feat(checker): add CEL for matching complicated expressions (#421)

* feat(lib/policy): add support for CEL checkers

This adds the ability for administrators to use Common Expression
Language[0] (CEL) for more advanced check logic than Anubis previously
offered.

These can be as simple as:

```yaml
- name: allow-api-routes
  action: ALLOW
  expression:
    and:
    - '!(method == "HEAD" || method == "GET")'
    - path.startsWith("/api/")
```

or get as complicated as:

```yaml
- name: allow-git-clients
  action: ALLOW
  expression:
    and:
    - userAgent.startsWith("git/") || userAgent.contains("libgit") || userAgent.startsWith("go-git") || userAgent.startsWith("JGit/") || userAgent.startsWith("JGit-")
    - >
      "Git-Protocol" in headers && headers["Git-Protocol"] == "version=2"
```

Internally these are compiled and evaluated with cel-go[1]. This also
leaves room for extensibility should that be desired in the future. This
will intersect with #338 and eventually intersect with TLS fingerprints
as in #337.

[0]: https://cel.dev/
[1]: https://github.com/google/cel-go

Signed-off-by: Xe Iaso <me@xeiaso.net>

* feat(data/apps): add API route allow rule for non-HEAD/GET

Signed-off-by: Xe Iaso <me@xeiaso.net>

* docs: document expression syntax

Signed-off-by: Xe Iaso <me@xeiaso.net>

* fix: fixes in review

Signed-off-by: Xe Iaso <me@xeiaso.net>

---------

Signed-off-by: Xe Iaso <me@xeiaso.net>

lib/policy/config/config.go

@@ -55,9 +55,11 @@ type BotConfig struct {
 	UserAgentRegex *string           `json:"user_agent_regex"`
 	PathRegex      *string           `json:"path_regex"`
 	HeadersRegex   map[string]string `json:"headers_regex"`
-	Action         Rule              `json:"action"`
 	RemoteAddr     []string          `json:"remote_addresses"`
-	Challenge      *ChallengeRules   `json:"challenge,omitempty"`
+	Expression     *ExpressionOrList `json:"expression"`
+
+	Action    Rule            `json:"action"`
+	Challenge *ChallengeRules `json:"challenge,omitempty"`
 }
 
 func (b BotConfig) Zero() bool {
@@ -85,7 +87,12 @@ func (b BotConfig) Valid() error {
 		errs = append(errs, ErrBotMustHaveName)
 	}
 
-	if b.UserAgentRegex == nil && b.PathRegex == nil && len(b.RemoteAddr) == 0 && len(b.HeadersRegex) == 0 {
+	allFieldsEmpty := b.UserAgentRegex == nil &&
+		b.PathRegex == nil &&
+		len(b.RemoteAddr) == 0 &&
+		len(b.HeadersRegex) == 0
+
+	if allFieldsEmpty && b.Expression == nil {
 		errs = append(errs, ErrBotMustHaveUserAgentOrPath)
 	}
 
@@ -137,6 +144,12 @@ func (b BotConfig) Valid() error {
 		}
 	}
 
+	if b.Expression != nil {
+		if err := b.Expression.Valid(); err != nil {
+			errs = append(errs, err)
+		}
+	}
+
 	switch b.Action {
 	case RuleAllow, RuleBenchmark, RuleChallenge, RuleDeny:
 		// okay

lib/policy/config/expressionorlist.go

@@ -0,0 +1,62 @@
+package config
+
+import (
+	"encoding/json"
+	"errors"
+	"slices"
+)
+
+var (
+	ErrExpressionOrListMustBeStringOrObject = errors.New("config: this must be a string or an object")
+	ErrExpressionEmpty                      = errors.New("config: this expression is empty")
+	ErrExpressionCantHaveBoth               = errors.New("config: expression block can't contain multiple expression types")
+)
+
+type ExpressionOrList struct {
+	Expression string   `json:"-"`
+	All        []string `json:"all"`
+	Any        []string `json:"any"`
+}
+
+func (eol ExpressionOrList) Equal(rhs *ExpressionOrList) bool {
+	if eol.Expression != rhs.Expression {
+		return false
+	}
+
+	if !slices.Equal(eol.All, rhs.All) {
+		return false
+	}
+
+	if !slices.Equal(eol.Any, rhs.Any) {
+		return false
+	}
+
+	return true
+}
+
+func (eol *ExpressionOrList) UnmarshalJSON(data []byte) error {
+	switch string(data[0]) {
+	case `"`: // string
+		return json.Unmarshal(data, &eol.Expression)
+	case "{": // object
+		type RawExpressionOrList ExpressionOrList
+		var val RawExpressionOrList
+		if err := json.Unmarshal(data, &val); err != nil {
+			return err
+		}
+		eol.All = val.All
+		eol.Any = val.Any
+
+		return nil
+	}
+
+	return ErrExpressionOrListMustBeStringOrObject
+}
+
+func (eol *ExpressionOrList) Valid() error {
+	if len(eol.All) != 0 && len(eol.Any) != 0 {
+		return ErrExpressionCantHaveBoth
+	}
+
+	return nil
+}

lib/policy/config/expressionorlist_test.go

@@ -0,0 +1,73 @@
+package config
+
+import (
+	"encoding/json"
+	"errors"
+	"testing"
+)
+
+func TestExpressionOrListUnmarshal(t *testing.T) {
+	for _, tt := range []struct {
+		name     string
+		inp      string
+		err      error
+		validErr error
+		result   *ExpressionOrList
+	}{
+		{
+			name: "simple",
+			inp:  `"\"User-Agent\" in headers"`,
+			result: &ExpressionOrList{
+				Expression: `"User-Agent" in headers`,
+			},
+		},
+		{
+			name: "object-and",
+			inp: `{
+			"all": ["\"User-Agent\" in headers"]
+			}`,
+			result: &ExpressionOrList{
+				All: []string{
+					`"User-Agent" in headers`,
+				},
+			},
+		},
+		{
+			name: "object-or",
+			inp: `{
+			"any": ["\"User-Agent\" in headers"]
+			}`,
+			result: &ExpressionOrList{
+				Any: []string{
+					`"User-Agent" in headers`,
+				},
+			},
+		},
+		{
+			name: "both-or-and",
+			inp: `{
+			"all": ["\"User-Agent\" in headers"],
+			"any": ["\"User-Agent\" in headers"]
+			}`,
+			validErr: ErrExpressionCantHaveBoth,
+		},
+	} {
+		t.Run(tt.name, func(t *testing.T) {
+			var eol ExpressionOrList
+
+			if err := json.Unmarshal([]byte(tt.inp), &eol); !errors.Is(err, tt.err) {
+				t.Errorf("wanted unmarshal error: %v but got: %v", tt.err, err)
+			}
+
+			if tt.result != nil && !eol.Equal(tt.result) {
+				t.Logf("wanted: %#v", tt.result)
+				t.Logf("got:    %#v", &eol)
+				t.Fatal("parsed expression is not what was expected")
+			}
+
+			if err := eol.Valid(); !errors.Is(err, tt.validErr) {
+				t.Errorf("wanted validation error: %v but got: %v", tt.err, err)
+			}
+		})
+	}
+}

lib/policy/config/testdata/bad/multiple_expression_types.json

@@ -0,0 +1,17 @@
+{
+  "bots": [
+    {
+      "name": "multiple-expression-types",
+      "action": "ALLOW",
+      "expression": {
+        "all": [
+          "userAgent.startsWith(\"git/\") || userAgent.contains(\"libgit\")",
+          "\"Git-Protocol\" in headers && headers[\"Git-Protocol\"] == \"version=2\"\n"
+        ],
+        "any": [
+          "userAgent.startsWith(\"evilbot/\")"
+        ]
+      }
+    }
+  ]
+}

lib/policy/config/testdata/bad/multiple_expression_types.yaml

@@ -0,0 +1,10 @@
+bots:
+- name: multiple-expression-types
+  action: ALLOW
+  expression:
+    all:
+    - userAgent.startsWith("git/") || userAgent.contains("libgit")
+    - >
+      "Git-Protocol" in headers && headers["Git-Protocol"] == "version=2"
+    any:
+    - userAgent.startsWith("evilbot/")

lib/policy/config/testdata/good/git_client.json

@@ -0,0 +1,14 @@
+{
+  "bots": [
+    {
+      "name": "allow-git-clients",
+      "action": "ALLOW",
+      "expression": {
+        "all": [
+          "userAgent.startsWith(\"git/\") || userAgent.contains(\"libgit\")",
+          "\"Git-Protocol\" in headers && headers[\"Git-Protocol\"] == \"version=2\""
+        ]
+      }
+    }
+  ]
+}

lib/policy/config/testdata/good/git_client.yaml

@@ -0,0 +1,8 @@
+bots:
+- name: allow-git-clients
+  action: ALLOW
+  expression:
+    all:
+    - userAgent.startsWith("git/") || userAgent.contains("libgit")
+    - >
+      "Git-Protocol" in headers && headers["Git-Protocol"] == "version=2"