Cyclic0007/python-markdownify
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

use a conversion function cache to improve runtime (#196)

Browse Source 
Signed-off-by: chrispy <chrispy@synopsys.com>
This commit is contained in:
Chris Papademetrious
2025-02-24 11:48:40 -05:00
committed by GitHub
parent 24977fd192
commit 6984dca7ab
1 changed files with 31 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

48
markdownify/__init__.py
View File

@@ -4,12 +4,11 @@ import re
import six import six
re_convert_heading = re.compile(r'convert_h(\d+)')
re_line_with_content = re.compile(r'^(.*)', flags=re.MULTILINE) re_line_with_content = re.compile(r'^(.*)', flags=re.MULTILINE)
re_whitespace = re.compile(r'[\t ]+') re_whitespace = re.compile(r'[\t ]+')
re_all_whitespace = re.compile(r'[\t \r\n]+') re_all_whitespace = re.compile(r'[\t \r\n]+')
re_newline_whitespace = re.compile(r'[\t \r\n]*[\r\n][\t \r\n]*') re_newline_whitespace = re.compile(r'[\t \r\n]*[\r\n][\t \r\n]*')
re_html_heading = re.compile(r'h[1-6]') re_html_heading = re.compile(r'h(\d+)')
# extract (leading_nl, content, trailing_nl) from a string # extract (leading_nl, content, trailing_nl) from a string
# (functionally equivalent to r'^(\n*)(.*?)(\n*)$', but greedy is faster than reluctant here) # (functionally equivalent to r'^(\n*)(.*?)(\n*)$', but greedy is faster than reluctant here)
@@ -165,6 +164,9 @@ class MarkdownConverter(object):
raise ValueError('You may specify either tags to strip or tags to' raise ValueError('You may specify either tags to strip or tags to'
' convert, but not both.') ' convert, but not both.')
# Initialize the conversion function cache
self.convert_fn_cache = {}
def convert(self, html): def convert(self, html):
soup = BeautifulSoup(html, 'html.parser') soup = BeautifulSoup(html, 'html.parser')
return self.convert_soup(soup) return self.convert_soup(soup)
@@ -266,9 +268,8 @@ class MarkdownConverter(object):
text = ''.join(child_strings) text = ''.join(child_strings)
# apply this tag's final conversion function # apply this tag's final conversion function
convert_fn_name = "convert_%s" % re.sub(r"[\[\]:-]", "_", node.name) convert_fn = self.get_conv_fn_cached(node.name)
convert_fn = getattr(self, convert_fn_name, None) if convert_fn is not None:
if convert_fn and self.should_convert_tag(node.name):
text = convert_fn(node, text, parent_tags=parent_tags) text = convert_fn(node, text, parent_tags=parent_tags)
return text return text
@@ -321,23 +322,36 @@ class MarkdownConverter(object):
return text return text
def __getattr__(self, attr): def get_conv_fn_cached(self, tag_name):
# Handle headings """Given a tag name, return the conversion function using the cache."""
m = re_convert_heading.match(attr) # If conversion function is not in cache, add it
if m: if tag_name not in self.convert_fn_cache:
n = int(m.group(1)) self.convert_fn_cache[tag_name] = self.get_conv_fn(tag_name)
def convert_tag(el, text, parent_tags): # Return the cached entry
return self._convert_hn(n, el, text, parent_tags) return self.convert_fn_cache[tag_name]
convert_tag.__name__ = 'convert_h%s' % n def get_conv_fn(self, tag_name):
setattr(self, convert_tag.__name__, convert_tag) """Given a tag name, find and return the conversion function."""
return convert_tag tag_name = tag_name.lower()
raise AttributeError(attr) # Handle strip/convert exclusion options
if not self.should_convert_tag(tag_name):
return None
# Handle headings with _convert_hn() function
match = re_html_heading.match(tag_name)
if match:
n = int(match.group(1))
return lambda el, text, parent_tags: self._convert_hn(n, el, text, parent_tags)
# For other tags, look up their conversion function by tag name
convert_fn_name = "convert_%s" % re.sub(r"[\[\]:-]", "_", tag_name)
convert_fn = getattr(self, convert_fn_name, None)
return convert_fn
def should_convert_tag(self, tag): def should_convert_tag(self, tag):
tag = tag.lower() """Given a tag name, return whether to convert based on strip/convert options."""
strip = self.options['strip'] strip = self.options['strip']
convert = self.options['convert'] convert = self.options['convert']
if strip is not None: if strip is not None: