search locally

README.md

@@ -6,17 +6,14 @@ This plugin allows to quickly look up an acronym on [Acronomy](https://acronomy.
 
 #### Installation
 
-- copy the `plasma-runner-acronomy.desktop` file to `~/.local/share/kservices5/`
+- copy the `plasma-runner-acronomy.desktop` file to `~/.loc al/share/kservices5/`
 - restart krunner: `kquitapp5 krunner; kstart5 krunner`
 
 The `acronomy.py` script needs to run permanently in the background to fetch the results.
 
 One easy way to accomplish this is to copy the `acronomy.desktop` to `~/.config/autostart/`.
 
-You can replace the search keyword (default `acr`) with the `--keyword` parameter.
-
-If you enable the `--less-privacy` mode, you don't need to specify any search keyword, but as a result all krunner searches are sent to the Astroacro API. (instead of only the ones starting with the search keyword)
-
+No search queries are sent to the Acronomy server. Instead a list of all acronyms is fetched and searched locally.
 
 This plugin uses the [Astroacro API](https://acronomy.lw1.at/api/). ([Privacy Policy](http://lw1.at/i))
 

acronomy.py

@@ -1,6 +1,7 @@
 #!/bin/env python3
 import argparse
 import webbrowser
+from datetime import datetime, timedelta
 
 import dbus.service
 import requests
@@ -12,33 +13,51 @@ DBusGMainLoop(set_as_default=True)
 objpath = "/acronomy"
 
 iface = "org.kde.krunner1"
-
 s = requests.Session()
+s.headers.update({'User-Agent': 'Acronomy Krunner'})
+
+
+class LocalData:
+    acronyms = {}
+    last_updated = datetime.now()
+
+    def __init__(self):
+        self.fetch_data()
+
+    def fetch_data(self):
+        print("fetching data")
+        r = s.get("https://acronomy.lw1.at/api/acronym/")
+        self.acronyms = {}
+        self.last_updated = datetime.now()
+        for acro in r.json():
+            self.acronyms[acro["name"].lower()] = acro
+
+    def search(self, query: str):
+        age = datetime.now() - self.last_updated
+        if age > timedelta(hours=3):
+            self.fetch_data()
+
+        query = query.lower()
+        for name, acro in self.acronyms.items():
+            if query in name:
+                yield acro
 
 
 class Runner(dbus.service.Object):
     def __init__(self, args):
         self.args = args
+        self.data = LocalData()
         dbus.service.Object.__init__(self, dbus.service.BusName("net.acronomy", dbus.SessionBus()), objpath)
 
     @dbus.service.method(iface, in_signature="s", out_signature="a(sssida{sv})")
     def Match(self, query: str):
-
-        if not self.args.less_privacy:
-            if not query.startswith(self.args.keyword):
-                return []
-            query = query.replace(self.args.keyword + " ", "")
-
-        if " " in query:
-            return []
-
         runners = []
-        r = s.get("https://acronomy.lw1.at/api/acronym/", params={"search": query})
+
         icon = "plasmagik"
         type = 100  # (Plasma::QueryType)
         relevance = 0.2  # 0-1
 
-        for result in r.json():
+        for result in self.data.search(query):
             data = result["slug"]
             display_text = result["name"] + ": " + result["full_name"]
             properties = {
@@ -61,8 +80,6 @@ class Runner(dbus.service.Object):
 
 def main():
     parser = argparse.ArgumentParser(description="acrronomy krunner background task")
-    parser.add_argument("-k", "--keyword", action="store", default="acr")
-    parser.add_argument("-l", "--less-privacy", action="store_true", default=False)
 
     args = parser.parse_args()