HTML-style formatting #582

Activity

1. santerioksanen

   added

   Request for enhancement

   on Jan 28, 2025
2. PHvL commented on Sep 5, 2025

   

   PHvL

   on Sep 5, 2025

   More actions

   Since we prepare the context for the template as pydantic schema anyway, we've added a BeforeValidator doing the conversion of HTML tags.

   from typing import Annotated
   from html.parser import HTMLParser
   import html.entities
   import re
   
   from docxtpl import RichText
   from pydantic import BaseModel, ConfigDict, BeforeValidator
   
   
   TEXT_PROPS_BOOL = {
   	'sub': 'subscript',
   	'sup': 'superscript',
   	'i': 'italic',
   	'b': 'bold',
   }
   
   
   class HTML_RICHTEXT(HTMLParser):
   	def __init__(self):
   		super().__init__()
   		self.reset()
   		self.fed = RichText('')
   		self.text_prop = {}
   
   	def handle_data(self, d):
   		self.fed.add(d, **self.text_prop)
   
   	def handle_starttag(self, tag, attrs):
   		if tag in TEXT_PROPS_BOOL:
   			self.text_prop[TEXT_PROPS_BOOL[tag]] = True
   
   	def handle_startendtag(self, tag, attrs):
   		if tag == 'br':
   			self.fed.add('\n')
   
   	def handle_endtag(self, tag):
   		if tag in ['h1', 'h2', 'h3', 'h4', 'h5', 'h6', 'p']:
   			self.fed.add('\n')
   		elif tag in TEXT_PROPS_BOOL:
   			if TEXT_PROPS_BOOL[tag] in self.text_prop:
   				del self.text_prop[TEXT_PROPS_BOOL[tag]]
   
   	def handle_charref(self, name):
   		self.fed.add(chr(int(name[1:])))
   
   	def handle_entityref(self, name):
   		self.fed.add(chr(html.entities.name2codepoint[name]))
   
   	def get_data(self):
   		return self.fed
   
   
   def str_to_richtext(html: str|RichText) -> RichText:
   	if isinstance(html, RichText):
   		return html
   	if '<' in html:
   		s = HTML_RICHTEXT()
   		html = re.sub(r'[\r\n]+', '\n', html)
   		html = re.sub(r'\s+', ' ', html)
   		s.feed(html)
   		return s.get_data()
   	else:
   		return RichText(html)
   
   
   class Letter(BaseModel):
   	model_config = ConfigDict(arbitrary_types_allowed=True)
   	body: Annotated[RichText, BeforeValidator(str_to_richtext)]
   
   context = Letter(body='I am <b>bolded</b> text.').model_dump()