util.jsonschema: Fix validation to not assume presence of "type" field

MattJ reported a curious issue where validation did not work as
expected. Primarily that the "type" field was expected to be mandatory,
and thus leaving it out would result in no checks being performed.
This was likely caused by misreading during initial development.

Spent some time testing against
https://github.com/json-schema-org/JSON-Schema-Test-Suite.git and
discovered a multitude of issues, far too many to bother splitting into
separate commits.

More than half of them fail. Many because of features not implemented,
which have been marked NYI. For example, some require deep comparisons
e.g. when objects or arrays are present in enums fields.

Some because of quirks with how Lua differs from JavaScript, e.g. no
distinct array or object types. Tests involving fractional floating
point numbers. We're definitely not going to follow references to remote
resources. Or deal with UTF-16 sillyness. One test asserted that 1.0 is
an integer, where Lua 5.3+ will disagree.

1 files changed, 152 insertions, 147 deletions

diff --git a/teal-src/util/jsonschema.tl b/teal-src/util/jsonschema.tl
index 5c5fd4ce..160c164c 100644
--- a/teal-src/util/jsonschema.tl
+++ b/teal-src/util/jsonschema.tl
@@ -16,13 +16,13 @@ local pointer = require "util.jsonpointer"
 local type json_type_name = json.json_type_name
 
 -- json_type_name here is non-standard
-local type schema_t = boolean | json_type_name | json_schema_object
+local type schema_t = boolean | json_schema_object
 
 local record json_schema_object
 	type json_type_name = json.json_type_name
 	type schema_object = json_schema_object
 
-	type : json_type_name
+	type : json_type_name | { json_type_name }
 	enum : { any }
 	const : any
 
@@ -47,7 +47,7 @@ local record json_schema_object
 	-- strings
 	maxLength : integer
 	minLength : integer
-	pattern : string
+	pattern : string -- NYI
 	format : string
 
 	-- arrays
@@ -57,17 +57,17 @@ local record json_schema_object
 	maxItems : integer
 	minItems : integer
 	uniqueItems : boolean
-	maxContains : integer
-	minContains : integer
+	maxContains : integer -- NYI
+	minContains : integer -- NYI
 
 	-- objects
 	properties : { string : schema_t }
-	maxProperties : integer
-	minProperties : integer
+	maxProperties : integer -- NYI
+	minProperties : integer -- NYI
 	required : { string }
 	dependentRequired : { string : { string } }
 	additionalProperties: schema_t
-	patternProperties: schema_t
+	patternProperties: schema_t -- NYI
 	propertyNames : schema_t
 
 	-- xml
@@ -99,10 +99,10 @@ end
 
 -- TODO validator function per schema property
 
-local type_validators : { json_type_name : function (schema_t, any, json_schema_object) : boolean } = {}
-
-local function simple_validate(schema : json_type_name, data : any) : boolean
-	if schema == "object" and data is table then
+local function simple_validate(schema : json_type_name | { json_type_name }, data : any) : boolean
+	if schema == nil then
+		return true
+	elseif schema == "object" and data is table then
 		return type(data) == "table" and (next(data)==nil or type((next(data, nil))) == "string")
 	elseif schema == "array" and data is table then
 		return type(data) == "table" and (next(data)==nil or type((next(data, nil))) == "number")
@@ -110,144 +110,171 @@ local function simple_validate(schema : json_type_name, data : any) : boolean
 		return math.type(data) == schema
 	elseif schema == "null" then
 		return data == null
+	elseif schema is { json_type_name } then
+		for _, one in ipairs(schema as { json_type_name }) do
+			if simple_validate(one, data) then
+				return true
+			end
+		end
+		return false
 	else
 		return type(data) == schema
 	end
 end
 
-type_validators.string = function (schema : json_schema_object, data : any) : boolean
-	-- XXX this is measured in byte, while JSON measures in ... bork
-	-- TODO use utf8.len?
-	if data is string then
-		if schema.maxLength and #data > schema.maxLength then
-			return false
-		end
-		if schema.minLength and #data < schema.minLength then
-			return false
-		end
-		return true
+local complex_validate : function ( json_schema_object, any, json_schema_object ) : boolean
+
+local function validate (schema : schema_t, data : any, root : json_schema_object) : boolean
+	if schema is boolean then
+		return schema
+	else
+		return complex_validate(schema, data, root)
 	end
-	return false
 end
 
-type_validators.number = function (schema : json_schema_object, data : number) : boolean
-	if schema.multipleOf and data % schema.multipleOf ~= 0 then
-		return false
-	end
+function complex_validate (schema : json_schema_object, data : any, root : json_schema_object) : boolean
 
-	if schema.maximum and not ( data <= schema.maximum ) then
-		return false
+	if root == nil then
+		root = schema
 	end
 
-	if schema.exclusiveMaximum and not ( data < schema.exclusiveMaximum ) then
-		return false
+	if schema["$ref"] and schema["$ref"]:sub(1,1) == "#" then
+		local referenced = pointer.resolve(root as table, schema["$ref"]:sub(2)) as schema_t
+		if referenced ~= nil and referenced ~= root and referenced ~= schema then
+			if not validate(referenced, data, root) then
+				return false;
+			end
+		end
 	end
 
-	if schema.minimum and not ( data >= schema.minimum ) then
-		return false
+	if not simple_validate(schema.type, data) then
+		return false;
 	end
 
-	if schema.exclusiveMinimum and not ( data > schema.exclusiveMinimum ) then
-		return false
+	if schema.type == "object" then
+		if data is table then
+			-- just check that there the keys are all strings
+			for k in pairs(data) do
+				if not k is string then
+					return false
+				end
+			end
+		end
 	end
 
-	return true
-end
-
-type_validators.integer = type_validators.number
+	if schema.type == "array" then
+		if data is table then
+			-- just check that there the keys are all numbers
+			for i in pairs(data) do
+				if not i is integer then
+					return false
+				end
+			end
+		end
+	end
 
-local function validate(schema : schema_t, data : any, root : json_schema_object) : boolean
-	if schema is boolean then
-		return schema
+	if schema["enum"] ~= nil then
+		local match = false
+		for _, v in ipairs(schema["enum"]) do
+			if v == data then
+				-- FIXME supposed to do deep-compare
+				match = true
+				break
+			end
+		end
+		if not match then
+			return false
+		end
 	end
-	if schema is json_type_name then
-		return simple_validate(schema, data)
+
+	-- XXX this is measured in byte, while JSON measures in ... bork
+	-- TODO use utf8.len?
+	if data is string then
+		if schema.maxLength and #data > schema.maxLength then
+			return false
+		end
+		if schema.minLength and #data < schema.minLength then
+			return false
+		end
 	end
-	if schema is json_schema_object then
-		if root == nil then
-			root = schema
+
+	if data is number then
+		if schema.multipleOf and (data == 0 or data % schema.multipleOf ~= 0) then
+			return false
 		end
-		if schema["$ref"] and schema["$ref"]:sub(1,1) == "#" then
-			local referenced = pointer.resolve(root as table, schema["$ref"]:sub(2)) as schema_t
-			if referenced ~= nil then
-				return validate(referenced, data, root);
-			end
+
+		if schema.maximum and not ( data <= schema.maximum ) then
+			return false
 		end
 
-		if schema.allOf then
-			for _, sub in ipairs(schema.allOf) do
-				if not validate(sub, data, root) then
-					return false
-				end
-			end
-			return true
+		if schema.exclusiveMaximum and not ( data < schema.exclusiveMaximum ) then
+			return false
 		end
 
-		if schema.oneOf then
-			local valid = 0
-			for _, sub in ipairs(schema.oneOf) do
-				if validate(sub, data, root) then
-					valid = valid + 1
-				end
-			end
-			return valid == 1
+		if schema.minimum and not ( data >= schema.minimum ) then
+			return false
 		end
 
-		if schema.anyOf then
-			for _, sub in ipairs(schema.anyOf) do
-				if validate(sub, data, root) then
-					return true
-				end
-			end
+		if schema.exclusiveMinimum and not ( data > schema.exclusiveMinimum ) then
 			return false
 		end
+	end
 
-		if schema["not"] then
-			if validate(schema["not"], data, root) then
+	if schema.allOf then
+		for _, sub in ipairs(schema.allOf) do
+			if not validate(sub, data, root) then
 				return false
 			end
 		end
+	end
 
-		if schema["if"] then
-			if validate(schema["if"], data, root) then
-				if schema["then"] then
-					return validate(schema["then"], data, root)
-				end
-			else
-				if schema["else"] then
-					return validate(schema["else"], data, root)
-				end
+	if schema.oneOf then
+		local valid = 0
+		for _, sub in ipairs(schema.oneOf) do
+			if validate(sub, data, root) then
+				valid = valid + 1
 			end
 		end
-
-		if schema.const ~= nil and schema.const ~= data then
+		if valid ~= 1 then
 			return false
 		end
+	end
 
-		if schema["enum"] ~= nil then
-			for _, v in ipairs(schema["enum"]) do
-				if v == data then
-					return true
-				end
+	if schema.anyOf then
+		local match = false
+		for _, sub in ipairs(schema.anyOf) do
+			if validate(sub, data, root) then
+				match = true
+				break
 			end
+		end
+		if not match then
 			return false
 		end
+	end
 
-		if schema.type then
-			if not simple_validate(schema.type, data) then
-				return false
-			end
+	if schema["not"] then
+		if validate(schema["not"], data, root) then
+			return false
+		end
+	end
 
-			local validator = type_validators[schema.type]
-			if validator then
-				return validator(schema, data, root)
+	if schema["if"] ~= nil then
+		if validate(schema["if"], data, root) then
+			if schema["then"] then
+				return validate(schema["then"], data, root)
+			end
+		else
+			if schema["else"] then
+				return validate(schema["else"], data, root)
 			end
 		end
-		return true
 	end
-end
 
-type_validators.table = function (schema : json_schema_object, data : any, root : json_schema_object) : boolean
+	if schema.const ~= nil and schema.const ~= data then
+		return false
+	end
+
 	if data is table then
 
 		if schema.maxItems and #data > schema.maxItems then
@@ -266,24 +293,28 @@ type_validators.table = function (schema : json_schema_object, data : any, root
 			end
 		end
 
-		if schema.properties then
-			local additional : schema_t = schema.additionalProperties or true
-			for k, v in pairs(data) do
-				if schema.propertyNames and not validate(schema.propertyNames, k, root) then
+		if schema.propertyNames ~= nil then
+			for k in pairs(data) do
+				if not validate(schema.propertyNames, k, root) then
 					return false
 				end
-				local s = schema.properties[k as string] or additional
-				if not validate(s, v, root) then
+			end
+		end
+
+		if schema.properties then
+			for k, sub in pairs(schema.properties) do
+				if data[k] ~= nil and not validate(sub, data[k], root) then
 					return false
 				end
 			end
-		elseif schema.additionalProperties then
+		end
+
+		if schema.additionalProperties ~= nil then
 			for k, v in pairs(data) do
-				if schema.propertyNames and not validate(schema.propertyNames, k, root) then
-					return false
-				end
-				if not validate(schema.additionalProperties, v, root) then
-					return false
+				if schema.properties == nil or schema.properties[k as string] == nil then
+					if not validate(schema.additionalProperties, v, root) then
+						return false
+					end
 				end
 			end
 		end
@@ -300,9 +331,11 @@ type_validators.table = function (schema : json_schema_object, data : any, root
 		end
 
 		local p = 0
-		if schema.prefixItems then
+		if schema.prefixItems ~= nil then
 			for i, s in ipairs(schema.prefixItems) do
-				if validate(s, data[i], root) then
+				if data[i] == nil then
+					break
+				elseif validate(s, data[i], root) then
 					p = i
 				else
 					return false
@@ -310,7 +343,7 @@ type_validators.table = function (schema : json_schema_object, data : any, root
 			end
 		end
 
-		if schema.items then
+		if schema.items ~= nil then
 			for i = p+1, #data do
 				if not validate(schema.items, data[i], root) then
 					return false
@@ -318,7 +351,7 @@ type_validators.table = function (schema : json_schema_object, data : any, root
 			end
 		end
 
-		if schema.contains then
+		if schema.contains ~= nil then
 			local found = false
 			for i = 1, #data do
 				if validate(schema.contains, data[i], root) then
@@ -330,39 +363,11 @@ type_validators.table = function (schema : json_schema_object, data : any, root
 				return false
 			end
 		end
-
-		return true
 	end
-	return false
-end
 
-type_validators.object = function (schema : schema_t, data : any, root : json_schema_object) : boolean
-	if data is table then
-		for k in pairs(data) do
-			if not k is string then
-				return false
-			end
-		end
-
-		return type_validators.table(schema, data, root)
-	end
-	return false
+	return true;
 end
 
-type_validators.array = function (schema : schema_t, data : any, root : json_schema_object) : boolean
-	if data is table then
-
-		-- just check that there the keys are all numbers
-		for i in pairs(data) do
-			if not i is number then
-				return false
-			end
-		end
-
-		return type_validators.table(schema, data, root)
-	end
-	return false
-end
 
 json_schema_object.validate = validate;