bench_utils.py

from functools import cache
import re
import random
import copy


def get_questions(x: dict, num_filters=None):
    if 'questions' in x:
        if num_filters is None:
            return x['questions']
        else:
            questions = copy.copy(x['questions'])
            random.seed(42)
            random.shuffle(questions)
            return questions[:num_filters]
    elif 'question' in x:
        if x['question'] is None:
            return []
        else:
            return [x['question']]
    raise ValueError(f'Unknown question format: {x}')
    
    
def calculate_latex_command_text_ratio(line_text: str) -> tuple[float, list, list]:
    pattern = r'\\[a-zA-Z]+\*?(?:\[.*?\])?(?:{([^{}]*(?:{[^{}]*})*[^{}]*)})?'
    
    matches = list(re.finditer(pattern, line_text))
    full_commands = [match.group(0) for match in matches]
    
    command_chars_length = sum(len(cmd) for cmd in full_commands)
    
    total_length = len(line_text)
    
    ratio = command_chars_length / total_length if total_length > 0 else 0.0
    
    return ratio


def extract_target_spans(latex_content: str) -> list[tuple[int, int]]:
    spans = []
    
    pattern = r'\\begin{abstract}(.*?)\\end{abstract}'
    matches = re.finditer(pattern, latex_content, re.DOTALL)
    for match in matches:
        span = match.span(0)
        spans.append(span)
    
    pattern = r'(\\section{[^}]*(?:Introduction|Evaluation|Experiment|Result|Discussion)[^}]*}.*?)(?:\\section|$)'
    matches = re.finditer(pattern, latex_content, re.DOTALL)
    for match in matches:
        span = match.span(1)
        spans.append(span)
    
    pattern = r'(\\subsection{[^}]*(?:Result|Evaluation|Discussion)[^}]*}.*?)(?:\\subsection|$)'
    matches = re.finditer(pattern, latex_content, re.DOTALL)
    for match in matches:
        span = match.span(1)
        spans.append(span)
    
    return spans


def extract_block_spans(latex_content: str) -> list[tuple[int, int]]:
    figure_pattern = r'\\begin{(?:.+?)\*?}(.*?)\\end{(?:.+?)\*?}'
    
    spans = []
    
    figure_matches = re.finditer(figure_pattern, latex_content, re.DOTALL)
    
    for match in figure_matches:
        span = match.span(0)
        spans.append(span)
    
    return spans


def check_span_overlapping(spans: list[tuple[int, int]], targets: list[tuple[int, int]]) -> bool:
    for span in spans:
        for target in targets:
            if span[0] < target[1] and span[1] > target[0]:
                return True
    return False