wiktionary_examples: drop cited example lines (WIP)

A usage example carrying {{źródło|…}}, {{zrodlo|…}} or <ref> is a quotation
from a third-party work, used on Wiktionary under the Art. 29 quotation right,
not contributor text licensed CC BY-SA. The fetcher now drops those lines and
records their citations in the audit side-car.

On the 2026-10-01 dump: 32,280 cited lines dropped, 45,547 uncited kept;
1,915 documents (558,105 chars) pass the 200-char gate.

Not yet done: filter for untemplated attributions such as "(z Internetu)" or
"(Autor: Tytuł)", rebuild of data/wiktionary_examples/, datasheet and registry.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

src/fetch_wiktionary_examples.py

@@ -9,10 +9,11 @@ Unlike most sources in this registry, this one is NOT routed through
 SpeakLeash's redistribution — it is parsed directly from the Wikimedia dump.
 The `provenance` field in sources.py records this.
 
-One document = one headword; all its example sentences are concatenated,
-newline-separated, in source order. Citation attribution found in
-`{{źródło|...}}` templates is stripped from the text and written to a
-side-car file for auditing (not part of the corpus).
+One document = one headword; its uncited example sentences are concatenated,
+newline-separated, in source order. A line carrying a `{{źródło|...}}` template
+or a `<ref>` is a quotation from a third-party work (used under the Art. 29
+quotation right, not licensed CC BY-SA by Wiktionary contributors), so it is
+dropped; its citation goes to a side-car file for auditing (not part of the corpus).
 
 Only the standard library is required for parsing; `zstandard` is used for
 output compression (falls back to plain .jsonl with a warning).
@@ -58,6 +59,7 @@ NUMBERED_LINE = re.compile(r"^:\s*\(([\d.]+)\)\s*(.+?)\s*$", re.M)
 SENSE_NUMBER = re.compile(r"\(\d+\.\d+\)")
 
 CITATION_FIELDS = ("autor", "tytuł", "tytul", "rozdział", "rozdzial", "rok", "isbn")
+CITATION_TEMPLATES = ("{{źródło", "{{zrodlo")
 
 
 def split_templates(text: str) -> tuple[str, list[str]]:
@@ -147,22 +149,32 @@ def polish_section(wikitext: str) -> tuple[str, str] | None:
     return headword, rest
 
 
-def examples_and_citations(section: str) -> tuple[list[str], list[str]]:
+def is_cited(raw: str, templates: list[str]) -> bool:
+    """A line with a source attached is a quotation from a third-party work (Art. 29
+    quotation right), not a contributor's own CC BY-SA text."""
+    return "<ref" in raw.lower() or any(t.startswith(CITATION_TEMPLATES) for t in templates)
+
+
+def examples_and_citations(section: str) -> tuple[list[str], list[str], int]:
+    """(uncited example sentences, citations of the dropped lines, number of dropped lines)."""
     match = EXAMPLES_BLOCK.search(section)
     if not match:
-        return [], []
-    examples, citations = [], []
+        return [], [], 0
+    examples, citations, dropped = [], [], 0
     for _, raw in NUMBERED_LINE.findall(match.group(1)):
         without_templates, templates = split_templates(raw)
-        for template in templates:
-            if template.startswith(("{{źródło", "{{zrodlo")):
-                described = describe_citation(clean(template))
-                if described:
-                    citations.append(described)
+        if is_cited(raw, templates):
+            dropped += 1
+            for template in templates:
+                if template.startswith(CITATION_TEMPLATES):
+                    described = describe_citation(clean(template))
+                    if described:
+                        citations.append(described)
+            continue
         sentence = clean(without_templates)
         if len(sentence) >= MIN_EXAMPLE_CHARS:
             examples.append(sentence)
-    return examples, sorted(set(citations))
+    return examples, sorted(set(citations)), dropped
 
 
 def iter_pages(dump_path: Path):
@@ -197,7 +209,7 @@ def main() -> None:
     out_dir = Path(args.out).expanduser()
     out_dir.mkdir(parents=True, exist_ok=True)
 
-    pages = entries = with_examples = with_citations = 0
+    pages = entries = with_examples = with_citations = cited_lines = uncited_lines = 0
     documents: list[tuple[str, str]] = []
     citation_index: list[dict] = []
 
@@ -213,13 +225,15 @@ def main() -> None:
             continue
         entries += 1
         headword, body = section
-        examples, citations = examples_and_citations(body)
-        if not examples:
-            continue
-        with_examples += 1
+        examples, citations, dropped = examples_and_citations(body)
+        cited_lines += dropped
         if citations:
             with_citations += 1
             citation_index.append({"headword": headword, "citations": citations})
+        if not examples:
+            continue
+        with_examples += 1
+        uncited_lines += len(examples)
         documents.append((headword, "\n".join(examples)))
 
     kept = [(h, t) for h, t in documents if len(t) >= args.min_chars]
@@ -230,6 +244,8 @@ def main() -> None:
     print(f"Polish entries:           {entries:,}")
     print(f"entries with examples:    {with_examples:,}")
     print(f"entries with attribution: {with_citations:,}")
+    print(f"cited lines dropped:      {cited_lines:,}")
+    print(f"uncited lines kept:       {uncited_lines:,}")
     print(f"documents >= {args.min_chars} chars:  {len(kept):,}")
     print(f"characters:               {chars:,}")
 

src/test_wiktionary_examples.py

@@ -0,0 +1,27 @@
+"""Cited usage examples are third-party quotations and must not reach the corpus."""
+
+from fetch_wiktionary_examples import examples_and_citations
+
+SECTION = """{{znaczenia}}
+''rzeczownik''
+: (1.1) [[dom]]
+{{przykłady}}
+: (1.1) Kupiliśmy '''dom''' pod miastem.
+: (1.1) ''Dom stał na wzgórzu.'' {{źródło|autor=Bolesław Prus|tytuł=Lalka|rok=1890}}
+: (1.2) Wróciłem do domu późno.<ref>Narodowy Korpus Języka Polskiego</ref>
+: (1.2) Nasz '''dom''' ma ogród. {{zrodlo|autor=Jan Nowak|tytuł={{kursywa|Ogrody}}}}
+: (1.3) To [[mieszkanie|mieszkanie]] jest jasne i ciche.
+{{składnia}}
+"""
+
+
+def test_cited_lines_are_dropped_and_recorded():
+    examples, citations, dropped = examples_and_citations(SECTION)
+    assert examples == ["Kupiliśmy dom pod miastem.", "To mieszkanie jest jasne i ciche."]
+    assert dropped == 3
+    assert "Bolesław Prus | Lalka | 1890" in citations
+    assert any(c.startswith("Jan Nowak") for c in citations)
+
+
+def test_section_without_examples():
+    assert examples_and_citations("{{znaczenia}}\n: (1.1) dom\n") == ([], [], 0)