import sys from bs4 import BeautifulSoup htmlfilepath = 'table.html' tsvfilepath =…
- User
import sys from bs4 import BeautifulSoup
html_file_path = 'table.html' tsv_file_path = 'table.tsv'
def process_svg_for_rects(svg_element): """ Process an SVG element to extract rect data associated with "-35" and "-10" text. Returns a dictionary with keys: w35, x35, w10, x10 """ rect_data = {'w35': "NONE", 'x35': "NONE", 'w10': "NONE", 'x10': "NONE"}
if not svg_element: # Guard clause if None is passed return rect_data rects = svg_element.find_all('rect') for rect in rects: next_text = rect.find_next('text') if next_text: text_content = next_text.get_text(strip=True) if text_content == "-35": rect_data['w35'] = rect.get('width', "NONE") rect_data['x35'] = rect.get('x', "NONE") elif text_content == "-10": rect_data['w10'] = rect.get('width', "NONE") rect_data['x10'] = rect.get('x', "NONE") return rect_datatry: with open(html_file_path, 'r', encoding='utf-8') as f: html_content = f.read() except FileNotFoundError: print(f"Error: Input file '{html_file_path}' not found.", file=sys.stderr) sys.exit(1) except Exception as e: print(f"Error reading file '{html_file_path}': {e}", file=sys.stderr) sys.exit(1)
try: soup = BeautifulSoup(html_content, 'lxml') except Exception as e: print(f"Error parsing HTML: {e}", file=sys.stderr) sys.exit(1)
output_tsv_lines = [] html_tables = soup.find_all('table')
if not html_tables: print("No elements found.", file=sys.stderr) sys.exit(1) # Consistent with original behavior
for table_idx, html_table_element in enumerate(html_tables): if len(html_tables) > 1: output_tsv_lines.append(f"# --- Data from Table {table_idx + 1} ---") else: output_tsv_lines.append(f"# --- Combined Table and Rect Data ---")
final_table_headers = [] # Stores tuples of (list_of_cell_texts, rect_data_dict_for_that_row) output_data_entries = [] max_cols_in_html_table = 0 # Tracks max columns found in HTML part (header or data) html_table_rows = html_table_element.find_all('tr') is_first_tr_being_processed = True for tr_element in html_table_rows: current_row_cells_text = [] is_current_tr_a_header_row = False cells = tr_element.find_all(['th', 'td']) # Process SVG for the current tr_element, regardless of cells svg_element_in_row = None for svg_tag in tr_element.find_all('svg'): if svg_tag.get('width') == '648' and svg_tag.get('height') == '50': svg_element_in_row = svg_tag break current_row_rect_data = process_svg_for_rects(svg_element_in_row) if not cells: # Row has no <th> or <td> cells # Treat as a data row with no HTML cell content. # It still contributes its rect_data. output_data_entries.append(([], current_row_rect_data)) # If this is the first row and it's empty, final_table_headers remains [] # and max_cols_in_html_table remains 0. These might be updated by subsequent rows. is_first_tr_being_processed = False continue for cell_element in cells: if cell_element.name == 'th': is_current_tr_a_header_row = True cell_text = cell_element.get_text(separator=" ", strip=True) cell_text = cell_text.replace('\n', ' ').replace('\r', '').replace('\t', ' ') current_row_cells_text.append(cell_text) if is_first_tr_being_processed and is_current_tr_a_header_row: # This is the first <tr> and it's a header row final_table_headers = list(current_row_cells_text) max_cols_in_html_table = len(final_table_headers) # Rect data from this header row (current_row_rect_data) is captured but # not added to output_data_entries as this is a header, not a data row. else: # This is a data row. Either: # 1. Not the first <tr> processed. # 2. Is the first <tr> processed, but not a header row (no <th>). if is_first_tr_being_processed: # Implies: first <tr> AND not a header row # No explicit headers from <th>, so create empty headers based on this first data row's columns final_table_headers = [''] * len(current_row_cells_text) max_cols_in_html_table = len(current_row_cells_text) output_data_entries.append((current_row_cells_text, current_row_rect_data)) max_cols_in_html_table = max(max_cols_in_html_table, len(current_row_cells_text)) # Ensure final_table_headers (if derived from an earlier, narrower header/data row, # or if it was empty because the first row was data) # is padded to match the widest data row encountered so far. if len(final_table_headers) < max_cols_in_html_table: final_table_headers.extend([''] * (max_cols_in_html_table - len(final_table_headers))) is_first_tr_being_processed = False # Combine HTML headers with SVG data headers combined_tsv_headers = list(final_table_headers) combined_tsv_headers.extend(["width-35", "x-35", "width-10", "x-10"]) output_tsv_lines.append('\t'.join(combined_tsv_headers)) # Create output data rows from collected entries for html_cell_list, rect_data_dict in output_data_entries: current_output_row_parts = [] # Add HTML table cell data current_output_row_parts.extend(html_cell_list) # Pad HTML part if it's shorter than max_cols_in_html_table if len(html_cell_list) < max_cols_in_html_table: current_output_row_parts.extend([''] * (max_cols_in_html_table - len(html_cell_list))) # Add Rect data part w35_val = rect_data_dict.get('w35', "NONE") x35_val = rect_data_dict.get('x35', "NONE") w10_val = rect_data_dict.get('w10', "NONE") x10_val = rect_data_dict.get('x10', "NONE") current_output_row_parts.extend([str(w35_val), str(x35_val), str(w10_val), str(x10_val)]) output_tsv_lines.append('\t'.join(current_output_row_parts)) output_tsv_lines.append("") # Add an empty line as a separator if multiple tablesWrite output
if not output_tsv_lines: # This path should not be taken if html_tables had content, as headers etc. are added. # For safety, ensure an empty file is created if, for some reason, it's empty. with open(tsv_file_path, 'w', encoding='utf-8') as f: pass print(f"Created empty file '{tsv_file_path}' (no content generated).") else: # Remove trailing empty lines (which are used as separators between tables) while output_tsv_lines and output_tsv_lines[-1] == "": output_tsv_lines.pop()
try: with open(tsv_file_path, 'w', encoding='utf-8') as f: if output_tsv_lines: f.write('\n'.join(output_tsv_lines)) f.write('\n') # Add a final newline character, consistent with original print(f"Successfully wrote to '{tsv_file_path}'.") except Exception as e: print(f"Error writing output file: {e}", file=sys.stderr) sys.exit(1)Make it so it can select input and output file. Keep defaults the same
Assistant
import sys from bs4 import BeautifulSoup import argparse def process_svg_for_rects(svg_element): """ Process an SVG element to extract rect data associated with "-35" and "-10" text. Returns a dictionary with keys: w35, x35, w10, x10 """ rect_data = {'w35': "NONE", 'x35': "NONE", 'w10': "NONE", 'x10': "NONE"} if not svg_element: # Guard clause if None is passed return rect_data rects = svg_element.find_all('rect') for rect in rects: next_text = rect.find_next('text') if next_text: text_content = next_text.get_text(strip=True) if text_content == "-35": rect_data['w35'] = rect.get('width', "NONE") rect_data['x35'] = rect.get('x', "NONE") elif text_content == "-10": rect_data['w10'] = rect.get('width', "NONE") rect_data['x10'] = rect.get('x', "NONE") return rect_data def main(): parser = argparse.ArgumentParser(description="Extract data from HTML tables and associated SVGs to a TSV file.") parser.add_argument("html_file", nargs='?', default='table.html', help="Path to the input HTML file (default: table.html)") parser.add_argument("tsv_file", nargs='?', default='table.tsv', help="Path to the output TSV file (default: table.tsv)") args = parser.parse_args() html_file_path = args.html_file tsv_file_path = args.tsv_file try: with open(html_file_path, 'r', encoding='utf-8') as f: html_content = f.read() except FileNotFoundError: print(f"Error: Input file '{html_file_path}' not found.", file=sys.stderr) sys.exit(1) except Exception as e: print(f"Error reading file '{html_file_path}': {e}", file=sys.stderr) sys.exit(1) try: soup = BeautifulSoup(html_content, 'lxml') except Exception as e: print(f"Error parsing HTML: {e}", file=sys.stderr) sys.exit(1) output_tsv_lines = [] html_tables = soup.find_all('table') if not html_tables: print("No <table> elements found.", file=sys.stderr) sys.exit(1) # Consistent with original behavior for table_idx, html_table_element in enumerate(html_tables): if len(html_tables) > 1: output_tsv_lines.append(f"# --- Data from Table {table_idx + 1} ---") else: output_tsv_lines.append(f"# --- Combined Table and Rect Data ---") final_table_headers = [] # Stores tuples of (list_of_cell_texts, rect_data_dict_for_that_row) output_data_entries = [] max_cols_in_html_table = 0 # Tracks max columns found in HTML part (header or data) html_table_rows = html_table_element.find_all('tr') is_first_tr_being_processed = True for tr_element in html_table_rows: current_row_cells_text = [] is_current_tr_a_header_row = False cells = tr_element.find_all(['th', 'td']) # Process SVG for the current tr_element, regardless of cells svg_element_in_row = None for svg_tag in tr_element.find_all('svg'): if svg_tag.get('width') == '648' and svg_tag.get('height') == '50': svg_element_in_row = svg_tag break current_row_rect_data = process_svg_for_rects(svg_element_in_row) if not cells: # Row has no <th> or <td> cells # Treat as a data row with no HTML cell content. # It still contributes its rect_data. output_data_entries.append(([], current_row_rect_data)) # If this is the first row and it's empty, final_table_headers remains [] # and max_cols_in_html_table remains 0. These might be updated by subsequent rows. is_first_tr_being_processed = False continue for cell_element in cells: if cell_element.name == 'th': is_current_tr_a_header_row = True cell_text = cell_element.get_text(separator=" ", strip=True) cell_text = cell_text.replace('\n', ' ').replace('\r', '').replace('\t', ' ') current_row_cells_text.append(cell_text) if is_first_tr_being_processed and is_current_tr_a_header_row: # This is the first <tr> and it's a header row final_table_headers = list(current_row_cells_text) max_cols_in_html_table = len(final_table_headers) # Rect data from this header row (current_row_rect_data) is captured but # not added to output_data_entries as this is a header, not a data row. else: # This is a data row. Either: # 1. Not the first <tr> processed. # 2. Is the first <tr> processed, but not a header row (no <th>). if is_first_tr_being_processed: # Implies: first <tr> AND not a header row # No explicit headers from <th>, so create empty headers based on this first data row's columns final_table_headers = [''] * len(current_row_cells_text) max_cols_in_html_table = len(current_row_cells_text) output_data_entries.append((current_row_cells_text, current_row_rect_data)) max_cols_in_html_table = max(max_cols_in_html_table, len(current_row_cells_text)) # Ensure final_table_headers (if derived from an earlier, narrower header/data row, # or if it was empty because the first row was data) # is padded to match the widest data row encountered so far. if len(final_table_headers) < max_cols_in_html_table: final_table_headers.extend([''] * (max_cols_in_html_table - len(final_table_headers))) is_first_tr_being_processed = False # Combine HTML headers with SVG data headers combined_tsv_headers = list(final_table_headers) combined_tsv_headers.extend(["width-35", "x-35", "width-10", "x-10"]) output_tsv_lines.append('\t'.join(combined_tsv_headers)) # Create output data rows from collected entries for html_cell_list, rect_data_dict in output_data_entries: current_output_row_parts = [] # Add HTML table cell data current_output_row_parts.extend(html_cell_list) # Pad HTML part if it's shorter than max_cols_in_html_table if len(html_cell_list) < max_cols_in_html_table: current_output_row_parts.extend([''] * (max_cols_in_html_table - len(html_cell_list))) # Add Rect data part w35_val = rect_data_dict.get('w35', "NONE") x35_val = rect_data_dict.get('x35', "NONE") w10_val = rect_data_dict.get('w10', "NONE") x10_val = rect_data_dict.get('x10', "NONE") current_output_row_parts.extend([str(w35_val), str(x35_val), str(w10_val), str(x10_val)]) output_tsv_lines.append('\t'.join(current_output_row_parts)) output_tsv_lines.append("") # Add an empty line as a separator if multiple tables # Write output if not output_tsv_lines: # This path should not be taken if html_tables had content, as headers etc. are added. # For safety, ensure an empty file is created if, for some reason, it's empty. with open(tsv_file_path, 'w', encoding='utf-8') as f: pass print(f"Created empty file '{tsv_file_path}' (no content generated).") else: # Remove trailing empty lines (which are used as separators between tables) while output_tsv_lines and output_tsv_lines[-1] == "": output_tsv_lines.pop() try: with open(tsv_file_path, 'w', encoding='utf-8') as f: if output_tsv_lines: f.write('\n'.join(output_tsv_lines)) f.write('\n') # Add a final newline character, consistent with original print(f"Successfully wrote to '{tsv_file_path}'.") except Exception as e: print(f"Error writing output file: {e}", file=sys.stderr) sys.exit(1) if __name__ == "__main__": main()Key changes:
argparseIntegration: The code now usesargparseto handle command-line arguments. This allows the user to specify the input HTML file and output TSV file when running the script.- Default Values: The
argparsesetup includesdefaultvalues for bothhtml_fileandtsv_file, ensuring that the script works as before if no arguments are provided. main()Function: The core logic is encapsulated within amain()function, which is good practice for organization and testability.if __name__ == "__main__":: This ensures that themain()function is only called when the script is executed directly (not when it's imported as a module).- Clearer Argument Handling: The
args = parser.parse_args()line parses the command-line arguments, and the file paths are then accessed usingargs.html_fileandargs.tsv_file.
How to run:
-
Save: Save the code as a Python file (e.g.,
html_to_tsv.py). -
Run from the command line:
- Using defaults:
python html_to_tsv.py(This will readtable.htmland write totable.tsv) - Specifying files:
python html_to_tsv.py input.html output.tsv(Replaceinput.htmlandoutput.tsvwith your desired file names). - Help:
python html_to_tsv.py -h(Displays help information about the script and its arguments).
- Using defaults:
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.