INNER CODE UNIT · Python

normalize_figure

gokayfem/awesome-vlm-architectures · scripts/extract_architecture_figures.py:46

def normalize_figure(value: str) -> str:
    return re.sub(
        r"[^0-9a-z]", "", value.lower().replace("figure", "").replace("fig", "")
    )


def find_caption_block(page: fitz.Page, figure: str) -> tuple[fitz.Rect, str]:
    wanted = normalize_figure(figure)
    patterns = [
        re.compile(rf"(?:Figure|Fig\.)\s*{re.escape(wanted)}\b", re.I),
        re.compile(
            rf"(?:Figure|Fig\.)\s*{re.escape(wanted.rstrip('abcdefghijklmnopqrstuvwxyz'))}\b",
            re.I,
        ),
    ]
    matches: list[tuple[fitz.Rect, str]] = []
    for block in page.get_text("blocks", sort=True):
        text = re.sub(r"\s+", " ", block[4]).strip()

View source record →