Extract table without vertical lines

[stevenwu2017]

Hi,
I encountered two types of problems when extracting table data from PDFs.
PDF File:
ETNSI696-201706_2.pdf

Case 1:
I want to extract the data from this table without vertical lines in the red box.
as follows:

python 3.10
pdfplumber 0.11.5

source code:

import pdfplumber

def is_thin_rectangle_exists(obj: dict) -> bool:
    ''' Advised by jsvine - https://github.com/jsvine/pdfplumber/discussions/1242
    '''
    if obj["object_type"] == "rect":
        return obj["height"] < 5 or obj["width"] < 4

    return True

def find_hline_position(key_line,horizontal_lines):
    if key_line:
        keyword_top_y = key_line[0]['top']

        lines_above_keyword = []
        for line in horizontal_lines:
            if line['bottom'] < keyword_top_y:
                lines_above_keyword.append(line)

        if lines_above_keyword:
            closest_line = min(lines_above_keyword, key=lambda line: keyword_top_y - line['bottom'])
            return closest_line

def extract_table_vertical_line(pdf_file):
    pdf = pdfplumber.open(pdf_file)
    page = pdf.pages[0]
    filtered_page = page.filter(is_thin_rectangle_exists)

    horizontal_lines = filtered_page.horizontal_edges
    key_line=page.search(r'Top 10 Positions')
    h_line=find_hline_position(key_line=key_line,horizontal_lines=horizontal_lines)

    table_settings = {
        "snap_y_tolerance": 6,
        'explicit_vertical_lines': [h_line['x0'],h_line['x1']],
    }

    im = filtered_page.to_image(resolution=300)

    im.debug_tablefinder(table_settings)
    im.show()

    tables = page.extract_tables(table_settings)
    print(tables[0])

if __name__ == '__main__':
    pdf_file_path = r'D:\temp\Fund_Allocation\ETNSI696-201706_2.pdf'
    extract_table_vertical_line(pdf_file_path)

debug_tablefinder:

The existing problems:

1. The vertical dividing lines in the middle of the table are missing.
2. The Chinese text extracted may have line - break issues.

Case 2:
I want to extract all the legend data within the red box.

I have no idea about how to extract the legend data. Could you please take a look?