chore: merge parser attestor #8
Conversation
[ERROR] snarkJS: Error: Scalar size does not match
at _multiExp (/opt/homebrew/lib/node_modules/snarkjs/node_modules/ffjavascript/build/main.cjs:4978:19)
at WasmCurve.multiExpAffine (/opt/homebrew/lib/node_modules/snarkjs/node_modules/ffjavascript/build/main.cjs:5015:22)
at groth16Verify$1 (/opt/homebrew/lib/node_modules/snarkjs/build/cli.cjs:6171:31)
at async Object.groth16Verify [as action] (/opt/homebrew/lib/node_modules/snarkjs/build/cli.cjs:13017:21)
at async clProcessor (/opt/homebrew/lib/node_modules/snarkjs/build/cli.cjs:483:27)
fix tests response tests write out the lock used optional logs on extractor renames more refactor codegen extracts http body CLI and doc Create bounty.md Created an ISSUE_TEMPLATE folder with the bounty.md file Update bounty.md Create LICENSE-MIT Create LICENSE-Apache Update README.md (#81) Updated README for consistency with Plonkathon and AES-proof, going to need some help to get it over the line, dependent on name change lots of cleanup place examples in respective dir basic circuit cleanup pending work fix tests
* Delete LICENSE-MIT * Rename LICENSE-Apache to LICENSE * Update README.md
* feat(cli): add extractor witness gen * fix: select subarray param bug * feat: add circuit config to `circuits.json`. Pending: http extraction * feat: add http extractor support * fix: tests * refactor files a bit * fix tests again, arghhhhhhhgstgst * simplify cli a bit more * more cli cleanup * add json proof docs * add http build docs --------- Co-authored-by: Colin Roberts <[email protected]>
* add input * fix array index codegen bug * refactor: tests utils reorganise * add spotify test * save work till now, start on integrated codegen * save work * fix: tests * refactor: codegen * add integrated codegen * test: add separate test for correctness * refactor: change `Syntax` and `StateUpdate` to prevent duplicate callable symbols in JSON parser * feat: accept inputs in bytes for `readJSONInputFile` * add extended witness gen in cli * feat: extended test works!!! * refactor: json cleanup * add tests * add test and lint CI * add http and extended codegen docs * add docs * rm old README info * Update http.rs (#92) --------- Co-authored-by: Colin Roberts <[email protected]>
* remove slice * remove syntax * feat: json improvements * feat(http): remove syntax use * feat(http): remove hasher substring matching * fix: delete extractor * fix: codegen * feat: remove hashing based substring matching * fix: spotify test * fix: tests
* init example * feat: add single value extractor * test: add example * feat: remove keyLen and add maxKeyLen * add optimisations from web-prover * `depth` as signal in `NextKVPairAtDepth` * fix broken tests
* simplify `JsonMaskObjectNIVC` * passing `JsonMaskArrayIndexNIVC` * bug: incorrect mask for "profile" key * bug: not extracting `"profile"` * add extra key index to match end quote * fix: masking tests pass * fix: final extraction * cleanup: unused signals, consistency Found unused signals with the circom-witnesscalc `build-circuit` binary. These are now removed. * feat: passing HTTP NIVC tests * fix: unallocated signals `HTTPParseAndLockStartLine` had unused signals. I set these properly now. * rename SubstringMatchWithIndex * add `HeaderFieldNameValueMatchPadded` * add a todo in json interpreter * change `LockHeader` to handle max key and value length * use circuit.compute to find witness and simplify test --------- Co-authored-by: lonerapier <[email protected]>
* http: `parse_and_lock_start_line` max arguments * add lock header tests * fix tests
|
@0xJepsen @lonerapier It feels like the strategy between these two PRs should be the same #8 => Is moving the parser-attestor into this repo Both seem viable to me, I'm personally lightly leaning toward waylon's approach here, because it helps to keep useful packages for people external to pluto in standalone repository (good for "github seo"). But ultimately I think you can both choose the approach, we should just pick one. |
|
I'm okay with either way to do it personally. But I do not want to bring in the circuit codegen to this repo. I just vote to have a minimal circuits only repository. In which case, maybe #11 is a better move -- checking it out now. |
|
i'm also in favor of just keeping only the circuits useful for web-prover and rest of the low level circuits in separate repos. But i'm not sure if the development workflow will be optimal. Any change in the web-prover will depend upon 2 releases in parser-attestor/aes-proof and then this repo. |
|
I'm of the opinion that we either completely combine these repos and deprecate/archive the old, or we do something super minimal like have just a Truth be told, I don't care much, but if I were to pick, I'd just combine repos and remove the fluff we have. E.g., we can archive the codegen. |
closes #9