#95 Baseline data update scripts that can print results of a query

Data/total_data.json

@@ -0,0 +1,28 @@
+{
+  "French": {
+    "nouns": 15710,
+    "verbs": 1241
+  },
+  "German": {
+    "nouns": 27526,
+    "verbs": 3125,
+    "prepositions": 188
+  },
+  "Portuguese": {
+    "nouns": 4530,
+    "verbs": 188
+  },
+  "Russian": {
+    "nouns": 194389,
+    "verbs": 11,
+    "prepositions": 12
+  },
+  "Spanish": {
+    "nouns": 8310,
+    "verbs": 87
+  },
+  "Swedish": {
+    "nouns": 41102,
+    "verbs": 4133
+  }
+}

Data/update_data.py

@@ -0,0 +1,66 @@
+"""
+Update Data
+-----------
+
+Updates all data for Scribe by running all WDQS queries and formatting scripts.
+"""
+
+
+import json
+import os
+
+from tqdm.auto import tqdm
+from wikidataintegrator import wdi_core
+
+data_dir_elements = []
+
+for path, _, files in os.walk("."):
+    for name in files:
+        data_dir_elements.append(os.path.join(path, name))
+
+data_dir_files = [f for f in os.listdir(".") if os.path.isfile(os.path.join(".", f))]
+
+data_dir_dirs = list(
+    {
+        f.split("./")[1].split("/")[0]
+        for f in data_dir_elements
+        if f.split("./")[1] not in data_dir_files
+    }
+)
+
+word_types = ["nouns", "verbs", "prepositions"]
+
+with open("total_data.json") as f:
+    current_data = json.load(f)
+current_languages = list(current_data.keys())
+
+possible_queries = []
+for d in data_dir_dirs:
+    for wt in word_types:
+        if "./" + d + "/" + wt in [
+            e[: len("./" + d + "/" + wt)] for e in data_dir_elements
+        ]:
+            possible_queries.append(d + "/" + wt)
+
+queries_to_run_lists = [
+    [q for q in possible_queries if q[: len(lang)] in current_languages]
+    for lang in current_languages
+]
+
+queries_to_run = list({q for sub in queries_to_run_lists for q in sub})
+
+for q in tqdm(queries_to_run[:1], desc="Data updated", unit="dirs",):
+    target_type = q.split("/")[1]
+    query_name = "query" + target_type.title() + ".sparql"
+    query_path = "./" + q + "/" + query_name
+
+    with open(query_path) as file:
+        query_lines = file.readlines()
+
+    # First format the lines into a multi-line string and then pass this to wikidataintegrator.
+    query_results = wdi_core.WDFunctionsEngine.execute_sparql_query(
+        """{}""".format("".join(query_lines))
+    )
+
+print(query_results)
+print(query_path)

Data/updated_data.txt

@@ -0,0 +1,6 @@
+- French (New): 15,710 nouns, 1,241 verbs (mostly infinitives), 67,609 translations
+- German: 401 nouns, 78 verbs, corrected many prepositions
+- Portuguese (New): 4,530 nouns, 188 verbs, 67,609 translations
+- Russian (New): 194,389 nouns, 11 verbs, 12 prepositions, 67,609 translations
+- Spanish: 180 nouns, 22 verbs
+- Swedish (New): 41,102 nouns, 4,133 verbs, 67,609 translations

README.md

@@ -74,10 +74,12 @@ The Scribe team has interest in creating keyboards for all languages of interest
 
 `*` Given the current **`beta`** status where words are machine translated.
 
-`†` Adjective-preposition support is in progress [(see issue)](https://github.com/scribe-org/Scribe-iOS/issues/86).
+`†` Adjective-preposition support is in progress [(see #86)](https://github.com/scribe-org/Scribe-iOS/issues/86).
 
 `‡` Only for languages for which preposition annotation is needed.
 
+WIP [#95](https://github.com/scribe-org/Scribe-iOS/issues/95): Updates to the above data can be done using [Data/update_data.py](https://github.com/scribe-org/Scribe-iOS/tree/main/Data/update_data.py).
+
 # Setup [`⇧`](#contents) <a id="setup"></a>
 
 Users access Scribe language keyboards through the following:

environment.yml

@@ -6,3 +6,5 @@ dependencies:
   - black>=19.10b0
   - transformers>=4.12
   - sentencepiece>=0.1.95
+  - pip:
+      - wikidataintegrator>=0.9.22