diff --git a/src/app.py b/src/app.py
index f8564335b8f08c2acdd3edfa68c95ad69f1f06d9..1c50351ed7a99622bd77fbab59ace14317affb36 100644
--- a/src/app.py
+++ b/src/app.py
@@ -1,10 +1,7 @@
-from flask import Flask, jsonify
-from flask import request
-from flask import Response
+from flask import Flask, Response, jsonify, request
from flask_cors import CORS
import config
-from get_wikicode import get_wikicode
from wiktextract_wrapper import Wiktextract
app = Flask(__name__)
@@ -23,31 +20,29 @@ def search(wiktlang, wordlang, word):
if wiktlang not in config.supported_wiktlangs:
return jsonify({"error": f"Language {wiktlang} not supported"}), 400
- wikicode = get_wikicode(word, wiktlang)
- if wikicode:
- wiktextractor = Wiktextract(wiktlang, wordlang)
- try:
- resp = wiktextractor.parse_page(word, wikicode)
+ wiktextractor = Wiktextract(wiktlang, wordlang)
+ try:
+ resp = wiktextractor.parse_page(word)
+ if resp:
return jsonify(resp)
-
- except Exception as e:
- print(e)
-
- return jsonify({"error": "Parsing page resulted in error: " + str(e)}), 500
- finally:
- wiktextractor.wxr.wtp.db_conn.close()
- if wiktextractor.wxr.thesaurus_db_conn:
- wiktextractor.wxr.thesaurus_db_conn.close()
-
- else:
- return (
- jsonify(
- {
- "error": f"{word} is unknown in “{wordlang}†in {wiktlang}.wiktionary.org."
- }
- ),
- 404,
- )
+ else:
+ return (
+ jsonify(
+ {
+ "error": f"{word} is unknown in “{wordlang}†in {wiktlang}.wiktionary.org."
+ }
+ ),
+ 404,
+ )
+
+ except Exception as e:
+ print(e)
+
+ return jsonify({"error": "Parsing page resulted in error: " + str(e)}), 500
+ finally:
+ wiktextractor.wxr.wtp.db_conn.close()
+ if wiktextractor.wxr.thesaurus_db_conn:
+ wiktextractor.wxr.thesaurus_db_conn.close()
if __name__ == "__main__":
diff --git a/src/get_wikicode.py b/src/get_wikicode.py
deleted file mode 100644
index 24291977bae17f9bfe128d43f80202aeb7b3eea7..0000000000000000000000000000000000000000
--- a/src/get_wikicode.py
+++ /dev/null
@@ -1,7 +0,0 @@
-import pywikibot
-
-
-def get_wikicode(title: str, wiktlang: str):
- site = pywikibot.Site(f"wiktionary:{wiktlang}")
- page = pywikibot.Page(site, title)
- return page.text
diff --git a/src/wiktextract_context.py b/src/wiktextract_context.py
index 79d08e0a0c8f2f56eb80e9ac5bfec60d07f7bfff..ce88f3158ab3fe36aa7dea6db4549f43e7524fd6 100644
--- a/src/wiktextract_context.py
+++ b/src/wiktextract_context.py
@@ -1,38 +1,11 @@
-from wiktextract import (
- WiktextractContext,
- WiktionaryConfig,
-)
-from wikitextprocessor import Wtp, Page
-
from typing import Optional
-from get_wikicode import get_wikicode
-
-
-class CustomWtp(Wtp):
- def get_page(
- self,
- title: str,
- namespace_id: Optional[int] = None,
- no_redirect: bool = False,
- ) -> Optional[Page]:
- # Call the original get_page method
- original_result = super().get_page(title, namespace_id, no_redirect)
-
- if original_result == None:
- # The db is often called with titles like "tracking/parameters/empty parameter". These seem to return None by design and are not present in Wiktionary. Skip these.
- if "/translations" in title and not "tracking" in title:
- print(f"Page '{title}' not found in db. Fetching from live wiktionary.")
-
- body = get_wikicode(title, self.lang_code)
-
- return Page(title, namespace_id, body=body)
-
- return original_result
+from wikitextprocessor import Wtp
+from wiktextract import WiktextractContext, WiktionaryConfig
-def get_wiktextract_context(wiktlang: str, wordlang: Optional[str] = None, mock_get_page: bool = True):
- db_path = f"./sqlite-{wiktlang}.db"
+def get_wiktextract_context(wiktlang: str, wordlang: Optional[str] = None):
+ db_path = f"./sqlite-{wiktlang}-all.db"
config = WiktionaryConfig(
dump_file_lang_code=wiktlang,
capture_language_codes=[wordlang] if wordlang else None,
@@ -47,7 +20,7 @@ def get_wiktextract_context(wiktlang: str, wordlang: Optional[str] = None, mock_
capture_inflections=True,
)
config.load_edition_settings()
- wtp = CustomWtp(db_path=db_path, lang_code=wiktlang) if mock_get_page else Wtp(db_path=db_path, lang_code=wiktlang)
+ wtp = Wtp(db_path=db_path, lang_code=wiktlang)
wxr = WiktextractContext(wtp, config)
return wxr
diff --git a/src/wiktextract_wrapper.py b/src/wiktextract_wrapper.py
index 0f072ec1d0a51548829adba9ccdd30f1915fddd0..c1eab07b27d58c1db1cbaf6576894ea1782ab3fd 100644
--- a/src/wiktextract_wrapper.py
+++ b/src/wiktextract_wrapper.py
@@ -1,13 +1,6 @@
from wiktextract.page import parse_page
from wiktextract_context import get_wiktextract_context
-
-db_path = "./sqlite.db"
-
-DEFAULT_PAGE_VALUES = {
- "namespace_id": 0,
- "model": "wikitext",
-}
class Wiktextract:
@@ -15,28 +8,12 @@ class Wiktextract:
self.wiktlang = wiktlang
self.wordlang = wordlang
- # self.page_handler = page_handler
- # self.page_handler.wxr : WiktextractContext =
self.wxr = get_wiktextract_context(wiktlang, wordlang)
- def parse_page(self, title: str, wikicode: str):
- # add page to the database (making it accessible to LUA templates)
- self.wxr.wtp.add_page(
- title=title,
- namespace_id=DEFAULT_PAGE_VALUES["namespace_id"],
- body=wikicode,
- model=DEFAULT_PAGE_VALUES["model"],
- )
-
- self.wxr.wtp.start_page(title)
-
- result = parse_page(self.wxr, title, wikicode)
-
- # remove the page from the database
- self.wxr.wtp.db_conn.execute(
- "DELETE FROM pages WHERE title = ? AND model = ?",
- (title, DEFAULT_PAGE_VALUES["model"]),
- )
- self.wxr.wtp.db_conn.commit()
+ def parse_page(self, title: str):
+ page = self.wxr.wtp.get_page(title)
+ if not page:
+ return None
+ result = parse_page(self.wxr, title, page.body)
return result