python: provide informative exception for trailing escapes in tables

[kieran-ryan]

🤔 What's changed?

With the Python implementation:

1. Whitespace either side of a table row is stripped.
2. The table row is passed to a function to strip and yields its cell values
3. Characters in each cell are iterated over to handle them appropriately, such as new columns (|), escape characters, etc.
4. As escape characters are typically followed by a special character (e.g. n, t), when an escape character is detected \\ an additional next call is made to grab the next character
5. As in step 1, whitespace after a trailing escape character (\\) would be stripped, there would not be a following character and a StopIteration error is raised.

gherkin/python/gherkin/gherkin_line.py

Lines 39 to 62 in b055261

	def split_table_cells(self, row):
	"""
	An iterator returning all the table cells in a row with their positions,
	accounting for escaping.
	"""
	row = iter(row)
	col = 0
	start_col = col + 1
	cell = ''
	first_cell = True
	while True:
	char = next(row, None)
	col += 1
	if char == '|':
	if first_cell:
	# First cell (content before the first |) is skipped
	first_cell = False
	else:
	yield (cell, start_col)
	cell = ''
	start_col = col + 1
	elif char == '\\':
	char = next(row)

By defaulting the iterator to an empty string, when there is a trailing escape character, the trailing escape character will not trigger a StopIteration error and subsequent code will appropriately raise an inconsistent cell count error message.

⚡️ What's your motivation?

• Fixes A space following Escape \ as the last character of a table line instead of | causes StopIterationError not CompositeError #223

🏷️ What kind of change is this?

• 🐛 Bug fix (non-breaking change which fixes a defect)

♻️ Anything particular you want feedback on?

📋 Checklist:

• I agree to respect and uphold the Cucumber Community Code of Conduct
• I've changed the behaviour of the code
  • I have added/updated tests to cover my changes.
• Users should know about my change
  • I have added an entry to the "Unreleased" section of the CHANGELOG, linking to this pull request.

---

This text was originally generated from a template, then edited by hand. You can modify the template here.

[mpkorstanje]

Cheers!

[kieran-ryan]

Thanks for the review! Will go with your latest preference following the considerations I have added

[mpkorstanje]

Alright, I had a look at the Javascript and Ruby implementations and they're just as confusing. Java was the odd one out. So this is a good solution.

What is weird though, they should have the same problem because both JS and Ruby do something equivalent to:

        elsif char == '\\'
          char = row[col]

And that should result in char being undefined. Which is rendered to string as undefined in JS atleast.

"hello " + "cucumber"[8] 

"hello undefined"

Fixing this it outside the scope of this PR though.

[mpkorstanje]

Not a problem in JS, while we do append undefined, this cell is never used because it is not terminated by a |.