File size: 20,808 Bytes
72effac
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
12c635e
 
 
 
 
 
 
 
 
 
 
 
72effac
12c635e
 
 
 
72effac
12c635e
 
 
 
 
 
 
 
 
 
 
 
 
 
72effac
12c635e
72effac
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
"""
Python Code Evaluator β€” SE-Group1
COS60011 Technology Design Project
"""

# ---------------------------------------------------------------------------
# IMPORTS (all at top)
# ---------------------------------------------------------------------------
import os
import re
import ast
import time
from collections import deque
from dotenv import load_dotenv
import gradio as gr
from google import genai

load_dotenv()


# ---------------------------------------------------------------------------
# MODULE 3 β€” Pre-processing Module
# ---------------------------------------------------------------------------
def pre_processing(description, code):
    prompt = f"""
#ROLE
You are a Python code reviewer. Your goal is to determine if the provided Python code strictly complies with the requirements mentioned by the user.
 
#CONTEXT
Developers may create code that does not entirely meet the specified requirements. You will examine the relationship between a natural-language description and a Python code sample and create a structured evaluation report.
 
#INPUT DATA
Description: {description}
Code: ```python
{code}
```
 
#TASK
1. Carefully read the requirements description.
2. Examine each line of the Python code.
3. Determine whether the code fulfils ALL requirements stated in the description.
4. Estimate the accuracy in percentages to assess how accurately the code matches the description.
5. List specific requirements that are either absent or incorrectly applied.
6. Base your evaluation solely on static code analysis and logical reasoning.
 
#CONSTRAINTS
1. Do not execute the code.
2. The output must be in the specified format.
3. The output should be clear and concise.
 
#OUTPUT FORMAT (to be strictly followed without any changes)
Result: either "Pass" or "Fail" (Pass if the code is functional and matches the description, Fail if it has bugs or doesn't match).
Accuracy: an integer from 0 to 100 representing how well the code matches the description and is bug-free.
Review: a short prose explanation of the result (e.g. "Code correctly handles all described requirements including edge cases.").    
Issues: The value should be a single string containing a numbered list. For each item, include the severity (Error, Warning, or Info) followed by a dash and description of the issue (e.g. "Missing type hints on function signature"). List each issue on a new line starting with a number. Use newline characters (\n) to separate each line.
...
 
"""
    return prompt


# ---------------------------------------------------------------------------
# MODULE 4 β€” Generation Module
# ---------------------------------------------------------------------------
MODEL_ID       = "gemma-4-31b-it"
GOOGLE_API_KEY = os.environ.get("GOOGLE_API_KEY")

def generate_response(prompt: str) -> str:
    if not GOOGLE_API_KEY:
        return (
            "⚠️ Error: API Key not found! "
            "Please configure GOOGLE_API_KEY in Settings β†’ Variables and secrets."
        )
    try:
        client   = genai.Client(api_key=GOOGLE_API_KEY)
        response = client.models.generate_content(model=MODEL_ID, contents=prompt)
        return response.text
    except Exception as e:
        return f"❌ An error occurred during generation: {str(e)}"


# ---------------------------------------------------------------------------
# MODULE 5 β€” Output Module
# ---------------------------------------------------------------------------
def parse_output(raw: str) -> dict:
    if "<start_of_turn>model" in raw:
        raw = raw.split("<start_of_turn>model")[-1]

    result_match   = re.search(r"RESULT:\s*(PASS|FAIL)", raw, re.IGNORECASE)
    accuracy_match = re.search(r"ACCURACY:\s*(\d{1,3})%?", raw, re.IGNORECASE)
    summary_match  = re.search(r"SUMMARY:\s*(.+)", raw, re.IGNORECASE)
    issues_match   = re.search(r"ISSUES:\s*([\s\S]+)", raw, re.IGNORECASE)

    result   = result_match.group(1).upper()  if result_match   else "UNKNOWN"
    accuracy = int(accuracy_match.group(1))   if accuracy_match else -1
    summary  = summary_match.group(1).strip() if summary_match  else "Could not extract summary."

    if issues_match:
        raw_issues = issues_match.group(1).strip()
        issues = [
            line.lstrip("-β€’* ").strip()
            for line in raw_issues.splitlines()
            if line.strip() and line.strip() not in ("-", "β€’")
        ]
    else:
        issues = ["Could not extract issues from model response."]

    return {
        "result":   result,
        "accuracy": accuracy,
        "summary":  summary,
        "issues":   issues,
        "raw":      raw.strip(),
    }


def format_for_display(parsed: dict) -> tuple[str, str, str]:
    emoji   = "βœ…" if parsed["result"] == "PASS" else ("❌" if parsed["result"] == "FAIL" else "⚠️")
    verdict = f"{emoji}  {parsed['result']}"
    acc_str = f"{parsed['accuracy']}%" if parsed["accuracy"] >= 0 else "N/A"
    metrics = f"Accuracy: {acc_str}\n\nSummary: {parsed['summary']}"
    issues_text = "\n".join(f"β€’ {issue}" for issue in parsed["issues"])
    return verdict, metrics, issues_text


# ---------------------------------------------------------------------------
# MODULE 2 β€” Validation & Flow Management Module
# ---------------------------------------------------------------------------
MIN_DESCRIPTION_CHARS = 20
MAX_DESCRIPTION_CHARS = 3000
MIN_CODE_CHARS        = 10
MAX_CODE_CHARS        = 8000
MAX_CODE_LINES        = 300

FORBIDDEN_PATTERNS = [
    r"ignore (all |previous |above )?instructions",
    r"disregard (all |previous |above )?instructions",
    r"you are now",
    r"act as (a |an )?",
    r"<\s*(script|iframe|object|embed)",
    r"system\s*prompt",
    r"jailbreak",
]

PYTHON_KEYWORDS = {
    "def", "class", "import", "from", "return", "if", "else", "elif",
    "for", "while", "try", "except", "with", "lambda", "yield", "pass",
    "raise", "assert", "in", "not", "and", "or", "True", "False", "None",
    "print", "len", "range", "self",
}

class RateLimiter:
    def __init__(self, max_calls: int = 5, window_seconds: int = 60):
        self.max_calls      = max_calls
        self.window_seconds = window_seconds
        self._timestamps: deque = deque()

    def is_allowed(self) -> tuple[bool, str]:
        now = time.time()
        while self._timestamps and now - self._timestamps[0] > self.window_seconds:
            self._timestamps.popleft()
        if len(self._timestamps) >= self.max_calls:
            wait = int(self.window_seconds - (now - self._timestamps[0])) + 1
            return False, (
                f"⏳ Rate limit reached β€” {self.max_calls} requests in "
                f"{self.window_seconds}s. Please wait ~{wait}s and try again."
            )
        self._timestamps.append(now)
        return True, ""

_rate_limiter = RateLimiter(max_calls=5, window_seconds=60)

def _check_forbidden(text: str) -> str | None:
    lower = text.lower()
    for pattern in FORBIDDEN_PATTERNS:
        if re.search(pattern, lower):
            return (
                "Input contains disallowed content. "
                "Please remove prompt-injection or HTML patterns and try again."
            )
    return None

def _looks_like_python(code: str) -> tuple[bool, str]:
    tokens = set(re.findall(r"[A-Za-z_]\w*", code))
    if not tokens.intersection(PYTHON_KEYWORDS):
        return False, (
            "🐍 The code doesn't appear to be Python β€” no recognisable Python "
            "keywords found (e.g. def, class, import, return). "
            "Please submit Python code only."
        )
    try:
        ast.parse(code)
    except SyntaxError as exc:
        line_hint = f" (line {exc.lineno})" if exc.lineno else ""
        return False, (
            f"🐍 Python syntax error{line_hint}: {exc.msg}. "
            "Please fix the syntax error before evaluating."
        )
    return True, ""

def validate_inputs(description: str, code: str) -> list[str]:
    errors: list[str] = []

    if not description or not description.strip():
        errors.append("πŸ“‹ Requirements description is required.")
    if not code or not code.strip():
        errors.append("🐍 Python code is required.")
    if errors:
        return errors

    desc, code_ = description.strip(), code.strip()

    if len(desc) < MIN_DESCRIPTION_CHARS:
        errors.append(f"πŸ“‹ Description too short ({len(desc)} chars) β€” minimum is {MIN_DESCRIPTION_CHARS} characters.")
    if len(code_) < MIN_CODE_CHARS:
        errors.append(f"🐍 Code too short ({len(code_)} chars) β€” minimum is {MIN_CODE_CHARS} characters.")
    if len(desc) > MAX_DESCRIPTION_CHARS:
        errors.append(f"πŸ“‹ Description too long ({len(desc):,} chars) β€” max is {MAX_DESCRIPTION_CHARS:,} characters.")
    if len(code_) > MAX_CODE_CHARS:
        errors.append(f"🐍 Code too long ({len(code_):,} chars) β€” max is {MAX_CODE_CHARS:,} characters.")
    if len(code_.splitlines()) > MAX_CODE_LINES:
        errors.append(f"🐍 Code has too many lines ({len(code_.splitlines())}) β€” max is {MAX_CODE_LINES} lines.")

    if err := _check_forbidden(desc):
        errors.append(f"πŸ“‹ {err}")
    if err := _check_forbidden(code_):
        errors.append(f"🐍 {err}")

    if not errors:
        is_python, py_error = _looks_like_python(code_)
        if not is_python:
            errors.append(py_error)

    if not errors:
        allowed, rate_msg = _rate_limiter.is_allowed()
        if not allowed:
            errors.append(rate_msg)

    return errors

def validate_and_evaluate(description: str, code: str):
    errors = validate_inputs(description, code)
    if errors:
        return "", "", "", "\n".join(f"{i+1}. {e}" for i, e in enumerate(errors))

    if not GOOGLE_API_KEY:
        return "", "", "", "❌ GOOGLE_API_KEY secret not set in Space Settings."

    prompt = build_prompt(description, code)
    try:
        raw_output = generate_response(prompt)
    except Exception as exc:
        return "", "", "", f"❌ Model error: {exc}"

    parsed = parse_output(raw_output)
    verdict, metrics, issues = format_for_display(parsed)
    return verdict, metrics, issues, ""


# ---------------------------------------------------------------------------
# MODULE 1 β€” UI Module
# ---------------------------------------------------------------------------
CUSTOM_CSS = """
@import url('https://fonts.googleapis.com/css2?family=Space+Mono:wght@400;700&family=DM+Sans:ital,wght@0,300;0,400;0,500;0,600;1,400&display=swap');
*, *::before, *::after { box-sizing: border-box; }
body, .gradio-container {
    font-family: 'DM Sans', sans-serif !important;
    background: #0D0F14 !important;
    color: #E8EAF0 !important;
}
.gradio-container {
    max-width: 1140px !important;
    margin: 0 auto !important;
    padding: 32px 24px !important;
}
.eval-header {
    display: flex;
    align-items: center;
    gap: 16px;
    padding: 0 0 28px;
    border-bottom: 1px solid #2E3140;
    margin-bottom: 28px;
}
.eval-header .logo {
    width: 44px; height: 44px;
    border-radius: 10px;
    background: linear-gradient(135deg, #534AB7 0%, #1D9E75 100%);
    display: flex; align-items: center; justify-content: center;
    flex-shrink: 0;
    font-size: 22px; line-height: 1;
}
.eval-header h1 {
    font-size: 20px !important;
    font-weight: 600 !important;
    letter-spacing: -0.3px !important;
    color: #E8EAF0 !important;
    margin: 0 !important;
}
.eval-header .model-badge {
    margin-left: auto;
    font-family: 'Space Mono', monospace;
    font-size: 10px;
    background: #1E2028;
    border: 1px solid #3A3E52;
    color: #6A6E80;
    padding: 4px 10px;
    border-radius: 4px;
    letter-spacing: 1.5px;
    white-space: nowrap;
}
.gradio-textbox textarea,
.gradio-code textarea,
.gradio-textbox input {
    background: #161820 !important;
    border: 1px solid #2E3140 !important;
    border-radius: 10px !important;
    color: #E8EAF0 !important;
    font-family: 'DM Sans', sans-serif !important;
    font-size: 13.5px !important;
    line-height: 1.7 !important;
    padding: 14px 16px !important;
    transition: border-color 0.15s !important;
    resize: vertical !important;
}
.gradio-textbox textarea:focus,
.gradio-code textarea:focus {
    border-color: #534AB7 !important;
    outline: none !important;
    box-shadow: 0 0 0 3px rgba(83, 74, 183, 0.15) !important;
}
.gradio-textbox textarea::placeholder {
    color: #4A4E60 !important;
}
#code-input textarea {
    font-family: 'Space Mono', monospace !important;
    font-size: 12.5px !important;
    line-height: 1.65 !important;
}
.gradio-textbox label span,
.gradio-code label span {
    font-family: 'DM Sans', sans-serif !important;
    font-size: 13px !important;
    font-weight: 700 !important;
    color: #E8EAF0 !important;
    text-transform: uppercase !important;
    letter-spacing: 0.8px !important;
}
#eval-btn {
    background: #534AB7 !important;
    border: none !important;
    color: #fff !important;
    font-family: 'DM Sans', sans-serif !important;
    font-size: 14px !important;
    font-weight: 500 !important;
    padding: 12px 32px !important;
    border-radius: 8px !important;
    cursor: pointer !important;
    transition: background 0.15s, transform 0.1s !important;
    letter-spacing: -0.1px !important;
}
#eval-btn:hover { background: #7F77DD !important; transform: translateY(-1px) !important; }
#eval-btn:active { transform: translateY(0) !important; }
#clear-btn {
    background: transparent !important;
    border: 1px solid #3A3E52 !important;
    color: #9DA0B0 !important;
    font-family: 'DM Sans', sans-serif !important;
    font-size: 13px !important;
    padding: 12px 20px !important;
    border-radius: 8px !important;
    cursor: pointer !important;
    transition: border-color 0.15s, color 0.15s !important;
}
#clear-btn:hover { border-color: #7F77DD !important; color: #E8EAF0 !important; }
.results-heading {
    font-size: 11px !important;
    font-weight: 500 !important;
    color: #6A6E80 !important;
    text-transform: uppercase !important;
    letter-spacing: 1px !important;
    padding: 0 0 16px !important;
    border-bottom: 1px solid #2E3140 !important;
    margin-bottom: 20px !important;
}
#verdict-out textarea {
    background: #161820 !important; border: 1px solid #2E3140 !important;
    border-radius: 10px !important; font-family: 'Space Mono', monospace !important;
    font-size: 26px !important; font-weight: 700 !important;
    text-align: center !important; padding: 20px !important;
    color: #E8EAF0 !important; cursor: default !important;
}
#accuracy-out textarea {
    background: #161820 !important; border: 1px solid #2E3140 !important;
    border-radius: 10px !important; font-family: 'Space Mono', monospace !important;
    font-size: 26px !important; font-weight: 700 !important;
    text-align: center !important; padding: 20px !important;
    color: #E8EAF0 !important; cursor: default !important;
}
#summary-out textarea {
    background: #161820 !important; border: 1px solid #2E3140 !important;
    border-radius: 10px !important; font-family: 'DM Sans', sans-serif !important;
    font-size: 13.5px !important; line-height: 1.7 !important;
    padding: 16px !important; color: #C0C3D0 !important; cursor: default !important;
}
#issues-out textarea {
    background: #161820 !important; border: 1px solid #2E3140 !important;
    border-radius: 10px !important; font-family: 'DM Sans', sans-serif !important;
    font-size: 13.5px !important; line-height: 1.8 !important;
    padding: 16px !important; color: #C0C3D0 !important;
    cursor: default !important; min-height: 120px !important;
}
#error-out textarea {
    background: #1A0E0E !important; border: 1px solid #5a2020 !important;
    border-radius: 10px !important; font-family: 'DM Sans', sans-serif !important;
    font-size: 13.5px !important; padding: 14px 16px !important;
    color: #F0997B !important; cursor: default !important;
}
.divider { height: 1px; background: #2E3140; margin: 20px 0; }
footer { display: none !important; }
::-webkit-scrollbar { width: 6px; height: 6px; }
::-webkit-scrollbar-track { background: transparent; }
::-webkit-scrollbar-thumb { background: #3A3E52; border-radius: 3px; }
::-webkit-scrollbar-thumb:hover { background: #534AB7; }
"""

HEADER_HTML = """
<div class="eval-header">
  <div class="logo">πŸ”</div>
  <div><h1>Python Code Evaluator</h1></div>
  <div class="model-badge">GEMMA-4 Β· 31B-IT</div>
</div>
"""

RESULTS_HEADING_HTML = """
<div class="divider"></div>
<div class="results-heading">⬑ &nbsp; Evaluation Results</div>
"""

INPUT_HINT_HTML = """
<div style="font-size:12px;color:#4A4E60;margin-top:-4px;padding-bottom:4px;">
  Tip β€” paste your description and code above, then click
  <strong style="color:#7F77DD">Evaluate</strong>.
</div>
"""

def build_ui(evaluate_fn):

    def _split_metrics(metrics_str: str):
        acc, summary = "", ""
        if not metrics_str:
            return acc, summary
        for line in metrics_str.splitlines():
            if line.startswith("Accuracy:"):
                acc = line.replace("Accuracy:", "").strip()
            elif line.startswith("Summary:"):
                summary = line.replace("Summary:", "").strip()
        return acc, summary

    def _ui_evaluate(description: str, code: str):
        verdict, metrics, issues, error = evaluate_fn(description, code)
        accuracy, summary = _split_metrics(metrics)
        if "PASS" in verdict.upper():
            verdict_display = "βœ…  PASS"
        elif "FAIL" in verdict.upper():
            verdict_display = "❌  FAIL"
        else:
            verdict_display = verdict or ""
        return verdict_display, accuracy, summary, issues, error

    with gr.Blocks(css=CUSTOM_CSS, title="Python Code Evaluator") as demo:

        gr.HTML(HEADER_HTML)

        with gr.Row(equal_height=True):
            description_input = gr.Textbox(
                label="Requirements Description",
                placeholder=(
                    "Describe what the Python code is supposed to do…\n\n"
                    "Example: Write a function that accepts a list of integers "
                    "and returns the sum of all even numbers."
                ),
                lines=10, max_lines=20, elem_id="desc-input",
            )
            code_input = gr.Textbox(
                label="Python Code",
                placeholder="# Paste your Python code here…\n\ndef example(data):\n    pass\n",
                lines=10, max_lines=30, elem_id="code-input",
            )

        gr.HTML(INPUT_HINT_HTML)

        with gr.Row():
            eval_btn  = gr.Button("Evaluate", variant="primary",  elem_id="eval-btn",  scale=0)
            clear_btn = gr.Button("Clear",    variant="secondary", elem_id="clear-btn", scale=0)

        gr.HTML(RESULTS_HEADING_HTML)

        with gr.Row(equal_height=True):
            verdict_out  = gr.Textbox(label="Verdict",   interactive=False, elem_id="verdict-out",  scale=1)
            accuracy_out = gr.Textbox(label="Accuracy",  interactive=False, elem_id="accuracy-out", scale=1)
            summary_out  = gr.Textbox(label="Summary",   interactive=False, elem_id="summary-out",  scale=2, lines=3)

        issues_out = gr.Textbox(label="Issues Detected", interactive=False, lines=5,  elem_id="issues-out")
        error_out  = gr.Textbox(label="Status / Errors", interactive=False, visible=True, elem_id="error-out")

        gr.HTML("""
        <div style="text-align:right;font-size:11px;color:#4A4E60;margin-top:8px;">
          Press <kbd style="background:#1E2028;border:1px solid #3A3E52;border-radius:4px;
          padding:2px 6px;font-family:'Space Mono',monospace;font-size:10px;color:#9DA0B0;">
          Ctrl+Enter</kbd> to evaluate
        </div>
        """)

        outputs = [verdict_out, accuracy_out, summary_out, issues_out, error_out]
        eval_btn.click(fn=_ui_evaluate, inputs=[description_input, code_input], outputs=outputs)
        description_input.submit(fn=_ui_evaluate, inputs=[description_input, code_input], outputs=outputs)
        code_input.submit(fn=_ui_evaluate, inputs=[description_input, code_input], outputs=outputs)

        def _clear():
            return "", "", "", "", "", ""

        clear_btn.click(
            fn=_clear, inputs=[],
            outputs=[description_input, code_input,
                     verdict_out, accuracy_out, summary_out, issues_out],
        )

    return demo


# ---------------------------------------------------------------------------
# ENTRY POINT
# ---------------------------------------------------------------------------
if __name__ == "__main__":
    app = build_ui(validate_and_evaluate)
    app.launch(server_name="0.0.0.0", server_port=7860, share=False, show_error=True)