Implement JS-esque `StateBase.srcCharCodeAt`

[hukkin]

Replaces use of StateBase.srcCharCode with the more JS-esque StateBase.srcCharCodeAt that can return None.

I didn't remove StateBase.srcCharCode for ease of migration because mdit-py-plugins use it. But did deprecate it.

Happy to hear what you think @chrisjsewell . Not sure if this is the solution or migration strategy you want, but it's something :smile:

I haven't measured performance yet. We may want to do that if this is something we want to proceed with otherwise

EDIT: This PR is in response to discussion in https://github.com/executablebooks/markdown-it-py/pull/186

[codecov[bot]]

Codecov Report

Merging #190 (6323f8b) into master (d6adf66) will decrease coverage by 0.06%. The diff coverage is 97.89%.

@@            Coverage Diff             @@
##           master     #190      +/-   ##
==========================================
- Coverage   96.17%   96.11%   -0.07%     
==========================================
  Files          61       61              
  Lines        3267     3267              
==========================================
- Hits         3142     3140       -2     
- Misses        125      127       +2     

Flag	Coverage Δ
pytests	96.11% <97.89%> (-0.07%)	:arrow_down:

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
markdown_it/ruler.py	88.88% <83.33%> (-0.77%)	:arrow_down:
markdown_it/helpers/parse_link_label.py	100.00% <100.00%> (ø)
markdown_it/rules_block/blockquote.py	100.00% <100.00%> (ø)
markdown_it/rules_block/fence.py	100.00% <100.00%> (ø)
markdown_it/rules_block/heading.py	100.00% <100.00%> (ø)
markdown_it/rules_block/hr.py	100.00% <100.00%> (ø)
markdown_it/rules_block/html_block.py	100.00% <100.00%> (ø)
markdown_it/rules_block/lheading.py	98.21% <100.00%> (+0.03%)	:arrow_up:
markdown_it/rules_block/list.py	98.86% <100.00%> (+0.01%)	:arrow_up:
markdown_it/rules_block/reference.py	97.70% <100.00%> (ø)
... and 15 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update d6adf66...6323f8b. Read the comment docs.

[hukkin]

Seeing that this PR only removes 4 instances of try: ... except IndexError: handling and that almost always we do an if pos < maximum: check or similar to ensure None is not an option, I actually still think it may be a fine decision to reject this change, and simply be more strict/explicit about out-of-range indices than upstream is. The current state is also better for performance. Arguably we'd be throwing away some work that has gone into making upstream more type safe.

The downsides for rejecting are diverging slightly from upstream logic, and the chance that there's still some place in the code where None handling is required but we don't know.