codecov.yml

# Codecov configuration for CODES. https://docs.codecov.com/docs/codecov-yaml
#
# Coverage is reported on PRs and commits but is INFORMATIONAL for now — it
# annotates without failing checks, so it can't block a merge before the project
# has an established baseline. Flip `informational: false` once coverage is
# trusted and you want it to gate.
coverage:
  status:
    project:
      default:
        informational: true
    patch:
      default:
        informational: true

# Files/paths excluded from the coverage metric. The coverage CI job also strips
# matching paths via `lcov --remove`; this is the belt-and-suspenders side so the
# Codecov UI agrees with the uploaded report.
ignore:
  - "tests/"             # the test harness itself
  - "**/configlex.*"     # generated flex lexer (modelconfig)
  - "**/configparser.*"  # generated bison parser (modelconfig)
  - "**/codeslexer.*"    # generated flex lexer (iokernellang)
  - "**/codesparser.*"   # generated bison parser (iokernellang; .y not committed)
  - "doc/"               # examples/tutorials (the ping-pong example IS tested,
                         # but it's teaching code, not the product — revisit if
                         # you'd rather measure it)

# PR comment from the Codecov bot. The defaults already post a comment; this just
# makes it explicit. require_base/require_changes:false mean it still comments on
# the first PRs, before master has an established coverage baseline to diff
# against.
comment:
  layout: "reach, diff, flags, files"
  behavior: default
  require_changes: false
  require_base: false
  require_head: true