Implement dataclass and Faker.seed

Makefile

@@ -0,0 +1,8 @@
+# Variables
+PYTHON := python
+PYTHONPATH := `pwd`
+
+# Protobuf
+.PHONY: protoc
+protoc:
+	protoc -I=./patient_data/ --python_out=./patient_data/ ./patient_data/protobuf_grpc.proto

Pipfile

@@ -13,11 +13,11 @@ fastapi = ">=0.105"
 brotli-asgi = ">=1.4"
 hypercorn = ">=0.15"
 httpx = {version = ">=0.25", extras = ["http2"]}
-
 # Optional production cloud specific packages.
 # Install using:  pipenv install [package] --catgeories cloud_logger cloud_secret
 #
 # DON'T install multiple vendors.  Install ONLY one.
+ujson = "*"
 
 [cloud_logger]
 #infisical = ">=1.2"
@@ -29,7 +29,6 @@ httpx = {version = ">=0.25", extras = ["http2"]}
 #google-cloud-secret-manager = ">=2.16"
 #boto3 = ">=1.26"
 #oci = ">=2.104"
-
 # Developement tools.
 # Install using:  pipenv install [package] --dev
 

patient_data/gen_data.py

@@ -0,0 +1,155 @@
+import random
+from faker import Faker
+from ujson import dumps
+from enum import Enum
+import sys
+from datetime import date
+from dateutil.relativedelta import relativedelta
+import protobuf_grpc as proto
+import patient_dataclass as pd
+from dataclasses import asdict
+
+faker = Faker('en_US')
+
+
+# Method to create a fake name
+def generate_name(seed, birth_date, gender=''):
+    Faker.seed(seed)
+    if gender == 'Non-Binary' or gender == 'Other':
+        full_name = faker.name_nonbinary()
+    elif gender == 'Male':
+        full_name = faker.name_male()
+    elif gender == 'Female':
+        full_name = faker.name_female()
+    else:
+        full_name = faker.name()
+    n = pd.Name(
+        faker.word(ext_word_list=['usual', 'official', 'temp', 'nickname', 'anonymous', 'old', 'maiden']),
+        full_name,
+        full_name.split(' ')[0],
+        full_name.split(' ')[1],
+        faker.date_between(start_date=birth_date)
+    )
+
+    return n
+
+
+# Method to create a fake telecom
+def generate_telecom(seed, rank, birth_date, names):
+    Faker.seed(seed)
+    t = pd.Telecom(
+        faker.word(ext_word_list=['phone', 'fax', 'email', 'sms']),
+        faker.word(ext_word_list=['home', 'work', 'temp', 'old', 'mobile']),
+        rank,
+        faker.date_between(start_date=birth_date)
+    )
+    if t.system == 'email':
+        t.value = f'{names[random.randint(0, len(names) - 1)].given}' \
+                  f'{names[random.randint(0, len(names) - 1)].family}@{faker.domain_name()}'
+    else:
+        t.value = faker.phone_number()
+
+    return t
+
+
+# Method to create a fake address
+def generate_address(seed, birth_date):
+    Faker.seed(seed)
+    a = pd.Address(
+        faker.word(ext_word_list=['home', 'work', 'temp', 'old', 'billing']),
+        faker.word(ext_word_list=['postal', 'physical', 'both']),
+        faker.street_address(),
+        faker.city(),
+        faker.state(),
+        faker.postcode(),
+        'USA',
+        faker.date_between(start_date=birth_date),
+    )
+    a.text = f'{a.line} {a.city}, {a.state} {a.zipcode} {a.country}'
+    if a.use in ['temp', 'old']:
+        a.period_end = faker.date_between(start_date=a.period_start + relativedelta(years=1))
+
+    return a
+
+
+# Format class for ease of selecting data format
+class Format(Enum):
+    Json = 1
+    Protobuf = 2
+
+    def __repr__(self):
+        return self.value
+
+    def __str__(self):
+        return str(self.value)
+
+
+def main(size, data_format: Format):
+    # Set-up
+    curr_size = 0
+    seed = 1
+    patient_list = []
+
+    # Set the maximum number of items for each repeated dataclass
+    max_names = 3
+    max_telecoms = 2
+    max_addresses = 2
+    max_contacts = 4
+
+    # Generate patients
+    while curr_size < size:
+        Faker.seed(seed)
+        random.seed(seed)
+        p = pd.Patient(
+            faker.pyint(min_value=11111, max_value=99999),
+            faker.boolean(chance_of_getting_true=90),
+            faker.word(ext_word_list=['Male', 'Female', 'Non-binary', 'Other']),
+            faker.date_of_birth(minimum_age=18, maximum_age=100),
+            faker.word(ext_word_list=['Married', 'Single', 'Separated', 'Divorced', 'Widowed'])
+        )
+
+        # Find date of death if applicable
+        if p.birth_date < date.today() - relativedelta(years=65) and \
+           faker.boolean(chance_of_getting_true=15):
+            p.deceased_on = faker.date_between(start_date=date.today()-relativedelta(years=10)).strftime('%m/%d/%Y')
+
+        # Generate names
+        for x in range(random.randint(1, max_names)):
+            seed += 1
+            p.names.append(generate_name(seed, p.birth_date, p.gender))
+
+        # Generate telecoms
+        for x in range(random.randint(1, max_telecoms)):
+            seed += 1
+            p.telecoms.append(generate_telecom(seed, x+1, p.birth_date, p.names))
+
+        # Generate addresses
+        for x in range(random.randint(1, max_addresses)):
+            seed += 1
+            p.addresses.append(generate_address(seed, p.birth_date))
+        # Generate contacts
+        for x in range(random.randint(1, max_contacts)):
+            c_birth_date = faker.date_of_birth(minimum_age=18, maximum_age=100)
+            c = pd.Contact(
+                faker.word(ext_word_list=['emergency', 'family', 'guardian', 'friend', 'partner', 'work',
+                                          'caregiver', 'agent', 'guarantor', 'parent']),
+                generate_name(seed+1, c_birth_date),
+                generate_address(seed+1, c_birth_date)
+            )
+            seed += 3
+            c.telecom = generate_telecom(seed, 1, c_birth_date, [c.name])
+            p.contacts.append(c)
+        # Add patient to list and keep track of size
+        patient_list.append(p)
+        curr_size += sys.getsizeof(p)
+        seed += 1
+
+    match data_format:
+        case Format.Json:
+            return dumps([asdict(p) for p in patient_list], indent=4, default=str)
+        case Format.Protobuf:
+            return proto.generate_patients(patient_list).all_patients
+
+
+if __name__ == '__main__':
+    print(main(128, Format.Protobuf))

patient_data/patient_dataclass.py

@@ -0,0 +1,56 @@
+from dataclasses import dataclass, field
+from datetime import date
+
+
+@dataclass
+class Name:
+    use: str
+    text: str
+    given: str
+    family: str
+    active: date
+
+
+@dataclass
+class Telecom:
+    system: str
+    use: str
+    rank: int
+    active: date
+    value: str = field(default='')
+
+
+@dataclass
+class Address:
+    use: str
+    type: str
+    line: str
+    city: str
+    state: str
+    zipcode: str
+    country: str
+    period_start: date
+    text: str = field(default='')
+    period_end: date = field(default=None)
+
+
+@dataclass
+class Contact:
+    relation: str
+    name: Name
+    address: Address
+    telecom: Telecom = field(default=Telecom)
+
+
+@dataclass
+class Patient:
+    id: int
+    active: bool
+    gender: str
+    birth_date: date
+    marital_status: str
+    deceased_on: str = field(default='')
+    names: list[Name] = field(default_factory=list)
+    telecoms: list[Telecom] = field(default_factory=list)
+    addresses: list[Address] = field(default_factory=list)
+    contacts: list[Contact] = field(default_factory=list)

patient_data/protobuf_grpc.proto

@@ -0,0 +1,54 @@
+syntax = 'proto2';
+
+message Name {
+  optional string use = 1;
+  optional string text = 2;
+  optional string given = 3;
+  optional string family = 4;
+  optional string active = 5;
+}
+
+message Telecom {
+  optional string system = 1;
+  optional string use = 2;
+  optional int32 rank = 3;
+  optional string active = 4;
+  optional string value = 5;
+}
+
+message Address {
+  optional string use = 1;
+  optional string type = 2;
+  optional string line = 3;
+  optional string city = 4;
+  optional string state = 5;
+  optional string zipcode = 6;
+  optional string country = 7;
+  optional string period_start = 8;
+  optional string text = 9;
+  optional string period_end = 10;
+}
+
+message Contact {
+  optional string relation = 1;
+  optional Name name = 2;
+  optional Address address = 3;
+  optional Telecom telecom = 4;
+}
+
+message Patient {
+  optional int32 id = 1;
+  optional bool active = 2;
+  optional string gender = 3;
+  optional string birth_date = 4;
+  optional string marital_status = 5;
+  optional string deceased_on = 6;
+  repeated Name names = 7;
+  repeated Telecom telecoms = 8;
+  repeated Address addresses = 9;
+  repeated Contact contacts = 10;
+}
+
+message Patients {
+  repeated Patient all_patients = 1;
+}