ditkrg/todo-to-issue-action
4
0
Fork 0
mirror of https://github.com/ditkrg/todo-to-issue-action.git synced 2026-01-22 22:06:43 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix: get correct line numbering of old and new lines

Browse Source 
track context of old and new lines separately to
get proper line numbers relative to old and
new version of a file

Closes GitHub #236
This commit is contained in:
Robert Alonso 2024-11-12 03:26:05 +00:00
parent f54fbebd3b
commit 46218738c9
2 changed files with 103 additions and 85 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

82
TodoParser.py
View File

@ -11,9 +11,7 @@ import itertools
class TodoParser(object): class TodoParser(object):
"""Parser for extracting information from a given diff file.""" """Parser for extracting information from a given diff file."""
FILE_HUNK_PATTERN = r'(?<=diff)(.*?)(?=diff\s--git\s)' FILE_HUNK_PATTERN = r'(?<=diff)(.*?)(?=diff\s--git\s)'
HEADER_PATTERN = r'(?<=--git).*?(?=$\n(index|new|deleted))' HEADERS_PATTERN = re.compile(r'(?<=--git) a/(.*?) b/(.*?)$\n(?=((new|deleted).*?$\n)?index ([0-9a-f]+)\.\.([0-9a-f]+))', re.MULTILINE)
LINE_PATTERN = r'^.*$'
FILENAME_PATTERN = re.compile(r'(?<=a/).+?(?=\sb/)')
LINE_NUMBERS_PATTERN = re.compile(r'@@[\d\s,\-+]*\s@@.*') LINE_NUMBERS_PATTERN = re.compile(r'@@[\d\s,\-+]*\s@@.*')
LINE_NUMBERS_INNER_PATTERN = re.compile(r'@@[\d\s,\-+]*\s@@') LINE_NUMBERS_INNER_PATTERN = re.compile(r'@@[\d\s,\-+]*\s@@')
ADDITION_PATTERN = re.compile(r'(?<=^\+).*') ADDITION_PATTERN = re.compile(r'(?<=^\+).*')
@ -154,15 +152,10 @@ class TodoParser(object):
# Iterate through each section extracted above. # Iterate through each section extracted above.
for hunk in extracted_file_hunks: for hunk in extracted_file_hunks:
# Extract the file information so we can figure out the Markdown language and comment syntax. # Extract the file information so we can figure out the Markdown language and comment syntax.
header_search = re.search(self.HEADER_PATTERN, hunk, re.MULTILINE) headers = self.HEADERS_PATTERN.search(hunk)
if not header_search: if not headers:
continue continue
files = header_search.group(0) curr_file = headers.group(2)
filename_search = re.search(self.FILENAME_PATTERN, files)
if not filename_search:
continue
curr_file = filename_search.group(0)
if self._should_ignore(curr_file): if self._should_ignore(curr_file):
continue continue
curr_markers, curr_markdown_language = self._get_file_details(curr_file) curr_markers, curr_markdown_language = self._get_file_details(curr_file)
@ -175,15 +168,18 @@ class TodoParser(object):
for i, line_numbers in enumerate(line_numbers_iterator): for i, line_numbers in enumerate(line_numbers_iterator):
line_numbers_inner_search = re.search(self.LINE_NUMBERS_INNER_PATTERN, line_numbers.group(0)) line_numbers_inner_search = re.search(self.LINE_NUMBERS_INNER_PATTERN, line_numbers.group(0))
line_numbers_str = line_numbers_inner_search.group(0).strip('@@ -') line_numbers_str = line_numbers_inner_search.group(0).strip('@@ -')
start_line = line_numbers_str.split(' ')[1].strip('+') deleted_start_line = line_numbers_str.split(' ')[0]
start_line = int(start_line.split(',')[0]) deleted_start_line = int(deleted_start_line.split(',')[0])
added_start_line = line_numbers_str.split(' ')[1].strip('+')
added_start_line = int(added_start_line.split(',')[0])
# Put this information into a temporary dict for simplicity. # Put this information into a temporary dict for simplicity.
block = { block = {
'file': curr_file, 'file': curr_file,
'markers': curr_markers, 'markers': curr_markers,
'markdown_language': curr_markdown_language, 'markdown_language': curr_markdown_language,
'start_line': start_line, 'deleted_start_line': deleted_start_line,
'added_start_line': added_start_line,
'hunk': hunk, 'hunk': hunk,
'hunk_start': line_numbers.end(), 'hunk_start': line_numbers.end(),
'hunk_end': None 'hunk_end': None
@ -209,8 +205,25 @@ class TodoParser(object):
# Now for each code block, check for comments, then those comments for TODOs. # Now for each code block, check for comments, then those comments for TODOs.
for block in code_blocks: for block in code_blocks:
# convert hunk string into newline-separated list (excluding first element which is always null and not actually first line of hunk) # for both the set of deleted lines and set of new lines, convert hunk string into
bl=block['hunk'].split('\n')[1:] # newline-separated list (excluding first element which is always null and not
# actually first line of hunk)
old=[]
new=[]
for line in block['hunk'].split('\n')[1:]:
if line: # if not empty
match line[0]:
case '-':
old.append(line)
case '+':
new.append(line)
case _:
if line != '\\ No newline at end of file':
old.append(line)
new.append(line)
elif line != '\\ No newline at end of file':
old.append(line)
new.append(line)
for marker in block['markers']: for marker in block['markers']:
# initialize list # initialize list
@ -250,13 +263,16 @@ class TodoParser(object):
# create regex object to search for comments # create regex object to search for comments
compiled_pattern=re.compile(comment_pattern) compiled_pattern=re.compile(comment_pattern)
# analyze the set of old lines and new lines separately, so that we don't, for example,
# accidentally treat deleted lines as if they were being added in this diff
for block_lines in [old, new]:
# for each element of list, enumerate it and if value is a regex match, include it in list that is returned, # for each element of list, enumerate it and if value is a regex match, include it in list that is returned,
# where each element of the list is a dictionary that is the start end end lines of the match (relative to # where each element of the list is a dictionary that is the start and end lines of the match (relative to
# start of the hunk) and the matching string itself # start of the hunk) and the matching string itself
comments_and_positions = [{'start': i, 'end': i, 'comment': x} for i, x in enumerate(bl) if compiled_pattern.search(x)] comments_and_positions = [{'start': i, 'end': i, 'comment': x} for i, x in enumerate(block_lines) if compiled_pattern.search(x)]
if len(comments_and_positions) > 0: if len(comments_and_positions) > 0:
# create filtered list which consolidates contiguous lines # append filtered list which consolidates contiguous lines
contiguous_comments_and_positions=[comments_and_positions[0]] contiguous_comments_and_positions.append(comments_and_positions[0])
for j, x in enumerate(comments_and_positions[1:]): for j, x in enumerate(comments_and_positions[1:]):
if x['start'] == (comments_and_positions[j]['end'] + 1): if x['start'] == (comments_and_positions[j]['end'] + 1):
contiguous_comments_and_positions[-1]['end']+=1 contiguous_comments_and_positions[-1]['end']+=1
@ -270,9 +286,14 @@ class TodoParser(object):
# compile above pattern # compile above pattern
compiled_pattern = re.compile(pattern, re.DOTALL) compiled_pattern = re.compile(pattern, re.DOTALL)
# analyze the set of old lines and new lines separately, so that we don't, for example,
# accidentally treat deleted lines as if they were being added in this diff
for block_lines in [old, new]:
# convert list to string
block_lines_str = '\n'.join(block_lines)
# search for the pattern within the hunk and # search for the pattern within the hunk and
# return a list of iterators to all of the matches # return a list of iterators to all of the matches
match_iters = compiled_pattern.finditer(block['hunk']) match_iters = compiled_pattern.finditer(block_lines_str)
# split off into overlapping pairs. i.e. ['A', 'B', C'] => [('A', 'B'), ('B', 'C')] # split off into overlapping pairs. i.e. ['A', 'B', C'] => [('A', 'B'), ('B', 'C')]
pairs = itertools.pairwise(match_iters) pairs = itertools.pairwise(match_iters)
@ -285,20 +306,19 @@ class TodoParser(object):
if i == 0: if i == 0:
# set start line and comment string of first section # set start line and comment string of first section
contiguous_comments_and_positions.append({ contiguous_comments_and_positions.append({
# -1 to ignore first newline, which isn't actually part of the hunk 'start': block_lines_str.count('\n', 0, prev_span[0]),
'start': block['hunk'].count('\n', 0, prev_span[0]) - 1,
'end': 0, 'end': 0,
'comment': pair[0].group(0) 'comment': pair[0].group(0)
}) })
# get number of lines in first section # get number of lines in first section
num_lines_in_first_section = block['hunk'].count('\n', prev_span[0], prev_span[1]) num_lines_in_first_section = block_lines_str.count('\n', prev_span[0], prev_span[1])
# set end line of first section relative to its start # set end line of first section relative to its start
contiguous_comments_and_positions[-1]['end'] = contiguous_comments_and_positions[-1]['start'] + num_lines_in_first_section contiguous_comments_and_positions[-1]['end'] = contiguous_comments_and_positions[-1]['start'] + num_lines_in_first_section
# get start/end index (within hunk) of current section # get start/end index (within hunk) of current section
curr_span = pair[1].span() curr_span = pair[1].span()
# determine number of lines between previous end and current start # determine number of lines between previous end and current start
num_lines_from_prev_section_end_line = block['hunk'].count('\n', prev_span[1], curr_span[0]) num_lines_from_prev_section_end_line = block_lines_str.count('\n', prev_span[1], curr_span[0])
# set start line of current section based on previous end # set start line of current section based on previous end
contiguous_comments_and_positions.append({ contiguous_comments_and_positions.append({
'start': contiguous_comments_and_positions[-1]['end'] + num_lines_from_prev_section_end_line, 'start': contiguous_comments_and_positions[-1]['end'] + num_lines_from_prev_section_end_line,
@ -306,7 +326,7 @@ class TodoParser(object):
'comment': pair[1].group(0) 'comment': pair[1].group(0)
}) })
# get number of lines in current section # get number of lines in current section
num_lines_in_curr_section = block['hunk'].count('\n', curr_span[0], curr_span[1]) num_lines_in_curr_section = block_lines_str.count('\n', curr_span[0], curr_span[1])
# set end line of current section relative to its start # set end line of current section relative to its start
contiguous_comments_and_positions[-1]['end'] = contiguous_comments_and_positions[-1]['start'] + num_lines_in_curr_section contiguous_comments_and_positions[-1]['end'] = contiguous_comments_and_positions[-1]['start'] + num_lines_in_curr_section
@ -315,20 +335,19 @@ class TodoParser(object):
if len(contiguous_comments_and_positions) == 0: if len(contiguous_comments_and_positions) == 0:
# redo the search, this time returning the # redo the search, this time returning the
# result directly rather than an iterator # result directly rather than an iterator
match = compiled_pattern.search(block['hunk']) match = compiled_pattern.search(block_lines_str)
if match: if match:
# get start/end index (within hunk) of this section # get start/end index (within hunk) of this section
span = match.span() span = match.span()
# set start line and comment string of first section # set start line and comment string of first section
contiguous_comments_and_positions.append({ contiguous_comments_and_positions.append({
# -1 to ignore first newline, which isn't actually part of the hunk 'start': block_lines_str.count('\n', 0, span[0]),
'start': block['hunk'].count('\n', 0, span[0]) - 1,
'end': 0, 'end': 0,
'comment': match.group(0) 'comment': match.group(0)
}) })
# get number of lines in first section # get number of lines in first section
num_lines_in_first_section = block['hunk'].count('\n', span[0], span[1]) num_lines_in_first_section = block_lines_str.count('\n', span[0], span[1])
# set end line of first section relative to its start # set end line of first section relative to its start
contiguous_comments_and_positions[-1]['end'] = contiguous_comments_and_positions[-1]['start'] + num_lines_in_first_section contiguous_comments_and_positions[-1]['end'] = contiguous_comments_and_positions[-1]['start'] + num_lines_in_first_section
@ -400,7 +419,8 @@ class TodoParser(object):
body=[], body=[],
hunk=hunk_info['hunk'], hunk=hunk_info['hunk'],
file_name=hunk_info['file'], file_name=hunk_info['file'],
start_line=hunk_info['start_line'] + comment_block['start'] + line_number_within_comment_block, start_line=((hunk_info['deleted_start_line'] if line_status == LineStatus.DELETED else hunk_info['added_start_line'])
+ comment_block['start'] + line_number_within_comment_block),
start_line_within_hunk=comment_block['start'] + line_number_within_comment_block + 1, start_line_within_hunk=comment_block['start'] + line_number_within_comment_block + 1,
num_lines=1, num_lines=1,
markdown_language=hunk_info['markdown_language'], markdown_language=hunk_info['markdown_language'],

2
tests/test_process_diff.py
View File

@ -80,8 +80,6 @@ class IssueUrlInsertionTest(unittest.TestCase):
self._setUp(['test_new.diff']) self._setUp(['test_new.diff'])
self._standardTest(80) self._standardTest(80)
# See GitHub issue #236
@unittest.expectedFailure
def test_line_numbering_with_deletions(self): def test_line_numbering_with_deletions(self):
self._setUp(['test_new_py.diff', 'test_edit_py.diff']) self._setUp(['test_new_py.diff', 'test_edit_py.diff'])
with self.subTest("Issue URL insertion"): with self.subTest("Issue URL insertion"):