text generator and test (#295)

* text generator and test

* fixed style and 256 to 255 error

* fixed eof

* new eof line

* one line eof

* updated num_rows test with multiple row counts

* removed whitespace in eof

* moved text_generator test into distinct_gen test folder

synthetic_data/distinct_generators/text_generator.py

@@ -0,0 +1,76 @@
+import numpy as np
+import string
+from numpy.random import Generator
+from typing import List, Optional
+
+
+def random_string(
+    rng: Generator,
+    chars: Optional[List[str]] = None,
+    num_rows: int = 1,
+    str_len_min: int = 1,
+    str_len_max: int = 256,
+    ) -> np.array:
+    """
+    Randomly generates an array of strings with length between a min and max value
+
+    :param rng: the np rng object used to generate random values
+    :type rng: numpy Generator
+    :param chars: a list of values that are allowed in a string or None
+    :type chars: List[str], None
+    :param num_rows: the number of rows in np array generated
+    :type num_rows: int, optional
+    :param str_len_min: the minimum length a string can be
+    :type str_len_min: int, optional
+    :param str_len_max: the maximum length a string can be
+    :type str_len_max: int, optional
+
+    :return: numpy array of strings
+    """
+    if chars is None:
+        chars = list(
+            string.ascii_uppercase
+            + string.ascii_lowercase
+            + string.digits
+            + " "
+            + string.punctuation
+        )
+    string_list = []
+
+    for _ in range(num_rows):
+        length = rng.integers(str_len_min, str_len_max)
+        string_entry = "".join(rng.choice(chars, (length,)))
+        string_list.append(string_entry)
+
+    return np.array(string_list)
+
+
+def random_text(
+    rng: Generator,
+    chars: Optional[str] = None,
+    num_rows: int = 1,
+    str_len_min: int = 256,
+    str_len_max: int = 1000,
+) -> np.array:
+    """
+    Randomly generates an array of text with length between a min and max value
+
+    :param rng: the np rng object used to generate random values
+    :type rng: numpy Generator
+    :param chars: a list of values that are allowed in a string or None
+    :type chars: List[str], None
+    :param num_rows: the number of rows in np array generated
+    :type num_rows: int, optional
+    :param str_len_min: the minimum length a string can be (must be larger than 255)
+    :type str_len_min: int, optional
+    :param str_len_max: the maximum length a string can be
+    :type str_len_max: int, optional
+
+    :return: numpy array of text
+    """
+    if str_len_min < 256:
+        raise ValueError(
+            f"str_len_min must be > 255. " f"Value provided: {str_len_min}."
+        )
+
+    return random_string(rng, chars, num_rows, str_len_min, str_len_max)

tests/distinct_generators/test_text_generator.py

@@ -0,0 +1,48 @@
+import unittest
+from unittest import mock
+import pandas as pd
+import numpy as np
+from synthetic_data.distinct_generators.text_generator import random_string, random_text
+
+
+class TestTextGeneratorFunctions(unittest.TestCase):
+    def setUp(self):
+        self.rng = np.random.default_rng(12345)
+
+    def test_return_type(self):
+        str_arr = random_string(self.rng)
+        txt_arr = random_text(self.rng)
+        for x in str_arr:
+            self.assertIsInstance(x, np.str_)
+        for x in txt_arr:
+            self.assertIsInstance(x, np.str_)
+
+    def test_str_length(self):
+        str_arr = random_string(self.rng, str_len_min=1, str_len_max=256)
+        txt_arr = random_text(self.rng, str_len_min=256, str_len_max=1000)
+        with self.assertRaises(ValueError):
+            random_text(self.rng, str_len_min=255)
+
+        self.assertLessEqual(len(str_arr[0]), 256)
+        self.assertGreaterEqual(len(str_arr[0]), 1)
+        self.assertLessEqual(len(txt_arr[0]), 1000)
+        self.assertGreaterEqual(len(txt_arr[0]), 256)
+
+    def test_num_rows(self):
+        num_rows = [1,5,10]
+        for nr in num_rows:
+            str_arr = random_string(self.rng, num_rows=nr)
+            txt_arr = random_text(self.rng, num_rows=nr)
+            self.assertEqual(str_arr.size, nr)
+            self.assertEqual(txt_arr.size, nr)
+
+    def test_chars(self):
+        chars_set = {"0","1"}
+        str_arr = random_string(self.rng, chars=["0","1"])
+        txt_arr = random_text(self.rng, chars=["0","1"])
+        for s in str_arr:
+            for char in s:
+                self.assertIn(char, chars_set)
+        for s in txt_arr:
+            for char in s:
+                self.assertIn(char, chars_set)