#!/usr/bin/env python3 import argparse import os import string import dataclasses @dataclasses.dataclass class Section: name: str rows: list = dataclasses.field(default_factory=lambda: [{}]) order: list = dataclasses.field(default_factory=list) ref: str = "" @dataclasses.dataclass class Directive: line_number: int type: str content: str HTML_TEMPLATE = string.Template(""" $title

$intro

$body
""") def parse_table(lines: list[str]) -> list[Section]: previus_column_type = None sections = [] current_section = None # Each nonblank matches this regular expression /(\S*)\s(.*)/ # where first capture is type of column (essentialy column id) and # second capture is given cell content. Where column type repeats not in row # we have new row. for line in lines: if not line: continue sep = line.find(' ') column_type, cell_content = line[:sep].strip(), line[sep:] if column_type == "SECTION": sections.append(Section(name=cell_content)) current_section = sections[-1] continue assert current_section is not None, "Define SECTION before table entries" if column_type not in current_section.order: current_section.order.append(column_type) if previus_column_type != column_type and column_type in current_section.rows[-1]: current_section.rows.append({}) cell = current_section.rows[-1].get(column_type, []) cell.append(cell_content) current_section.rows[-1][column_type] = cell previus_column_type = column_type # Eliminate common whitespace prefix in given column type (not all whitespace # prefix since examples in Python may have significant whitespace) for section in sections: for row in section.rows: for cell in row.values(): prefix_whitespace = min(len(s) - len(s.lstrip()) for s in cell) for i, s in enumerate(cell): cell[i] = s[prefix_whitespace:] return sections def compile_template(*, template_path: str, target_path: str): # Read template file and separate it into lines with open(template_path) as f: template = [line.strip() for line in f.readlines()] directives = [] for i, line in enumerate(template): s = line.split() if s and s[0] in ("BEGIN", "END", "TITLE"): directives.append(Directive(i, s[0], line[len(s[0]):].strip())) title, css_source, table_source, intro_source = 4 * [None] for directive in directives: if directive.type == "TITLE": title = directive.content continue if directive.type == "BEGIN": start_line = directive.line_number for end in directives: if end.type == "END" and end.content == directive.content: end_line = end.line_number break else: assert False, "Begin without matching end" span = template[start_line+1:end_line] if directive.content == "CSS": css_source = span elif directive.content == "TABLE": table_source = span elif directive.content == "INTRO": intro_source = span assert css_source is not None assert table_source is not None assert intro_source is not None assert title is not None sections = parse_table(lines=table_source) for section in sections: section.ref = section.name.strip().replace(" ", "_") nav = "" for section in sections: nav += f"
  • {section.name}
  • " body = "" for section in sections: table = f"

    {section.name}

    \n" for i, row in enumerate(section.rows): if i == 0: line = ["" for k in section.order] else: line = [] for column in section.order: val = row.get(column, []) line.append('') table += "\n" + '\n'.join(line) + "\n\n" table += "\n
    " + ' '.join(row[k]) + "
    ' + '
    '.join(val) + '

    \n" body += table final = HTML_TEMPLATE.substitute({ "body": body, "css": "\n".join(css_source), "intro": "\n".join(intro_source), "nav": nav, "title": title, }) with open(target_path, "w") as f: f.write(final) def main(): parser = argparse.ArgumentParser(description="Build language comparison chart") parser.add_argument(nargs='+', metavar="TEMPLATE", dest="templates", help="Template file that will be converted to HTML page") args = parser.parse_args() for template in args.templates: dst, _ = os.path.splitext(template) dst += ".html" compile_template(template_path=template, target_path=dst) if __name__ == "__main__": main()