Add XML parser

Author: nichtich

README.md

@@ -13,6 +13,8 @@ This web service implements a **[Data Validation API](#API)** being specified as
   - [From sources](#from-sources)
   - [With Docker](#with-docker)
 - [Configuration](#configuration)
+  - [Profiles](#profiles)
+  - [Checks](#checks)
 - [API](#api)
   - [GET /profiles](#get-profiles)
   - [GET /{profile}/validate](#get-profilevalidate)
@@ -65,6 +67,16 @@ The [default configuration](config.default.json) contains some base formats. To
 
 Each application profile is configured with a JSON object having a unique `id`, a list of `checks`, and additional metadata. See [profiles configuration JSON Schema](lib/validate/profiles-schema.json) for details.
 
+### Checks
+
+Each check is either a string, referencing another profile or a base format, or a JSON object for a more complex check.
+
+### Base formats
+
+- `json`
+- `xml`
+- ...
+
 ## API
 
 The webservice provides one endpoint to [list application profiles](#get-profiles) and one **Data Validation API** endpoint for each profile to validate data.

config.default.json

@@ -7,6 +7,10 @@
     {
       "id": "json",
       "checks": ["json"]
+    },
+    {
+      "id": "xml",
+      "checks": ["xml"]
     }
   ]
 }

lib/__init__.py

@@ -1,4 +1,5 @@
-from .validate import Validator, ValidationError, validateJSON, parseJSON
+from .validate import Validator, ValidationError, validateJSON, parseJSON, parseXML
 from .service import ValidationService
 
-__all__ = [ValidationService, Validator, ValidationError, validateJSON, parseJSON]
+__all__ = [ValidationService, Validator, ValidationError,
+           validateJSON, parseJSON, parseXML]

lib/validate/__init__.py

@@ -2,5 +2,6 @@
 from .error import ValidationError
 from .jsonschema import validateJSON
 from .json import parseJSON
+from .xml import parseXML
 
-__all__ = [Validator, ValidationError, validateJSON, parseJSON]
+__all__ = [Validator, ValidationError, validateJSON, parseJSON, parseXML]

lib/validate/validator.py

@@ -2,6 +2,7 @@
 from pathlib import Path
 from .json import parseJSON
 from .jsonschema import validateJSON
+from .xml import parseXML
 
 
 schema = json.load((Path(__file__).parent / 'profiles-schema.json').open())
@@ -35,3 +36,5 @@ def execute(self, profile, data=None, file=None):
         for check in checks:
             if check == "json":
                 parseJSON(data)
+            elif check == "xml":
+                parseXML(data)

lib/validate/xml.py

@@ -0,0 +1,21 @@
+from .error import ValidationError
+from xml.dom.minidom import parse, parseString, Document
+from xml.parsers.expat import ExpatError
+from xml.parsers.expat.errors import messages
+from io import IOBase
+
+
+def parseXML(data) -> Document:
+    try:
+        if isinstance(data, IOBase):
+            return parse(data)
+        else:
+            return parseString(data)
+    except ExpatError as e:
+        msg = messages[e.code]
+        col = e.offset + 1
+        pos = {
+            "line": str(e.lineno),
+            "linecol": f"{e.lineno}:{col}",
+        }
+        raise ValidationError(msg, pos)

tests/test_service.py

@@ -16,10 +16,14 @@ def test_config():
         "id": "json",
         "checks": ["json"],
         "url": "https://json.org/"
+    }, {
+        "id": "xml",
+        "checks": ["xml"]
     }]
 
     service = ValidationService(profiles=profiles)
-    assert service.profiles() == [{"id": "json", "url": "https://json.org/"}]
+    assert service.profiles() == [
+        {"id": "json", "url": "https://json.org/"}, {"id": "xml"}]
 
     with pytest.raises(Exception, match=r"This service does not support passing data via URL"):
         service.validate('json', url="http://example.org/")

tests/test_xml.py

@@ -0,0 +1,21 @@
+from lib import ValidationError, parseXML
+from xml.dom.minidom import Document
+from xml.parsers.expat.errors import codes
+
+not_wellformed = [
+    ('<a x="1"\n木="1" x="2"/>', {  # string
+        "message": 'duplicate attribute',
+        "position": {"line": "2", "linecol": "2:7"}}),
+    ('<?xml version="1.0"?>\n<木/>?'.encode("UTF-8"), {  # binary
+        "message": 'not well-formed (invalid token)',
+        "position": {"line": "2", "linecol": "2:5"}}),
+]
+
+
+def test_not_wellformed():
+    for (data, err) in not_wellformed:
+        try:
+            assert isinstance(parseXML(data), Document)
+            assert 0 == "ValidationError should have been thrown!"  # pragma: no cover
+        except ValidationError as e:
+            assert e.to_dict() == err