checker-blacklist/checker/source.go

package checker

import (
	"context"
	"encoding/json"
	"html/template"
	"sync"

	sdk "git.happydns.org/checker-sdk-go/checker"
)

// Severity strings shared between sources, rules, and the HTML report.
const (
	SeverityCrit = "crit"
	SeverityWarn = "warn"
	SeverityInfo = "info"
	SeverityOK   = "ok"
)

// Source is the contract every reputation source implements. The
// registry collects one Source per backend (DNSBL family, Safe
// Browsing, URLhaus, VirusTotal, OpenPhish, …); Collect fans out over
// the registry concurrently and folds the per-source results into the
// observation payload. Adding a new source is a single file plus a
// `Register(...)` call in init().
//
// A Source returns *one or more* SourceResult values. Most sources
// return exactly one (`{Listed, Reasons, …}`); the DNSBL family returns
// one result per zone. Returning many results from one Source keeps the
// definition tidy (one ID, one set of options, one rule entry) while
// still surfacing per-zone detail in the report.
type Source interface {
	ID() string
	Name() string

	// Options contributes the option fields the source needs. They are
	// merged into the global CheckerDefinition at startup.
	Options() SourceOptions

	// Query runs the source against `registered` (the eTLD+1 of the
	// target domain) and returns one result per logical sub-target. The
	// implementation should never return nil: when the source is
	// disabled, return a single SourceResult with Enabled=false.
	Query(ctx context.Context, domain, registered string, opts sdk.CheckerOptions) []SourceResult

	// Diagnose produces the action-required card for a *listed* result.
	// Implementations should focus on the operator's next step; the
	// generic report wraps it with the title bar and severity styling.
	// Called only when SourceResult.Listed is true.
	Diagnose(res SourceResult) Diagnosis

	// Evaluate inspects an already-collected SourceResult and returns
	// whether the domain is considered listed and at what severity.
	// Implementations must read observation fields only (Evidence,
	// Reasons, Error, Enabled, BlockedQuery, Details) and must never
	// consult r.Listed or r.Severity.
	Evaluate(r SourceResult) (listed bool, severity string)
}

// DetailRenderer is an optional interface a Source can implement when
// the generic SourceResult shape (Reasons + Evidence + URLs) cannot
// fully express its output. Examples: VirusTotal's per-vendor verdict
// table, URLhaus' URL list with online/offline status. The returned
// HTML fragment is dropped into the source's section verbatim and is
// expected to be safe (use html/template or template.HTMLEscape).
type DetailRenderer interface {
	Source
	RenderDetail(res SourceResult) (template.HTML, error)
}

// SourceOptions describes the option fields a source contributes to the
// CheckerDefinition. Audiences map directly to the SDK's
// CheckerOptionsDocumentation buckets.
type SourceOptions struct {
	Admin []sdk.CheckerOptionField
	User  []sdk.CheckerOptionField
}

// SourceResult is the unified envelope every source produces. Source-
// specific structured data lives in Details (json.RawMessage), so the
// generic code (rules, headline, base diagnosis card, summary table)
// can operate on the envelope without source-specific switches; the
// rich report sections fish Details back through DetailRenderer.
type SourceResult struct {
	SourceID   string          `json:"source_id"`
	SourceName string          `json:"source_name"`
	Subject    string          `json:"subject,omitempty"`  // e.g. zone label for DNSBL
	Enabled    bool            `json:"enabled"`
	Listed      bool            `json:"listed"`
	BlockedQuery bool           `json:"blocked_query,omitempty"` // resolver blocked, not a real listing
	Severity    string          `json:"severity,omitempty"` // when Listed
	Reasons    []string        `json:"reasons,omitempty"`
	Evidence   []Evidence      `json:"evidence,omitempty"`
	LookupURL  string          `json:"lookup_url,omitempty"`
	RemovalURL string          `json:"removal_url,omitempty"`
	Reference  string          `json:"reference,omitempty"`
	Error      string          `json:"error,omitempty"`
	Details    json.RawMessage `json:"details,omitempty"`
}

// Evidence is a single observation that supports a verdict. Keeping it
// loosely typed (Label/Value/Status + free-form Extra) covers DNSBL
// return codes, OpenPhish URLs, URLhaus URLs, VT engine verdicts, …
// without growing the schema for each source.
type Evidence struct {
	Label  string            `json:"label"`
	Value  string            `json:"value"`
	Status string            `json:"status,omitempty"`
	Extra  map[string]string `json:"extra,omitempty"`
}

// Diagnosis is the action-required card surfaced at the top of the
// report. Sources build it in their Diagnose method.
type Diagnosis struct {
	Severity   string
	Title      string
	Detail     string
	Fix        string
	FixIsURL   bool
	LookupURL  string
	RemovalURL string
}

// ---------- registry ----------

var (
	registryMu sync.RWMutex
	registry   []Source
)

// Register adds a Source to the global registry. Intended to be called
// from init(). Panics on duplicate IDs so misconfigurations fail loudly
// at startup rather than producing silently-overlapping rules/options.
func Register(s Source) {
	registryMu.Lock()
	defer registryMu.Unlock()
	for _, existing := range registry {
		if existing.ID() == s.ID() {
			panic("checker-blacklist: duplicate source ID " + s.ID())
		}
	}
	registry = append(registry, s)
}

// Sources returns a snapshot of the registered sources, in registration
// order. Callers must not mutate the slice.
func Sources() []Source {
	registryMu.RLock()
	defer registryMu.RUnlock()
	out := make([]Source, len(registry))
	copy(out, registry)
	return out
}

// disabledResult returns the standard "source is disabled" sentinel slice.
func disabledResult(id, name string) []SourceResult {
	return []SourceResult{{SourceID: id, SourceName: name, Enabled: false}}
}

// evidenceEval is the common Evaluate body: listed when there is at least
// one Evidence entry and no error.
func evidenceEval(r SourceResult, severity string) (bool, string) {
	if r.Enabled && r.Error == "" && len(r.Evidence) > 0 {
		return true, severity
	}
	return false, ""
}

// EvaluateResult looks up the source that produced r from the registry
// and delegates to its Evaluate method. Returns (false, "") when the
// source is not found — a safe default that never promotes a stale
// Listed=true value.
func EvaluateResult(r SourceResult) (bool, string) {
	registryMu.RLock()
	defer registryMu.RUnlock()
	for _, s := range registry {
		if s.ID() == r.SourceID {
			return s.Evaluate(r)
		}
	}
	return false, ""
}