add splits

Author: philimon-reset

.github/workflows/l10n-write-mappings.yml

@@ -0,0 +1,23 @@
+---
+name: L10N output mapping splits
+
+run-name: ${{ github.actor }} is writing L10N mapping splits
+on:
+  schedule:
+    - cron: "0 0 * * 1"
+
+permissions:
+  contents: "write"
+
+jobs:
+  Select-Channels:
+    runs-on: ubuntu-latest
+    outputs:
+      channels: ${{ steps.splits.outputs.channels }}
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+      - name: Select test channels
+        id: splits
+        run: |
+          python3 choose_l10n_beta_split.py

choose_l10n_beta_split.py

@@ -0,0 +1,38 @@
+import json
+import logging
+import sys
+from collections import defaultdict
+
+from choose_l10n_ci_set import valid_l10n_mappings
+
+
+def distribute_mappings_evenly(mappings):
+    """
+    Distribute the selected mappings into 3 splits and return the container.
+
+    Args:
+        mappings (dict): A dictionary of mappings, where the keys are sites and the values are sets of regions.
+    """
+    if not mappings:
+        return {}
+    # sort the mappings by the length of the regions per site
+    mappings = dict(sorted(mappings.items(), key=lambda val: len(val[1]), reverse=True))
+    # place the mappings into 3 containers evenly according to the load
+    loads = [0, 0, 0]
+    balanced_splits = [defaultdict(list) for _ in range(3)]
+    for site, regions in mappings.items():
+        min_idx = loads.index(min(loads))
+        balanced_splits[min_idx][site] = list(regions)
+        loads[min_idx] += len(regions)
+    return balanced_splits
+
+
+if __name__ == "__main__":
+    l10n_mappings = valid_l10n_mappings()
+    all_splits = distribute_mappings_evenly(l10n_mappings)
+    if not all_splits:
+        logging.warning("No valid l10n mappings")
+        sys.exit(1)
+    for idx, split in enumerate(all_splits, start=1):
+        with open(f"l10n_CM/beta_run_splits/l10n_split_{idx}.json", "w") as f:
+            json.dump(split, f)

l10n_CM/beta_run_splits/l10n_split_1.json

@@ -0,0 +1 @@
+{"demo": ["US", "GB", "DE", "FR", "PL", "ES", "CA"], "bijoubrigitte": ["DE"], "calvinklein": ["US"], "aldoshoes": ["US"], "peacocks": ["GB"], "assos": ["GB"], "fnac": ["FR"], "canadatire": ["CA"]}

l10n_CM/beta_run_splits/l10n_split_2.json

@@ -0,0 +1 @@
+{"amazon": ["US", "CA", "DE", "FR"], "walmart": ["US", "CA"], "mediamarkt": ["DE"], "zalando": ["PL"], "bestbuy": ["US"], "apple": ["US"], "whittard": ["GB"], "tiffany": ["IT"], "yellowkorner": ["FR"], "burtsbees": ["CA"]}

l10n_CM/beta_run_splits/l10n_split_3.json

@@ -0,0 +1 @@
+{"ebay": ["ES", "IT", "CA", "GB"], "etsy": ["US", "CA"], "justspices": ["DE"], "lowes": ["US"], "newegg": ["US"], "wish": ["US"], "diy": ["GB"], "artsper": ["FR"], "newbalance": ["CA"]}

modules/util.py

@@ -104,7 +104,11 @@ def __init__(self):
             "Yukon": "YT",
         }
         # temporary fix until faker issue is resolved
-        self.country_local_translation = {"Germany": "Deutschland", "Italy": "Italia", "Spain": "España"}
+        self.country_local_translation = {
+            "Germany": "Deutschland",
+            "Italy": "Italia",
+            "Spain": "España",
+        }
         self.fake = None
         self.locale = None
 
@@ -557,7 +561,7 @@ def normalize_regional_phone_numbers(self, phone: str, region: str) -> str:
             "GB": "44",
             "IT": "39",
             "PL": "48",
-            "ES": "34"
+            "ES": "34",
         }
 
         # Sub out anything that matches this regex statement with an empty string to get rid of extensions in generated phone numbers