Nimcheck — Project Status

Version: 0.1.0
Default branch: master
Remote: https://retoor.molodetz.nl/retoor/nimcheck.git
Last verified: 2026-07-15 — nimble build and nimble test pass (~11s, no crashes)

Summary

Nimcheck is feature-complete for v0.1.0. All planned languages have tokenizers, validators, fixtures, and tests. Critical crash bugs (infinite tokenizer loops → OOM) are fixed. The project is ready to use as a library, CLI, or CI gate.

What Is Done

Core framework

Component Status Notes
Tokenizer base (tokenizerbase.nim) Done Position tracking, bracket stack, closeBracket advance, finishTokenizeStep stall guard
Validator base (validatorbase.nim) Done Pipeline, factory, registry for all flavors
Detector (detector.nim) Done Extension, shebang, content scoring; false-positive filters (entropy, margin)
Public API (nimcheck.nim) Done validateSource, validateFile, inspectSource, detectFlavor, CLI
Error reporting (reporting/errors.nim) Done Human + JSON output
Never-crash guarantee Done All paths wrapped; fuzz-tested

Languages with full validator support (28 flavors, 40+ registry aliases)

Group Flavors Tokenizer module Validator module
Nim nim, nims, nimble nim_tokenizer.nim nim_validator.nim
Shell bash, sh, shell bash_tokenizer.nim bash_validator.nim
Python python, py python_tokenizer.nim python_validator.nim
JS / TS javascript, js, typescript, ts javascript_tokenizer.nim, type_xml_tokenizer.nim javascript_validator.nim, type_xml_validator.nim
PHP php, phtml php_tokenizer.nim php_validator.nim
HTML / XML html, htm, xml, xsd, xslt, svg html_tokenizer.nim, type_xml_tokenizer.nim html_validator.nim, type_xml_validator.nim
Jinja jinja, jinja2, j2 jinja_tokenizer.nim jinja_validator.nim
Config json, jsonc, yaml, yml, toml (parser / line-based) config_validators.nim
C family c, h, cpp, cxx, hpp, java, csharp, cs cfamily_tokenizers.nim cfamily_validators.nim
Systems / scripting go, rust, ruby, lua, swift, kotlin, kt extended_tokenizers.nim, lang_tokenizers.nim extended_validators.nim, lang_validators.nim
Web / data / ops css, sql, markdown, md, dockerfile, makefile extended_tokenizers.nim extended_validators.nim

Tests

Suite Count Status
Per-language basic + exhaustive 30+ modules All pass
Config (test_config.nim) JSON/YAML/TOML Pass
Mixed files (test_mixed.nim) HTML+Jinja etc. Pass
Fuzz (test_fuzz.nim) Null bytes, 10k strings, 100-level nesting, concurrency Pass
Fixtures tests/fixtures/<lang>/ valid, invalid, edge_cases Present for all new langs

CI / Gitea

Item Status
Workflow .gitea/workflows/ci.yml Done — lint, test matrix (Nim 2.0.8 + stable), build artifact
Default branch triggers master (primary), main and develop also accepted
act_runner requirement ubuntu-latest label — see GITEA.md

Bug fixes shipped in v0.1.0

  1. closeBracket() did not advance pos — caused infinite loops on ), ], } in CLike tokenizers → OOM kill.
  2. Jinja lone { — text collector stalled without advancing.
  3. JS/TS double-advance — caller + closeBracket both advanced.
  4. finishTokenizeStep — safety net if tokenizer position stalls.
  5. Lua for … do — do incorrectly opened nested block.
  6. Ruby postfix unless — modifier form incorrectly opened block.
  7. Makefile test tabs — \t in triple-quoted strings was backslash+t, not TAB.
  8. SQL trailing comma — ,\n) not detected (whitespace between comma and paren).
  9. detectFlavor scoring — C++, Kotlin, Swift, CSS, Markdown samples now detect correctly.

How To Verify (local)

# From repo root
nimble build          # or: make build
nimble test           # or: make test     (~11 seconds)
make lint             # strict compile check

Expected: exit code 0, no [FAILED] lines.

How To Use

Library

import nimcheck
let r = validateFile("src/main.nim")
echo r.valid

CLI

make build
bin/nimcheck validate --file=script.py
bin/nimcheck detect --file=unknown.txt
bin/nimcheck inspect --file=main.go --json

See docs/API.md for API and JSON schemas; README.md for overview.

What Is NOT Done (optional / future)

These are not blockers for v0.1.0:

Item Priority Notes
Compiler-grade semantics Low Validators are tokenizer/structure checks, not full parsers
finishTokenizeStep on every tokenizer Low Defense in depth; core paths covered
Semantic analysis (undefined vars, types) Future Error codes E1004–E1009 reserved
Coverage job in CI Optional Requires nimcoverage; manual dispatch only
Rename master → main Optional Repo uses master; CI supports both
Published nimble package Future nimble install from git works today

Repository Hygiene

Ignored by .gitignore: compiled test binaries, bin/, *.log, debug_*, test_hang*, editor temp files.

Do not commit: local debug scripts, test_run*.log, choosenim caches.

Release Checklist (when bumping version)

  1. Update version in nimcheck.nimble and nimcheckVersion in src/nimcheck.nim
  2. Update version line in README.md and this file
  3. nimble build && nimble test
  4. Tag: git tag v0.2.0 && git push origin v0.2.0
  5. Confirm Gitea Actions green on master

Quick Reference

Task Command
Build make build / nimble build
All tests make test / nimble test
One language nim c -r tests/test_go.nim
Lint make lint
Clean make clean
CI locally Same as lint + test
Push git push origin master

Documentation layout

File Purpose
README.md Product overview (concise)
docs/API.md API reference, types, JSON, CLI
docs/ERRORS.md Error codes, per-language table
docs/ARCHITECTURE.md Pipeline design
docs/GITEA.md CI setup
CONTRIBUTING.md Dev workflow