Merge pull request #1008 from xxyzz/ku

[ku] extract other forms sections

src/wiktextract/extractor/ku/linkage.py

@@ -0,0 +1,59 @@
+from wikitextprocessor import NodeKind, TemplateNode, WikiNode
+
+from ...page import clean_node
+from ...wxr_context import WiktextractContext
+from .models import Form, WordEntry
+
+
+def extract_other_form_section(
+    wxr: WiktextractContext,
+    word_entry: WordEntry,
+    level_node: WikiNode,
+) -> None:
+    for list_node in level_node.find_child(NodeKind.LIST):
+        for list_item in list_node.find_child(NodeKind.LIST_ITEM):
+            for t_node in list_item.find_child(NodeKind.TEMPLATE):
+                if t_node.template_name.startswith("ku-"):
+                    extract_ku_form_template(wxr, word_entry, t_node)
+                elif t_node.template_name == "g":
+                    extract_g_template(wxr, word_entry, t_node)
+
+
+def extract_ku_form_template(
+    wxr: WiktextractContext,
+    word_entry: WordEntry,
+    t_node: TemplateNode,
+) -> None:
+    expanded_node = wxr.wtp.parse(
+        wxr.wtp.node_to_wikitext(t_node), expand_all=True
+    )
+    form = Form(form="")
+    for index, span_tag in enumerate(expanded_node.find_html("span")):
+        if index == 0:
+            form.raw_tags.append(clean_node(wxr, None, span_tag))
+        elif index == 1:
+            form.form = clean_node(wxr, None, span_tag)
+    if form.form != "":
+        word_entry.forms.append(form)
+
+
+def extract_g_template(
+    wxr: WiktextractContext,
+    word_entry: WordEntry,
+    t_node: TemplateNode,
+) -> None:
+    form = Form(
+        form=clean_node(
+            wxr,
+            None,
+            t_node.template_parameters.get(
+                2, t_node.template_parameters.get("cuda", "")
+            ),
+        ),
+        roman=clean_node(wxr, None, t_node.template_parameters.get("tr", "")),
+        translation=clean_node(
+            wxr, None, t_node.template_parameters.get("w", "")
+        ),
+    )
+    if form.form != "":
+        word_entry.forms.append(form)

src/wiktextract/extractor/ku/models.py

@@ -37,6 +37,7 @@ class Form(KurdishBaseModel):
     tags: list[str] = []
     raw_tags: list[str] = []
     roman: str = ""
+    translation: str = ""
 
 
 class Translation(KurdishBaseModel):

src/wiktextract/extractor/ku/page.py

@@ -6,6 +6,7 @@
 from ...page import clean_node
 from ...wxr_context import WiktextractContext
 from .etymology import extract_etymology_section
+from .linkage import extract_other_form_section
 from .models import Sense, WordEntry
 from .pos import extract_pos_section
 from .section_titles import POS_DATA
@@ -31,6 +32,10 @@ def parse_section(
         extract_translation_section(
             wxr, page_data[-1] if len(page_data) > 0 else base_data, level_node
         )
+    elif title_text in ["Bi alfabeyên din", "Herwiha", "Bide ber"]:
+        extract_other_form_section(
+            wxr, page_data[-1] if len(page_data) > 0 else base_data, level_node
+        )
 
     for next_level in level_node.find_child(LEVEL_KIND_FLAGS):
         parse_section(wxr, page_data, base_data, next_level)

tests/test_ku_linkage.py

@@ -0,0 +1,43 @@
+from unittest import TestCase
+
+from wikitextprocessor import Wtp
+
+from wiktextract.config import WiktionaryConfig
+from wiktextract.extractor.ku.page import parse_page
+from wiktextract.wxr_context import WiktextractContext
+
+
+class TestKuLinkage(TestCase):
+    maxDiff = None
+
+    def setUp(self) -> None:
+        self.wxr = WiktextractContext(
+            Wtp(lang_code="ku"),
+            WiktionaryConfig(
+                dump_file_lang_code="ku", capture_language_codes=None
+            ),
+        )
+
+    def tearDown(self):
+        self.wxr.wtp.close_db_conn()
+
+    def test_ku_ar(self):
+        self.wxr.wtp.add_page("Şablon:ziman", 10, "Kurmancî")
+        self.wxr.wtp.add_page(
+            "Şablon:ku-ar",
+            10,
+            """<span class="Latn" lang="ku">[[kurdî-erebî#Kurmancî|kurdî-erebî]]</span>: <span class="Arab" lang="ku">[[کووچک#Kurmancî|کووچک]]</span>&lrm;""",
+        )
+        page_data = parse_page(
+            self.wxr,
+            "kûçik",
+            """== {{ziman|ku}} ==
+=== Navdêr ===
+# [[heywan|Heywanek]]
+==== Bi alfabeyên din ====
+* {{ku-ar|کووچک}}""",
+        )
+        self.assertEqual(
+            page_data[0]["forms"],
+            [{"form": "کووچک", "raw_tags": ["kurdî-erebî"]}],
+        )