format_links: refactor; support multiple punctuation

This commit is contained in:
Adeodato Simó 2023-10-09 21:41:22 -03:00
parent 17d741039c
commit 294788aa1a
No known key found for this signature in database
GPG key ID: CDF447845F1A986F
2 changed files with 28 additions and 47 deletions

View file

@ -427,6 +427,10 @@ http://www.fish.com/"""
views.status.format_links(f"{url}."),
f'<a href="{url}">www.fish.com/</a>.',
)
self.assertEqual(
views.status.format_links(f"{url}!?!"),
f'<a href="{url}">www.fish.com/</a>!?!',
)
def test_format_links_punctuation_parens(self, *_):
"""ignore trailing punctuation and brackets combined"""

View file

@ -1,7 +1,6 @@
""" what are we here for if not for posting """
import re
import logging
from urllib.parse import urlparse
from django.contrib.auth.decorators import login_required
from django.core.validators import URLValidator
@ -297,67 +296,45 @@ def find_or_create_hashtags(content):
def format_links(content):
"""detect and format links"""
validator = URLValidator()
formatted_content = ""
validator = URLValidator(["http", "https"])
schema_re = re.compile(r"\bhttps?://")
split_content = re.split(r"(\s+)", content)
for potential_link in split_content:
if not potential_link:
for i, potential_link in enumerate(split_content):
if not schema_re.search(potential_link):
continue
# FIXME: allow for multiple punctuation characters, e.g. `...` and `!?`.
ends_with_punctuation = _ends_with_punctuation(potential_link)
if ends_with_punctuation:
punctuation_glyph = potential_link[-1]
potential_link = potential_link[0:-1]
wrapped = _wrapped(potential_link)
if wrapped:
wrapper_close = potential_link[-1]
formatted_content += potential_link[0]
potential_link = potential_link[1:-1]
# Strip surrounding brackets and trailing punctuation.
prefix, potential_link, suffix = _unwrap(potential_link)
try:
# raises an error on anything that's not a valid link
validator(potential_link)
# use everything but the scheme in the presentation of the link
url = urlparse(potential_link)
link = url.netloc + url.path + url.params
if url.query != "":
link += "?" + url.query
if url.fragment != "":
link += "#" + url.fragment
formatted_content += f'<a href="{potential_link}">{link}</a>'
link = schema_re.sub("", potential_link)
split_content[i] = f'{prefix}<a href="{potential_link}">{link}</a>{suffix}'
except (ValidationError, UnicodeError):
formatted_content += potential_link
pass
if wrapped:
formatted_content += wrapper_close
if ends_with_punctuation:
formatted_content += punctuation_glyph
return formatted_content
return "".join(split_content)
def _wrapped(text):
"""check if a line of text is wrapped"""
wrappers = ["()", "[]", "{}"]
for wrapper in wrappers:
def _unwrap(text):
"""split surrounding brackets and trailing punctuation from a string of text"""
punct = re.compile(r'([.,;:!?"’”»]+)\Z')
prefix = suffix = ""
if punct.search(text):
# Move punctuation to suffix segment.
text, suffix, _ = punct.split(text)
for wrapper in ("()", "[]", "{}"):
if text[0] == wrapper[0] and text[-1] == wrapper[-1]:
return True
return False
# Split out wrapping chars.
suffix = text[-1] + suffix
prefix, text = text[:1], text[1:-1]
def _ends_with_punctuation(text):
"""check if a line of text ends with a punctuation glyph"""
glyphs = [".", ",", ";", ":", "!", "?", "", "", '"', "»"]
for glyph in glyphs:
if text[-1] == glyph:
return True
return False
return prefix, text, suffix
def to_markdown(content):