analyzer request

presidio-analyzer/Pipfile.lock

@@ -107,6 +107,14 @@
             "markers": "python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'",
             "version": "==2.10"
         },
+        "importlib-metadata": {
+            "hashes": [
+                "sha256:ace61d5fc652dc280e7b6b4ff732a9c2d40db2c0f92bc6cb74e07b73d53a1771",
+                "sha256:fa5daa4477a7414ae34e95942e4dd07f62adf589143c875c133c1e53c4eff38d"
+            ],
+            "markers": "python_version < '3.8'",
+            "version": "==3.4.0"
+        },
         "itsdangerous": {
             "hashes": [
                 "sha256:321b033d07f2a4136d3ec762eac9f16a10ccd60f53c0c91af90217ace7ba1f19",
@@ -320,7 +328,7 @@
                 "sha256:30639c035cdb23534cd4aa2dd52c3bf48f06e5f4a941509c8bafd8ce11080259",
                 "sha256:8b74bedcbbbaca38ff6d7491d76f2b06b3592611af620f8426e82dddb04a5ced"
             ],
-            "markers": "python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2'",
+            "markers": "python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'",
             "version": "==1.15.0"
         },
         "spacy": {
@@ -390,6 +398,15 @@
             "markers": "python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'",
             "version": "==4.56.0"
         },
+        "typing-extensions": {
+            "hashes": [
+                "sha256:7cb407020f00f7bfc3cb3e7881628838e69d8f3fcab2f64742a5e76b2f841918",
+                "sha256:99d4073b617d30288f569d3f13d2bd7548c3a7e4c8de87db09a9d29bb3a4a60c",
+                "sha256:dafc7639cde7f1b6e1acc0f457842a83e722ccca8eef5270af2d74792619a89f"
+            ],
+            "markers": "python_version < '3.8'",
+            "version": "==3.7.4.3"
+        },
         "urllib3": {
             "hashes": [
                 "sha256:19188f96923873c92ccb987120ec4acaa12f0461fa9ce5d3d0772bc965a39e08",
@@ -412,6 +429,14 @@
             ],
             "markers": "python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'",
             "version": "==1.0.1"
+        },
+        "zipp": {
+            "hashes": [
+                "sha256:102c24ef8f171fd729d46599845e95c7ab894a4cf45f5de11a44cc7444fb1108",
+                "sha256:ed5eee1974372595f9e416cc7bbeeb12335201d8081ca8a0743c954d4446e5cb"
+            ],
+            "markers": "python_version >= '3.6'",
+            "version": "==3.4.0"
         }
     },
     "develop": {
@@ -454,6 +479,14 @@
             ],
             "version": "==1.0.2"
         },
+        "importlib-metadata": {
+            "hashes": [
+                "sha256:ace61d5fc652dc280e7b6b4ff732a9c2d40db2c0f92bc6cb74e07b73d53a1771",
+                "sha256:fa5daa4477a7414ae34e95942e4dd07f62adf589143c875c133c1e53c4eff38d"
+            ],
+            "markers": "python_version < '3.8'",
+            "version": "==3.4.0"
+        },
         "iniconfig": {
             "hashes": [
                 "sha256:011e24c64b7f47f6ebd835bb12a743f2fbe9a26d4cecaa7f53bc4f35ee9da8b3",
@@ -529,7 +562,7 @@
                 "sha256:c203ec8783bf771a155b207279b9bccb8dea02d8f0c9e5f8ead507bc3246ecc1",
                 "sha256:ef9d7589ef3c200abe66653d3f1ab1033c3c419ae9b9bdb1240a85b024efc88b"
             ],
-            "markers": "python_version >= '2.6' and python_version not in '3.0, 3.1, 3.2'",
+            "markers": "python_version >= '2.6' and python_version not in '3.0, 3.1, 3.2, 3.3'",
             "version": "==2.4.7"
         },
         "pytest": {
@@ -552,8 +585,25 @@
                 "sha256:806143ae5bfb6a3c6e736a764057db0e6a0e05e338b5630894a5f779cabb4f9b",
                 "sha256:b3bda1d108d5dd99f4a20d24d9c348e91c4db7ab1b749200bded2f839ccbe68f"
             ],
-            "markers": "python_version >= '2.6' and python_version not in '3.0, 3.1, 3.2'",
+            "markers": "python_version >= '2.6' and python_version not in '3.0, 3.1, 3.2, 3.3'",
             "version": "==0.10.2"
+        },
+        "typing-extensions": {
+            "hashes": [
+                "sha256:7cb407020f00f7bfc3cb3e7881628838e69d8f3fcab2f64742a5e76b2f841918",
+                "sha256:99d4073b617d30288f569d3f13d2bd7548c3a7e4c8de87db09a9d29bb3a4a60c",
+                "sha256:dafc7639cde7f1b6e1acc0f457842a83e722ccca8eef5270af2d74792619a89f"
+            ],
+            "markers": "python_version < '3.8'",
+            "version": "==3.7.4.3"
+        },
+        "zipp": {
+            "hashes": [
+                "sha256:102c24ef8f171fd729d46599845e95c7ab894a4cf45f5de11a44cc7444fb1108",
+                "sha256:ed5eee1974372595f9e416cc7bbeeb12335201d8081ca8a0743c954d4446e5cb"
+            ],
+            "markers": "python_version >= '3.6'",
+            "version": "==3.4.0"
         }
     }
 }

presidio-analyzer/app.py

@@ -1,5 +1,6 @@
 from presidio_analyzer.presidio_logger import PresidioLogger
 from presidio_analyzer.analyzer_engine import AnalyzerEngine
+from presidio_analyzer.analyzer_request import AnalyzerRequest
 from flask import Flask, request
 import json
 import os
@@ -40,22 +41,15 @@ class Server:
             Executes the analyzer function
             """
             # Parse the request params
-            req_data = request.get_json()
-
-            text = req_data.get("text")
-            language = req_data.get("language")
-            entities = req_data.get("entities")
-            correlation_id = req_data.get("correlation_id")
-            score_threshold = req_data.get("score_threshold")
-            trace = req_data.get("trace")
+            req_data = AnalyzerRequest(request.get_json())
             try:
                 recognizer_result_list = self.engine.analyze(
-                    text,
-                    language,
-                    correlation_id=correlation_id,
-                    score_threshold=score_threshold,
-                    entities=entities,
-                    trace=trace)
+                    req_data.text,
+                    req_data.language,
+                    correlation_id=req_data.correlation_id,
+                    score_threshold=req_data.score_threshold,
+                    entities=req_data.entities,
+                    trace=req_data.trace)
 
                 return json.dumps(recognizer_result_list, default=lambda o: o.to_json(),
                                   sort_keys=True, indent=4)

presidio-analyzer/presidio_analyzer/__init__.py

@@ -8,6 +8,7 @@ from presidio_analyzer.pattern_recognizer import PatternRecognizer
 from presidio_analyzer.remote_recognizer import RemoteRecognizer
 from presidio_analyzer.recognizer_registry import RecognizerRegistry
 from presidio_analyzer.analyzer_engine import AnalyzerEngine
+from presidio_analyzer.analyzer_request import AnalyzerRequest
 
 
 __all__ = [
@@ -21,4 +22,5 @@ __all__ = [
     "RemoteRecognizer",
     "RecognizerRegistry",
     "AnalyzerEngine",
+    "AnalyzerRequest"
 ]

presidio-analyzer/presidio_analyzer/analyzer_request.py

@@ -0,0 +1,11 @@
+class AnalyzerRequest:
+    """
+    Analyzer request data
+    """
+    def __init__(self, req_data):
+        self.text = req_data.get("text")
+        self.language = req_data.get("language")
+        self.entities = req_data.get("entities")
+        self.correlation_id = req_data.get("correlation_id")
+        self.score_threshold = req_data.get("score_threshold")
+        self.trace = req_data.get("trace")