| |
| |
|
|
| """Decode known rich inline commands for the legacy tabular fallback. |
| |
| Used only after strict rich-grid parsing fails. This preserves unknown commands |
| and malformed arguments; it never invents cell contents or repairs table spans. |
| """ |
| import html |
| import re |
| try: |
| from .hybrid_cleanup import decode_styles |
| except ImportError: |
| from hybrid_cleanup import decode_styles |
|
|
| try: |
| from .rich_table_reader import TAGS, LITERALS, argument |
| except ImportError: |
| from rich_table_reader import TAGS, LITERALS, argument |
|
|
|
|
| def decode_known_inline(text, _depth=0): |
| """Decode balanced, recognized inline arguments without exposing delimiters.""" |
| if _depth > 64: |
| return text |
| out = [] |
| i = 0 |
| known = set(TAGS) | set(LITERALS) | {'thead', 'linebreak', 'mbox', 'hl', 'colorbox'} |
| while i < len(text): |
| |
| |
| if text[i] == '\\' and i + 1 < len(text) and not text[i + 1].isalpha(): |
| literal = text[i + 1] |
| out.append('\\&#' + str(ord(literal)) + ';' if literal in '&%$#_*{}' else text[i:i + 2]) |
| i += 2 |
| continue |
| match = re.match(r'\\([A-Za-z]+)\{', text[i:]) |
| if not match or match[1] not in known: |
| out.append(text[i]) |
| i += 1 |
| continue |
| command = match[1] |
| try: |
| body, end = argument(text, i + len(command) + 1) |
| except ValueError: |
| out.append(text[i]) |
| i += 1 |
| continue |
| if command in {'hl', 'colorbox'}: |
| try: |
| from .semantic_styles import mark_open |
| except ImportError: |
| from semantic_styles import mark_open |
| from bs4 import BeautifulSoup |
| node = BeautifulSoup('<mark></mark>', 'html.parser').mark |
| if command == 'colorbox': |
| node['background'] = body |
| try:body, end = argument(text, end) |
| except ValueError: |
| out.append(text[i:end]); i = end; continue |
| out.append(mark_open(node) + decode_known_inline(body, _depth + 1) + '</mark>') |
| elif command in TAGS or command == 'thead': |
| inner = decode_known_inline(body, _depth + 1) |
| tag = 'nvp-header-cell' if command == 'thead' else TAGS[command] |
| out.append('<' + tag + '>' + inner + '</' + tag + '>') |
| elif command in LITERALS and not body: |
| |
| out.append('\\&#' + str(ord(LITERALS[command])) + ';') |
| elif command == 'linebreak' and not body: |
| out.append('<nvp-line-break></nvp-line-break>') |
| elif command == 'mbox' and not body: |
| pass |
| else: |
| out.append(text[i:end]) |
| i = end |
| return decode_styles(''.join(out)) |
|
|
|
|
| def restore_header_cells(html_text): |
| from bs4 import BeautifulSoup |
| |
| |
| html_text = html_text.replace('<nvp-line-break></nvp-line-break>', '<br>') |
| if '<nvp-header-cell>' not in html_text: |
| return html_text |
| soup = BeautifulSoup(html_text, 'html.parser') |
| for cell in soup.find_all(['td', 'th']): |
| markers = cell.find_all('nvp-header-cell') |
| if markers: |
| cell.name = 'th' |
| for marker in markers: |
| marker.unwrap() |
| return str(soup) |
|
|
