add regexMatch

adtl/__init__.py

@@ -207,6 +207,8 @@ def parse_if(
             return cast_value != value
         elif cmp in ["=", "=="]:
             return cast_value == value
+        elif cmp == "=~":
+            return bool(re.match(value, cast_value, re.IGNORECASE))
         else:
             raise ValueError(f"Unrecognized operand: {cmp}")
     elif isinstance(rule[key], set):  # common error, missed colon to make it a dict

docs/specification.md

@@ -165,7 +165,7 @@ if = { foobar_type = 4 }
 ```
 
 Operations other than equals can be specified as `{ field_name = {op = value} }`
-where *op* is one of `< | > | <= | >= | !=`. Logical operations (and, or, not) are
+where *op* is one of `< | > | <= | >= | != | =~`. Logical operations (and, or, not) are
 supported with `any = [ condition-list ]` (or), `all = [ condition-list ]` (and),
 `not = { condition }` (not).
 In the above example, if we wanted to set from field *foobar* only if
@@ -180,6 +180,15 @@ if.all = [  # in TOML this is a nested key, like { "if": { "all": [ ... ] } }
 ]
 ```
 
+The `=~` operator allows matching with regular expressions, similar to the Bash
+and Perl operators. The following will match strings `SARS-CoV 2`, `SARS COV 2`
+and `sars-cov-2`. Case is ignored when matching.
+
+```toml
+field = "foobar"
+if.foobar."=~" = ".*SARS[- ]CoV[- ]2.*"
+```
+
 The **oneToMany** table has default conditional behaviour so that rows are only shown
 if the row is not empty, and contains values which can be mapped correctly if maps are
 provided. For example, an observation recording the presence/absence of vomiting should only

pyproject.toml

@@ -7,9 +7,12 @@ packages = ["adtl"]
 
 [project]
 name = "adtl"
-version = "0.4.0"
+version = "0.5.0"
 description = "Another data transformation language"
-authors = [{name = "Abhishek Dasgupta", email = "abhishek.dasgupta@dtc.ox.ac.uk"}]
+authors = [
+  {name = "Abhishek Dasgupta", email = "abhishek.dasgupta@dtc.ox.ac.uk"},
+  {name = "Pip Liggins", email = "philippa.liggins@dtc.ox.ac.uk"}
+]
 license = {file = "LICENSE"}
 requires-python = ">=3.8"
 readme = "README.md"

tests/test_parser.py

@@ -489,6 +489,20 @@ def test_get_value(row_rule, expected):
 @pytest.mark.parametrize(
     "row_rule,expected",
     [
+        (({"pathogen": "covid 19"}, {"pathogen": {"=~": ".*covid.*"}}), True),
+        (({"pathogen": "covid 19"}, {"pathogen": {"=~": ".*SARS-?CoV-?2.*"}}), False),
+        (
+            ({"pathogen": "sars cov 2"}, {"pathogen": {"=~": ".*SARS[- ]CoV[- ]2.*"}}),
+            True,
+        ),
+        (
+            ({"pathogen": "sars-cov 2"}, {"pathogen": {"=~": ".*SARS[- ]CoV[- ]2.*"}}),
+            True,
+        ),
+        (
+            ({"pathogen": "coronavírus"}, {"pathogen": {"=~": ".*coronav[ií]rus.*"}}),
+            True,
+        ),
         ((ROW_CONDITIONAL, {"outcome_type": 4}), True),
         ((ROW_CONDITIONAL, {"not": {"outcome_type": 4}}), False),
         ((ROW_CONDITIONAL, {"outcome_type": {"==": 4}}), True),