted"
        elif action.action_type == ActionType.FLAG_ISSUE:
            if match and match.id in self.matched_issue_ids and reward > 0:
                reward_reason = f"Correctly identified issue: {match.description[:60]}"
            elif match and reward < 0:
                reward_reason = "Duplicate issue flagged"
            elif not match:
                reward_reason = "False positive: no matching ground truth issue"
            else:
                reward_reason = f"Matched issue {match.id}" if match else "No match"
        else:
            reward_reason = "Non-scoring action"

        # Record action
        record = ActionRecord(
            action_type=action.action_type,
            body=action.body,
            filename=action.filename,
            line_number=action.line_number,
            category=action.category,
            severity=action.severity,
            verdict=action.verdict,
            reward=float(reward),
            timestamp=datetime.now(timezone.utc).isoformat()
        )
        self.history.append(record)

        return StepResult(
            observation=self._build_observation(),
            reward=float(reward),
            reward_info=Reward(
                value=float(max(0.0, reward)),
                reason=reward_reason,
                is_terminal=self.done
            ),
            done=self.done,
            info={"terminated_reason": self.terminated_reason}
        )

    def _is_match(self, action: Action, issue: GroundTruthIssue) -> bool:
        if action.filename != issue.filename:
            return False
        if action.line_number is None:
            return False
        if abs(action.line_number - issue.line_number) > 3:
            return False
        if action.category != issue.category:
            return False
        
        body_lower = (action.body or "").lower()
        return any(kw.lower() in body_lower for kw in issue.keywords)

    def _build_observation(self) -> Observation:
        max_steps = self.TASK_MAX_STEPS.get(self.task_id, 10)
        diff = "\n".join(f.patch for f in self.scenario.files_changed)
        
        return Observation(
            task_id=self.task_id,
            scenario_hash=self.scenario.hash,
            pr_title=self.scenario.pr_title,
            pr_description=self.scenario.pr_description,
            diff=diff,
            files_changed=self.scenario.files_changed,
            step_count=self.step_count,
            max_steps=max_steps,
            noise_budget=self.noise_budget,
            max_noise_budget=self.MAX_NOISE_BUDGET,
            issues_flagged=len(self.matched_issue_ids),
        )

    def state(self) -> Observation:
        return self._build_observation()

    def get_final_result(self) -> EpisodeResult:
        if self.task_id == TaskId.BUG_DETECTION:
            final_score = grade_bug_detection(self.scenario, self.history)
        elif self.task_id == TaskId.SECURITY_AUDIT:
            final_score = grade_security_audit(self.scenario, self.history)
        else:
            final_score = grade_architectural_review(self.scenario, self.history)
            
        return EpisodeResult(
            episode_id=self.episode_id,
            task_id=self.task_id,
            scenario_hash=self.scenario.hash,
            seed=self.seed,
            final_score=round(final_score, 4),
            steps_taken=self.step_count,
            issues_found=len(self.matched_issue_ids),
            issues_total=len(self.scenario.ground_truth_issues),
            noise_penalties=self.MAX_NOISE_BUDGET - self.noise_budget,
            history=self.history,
            terminated_reason=self.terminated_reason
        )