How to exclude files from lint rules

Some files legitimately fail a rule — generated metadata, vendored code, descriptions Salesforce writes for you. Pick the exclusion mechanism by scope — a single violation, one rule across file types, or whole files — then regenerate the baseline.

Suppress a single violation in source

Code Analyzer's inline markers work in any text file and for every engine — the marker is a plain comment in whatever comment syntax the file has:

// Code-analyzer-suppress(pmd:AvoidDebugStatements)
System.debug(LoggingLevel.INFO, 'probe');
// Code-analyzer-unsuppress(pmd:AvoidDebugStatements)

<!-- Code-analyzer-suppress(pmd:DescriptionWrongFormat) -->
<description>a description Salesforce generated</description>

A suppression runs from its marker line to the end of the file unless closed with the matching Code-analyzer-unsuppress(...) — close it, or the rest of the file is exempt too. The selector takes a rule (engine:rule), a whole engine (pmd), or all. See the Code Analyzer suppression docs for the full syntax.

Ruleset <exclude-pattern> does not work

It has no effect under Code Analyzer (pmd/pmd#5758). Do not use it — pick one of the mechanisms below.

Exclude files from a single PMD rule

Add a violationSuppressXPath property to the rule in the ruleset under .pmd/, matching file names with pmd:fileName():

<property
    name="violationSuppressXPath"
    value=".[ends-with(pmd:fileName(), '.email-meta.xml') or ends-with(pmd:fileName(), '.bot-meta.xml')]"
/>

pmd:fileName() returns the bare file name, not the path — match on suffixes (file type) or exact names, not folders. The DescriptionWrongFormat rule in .pmd/metadata-ruleset.xml carries a working example.

Exclude files from every engine and rule

Add the paths to ignores.files in code-analyzer/code-analyzer.yml (glob syntax):

ignores:
    files:
        - "force-app/main/default/pages/CommunitiesLanding.page"

This removes the files from the whole analysis — PMD, CPD, eslint, all of it. Use it for vendored or generated content; prefer the per-rule suppression when only one rule misfires.

What the analysis scans in the first place is in related ignore files.

For eslint-specific scoping, use the ignores entries of the root eslint.config.js instead.

Regenerate the baseline

A new exclusion changes what the scan reports, so the recorded caps drift:

npm run lint:createbaseline

Review the diff — entries for the now-excluded files should disappear — and commit the baseline together with the exclusion.

The baseline is not an exclusion mechanism — see how to baseline the code analyzer.