matrix/weechat-matrix
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Don't process markdown in URL's

Browse source
This commit is contained in:
Tom Smeding 2020-06-07 19:32:26 +02:00 committed by Denis Kasak
parent 172826ea6b
commit d37e08cce2
1 changed files with 36 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

44
matrix/colors.py
View file

@ -88,11 +88,21 @@ class Formatted(object):
substrings = [] # type: List[FormattedString] substrings = [] # type: List[FormattedString]
attributes = DEFAULT_ATTRIBUTES.copy() attributes = DEFAULT_ATTRIBUTES.copy()
# Disallow backticks in URL's so that code blocks are unaffected by the
# URL handling
url_regex = r"\b[a-z]+://[^\s`]+"
# Escaped things are not markdown delimiters, so substitute them away # Escaped things are not markdown delimiters, so substitute them away
# when (quickly) looking for the last delimiters in the line. Note that # when (quickly) looking for the last delimiters in the line.
# the replacement needs to be the same length as the original for the # Additionally, URL's are ignored for the purposes of markdown
# indices to be correct. # delimiters.
escaped_masked = re.sub(r"\\[\\*_`]", "aa", line) # Note that the replacement needs to be the same length as the original
# for the indices to be correct.
escaped_masked = re.sub(
r"\\[\\*_`]|(?:" + url_regex + ")",
lambda m: "a" * len(m[0]),
line
)
def last_match_index(regex, offset_in_match): def last_match_index(regex, offset_in_match):
matches = list(re.finditer(regex, escaped_masked)) matches = list(re.finditer(regex, escaped_masked))
@ -140,13 +150,30 @@ class Formatted(object):
escapable_chars = wrapper_init_chars.copy() escapable_chars = wrapper_init_chars.copy()
escapable_chars.add("\\") escapable_chars.add("\\")
# Collect URL spans
url_spans = [m.span() for m in re.finditer(url_regex, line)]
url_spans.reverse() # we'll be popping from the end
# Whether we are currently in a URL
in_url = False
i = 0 i = 0
while i < len(line): while i < len(line):
# Update the 'in_url' flag. The first condition is not a while loop
# because URL's must contain '://', ensuring that we will not skip
# 2 URL's in one iteration.
if url_spans and i >= url_spans[-1][1]:
in_url = False
url_spans.pop()
if url_spans and i >= url_spans[-1][0]:
in_url = True
# Markdown escape # Markdown escape
if i + 1 < len(line) and line[i] == "\\" \ if i + 1 < len(line) and line[i] == "\\" \
and (line[i + 1] in escapable_chars and (line[i + 1] in escapable_chars
if not attributes["code"] if not attributes["code"]
else line[i + 1] == "`"): else line[i + 1] == "`") \
and not in_url:
text += line[i + 1] text += line[i + 1]
i = i + 2 i = i + 2
@ -208,7 +235,7 @@ class Formatted(object):
attributes["bgcolor"] = None attributes["bgcolor"] = None
# Markdown wrapper (emphasis/bold/code) # Markdown wrapper (emphasis/bold/code)
elif line[i] in wrapper_init_chars: elif line[i] in wrapper_init_chars and not in_url:
for l in range(wrapper_max_len, 0, -1): for l in range(wrapper_max_len, 0, -1):
if i + l <= len(line) and line[i : i + l] in wrappers: if i + l <= len(line) and line[i : i + l] in wrappers:
descriptor = wrappers[line[i : i + l]] descriptor = wrappers[line[i : i + l]]
@ -255,8 +282,9 @@ class Formatted(object):
break break
else: else:
# No wrapper matched here (NOTE: cannot happen if "*" and # No wrapper matched here (NOTE: cannot happen since all
# "_" are both in wrappers, but for completeness' sake) # wrapper prefixes are also wrappers, but for completeness'
# sake)
text = text + line[i] text = text + line[i]
i = i + 1 i = i + 1