Function parse_poster_xml

PosterAgent/tree_split_layout.py:15–90

Parse an XML describing a single poster layout, e.g.: Introduction ... </P

(xml_file)

    return tree.getroot()

def parse_poster_xml(xml_file):
    """
    Parse an XML describing a single poster layout, e.g.:

    <Poster Width="685" Height="968">
      <Panel left="5" right="160" width="674" height="123">
        <Text>Introduction</Text>
        <Figure left="567" right="178" width="81" height="99" no="1" ... />
      </Panel>
      ...
    </Poster>

    Returns a dict with:
      {
        'poster_width': float,
        'poster_height': float,
        'panels': [
          {
            'x': float,
            'y': float,
            'width': float,
            'height': float,
            'text_blocks': [string, string, ...],
            'figure_blocks': [(fx, fy, fw, fh), ...]
          },
          ...
        ]
      }
    """
    root = parse_xml_with_recovery(xml_file)

    # Poster dimensions
    poster_w = float(root.get("Width", "1"))
    poster_h = float(root.get("Height", "1"))

    panels_data = []

    # Iterate <Panel> elements
    for panel_node in root.findall("Panel"):
        x = float(panel_node.get("left", "0"))
        y = float(panel_node.get("right", "0"))
        w = float(panel_node.get("width", "0"))
        h = float(panel_node.get("height", "0"))

        # Gather text blocks
        text_blocks = []
        for text_node in panel_node.findall("Text"):
            txt = text_node.text or ""
            txt = txt.strip()
            if txt:
                text_blocks.append(txt)

        # Gather figure blocks
        figure_blocks = []
        for fig_node in panel_node.findall("Figure"):
            fx = float(fig_node.get("left", "0"))
            fy = float(fig_node.get("right", "0"))
            fw = float(fig_node.get("width", "0"))