Fix parsing of leading zeros in named HTML entities (fixes #288)

CHANGELOG

@@ -1,6 +1,6 @@
 v0.7 (unreleased):
 
-- ...
+- Fixed parsing of leading zeros in named HTML entities. (#288)
 
 v0.6.4 (released February 14, 2022):
 

docs/changelog.rst

@@ -7,7 +7,8 @@ v0.7
 Unreleased
 (`changes <https://github.com/earwig/mwparserfromhell/compare/v0.6.4...main>`__):
 
-- ...
+- Fixed parsing of leading zeros in named HTML entities.
+  (`#288 <https://github.com/earwig/mwparserfromhell/issues/288>`_)
 
 v0.6.4
 ------

src/mwparserfromhell/parser/ctokenizer/tok_parse.c

@@ -1099,7 +1099,7 @@ Tokenizer_really_parse_entity(Tokenizer *self)
             }
             break;
         }
-        if (i == 0 && this == '0') {
+        if (i == 0 && numeric && this == '0') {
             zeroes++;
             self->head++;
             continue;

tests/tokenizer/html_entities.mwtest

@@ -156,3 +156,10 @@ name:   invalid_partial_amp_pound_x_semicolon
 label:  invalid entities: an ampersand, pound sign, and x
 input:  "&#x"
 output: [Text(text="&#x")]
+
+---
+
+name:   invalid_zeros_before_named
+label:  invalid entities: zeros before a valid named entity
+input:  "&000nbsp;"
+output: [Text(text="&000nbsp;")]