revise SQL

jimm.py

@@ -1,30 +1,22 @@
+for i, c in enumerate('RGYBMC'): globals()[c] = lambda s, i=i: f'\x1b[{91+i}m{s}\x1b[0m'
 Soup = lambda html: __import__('bs4').BeautifulSoup(html, 'lxml')
 unsafe = __import__('contextlib').suppress(Exception)
-for i, c in enumerate('RGYBMC'): globals()[c] = lambda s, i=i: f'\x1b[{91+i}m{s}\x1b[0m'
 
 def SQL():
-    import sqlite3, os, hashlib
+    import sqlite3, hashlib, os;  os.makedirs('.db-blob', exist_ok=True)
     (con := sqlite3.connect('.db', isolation_level=None)).row_factory = sqlite3.Row
-    con.execute('PRAGMA journal_mode=wal')
-    con.execute('PRAGMA busy_timeout='f'{1e9}')
+    con.execute('PRAGMA journal_mode=wal'); con.execute('PRAGMA busy_timeout='f'{1e9}')
     con.execute('CREATE TABLE IF NOT EXISTS kv(k, v, t DEFAULT CURRENT_TIMESTAMP)')
-    os.makedirs('.db-blob', exist_ok=True)
-    def put(sql, filename, blob):
-        sha1 = hashlib.sha1(blob).hexdigest()
-        if not sql('SELECT 1 FROM kv WHERE v=?', sha1):
+    def setitem(sql, filename, blob):
+        if not sql('SELECT 1 FROM kv WHERE v=?', sha1 := hashlib.sha1(blob).hexdigest()):
             with open(f'.db-blob/{sha1}', 'xb') as f: f.write(blob)
             print(f'{G(len(blob)):>16} {filename}')
-        sql[filename] = sha1
-    def get(sql, filename):
-        return open(f'.db-blob/{sql[filename]}', 'rb').read()
-
-    return type('', (), dict(put=put, get=get,
-        __call__=lambda _, q, *p: list(map(dict, con.execute(q, p))),
-        __setitem__=lambda sql, k, v: sql('INSERT INTO kv(k,v) VALUES(?,?)', k, v),
-        __getitem__ = lambda sql, k: sql(
-            'SELECT v FROM kv WHERE k=? ORDER BY t DESC LIMIT 1', k)[0]['v'],
-        __iter__=lambda sql: (kv.values() for kv in sql(
-            'SELECT k, v FROM kv GROUP BY k HAVING t = MAX(t)'))))()
+        sql('INSERT INTO kv(k,v) VALUES(?,?)', filename, sha1)
+    def getitem(sql, filename):
+        if sha1 := sql('SELECT v FROM kv WHERE k=? ORDER BY t DESC LIMIT 1', filename):
+            return open(f'.db-blob/{sha1[0]["v"]}', 'rb').read()
+    return type('', (), dict(__setitem__=setitem, __getitem__=getitem,
+        __call__=lambda _, q, *p: list(map(dict, con.execute(q, p)))))()
 sql = SQL()
 
 def sync(coro):
@@ -42,26 +34,23 @@ def sync(coro):
     return wrapper
 
 @sync
-async def Page(headless=True):
+async def Page(headless=True, wait_until='networkidle', timeout=0):
     from playwright.async_api import async_playwright
     browser = await (await async_playwright().start()).firefox.launch(headless=headless)
-    (page := await browser.new_page()).set_default_timeout(0)
+    (page := await browser.new_page()).set_default_timeout(timeout)
     for attr in dir(page):
         if callable(method := getattr(page, attr)): setattr(page, attr, sync(method))
     async def handle(route):
-        with unsafe:
-            if route.request.method == 'GET' and (response := await route.fetch()).ok:
-                sql.put(route.request.url, await response.body())
+        if route.request.method == 'GET' and (response := await route.fetch()).ok:
+            sql[route.request.url] = await response.body()
         await route.continue_()
-    page.route('**/*', handle)
-
+    page.route('**', handle)
     def goto(url, goto=page.goto):
-        goto(url, wait_until='networkidle')
-        sql.put(url.split('://')[1], page.content().encode())
+        goto(url, wait_until=wait_until)
+        sql[url.split('://')[1]] = page.content().encode()
         from IPython.display import Image
         return Image(page.screenshot())
     page.goto = goto
     return page
 page = Page()
-
-page.goto('https://naver.com/')
+page.goto('https://naver.com/')