Update validation pipeline to use a json schema

kingfisher_scrapy/exceptions.py

@@ -8,7 +8,3 @@ class AuthenticationError(KingfisherScrapyError):
 
 class SpiderArgumentError(KingfisherScrapyError):
     """Raised when a spider argument's value is invalid"""
-
-
-class MissingRequiredFieldError(KingfisherScrapyError, KeyError):
-    """Raised when an item is missing a required field"""

kingfisher_scrapy/item_schema/file.json

@@ -0,0 +1,28 @@
+{
+  "$schema": "http://json-schema.org/draft-04/schema#",
+  "allOf": [
+    {
+      "$ref": "item.json#/definitions/KingfisherFileItem"
+    }
+  ],
+  "title": "File",
+  "type": "object",
+  "properties": {
+    "post_to_api": {
+      "type": [
+        "boolean"
+      ]
+    },
+    "path": {
+      "description": "For the KingfisherProcessAPI extension to read the file.",
+      "type": [
+        "string"
+      ]
+    },
+    "files_store": {
+      "type": [
+        "string"
+      ]
+    }
+  }
+}

kingfisher_scrapy/item_schema/file_error.json

@@ -0,0 +1,19 @@
+{
+  "$schema": "http://json-schema.org/draft-04/schema#",
+  "title": "File Error",
+  "type": "object",
+  "allOf": [
+    {
+      "$ref": "item.json#/definitions/KingfisherItem"
+    }
+  ],
+  "properties": {
+    "errors": {
+      "type": "string",
+      "minLength": 1
+    }
+  },
+  "required": [
+    "errors"
+  ]
+}

kingfisher_scrapy/item_schema/file_item.json

@@ -0,0 +1,19 @@
+{
+  "$schema": "http://json-schema.org/draft-04/schema#",
+  "allOf": [
+    {
+      "$ref": "item.json#/definitions/KingfisherFileItem"
+    }
+  ],
+  "title": "File Item",
+  "type": "object",
+  "properties": {
+    "number": {
+      "type": "integer",
+      "minimum": 1
+    }
+  },
+  "required": [
+    "number"
+  ]
+}

kingfisher_scrapy/item_schema/item.json

@@ -0,0 +1,69 @@
+{
+  "$schema": "http://json-schema.org/draft-04/schema#",
+  "definitions": {
+    "KingfisherItem": {
+      "title": "Kingfisher Item",
+      "description": "A generic item with file_name and url to be extended by other items",
+      "type": "object",
+      "properties": {
+        "file_name": {
+          "type": "string",
+          "pattern": "^[^/]+$"
+        },
+        "url": {
+          "type": "string",
+          "format": "uri"
+        }
+      },
+      "required": [
+        "file_name",
+        "url"
+      ]
+    },
+    "KingfisherFileItem": {
+      "title": "Kingfisher Item",
+      "description": "A base object to be extended by other File type items",
+      "type": "object",
+      "allOf": [
+        {
+          "$ref": "#/definitions/KingfisherItem"
+        }
+      ],
+      "properties": {
+        "data_type": {
+          "type": "string",
+          "enum": [
+            "record",
+            "release",
+            "record_list",
+            "release_list",
+            "compiled_release",
+            "record_package",
+            "release_package",
+            "record_package_list",
+            "release_package_list",
+            "record_package_list_in_results",
+            "release_package_list_in_results",
+            "release_package_json_lines",
+            "record_package_json_lines",
+            "release_package_in_ocdsReleasePackage_in_list_in_results",
+            "release_in_Release"
+          ]
+        },
+        "encoding": {
+          "type": [
+            "string"
+          ]
+        },
+        "data": {
+          "type": "string",
+          "minLength": 1
+        }
+      },
+      "required": [
+        "data",
+        "data_type"
+      ]
+    }
+  }
+}

kingfisher_scrapy/items.py

@@ -1,23 +1,12 @@
 # https://docs.scrapy.org/en/latest/topics/items.html
-import scrapy
 
-from kingfisher_scrapy.exceptions import MissingRequiredFieldError
+import scrapy
 
 
 class KingfisherItem(scrapy.Item):
     file_name = scrapy.Field()
     url = scrapy.Field()
-
-    def validate(self):
-        """
-        Raises an error if any required field is missing.
-
-        :raises kingfisher_scrapy.extensions.MissingRequiredFieldError: if any required field is missing
-        """
-        if hasattr(self, 'required'):
-            for field in self.required:
-                if field not in self:
-                    raise MissingRequiredFieldError(field)
+    validate = True
 
 
 class File(KingfisherItem):
@@ -32,34 +21,13 @@ class File(KingfisherItem):
     path = scrapy.Field()
     files_store = scrapy.Field()
 
-    required = [
-        'file_name',
-        'url',
-        'data',
-        'data_type',
-    ]
-
 
 class FileItem(KingfisherItem):
     number = scrapy.Field()
     data = scrapy.Field()
     data_type = scrapy.Field()
     encoding = scrapy.Field()
 
-    required = [
-        'number',
-        'file_name',
-        'url',
-        'data',
-        'data_type',
-    ]
-
 
 class FileError(KingfisherItem):
     errors = scrapy.Field()
-
-    required = [
-        'file_name',
-        'url',
-        'errors',
-    ]

kingfisher_scrapy/pipelines.py

@@ -1,12 +1,25 @@
 # https://docs.scrapy.org/en/latest/topics/item-pipeline.html
 # https://docs.scrapy.org/en/latest/topics/signals.html#item-signals
+import os
+import pathlib
+
+import jsonref as jsonref
+from jsonschema import FormatChecker
+from jsonschema.validators import Draft4Validator
 
 
 class Validate:
+    def __init__(self):
+        self.validators = {}
+        schema_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'item_schema')
+        for item in ['file', 'file_error', 'file_item']:
+            filename = os.path.join(schema_path, f'{item}.json')
+            with open(filename) as f:
+                schema = jsonref.load(f, base_uri=pathlib.Path(os.path.join(schema_path), 'item_schema').as_uri())
+            class_name = ''.join(word.title() for word in item.split('_'))
+            self.validators[class_name] = Draft4Validator(schema, format_checker=FormatChecker())
+
     def process_item(self, item, spider):
         if hasattr(item, 'validate'):
-            # We call this in the item pipeline to guarantee that all items are validated. However, its backtrace isn't
-            # as helpful for debugging, so we could also call it in ``BaseSpider`` if this becomes an issue.
-            item.validate()
-
+            self.validators.get(item.__class__.__name__).validate(dict(item))
         return item

requirements.in

@@ -2,8 +2,11 @@
 # https://github.com/open-contracting/deploy/blob/master/salt/ocdskingfishercollect/scrapyd-requirements.txt
 
 jsonpointer
+jsonref
+jsonschema
 rarfile
 requests
+rfc3987
 Scrapy
 scrapyd-client
 ijson>=3

requirements.txt

@@ -4,7 +4,7 @@
 #
 #    pip-compile
 #
-attrs==19.3.0             # via automat, service-identity, twisted
+attrs==19.3.0             # via automat, jsonschema, service-identity, twisted
 automat==0.8.0            # via twisted
 certifi==2019.11.28       # via requests, sentry-sdk
 cffi==1.13.2              # via cryptography
@@ -15,8 +15,11 @@ cssselect==1.1.0          # via parsel, scrapy
 hyperlink==19.0.0         # via twisted
 idna==2.8                 # via hyperlink, requests
 ijson==3.0.3
+importlib-metadata==1.6.1  # via jsonschema
 incremental==17.5.0       # via twisted
 jsonpointer==2.0
+jsonref==0.2
+jsonschema==3.2.0
 lxml==4.4.2               # via parsel, scrapy
 parsel==1.5.2             # via scrapy
 protego==0.1.16           # via scrapy
@@ -26,17 +29,20 @@ pycparser==2.19           # via cffi
 pydispatcher==2.0.5       # via scrapy
 pyhamcrest==1.9.0         # via twisted
 pyopenssl==19.1.0         # via scrapy
+pyrsistent==0.16.0        # via jsonschema
 queuelib==1.5.0           # via scrapy
 rarfile==3.1
 requests==2.22.0
+rfc3987==1.3.8
 scrapy==1.8.0
 scrapyd-client==1.1.0
 sentry-sdk==0.14.4
 service-identity==18.1.0  # via scrapy
-six==1.13.0               # via automat, cryptography, parsel, protego, pyhamcrest, pyopenssl, scrapy, scrapyd-client, w3lib
+six==1.13.0               # via automat, cryptography, jsonschema, parsel, protego, pyhamcrest, pyopenssl, pyrsistent, scrapy, scrapyd-client, w3lib
 twisted==20.3.0           # via scrapy
 urllib3==1.25.7           # via requests, sentry-sdk
 w3lib==1.21.0             # via parsel, scrapy
+zipp==3.1.0               # via importlib-metadata
 zope.interface==4.7.1     # via scrapy, twisted
 
 # The following packages are considered to be unsafe in a requirements file:

requirements_dev.txt

@@ -21,13 +21,15 @@ flake8==3.7.9
 hyperlink==19.0.0
 idna==2.8
 ijson==3.0.3
-importlib-metadata==1.3.0  # via pluggy, pytest
+importlib-metadata==1.6.1
 incremental==17.5.0
 isort==4.3.21
 jsonpointer==2.0
+jsonref==0.2
+jsonschema==3.2.0
 lxml==4.4.2
 mccabe==0.6.1             # via flake8
-more-itertools==8.0.2     # via pytest, zipp
+more-itertools==8.0.2     # via pytest
 packaging==19.2           # via pytest
 parsel==1.5.2
 pip-tools==5.1.0
@@ -43,11 +45,13 @@ pyflakes==2.1.1           # via flake8
 pyhamcrest==1.9.0
 pyopenssl==19.1.0
 pyparsing==2.4.5          # via packaging
+pyrsistent==0.16.0
 pytest-cov==2.8.1
 pytest==5.3.2
 queuelib==1.5.0
 rarfile==3.1
 requests==2.22.0
+rfc3987==1.3.8
 scrapy==1.8.0
 scrapyd-client==1.1.0
 sentry-sdk==0.14.4
@@ -57,7 +61,7 @@ twisted==20.3.0
 urllib3==1.25.7
 w3lib==1.21.0
 wcwidth==0.1.7            # via pytest
-zipp==0.6.0               # via importlib-metadata
+zipp==3.1.0
 zope.interface==4.7.1
 
 # The following packages are considered to be unsafe in a requirements file: