Pull request #33: fix: improve unit recognition

Merge in FRAG/50ohm.de from fix/improve-unit-recognition to main

* commit 'b183ef39f1d200b7ebb299aece1cc66065eb4729':
  fix: parse percentage as unit
  fix: improve unit recognition

renderer/unit.py

@@ -35,13 +35,15 @@ class Unit(SpanToken):
     prefixes = ["p", "n", "μ", "m", "k", "M", "G"]
 
     pattern = re.compile(
-        r"(?P<value>[0-9,]+) ?(?P<prefix>"
+        # Check for word boundary at start of value ensures only numeric values are caught.
+        r"\b(?P<value>[0-9,]+) ?(?P<prefix>"
         # Group all prefixes together in a capture group
         # Empty string at end of list is required so the group always exists in the resulting match
         + "|".join(prefixes + [""])
         + ")(?P<unit>"
         # Group all units together in a capture group.
         + "|".join(units)
+        # Word boundary check at end of unit to ensure proper handling of units like A, Ah etc.
         + r")(?!\w)"
     )
 
@@ -50,6 +52,8 @@ class Unit(SpanToken):
     unit: str
 
     def __init__(self, match: re.Match):
+        print("pattern", self.pattern)
+
         self.value = match.group("value")
         self.prefix = match.group("prefix") or ""
         self.unit = match.group("unit")

test/test_unit.py

@@ -15,6 +15,9 @@ def test_basic_units_html():
         "0,75 kW": "0,75 kW",
         "33 A": "33 A",
         "1 Ah": "1 Ah",  # Ah can be a tricky case if the RegEx is wrong, as A might match before Ah, thus rendering the output with an "A" unit
+        "10 A/mm²": "10 A/mm²",
+        "DL9MJ": "DL9MJ",  # Ensure other strings, like callsings, aren‘t mistaken for units. ;)
+        "Ein Satz mit 200 V.": "Ein Satz mit 200 V.",  # Support units before sentence ends.
     }
 
     for input, output in assertions.items():
@@ -38,6 +41,7 @@ def test_special_units_html():
     """Test special units: Don't put space before ° and %."""
     assertions = {
         "100%": "100%",
+        "100 %": "100%",
         "42,24°": "42,24°",
     }