Skip to content

Releases: sysevol-ai/CodeNib

Release list

CodeNib v0.1.0

CodeNib v0.1.0 Pre-release
Pre-release

Choose a tag to compare

@fishmingyu fishmingyu released this 29 Jul 10:51

What's Changed

  • add readme by @fishmingyu in #1
  • [Debug] add module attributes and nested import by @fishmingyu in #2
  • Add support for C++ by @fishmingyu in #3
  • Syntax debug by @fishmingyu in #4
  • [Debug] add support for nested function by @fishmingyu in #5
  • Nested type inference by @fishmingyu in #6
  • Update igraph as backbone by @fishmingyu in #7
  • Codegraph update by @fishmingyu in #8
  • Stz dev by @fishmingyu in #9
  • search API update by @fishmingyu in #10
  • [Update] refactor SimilarityAPI with start() and query() by @stzoozz in #11
  • Update Roi Subgraph by @fishmingyu in #12
  • add modular BaseAPI framework & SimilarityAPI by @stzoozz in #13
  • [Update] add roi subgraph filtering by @fishmingyu in #14
  • [Update] add Similarity API with model & device switching by @stzoozz in #15
  • [Code Chunk] Support by @fishmingyu in #16
  • Agent + tool by @fishmingyu in #17
  • transgraph tool by @fishmingyu in #19
  • [Agent] Initialize search code snippets by @stzoozz in #18
  • [Update] Add BM25-transgraph compatibility by @stzoozz in #20
  • Transgraph debug for nested function and hierarchy by @fishmingyu in #21
  • Graph file dependency rebase by @fishmingyu in #22
  • Agent part utils update by @fishmingyu in #23
  • [Framework] refactor from llama-index to langchain by @fishmingyu in #25
  • [Update] Add node_id for dense chunk to align with code graph by @stzoozz in #24
  • [VecDB] refactor framework from llama_index to langchain by @fishmingyu in #26
  • Split chunks with max_lines_per_chunk by @stzoozz in #27
  • Agent VLLM support by @fishmingyu in #28
  • Build bm25 index from chunk by @fishmingyu in #29
  • [Chunk] Add class name in node_id for method by @stzoozz in #30
  • Fishming.dev by @fishmingyu in #31
  • [Env] add locbench, adjust swebench env path by @fishmingyu in #32
  • Add Search + Rerank (S+R) baseline to examples by @stzoozz in #33
  • [S+R] Support LocBench in test by @stzoozz in #34
  • Exclude config for using scip python index by @fishmingyu in #35
  • [SCIP] add third party and update readme by @fishmingyu in #36
  • [Examples] Update S+R pipeline & Add Agentless pipeline by @stzoozz in #37
  • [Commit] add test repo chunker by @fishmingyu in #38
  • SWEBench verified update by @fishmingyu in #39
  • [CI/CD] add PR template by @fishmingyu in #40
  • Extract agent update and SCIP doc by @fishmingyu in #41
  • [Core] update c++ version graph builder by @fishmingyu in #43
  • profiler + graph index preprocess by @fishmingyu in #44
  • Add Regex node indexing and enhance example outputs by @stzoozz in #42
  • Swebench-verified location ground-truth generator with its test (for manually checking) by @siriuxyu in #45
  • Ops and Plans initiative, rerank model revise by @fishmingyu in #46
  • Rerank refactor and change llm config by @fishmingyu in #48
  • Add build_embeddings script and fix with pre-commit by @stzoozz in #47
  • [VectorDB] optimize index by @fishmingyu in #49
  • Refactor langchain and retrieve ops by @fishmingyu in #50
  • [swebench-verified] FIX - set work directory to ~/.codeminer/tmp/ by @siriuxyu in #51
  • Update retrieve and rerank the model by @fishmingyu in #52
  • Align the node metadata between eval and index by @stzoozz in #53
  • [VectorDB] Support index metrics (l2, ip) in the FAISS vector store by @stzoozz in #54
  • Align python and cpp scip_decoder with tests by @siriuxyu in #55
  • [core] 2 Optimization Attempts by @siriuxyu in #56
  • Align the chunker and eval for retrieval rerank pipeline by @stzoozz in #57
  • Code Chunking update with new lang and structure by @fishmingyu in #58
  • Refactor code chunking and improve retrieval evaluation by @stzoozz in #59
  • Fishming.dev by @fishmingyu in #61
  • Fishming.dev by @fishmingyu in #62
  • Fix CI/CD and tests by @fishmingyu in #63
  • fix bug for ci that affect original env and profile by @fishmingyu in #64
  • [CI] fix bug that affect original home env by @fishmingyu in #65
  • [CI] fix bug that affect original home env by @fishmingyu in #66
  • Add skeleton mode; fix CI env by @fishmingyu in #67
  • chore: add default flake8, git attributes, clang-format; align with p… by @fishmingyu in #69
  • [VectorDB] Implement hierarchical vector store (l0 skeleton + l2) by @stzoozz in #68
  • [Examples] Update retrieval rerank pipeline to support hierarchical vector store by @stzoozz in #70
  • Added parallel cpp scip decoder & Clean gt-script's test by @siriuxyu in #60
  • Fishming.dev by @fishmingyu in #71
  • [Emmbedding] fixretrieve rerank pipeline with L0 by @fishmingyu in #72
  • New JS/TS chunker, hybrid retrieval by @fishmingyu in #74
  • [Rerank] Add embedding-based rerank and SWE-bench dev eval scripts by @fishmingyu in #75
  • refactor dataset usage and logic by @fishmingyu in #76
  • fix issues for dataset loading by @fishmingyu in #78
  • Multilingual CodeGraph with SCIP by @siriuxyu in #77
  • Introduce SWE-bench Multilingual support and a query synthesis pipeline for code search QA datasets, including utilities for sampling instances and generating multi-level natural language queries. by @fishmingyu in #79
  • Fix shallow-clone checkout for SWE-bench + split collect/synthesize pipeline scripts by @fishmingyu in #80
  • [CI/CD] Stabilize SCIP multilingual CI: centralize non-Python deps + migrate test to pytest by @fishmingyu in #81
  • SCIP multi-lang robustness + deterministic/consensus SWE-bench synthesis by @fishmingyu in #82
  • debug query synthesizer for swebench by @boqiny in #83
  • [SCIP] Debug and refactor the SCIP indexing pipeline for multilingual codegraph by @siriuxyu in #84
  • Improve retrieval/rerank profiler outputs and add tagged profiling to eval scripts by @fishmingyu in #85
  • [Refactor] Unified LS-router for both SCIP and Clangd by @siriuxyu in #87
  • Add Go chunker support and stabilize test fixtures by @fishmingyu in #88
  • Add BM25, Embedding, and Graph Retrieval Pipelines with Baselines by @imaryandokania in #86
  • Dataset pipeline: sampling, GT locator, and HF dataset builder by @fishmingyu in #89
  • Add agent compiler infrastructure and migrate LLM layer to LiteLLM by @fishmingyu in #90
  • Add agent runner, graph expand skill and ops by @fishmingyu in #93
  • Add test for skill/embedding_search by @Turquoise-T in #91
  • [Test] Add BM25 search skill tests...
Read more