Merge pull request #473 from xxyzz/fr

Find alternative form of the "typographic variant" POS word

src/wiktextract/data/fr/pos_subtitles.json

@@ -444,13 +444,22 @@
     "pos": "num"
   },
   "onom": {
-    "pos": "onomatopoeia"
+    "pos": "onomatopoeia",
+    "tags": [
+      "onomatopoeic"
+    ]
   },
   "onoma": {
-    "pos": "onomatopoeia"
+    "pos": "onomatopoeia",
+    "tags": [
+      "onomatopoeic"
+    ]
   },
   "onomatopée": {
-    "pos": "onomatopoeia"
+    "pos": "onomatopoeia",
+    "tags": [
+      "onomatopoeic"
+    ]
   },
   "part": {
     "pos": "particle"
@@ -744,16 +753,28 @@
     "pos": "symbol"
   },
   "var-typo": {
-    "pos": "typographic variant"
+    "pos": "typographic variant",
+    "tags": [
+      "alt-of"
+    ]
   },
   "variante par contrainte typographique": {
-    "pos": "typographic variant"
+    "pos": "typographic variant",
+    "tags": [
+      "alt-of"
+    ]
   },
   "variante typo": {
-    "pos": "typographic variant"
+    "pos": "typographic variant",
+    "tags": [
+      "alt-of"
+    ]
   },
   "variante typographique": {
-    "pos": "typographic variant"
+    "pos": "typographic variant",
+    "tags": [
+      "alt-of"
+    ]
   },
   "verb pr": {
     "pos": "verb",
@@ -776,4 +797,4 @@
       "pronominal"
     ]
   }
-}
+}

src/wiktextract/extractor/fr/gloss.py

@@ -1,11 +1,12 @@
 from collections import defaultdict
+from typing import Union
 
 from wikitextprocessor import NodeKind, WikiNode
 from wikitextprocessor.parser import TemplateNode
 from wiktextract.page import clean_node
 from wiktextract.wxr_context import WiktextractContext
 
-from .models import Example, Sense, WordEntry
+from .models import AltForm, Example, Sense, WordEntry
 
 
 def extract_gloss(
@@ -65,6 +66,9 @@ def extract_gloss(
                 and gloss_only_nodes[index].template_name == "note"
             ):
                 note_index = index
+        find_alt_of_form(
+            wxr, gloss_only_nodes[:note_index], page_data[-1].pos, gloss_data
+        )
         gloss_text = clean_node(
             wxr, gloss_data, gloss_only_nodes[:note_index]
         ).strip(" ()")
@@ -146,3 +150,28 @@ def process_exemple_template(
     )
     if len(example_data.text) > 0:
         gloss_data.examples.append(example_data)
+
+
+def find_alt_of_form(
+    wxr: WiktextractContext,
+    gloss_nodes: list[Union[str, WikiNode]],
+    pos_type: str,
+    gloss_data: Sense,
+):
+    if pos_type == "typographic variant":
+        alt_of = ""
+        for gloss_node in filter(
+            lambda n: isinstance(n, WikiNode), gloss_nodes
+        ):
+            # use the last link
+            if gloss_node.kind == NodeKind.LINK:
+                alt_of = clean_node(wxr, None, gloss_node)
+            if isinstance(gloss_node, TemplateNode):
+                gloss_node = wxr.wtp.parse(
+                    wxr.wtp.node_to_wikitext(gloss_node), expand_all=True
+                )
+            for link in gloss_node.find_child_recursively(NodeKind.LINK):
+                alt_of = clean_node(wxr, None, link)
+        if len(alt_of) > 0:
+            gloss_data.alt_of.append(AltForm(word=alt_of))
+            gloss_data.tags.append("alt-of")

src/wiktextract/extractor/fr/models.py

@@ -76,12 +76,17 @@ class Linkage(FrenchBaseModel):
     lang_code: str = Field("", description="Wiktionary language code")
 
 
+class AltForm(FrenchBaseModel):
+    word: str
+
+
 class Sense(FrenchBaseModel):
     glosses: list[str] = []
     tags: list[str] = []
     categories: list[str] = []
     examples: list[Example] = []
     note: str = ""
+    alt_of: list[AltForm] = []
 
 
 class WordEntry(FrenchBaseModel):

src/wiktextract/extractor/fr/page.py

@@ -113,11 +113,13 @@ def process_pos_block(
     pos_argument: str,
     pos_title: str,
 ):
-    pos_type = wxr.config.POS_SUBTITLES[pos_argument]["pos"]
+    pos_data = wxr.config.POS_SUBTITLES[pos_argument]
+    pos_type = pos_data["pos"]
     if len(page_data) == 0 or "pos" in page_data[-1].model_fields_set:
         page_data.append(base_data.model_copy(deep=True))
     page_data[-1].pos = pos_type
     page_data[-1].pos_title = pos_title
+    page_data[-1].tags.extend(pos_data.get("tags", []))
     child_nodes = list(pos_title_node.filter_empty_str_child())
     form_line_start = 0  # Ligne de forme
     gloss_start = len(child_nodes)

tests/test_fr_gloss.py

@@ -308,3 +308,60 @@ def test_gloss_note_template(self):
                 }
             ],
         )
+
+    def test_typographic_variant_alt_of_template(self):
+        self.wxr.wtp.start_page("abajhuro")
+        self.wxr.wtp.add_page(
+            "Modèle:eo-sys-h",
+            10,
+            """''Orthographe par contrainte typographique par [[Annexe:Systèmes h et x en espéranto#Système h|système h]] de'' <bdi>[[abaĵuro#eo|abaĵuro]]</bdi>""",
+        )
+        root = self.wxr.wtp.parse("# {{eo-sys-h|abaĵuro}}.")
+        page_data = [
+            WordEntry(
+                word="abajhuro",
+                lang_code="fr",
+                lang="Français",
+                pos="typographic variant",
+            )
+        ]
+        extract_gloss(self.wxr, page_data, root.children[0])
+        self.assertEqual(
+            [d.model_dump(exclude_defaults=True) for d in page_data[-1].senses],
+            [
+                {
+                    "glosses": [
+                        "Orthographe par contrainte typographique par système h de abaĵuro."
+                    ],
+                    "alt_of": [{"word": "abaĵuro"}],
+                    "tags": ["alt-of"],
+                }
+            ],
+        )
+
+    def test_typographic_variant_alt_of_text(self):
+        self.wxr.wtp.start_page("alphoenix")
+        root = self.wxr.wtp.parse(
+            "# ''Variante par contrainte typographique de'' [[alphœnix]]."
+        )
+        page_data = [
+            WordEntry(
+                word="alphoenix",
+                lang_code="fr",
+                lang="Français",
+                pos="typographic variant",
+            )
+        ]
+        extract_gloss(self.wxr, page_data, root.children[0])
+        self.assertEqual(
+            [d.model_dump(exclude_defaults=True) for d in page_data[-1].senses],
+            [
+                {
+                    "glosses": [
+                        "Variante par contrainte typographique de alphœnix."
+                    ],
+                    "alt_of": [{"word": "alphœnix"}],
+                    "tags": ["alt-of"],
+                }
+            ],
+        )