perf: fix race conditions with file locking and optimize backend latency

This commit is contained in:
a2nr 2026-05-05 22:57:19 +07:00
parent 7a3a2ecc13
commit 2b6ec21633
3 changed files with 153 additions and 92 deletions

View File

@ -25,7 +25,6 @@ def login():
token = (data.get('token') or '').strip()
if not token:
time.sleep(1.5) # Tarpitting for empty tokens
return jsonify({'success': False, 'message': 'Token is required'})
student_info = validate_token(token)
@ -42,7 +41,6 @@ def login():
)
return response
else:
time.sleep(1.5) # Tarpitting for invalid tokens
return jsonify({'success': False, 'message': 'Invalid token'})
except Exception as e:

View File

@ -53,6 +53,9 @@ def track_progress():
def api_progress_report():
"""Return progress report data as JSON."""
token = request.args.get('token', '').strip()
if not token:
token = request.cookies.get('student_token', '').strip()
if not token or not validate_token(token):
return jsonify({'success': False, 'message': 'Unauthorized'}), 401
@ -70,6 +73,9 @@ def api_progress_report():
def export_progress_csv():
"""Export the progress report as CSV."""
token = request.args.get('token', '').strip()
if not token:
token = request.cookies.get('student_token', '').strip()
if not token or not validate_token(token):
return jsonify({'success': False, 'message': 'Unauthorized'}), 401

View File

@ -1,107 +1,166 @@
"""
Token and student progress operations backed by CSV file.
Optimized with in-memory caching and cross-process file locking.
"""
import csv
import logging
import os
import fcntl
import threading
from typing import Dict, List, Optional, Tuple
from config import TOKENS_FILE
# Global cache and synchronization for the current process
_cache_lock = threading.Lock()
_cached_tokens: Dict[str, dict] = {}
_cached_mtime: float = 0.0
_cached_fieldnames: List[str] = []
def _load_tokens_safely() -> Tuple[Dict[str, dict], List[str]]:
"""Load tokens from CSV with file locking and return (tokens_dict, fieldnames)."""
if not os.path.exists(TOKENS_FILE):
return {}, []
try:
with open(TOKENS_FILE, 'r', newline='', encoding='utf-8') as f:
# Acquire shared lock for reading (blocks if an exclusive lock is held)
fcntl.flock(f.fileno(), fcntl.LOCK_SH)
try:
reader = csv.DictReader(f, delimiter=';')
fieldnames = reader.fieldnames or []
# Use a dictionary for O(1) lookups by token
tokens = {row['token']: row for row in reader if 'token' in row}
return tokens, fieldnames
finally:
fcntl.flock(f.fileno(), fcntl.LOCK_UN)
except Exception as e:
logging.error(f"Error loading tokens file: {e}")
return {}, []
def _get_tokens() -> Tuple[Dict[str, dict], List[str]]:
"""Get tokens from cache or reload if file has changed on disk."""
global _cached_tokens, _cached_mtime, _cached_fieldnames
if not os.path.exists(TOKENS_FILE):
return {}, []
try:
current_mtime = os.path.getmtime(TOKENS_FILE)
except OSError:
return {}, []
with _cache_lock:
if current_mtime > _cached_mtime:
# File has been modified by this or another process
new_tokens, new_fieldnames = _load_tokens_safely()
if new_tokens: # Only update if we successfully read something
_cached_tokens = new_tokens
_cached_fieldnames = new_fieldnames
_cached_mtime = current_mtime
logging.debug(f"Reloaded {len(_cached_tokens)} tokens from {TOKENS_FILE}")
return _cached_tokens, _cached_fieldnames
def get_teacher_token():
"""Return the teacher token (first data row in CSV)."""
if not os.path.exists(TOKENS_FILE):
tokens, _ = _get_tokens()
if not tokens:
return None
with open(TOKENS_FILE, 'r', newline='', encoding='utf-8') as csvfile:
reader = csv.DictReader(csvfile, delimiter=';')
for row in reader:
return row['token']
return None
# Python 3.7+ preserves insertion order in dicts
return next(iter(tokens.keys()))
def is_teacher_token(token):
"""Check if the given token belongs to the teacher (first row)."""
"""Check if the given token belongs to the teacher."""
return token == get_teacher_token()
def validate_token(token):
"""Validate if a token exists in the CSV file and return student info."""
if not os.path.exists(TOKENS_FILE):
"""Validate if a token exists and return student info."""
if not token:
return None
with open(TOKENS_FILE, 'r', newline='', encoding='utf-8') as csvfile:
reader = csv.DictReader(csvfile, delimiter=';')
for row in reader:
if row['token'] == token:
return {
'token': row['token'],
'student_name': row['nama_siswa'],
'is_teacher': is_teacher_token(token),
}
tokens, _ = _get_tokens()
row = tokens.get(token)
if row:
return {
'token': row['token'],
'student_name': row['nama_siswa'],
'is_teacher': is_teacher_token(token),
}
return None
def get_student_progress(token):
"""Get the progress of a student based on their token."""
if not os.path.exists(TOKENS_FILE):
if not token:
return None
with open(TOKENS_FILE, 'r', newline='', encoding='utf-8') as csvfile:
reader = csv.DictReader(csvfile, delimiter=';')
for row in reader:
if row['token'] == token:
return row
return None
tokens, _ = _get_tokens()
return tokens.get(token)
def update_student_progress(token, lesson_name, status="completed"):
"""Update the progress of a student for a specific lesson."""
"""Update the progress of a student for a specific lesson with cross-process locking."""
if not os.path.exists(TOKENS_FILE):
logging.warning(f"Tokens file {TOKENS_FILE} does not exist")
return False
rows = []
with open(TOKENS_FILE, 'r', newline='', encoding='utf-8') as csvfile:
reader = csv.DictReader(csvfile, delimiter=';')
fieldnames = reader.fieldnames
rows = list(reader)
updated = False
for row in rows:
if row['token'] == token:
if lesson_name in fieldnames:
row[lesson_name] = status
updated = True
logging.info(f"Updating progress for token {token}, lesson {lesson_name}, status {status}")
else:
logging.warning(f"Lesson '{lesson_name}' not found in CSV columns: {fieldnames}")
break
if updated:
with open(TOKENS_FILE, 'w', newline='', encoding='utf-8') as csvfile:
writer = csv.DictWriter(csvfile, fieldnames=fieldnames, delimiter=';')
writer.writeheader()
writer.writerows(rows)
logging.info(f"Updated progress for token {token}, lesson {lesson_name}, status {status}")
else:
logging.warning(f"Failed to update progress for token {token}, lesson {lesson_name}")
return updated
# Use 'r+' to allow reading and writing in the same file handle with one lock
try:
with open(TOKENS_FILE, 'r+', newline='', encoding='utf-8') as f:
# Acquire exclusive lock for the whole read-modify-write cycle
fcntl.flock(f.fileno(), fcntl.LOCK_EX)
try:
reader = csv.DictReader(f, delimiter=';')
fieldnames = reader.fieldnames
rows = list(reader)
updated = False
for row in rows:
if row['token'] == token:
if lesson_name in fieldnames:
row[lesson_name] = status
updated = True
logging.info(f"Updating progress for token {token}, lesson {lesson_name}, status {status}")
else:
logging.warning(f"Lesson '{lesson_name}' not found in CSV columns")
break
if updated:
# Clear and write back
f.seek(0)
f.truncate()
writer = csv.DictWriter(f, fieldnames=fieldnames, delimiter=';')
writer.writeheader()
writer.writerows(rows)
f.flush()
os.fsync(f.fileno()) # Guarantee write to disk
return True
else:
return False
finally:
fcntl.flock(f.fileno(), fcntl.LOCK_UN)
except Exception as e:
logging.error(f"Error updating student progress: {e}")
return False
def initialize_tokens_file(lesson_names):
"""Initialize the tokens CSV file with headers and lesson columns."""
if not os.path.exists(TOKENS_FILE):
headers = ['token', 'nama_siswa'] + lesson_names
with open(TOKENS_FILE, 'w', newline='', encoding='utf-8') as csvfile:
writer = csv.writer(csvfile, delimiter=';')
writer.writerow(headers)
print(f"Created new tokens file: {TOKENS_FILE} with headers: {headers}")
try:
with open(TOKENS_FILE, 'w', newline='', encoding='utf-8') as csvfile:
# No lock needed for initial creation
writer = csv.writer(csvfile, delimiter=';')
writer.writerow(headers)
print(f"Created new tokens file: {TOKENS_FILE} with headers: {headers}")
except Exception as e:
logging.error(f"Error initializing tokens file: {e}")
def calculate_student_completion(student_data, all_lessons):
@ -119,40 +178,38 @@ def calculate_student_completion(student_data, all_lessons):
def get_all_students_progress(all_lessons_func):
"""Get all students' progress data for the progress report.
Returns (all_students_progress, ordered_lessons).
"""
"""Get all students' progress data for the progress report."""
all_students_progress = []
ordered_lessons = []
lesson_headers = []
if not os.path.exists(TOKENS_FILE):
tokens, fieldnames = _get_tokens()
if not tokens:
return all_students_progress, ordered_lessons
with open(TOKENS_FILE, 'r', newline='', encoding='utf-8') as csvfile:
reader = csv.DictReader(csvfile, delimiter=';')
lesson_headers = [field for field in reader.fieldnames if field not in ['token', 'nama_siswa']]
lesson_headers = [field for field in fieldnames if field not in ['token', 'nama_siswa']]
all_lessons_dict = {}
for lesson in all_lessons_func():
lesson_key = lesson['filename'].replace('.md', '')
all_lessons_dict[lesson_key] = lesson
all_lessons_dict = {}
for lesson in all_lessons_func():
lesson_key = lesson['filename'].replace('.md', '')
all_lessons_dict[lesson_key] = lesson
for lesson_header in lesson_headers:
if lesson_header in all_lessons_dict:
ordered_lessons.append(all_lessons_dict[lesson_header])
else:
ordered_lessons.append({
'filename': f"{lesson_header}.md",
'title': lesson_header.replace('_', ' ').title(),
'description': 'Lesson information not available',
})
for lesson_header in lesson_headers:
if lesson_header in all_lessons_dict:
ordered_lessons.append(all_lessons_dict[lesson_header])
else:
ordered_lessons.append({
'filename': f"{lesson_header}.md",
'title': lesson_header.replace('_', ' ').title(),
'description': 'Lesson information not available',
})
for row in reader:
student_data = dict(row)
del student_data['token']
student_data['completed_count'] = calculate_student_completion(student_data, ordered_lessons)
all_students_progress.append(student_data)
for row in tokens.values():
student_data = dict(row)
# Don't delete 'token' from the original dict in cache!
student_data_copy = student_data.copy()
if 'token' in student_data_copy:
del student_data_copy['token']
student_data_copy['completed_count'] = calculate_student_completion(student_data_copy, ordered_lessons)
all_students_progress.append(student_data_copy)
return all_students_progress, ordered_lessons