Skip to content

Conversation

@schneems
Copy link
Collaborator

Ripper is the built in parsing library from Ruby. It is much faster than Parser and it has everything we need.

This PR replaces the parser gem with ripper.

Ripper is the built in parsing library from Ruby. It is much faster than Parser and it has everything we need.

This PR replaces the parser gem with ripper.
@schneems
Copy link
Collaborator Author

This PR actually decreases test time noticeably:

Before: Finished in 2.4 seconds (files took 0.29151 seconds to load)
After: Finished in 1.6 seconds (files took 0.14218 seconds to load)

@schneems schneems merged commit 7bac04a into main Nov 23, 2020
@schneems schneems deleted the schneems/ripper branch December 4, 2020 15:27
schneems added a commit that referenced this pull request Jan 15, 2022
# This is the 1st commit message:

Move queue logic and engulf logic to a class

The queuing and engulfing logic are tied together. We can bundle the two of them into a single class with a specific interface.
# This is the commit message #2:

WIP A*ish

# This is the commit message #3:

WIP with lex_diff OMG

# This is the commit message #4:

WIP Pull out "grab equal" logic of UpDownExpand

# This is the commit message #5:

WIP move class to proper file location

# This is the commit message #6:

WIP Fix spelling of method

# This is the commit message #7:

WIP LOL search

# This is the commit message #8:

Runs, finds the large code blocks, but is too greedy

# This is the commit message #9:

Move LexDiff to own class, refactor to an array

We're never using the hash keys so we don't need them. We can save memory and time by storing the values as an array.

# This is the commit message #10:

WIP LOL spec

# This is the commit message #11:

Micro optimize lex diff

# This is the commit message #12:

UpDownExpand fix major bug with internal mutability

When calling `line.lex_detect.dup` it was duplicating the outer memory, but not the inner memory. Internally we're using an array and the old array was still being referenced and mutated.

# This is the commit message #13:

LexDiff Huge perf upgrade

Store values in the object instead of an array.

# This is the commit message #14:

WIP playing with fuzzing this into the old algorithm

# This is the commit message #15:

WORKS and is hella fast

# This is the commit message #16:

WIP docs

# This is the commit message #17:

Put a max band on number of iterations

Unbounded while loops scare me, in theory this class should never iterate more than line times, but mistakes happen. When they do I want feedback for the user without disrupting the rest of their program.

# This is the commit message #18:

WIP Cleanup and docs

# This is the commit message #19:

WIP Cleanup and docs

# This is the commit message #20:

Standardrb

# This is the commit message #21:

Standardrb

# This is the commit message #22:

Refactor LexDiff, remove unused methods, add docs
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants