Skip to content

Commit

Permalink
Merge pull request #870 from Oisov/linter_v4.2
Browse files Browse the repository at this point in the history 
Linter v4.2 - More fixes to codeblock formating
  • Loading branch information
@Oisov
Oisov authored Aug 12, 2018
2 parents c5e3239 + edaa551 commit b8850ef
Showing 1 changed file with 67 additions and 31 deletions.
98 changes: 67 additions & 31 deletions linter/linter_lib/linter_auto_md.py
View file
Original file line number Diff line number Diff line change
@@ -1,5 +1,6 @@
from linter_defaults import *
import textwrap
import collections

# REGEX_FIND_YML = re.compile(r"(?<=^---\s)[\s\S]+?(?=\r?\n---)", re.DOTALL)
REGEX_FIND_YML = re.compile(r"^---[\s\S]*?---", re.DOTALL)
Expand Down Expand Up @@ -248,6 +249,7 @@ def append_if_new(md_data_, line, current, nxt=''):

for i, line in enumerate(md_data_lst):

orig_line = line
if line:
line = fix_leading_trailing_whitespace(line)

Expand Down Expand Up @@ -287,30 +289,38 @@ def append_if_new(md_data_, line, current, nxt=''):
if not (md_data_['last'] == 'html'):

is_codeblock = md_data_['last'] == 'codeblocks'
if is_codeblock:
codeblock_start_end = re.search('^ *((`{3,}([^`\n]+)`{3,})|(([^`\n\r]*)(`{3,})(.*)))', line)
if not is_codeblock:
# This regex checks for the START of a codeblock or a oneliner
codeblock_start = re.search('^( *`{1,}[^\r\n`]+`{1,} *$)|([^\r\n`]+(`{2,}([^\r\n`]+)`{2,}))|( *(`{3,}).*)', line)
if codeblock_start:
oneliner = codeblock_start.group(1)
inline_oneliner = codeblock_start.group(3)
start_multiline_codeblock = codeblock_start.group(5)
if oneliner:
md_data_ = append_if_new(md_data_, line, 'codeblocks', 'textblocks')
elif inline_oneliner:
remove_mutliple_backticks = re.sub(r'`{2,}(.*?)`{2,}', r'`\1`', line)
md_data_ = append_if_new(md_data_, remove_mutliple_backticks, 'textblocks')
elif start_multiline_codeblock:
backticks_start = codeblock_start.group(6)
md_data_ = append_if_new(md_data_, line, 'codeblocks')
continue
else:
codeblock_start_end = re.search('^ *((`{3,}([^`\n]+)`{3,})|(( *)(`{3,})(.*)))', line)
# The regex above checks both for starting and ending backticks
if codeblock_start_end or is_codeblock:
md_data_ = append_if_new(md_data_, line, 'codeblocks')
if codeblock_start_end:
current_backticks = codeblock_start_end.group(6)
is_oneliner = True if codeblock_start_end.group(2) else False
if is_oneliner or (is_codeblock and current_backticks == backticks_start):
# print(line)
backticks_start = ''
md_data_ = append_if_new(md_data_, orig_line, 'codeblocks')
# This regex checks for the END of a codeblock or a oneliner
codeblock_end = re.search('[^`\n\r]*(`{3,}).*', line)
if codeblock_end:
current_backticks = codeblock_end.group(1)
if current_backticks == backticks_start:
md_data_ = append_if_new(md_data_, '', 'textblocks')
elif not backticks_start:
backticks_start = current_backticks
continue


# If we are in a html block, use regex to check for end of block, else append
if md_data_['last'] == 'html':
# Searches for expressions of the form: </hide>
match = re.search(r'( *< */ *({}).*?> *$)'.format(html_key), line)
md_data_ = append_if_new(md_data_, line, 'html', 'textblocks' if match else '')
md_data_ = append_if_new(md_data_, orig_line, 'html', 'textblocks' if match else '')
continue

# This regex searches for html
Expand Down Expand Up @@ -501,7 +511,7 @@ def is_list_symbol(line, i, lines, number_of_lines):
return False
symbol = lst.group(1)
spacing = lst.group(2)
if symbol == '*' and re.search(r'^\*.*\*', line):
if symbol == '*' and len(re.findall(r'\*', line)) % 2 == 0:
return False
elif symbol == '-':
if line[1].isdigit():
Expand Down Expand Up @@ -586,11 +596,23 @@ def fix_md_text(text):
return '\n\n'.join(text_['text_new'])


def fix_codeblocks(codeblock, has_looped = False):
def fix_codeblocks(codeblock):
codelines = codeblock.split('\n')
is_oneliner = len(codelines) == 1

# print(codeblock)
if is_oneliner:
line = re.search(r'^( *)(`{1,})(((\w* )(.*?))|(.*?))`{1,}', codeblock)
indent = line.group(1)
backticks = '`'*max(len(line.group(2)), 3)
language = line.group(5)
content = line.group(6 if language else 7).strip()
return fix_codeblocks('{}{}{}\n{}{}\n{}{}'.format(indent,
backticks,
language.strip() if language else '',
indent,
content, indent, backticks))
first_line, last_line = codelines[0], codelines[-1]
rem_lines = fix_leading_trailing_newlines('\n'.join(codelines[1:-1]))

# This makes sure that the codeblocks has 3 backticks or more
first_line = re.sub(r'^( *)`` *(\w+|$)', r'\1```\2', first_line)
Expand All @@ -600,7 +622,7 @@ def fix_codeblocks(codeblock, has_looped = False):
# This block might be removed if many new languages emerges
# The supported languages can be changed in the PROGRAMMING_LANGUAGE
# variable. Located in 'linter_defaults'
has_language = re.search('( *)(`{3,}) *(\w*)(.*)', first_line)
has_language = re.search('( *)(`{3,}) *(\w*)([^`\r\n]*)', first_line)
indent = has_language.group(1)
backticks = has_language.group(2)
if has_language.group(3).strip():
Expand All @@ -612,7 +634,7 @@ def fix_codeblocks(codeblock, has_looped = False):
first_line)
first_line = '{}{}{}'.format(indent, backticks, language.strip())
if remaining:
first_line += '\n' + remaining.strip()
first_line += '\n' + indent + remaining.strip()

''' The following if sentence does the following transformation:
Expand All @@ -628,15 +650,10 @@ def fix_codeblocks(codeblock, has_looped = False):
code_text = has_trailing_text.group(1)
trailing = has_trailing_text.group(3).strip()
last_line = code_text + '\n' + indent + backticks if code_text.strip() else indent + backticks
last_line += '\n\n' + trailing if trailing else ''
last_line += '\n\n' + indent + trailing if trailing else ''

codelines[0] = first_line
codelines[-1] = last_line
codeblock = '{}\n{}\n{}'.format(first_line, rem_lines, last_line)

codeblock = '\n'.join(codelines)

if is_oneliner and not has_looped:
return fix_codeblocks(codeblock, True)
return codeblock


Expand All @@ -655,11 +672,30 @@ def add_newline_end_of_file(md_data):
return md_data


def is_codeblocks_closed(md_string):
''' The first line searches through the entire file for all lines starting with 3 ` or more
The second line count the number of occurences of each number of backticks
The third line returns True only if it finds an odd number of backticks (meaning it is not closed)
'''

# backticks = ['````', '```', ..., '```', '```']
backticks = [match.strip() for match in re.findall(r'`{3,}', md_string)]
# count_of_all_backticks = Counter({'```': 29, '````': 1})
count_of_all_backticks = collections.Counter(backticks)
return not any(backtick_count % 2 != 0 for backtick_count in count_of_all_backticks.values())


def update_md(md_data, filepath):
# Remove trailing whitespace if it exists
md_data = [line.rstrip() for line in md_data]
md_data_lst = [line.rstrip() for line in md_data]

md_string = '\n'.join(md_data_lst)
if not is_codeblocks_closed(md_string):
print("ERROR: Unbalanced codeblocks in:\n {} \n Please fix this before formating".format(color_word(filepath, MAIN_2_CLR)))
input("")
return md_string

md_data_ = split_md(md_data)
md_data_ = split_md(md_data_lst)

for i in md_data_['html']:
html = md_data_['lst'][i]
Expand All @@ -670,9 +706,9 @@ def update_md(md_data, filepath):

for i in md_data_['codeblocks']:
codeblock = md_data_['lst'][i]
# print(codeblock)
# print()
md_data_['lst'][i] = fix_codeblocks(codeblock)
# print(md_data_['lst'][i])
# print('===================================')

for i in md_data_['textblocks']:
text = md_data_['lst'][i]
Expand Down

0 comments on commit b8850ef

Please sign in to comment.