Handling edge cases: wtq-00080, wtq-00085, wtq-00113, wtq-00162, wtq-00168#10
Handling edge cases: wtq-00080, wtq-00085, wtq-00113, wtq-00162, wtq-00168#10P6rguVyrst wants to merge 9 commits into
Conversation
The encoder now quotes keys that match boolean/null keywords (t, f, true,
false, null, none, nil) to prevent the decoder from misinterpreting them.
The decoder now uses a new parse_key() function for dictionary keys that
preserves them as strings, unlike parse_value() which converts keywords.
This fixes a critical round-trip bug where {"f": 1} would decode as
{False: 1}.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Values in dictionary compression headers containing commas, colons, or other special characters were not being quoted, causing data corruption on decode (values truncated at delimiter). Now uses _format_value() for consistent quoting of dictionary values. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
|
Warning Rate limit exceeded
Your organization is not enrolled in usage-based pricing. Contact your admin to enable usage-based pricing to continue reviews beyond the rate limit, or try again in 11 minutes and 44 seconds. ⌛ How to resolve this issue?After the wait time has elapsed, a review can be triggered using the We recommend that you space out your commits to avoid hitting the rate limit. 🚦 How do rate limits work?CodeRabbit enforces hourly rate limits for each developer per organization. Our paid plans have higher rate limits than the trial, open-source and free plans. In all cases, we re-allow further reviews after a brief timeout. Please see our FAQ for further information. ℹ️ Review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (5)
✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
P6rguVyrst
changed the title
|
I have a 10$ Claude api token unused for a long time. If you want me to send it over you just let me know your email. |
…my tests failing +1 existing one.
Add add failure scenatios
Revert "Add add failure scenatios"
Delta encoding violated spec §2.3's MUST round-trip requirement for float columns: prev + (cur - prev) in IEEE-754 does not recover the original double's bit pattern for arbitrary values, and round(diff, 10) at the encode step compounded the loss. Benchmark data exposed this as e.g. 1865.43 decoding to 1865.4299999999994. Restrict SparseMode.DELTA eligibility to int-only columns. Float columns now fall through to standard value encoding, which round-trips exactly via Python's shortest-round-trip str(float). Int delta encoding (the common case: IDs, counts, timestamps) is unchanged. Regression test covers multiple precision regimes — benchmark values, math.pi/math.e, 0.1+0.2, extreme exponents, negatives — so a future round-to-N workaround cannot sneak through. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2 participants
The failing tests: #9
If the accuracy tests pass (close PRs)-- I'm wrong, but I have no way to validate them.
wtq-00080 Example
## 3 Distinct bugs
1 Fix boolean-like dictionary keys being converted to booleans
The encoder now quotes keys that match boolean/null keywords (t, f, true,
false, null, none, nil) to prevent the decoder from misinterpreting them.
The decoder now uses a new parse_key() function for dictionary keys that
preserves them as strings, unlike parse_value() which converts keywords.
This fixes a critical round-trip bug where {"f": 1} would decode as
{False: 1}.
2 Fix dictionary header values not being quoted
Values in dictionary compression headers containing commas, colons,
or other special characters were not being quoted, causing data
corruption on decode (values truncated at delimiter).
Now uses _format_value() for consistent quoting of dictionary values.
3 Exclude float columns from delta encoding
Delta encoding violated spec §2.3's MUST round-trip requirement for float
columns: prev + (cur - prev) in IEEE-754 does not recover the original
double's bit pattern for arbitrary values, and round(diff, 10) at the
encode step compounded the loss. Benchmark data exposed this as e.g.
1865.43 decoding to 1865.4299999999994.
Restrict SparseMode.DELTA eligibility to int-only columns. Float columns
now fall through to standard value encoding, which round-trips exactly
via Python's shortest-round-trip str(float). Int delta encoding (the
common case: IDs, counts, timestamps) is unchanged.
Regression test covers multiple precision regimes — benchmark values,
math.pi/math.e, 0.1+0.2, extreme exponents, negatives — so a future
round-to-N workaround cannot sneak through.