Add tests for validation (#13)

.github/workflows/lint_and_package.yml

@@ -1,4 +1,4 @@
-name: Lint
+name: Lint and tests
 on: [push, pull_request]
 jobs:
   lint:
@@ -14,6 +14,22 @@ jobs:
     - name: Flake8
       run: flake8 checkcel --ignore=E501,W504
 
+  pytest:
+    runs-on: ubuntu-latest
+    needs: lint
+    steps:
+    - name: Checkout
+      uses: actions/checkout@v2
+    - uses: actions/setup-python@v1
+      with:
+        python-version: 3.7
+    - name: Install pytest
+      run: pip install pytest
+    - name: Install checkcel
+      run: pip install .
+    - name: Test
+      run: pytest -v tests/
+
   pypi:
     runs-on: ubuntu-latest
     needs: lint

checkcel/checkcel.py

@@ -10,7 +10,8 @@
 class Checkcel(Checkplate):
     def __init__(
         self,
-        source,
+        source="",
+        data=None,
         format="spreadsheet",
         delimiter=",",
         sheet=0,
@@ -22,6 +23,7 @@ def __init__(
         self.missing_validators = None
         self.missing_fields = None
         self.source = source
+        self.data = data
         self.format = format
         self.delimiter = delimiter
         self.sheet = int(sheet)
@@ -31,6 +33,9 @@ def __init__(
         self.column_set = set()
         self.ignore_missing_validators = False
 
+        if not (self.source or self.data is not None):
+            raise Exception("Need to provide either a source or the data (as a pandas dataframe)")
+
         if format not in ["spreadsheet", "tabular"]:
             raise Exception("Type must be either spreadsheet or tabular")
 
@@ -84,21 +89,25 @@ def _log_missing(self, missing_items):
 
     def validate(self):
         self.logger.info(
-            "\nValidating {}(source={})".format(self.__class__.__name__, self.source)
+            "\nValidating {}{}".format(self.__class__.__name__, "(source={})".format(self.source) if self.source else "")
         )
 
-        if self.format == "spreadsheet":
-            with warnings.catch_warnings():
-                warnings.simplefilter("ignore")
-                df = pandas.read_excel(self.source, sheet_name=self.sheet, keep_default_na=False, skiprows=self.row)
-        else:
-            df = pandas.read_csv(self.source, sep=self.delimiter, skiprows=self.row)
+        if self.source:
+            if self.format == "spreadsheet":
+                with warnings.catch_warnings():
+                    warnings.simplefilter("ignore")
+                    df = pandas.read_excel(self.source, sheet_name=self.sheet, keep_default_na=False, skiprows=self.row)
+            else:
+                df = pandas.read_csv(self.source, sep=self.delimiter, skiprows=self.row)
+
+            if len(df) == 0:
+                self.logger.info(
+                    "\033[1;33m" + "Source file has no data" + "\033[0m"
+                )
+                return False
 
-        if len(df) == 0:
-            self.logger.info(
-                "\033[1;33m" + "Source file has no data" + "\033[0m"
-            )
-            return False
+        else:
+            df = self.data
 
         df = df.loc[:, ~df.columns.str.contains('^Unnamed')]
 

checkcel/checkplate.py

@@ -24,7 +24,7 @@ def __init__(self, validators={}, empty_ok=False, ignore_case=False, ignore_spac
         self.ignore_case = ignore_case
         self.ignore_space = ignore_space
         # self.trim_values = False
-        for validator in self.validators:
+        for validator in self.validators.values():
             validator._set_attributes(self.empty_ok, self.ignore_case, self.ignore_space)
 
     def load_from_python_file(self, file_path):

checkcel/validators.py

@@ -550,8 +550,14 @@ def validate(self, field, row_number, row={}):
         if self.ignore_space:
             field = field.strip()
 
-        if field == "" and self.empty_ok:
-            return
+        if not field:
+            if self.empty_ok:
+                return
+            else:
+                raise ValidationException(
+                    "Field cannot be empty"
+                )
+
         if self.unique_with and not self.unique_check:
             self._precheck_unique_with(row)
 
@@ -807,11 +813,11 @@ def validate(self, field, row_number, row={}):
 
         if self.format == "DD":
             regex_lat = r"[-+]?((90(\.0+)?)|([1-8]?\d(\.\d+)?))[NSns]?"
-            regex_long = r"[-+]?((180(\.0+)?)|([1-8]?\d(\.\d+)?))[wWeE]?"
+            regex_long = r"[-+]?((180(\.0+)?)|(((1[0-7]\d)|([1-9]?\d))(\.\d+)?))[wWeE]?"
 
         else:
-            regex_lat = r"((90[°|\s]\s*)(0{1,2}['|\s]\s*)(0{1,2}([.|,]0{1,20})?[\"|\s]\s*)|(([1-8]\d|\d)[°|\s]\s*)(([0-5]\d|\d)['|\s]\s*)(([0-5]\d|\d)([.|,]\d{1,20})?[\"|\s]\s*))[NSns]"
-            regex_long = r"((180[°|\s]\s*)(0{1,2}['|\s]\s*)(0{1,2}([.|,]0{1,20})?[\"|\s]\s*)|((1[0-7]\d|\d\d|\d)[°|\s]\s*)(([0-5]\d|\d)['|\s]\s*)(([0-5]\d|\d)([.|,]\d{1,20})?[\"|\s]\s*))[EWew]"
+            regex_lat = r"((([1-8]?\d)(°\s?|\s)([1-5]?\d|60)('\s?|\s)?([1-5]?\d(\.\d+)?|60)(\"\s?|\s)?)|(90(°\s?|\s)0('\s?|\s)0(\"\s?|\s)?))[NSns]?"
+            regex_long = r"((((1[0-7][0-9])|([0-9]{1,2}))(°\s?|\s)([1-5]?\d|60)('\s?|\s)([1-5]?\d(\.\d+)?|60)(\"\s?|\s)?)|(180(°\s?|\s)0('\s?|\s)0(\"\s?|\s)?))[EWew]?"
 
         if self.only_long:
             regex = r"^{}$".format(regex_long)

requirements.txt

@@ -3,3 +3,4 @@ odfpy
 pandas
 email-validator
 pyyaml
+requests