De-hardcode 13th of March.

[berlin-corona-table] / scrape.py

diff --git a/scrape.py b/scrape.py
index bbb4835c00fc5eedebe064ab540264d1470b0c05..f11f315f5e73e8f3d6b5b3b98c0f7fbfac63267a 100755 (executable)
--- a/scrape.py
+++ b/scrape.py
@@ -88,23 +88,6 @@ data = {
         'TK': {'growth': 14, 'total': 183},
         'sum': {'growth': 263, 'total': 3486}
     },
-   # This one has no press release but can be reconstructed from
-   # the neighbour ones.
-   datetime.datetime(2020, 3, 13): {
-        'CW': {'growth': 16, 'total': 47},
-        'FK': {'growth': 8, 'total': 22},
-        'Li': {'growth': 2, 'total': 8},
-        'MH': {'growth': 1, 'total': 4},
-        'Mi': {'growth': 9, 'total': 29},
-        'Ne': {'growth': 6, 'total': 16},
-        'Pa': {'growth': 11, 'total': 26},
-        'Re': {'growth': 0, 'total': 11},
-        'Sp': {'growth': 1, 'total': 9},
-        'SZ': {'growth': 0, 'total': 20},
-        'TS': {'growth': 1, 'total': 17},
-        'TK': {'growth': 3, 'total': 7},
-        'sum': {'growth': 58, 'total': 216}
-   },
    # Here the growth numbers needed to be reconstructed.
    datetime.datetime(2020, 3, 10): {
         'CW': {'growth': 2, 'total': 15},
@@ -169,6 +152,70 @@ data = {
         'TK': {'growth': 0, 'total': 0},
         'sum': {'growth': 12, 'total': 40}
    },
+   # Here the growth numbers needed to be reconstructed.
+   datetime.datetime(2020, 3, 6): {
+        'CW': {'growth': 1, 'total': 3},
+        'FK': {'growth': 0, 'total': 4},
+        'Li': {'growth': 0, 'total': 0},
+        'MH': {'growth': 0, 'total': 1},
+        'Mi': {'growth': 4, 'total': 7},
+        'Ne': {'growth': 1, 'total': 2},
+        'Pa': {'growth': 1, 'total': 4},
+        'Re': {'growth': 0, 'total': 3},
+        'Sp': {'growth': 0, 'total': 0},
+        'SZ': {'growth': 0, 'total': 0},
+        'TS': {'growth': 2, 'total': 4},
+        'TK': {'growth': 0, 'total': 0},
+        'sum': {'growth': 9, 'total': 28}
+   },
+   # Here the growth numbers needed to be reconstructed.
+   datetime.datetime(2020, 3, 5): {
+        'CW': {'growth': 2, 'total': 2},
+        'FK': {'growth': 0, 'total': 4},
+        'Li': {'growth': 0, 'total': 0},
+        'MH': {'growth': 0, 'total': 1},
+        'Mi': {'growth': 0, 'total': 3},
+        'Ne': {'growth': 0, 'total': 1},
+        'Pa': {'growth': 1, 'total': 3},
+        'Re': {'growth': 2, 'total': 3},
+        'Sp': {'growth': 0, 'total': 0},
+        'SZ': {'growth': 0, 'total': 0},
+        'TS': {'growth': 1, 'total': 2},
+        'TK': {'growth': 0, 'total': 0},
+        'sum': {'growth': 6, 'total': 19}
+   },
+   # Here the growth numbers needed to be reconstructed.
+   datetime.datetime(2020, 3, 4): {
+        'CW': {'growth': 0, 'total': 0},
+        'FK': {'growth': 2, 'total': 4},
+        'Li': {'growth': 0, 'total': 0},
+        'MH': {'growth': 0, 'total': 1},
+        'Mi': {'growth': 0, 'total': 3},
+        'Ne': {'growth': 0, 'total': 1},
+        'Pa': {'growth': 1, 'total': 2},
+        'Re': {'growth': 1, 'total': 1},
+        'Sp': {'growth': 0, 'total': 0},
+        'SZ': {'growth': 0, 'total': 0},
+        'TS': {'growth': 0, 'total': 1},
+        'TK': {'growth': 0, 'total': 0},
+        'sum': {'growth': 4, 'total': 13}
+   },
+   # Here the growth numbers needed to be reconstructed.
+   datetime.datetime(2020, 3, 3): {
+        'CW': {'growth': 0, 'total': 0},
+        'FK': {'growth': 2, 'total': 2},
+        'Li': {'growth': 0, 'total': 0},
+        'MH': {'growth': 0, 'total': 1},
+        'Mi': {'growth': 0, 'total': 3},
+        'Ne': {'growth': 0, 'total': 1},
+        'Pa': {'growth': 1, 'total': 1},
+        'Re': {'growth': 0, 'total': 0},
+        'Sp': {'growth': 0, 'total': 0},
+        'SZ': {'growth': 0, 'total': 0},
+        'TS': {'growth': 0, 'total': 1},
+        'TK': {'growth': 0, 'total': 0},
+        'sum': {'growth': 3, 'total': 9}
+   },
 }
 fixes = {
    # Here the official total is 215, while the summation of district
@@ -250,7 +297,8 @@ for path in day_urls:
     if date <= datetime.datetime(2020, 3, 13):
        date = date - datetime.timedelta(days=1)
     table = soup.find('table')
-    if table is None and date in data:
+    # For 13th of March we lack a press release.
+    if table is None and (date in data or date == datetime.datetime(2020, 3, 13)):
         continue
     data[date] = {}
     for tr in [tr for tr in table.children if type(tr) == bs4.element.Tag][1:]:
@@ -270,6 +318,21 @@ for path in day_urls:
         total = int(total_str.replace('.', ''))
         data[date][district_short] = {'growth': growth, 'total': total}
     first_run = False
+
+# Reconstruct data for 13th of March.
+day_target = datetime.datetime(2020, 3, 13)
+day_after = day_target + datetime.timedelta(days=1)
+day_before = day_target - datetime.timedelta(days=1)
+data[day_target] = {}
+for district in [d for d in districts_sorted]:
+   data[day_target][district] = {}
+   total_after = data[day_after][district]['total']
+   growth_after = data[day_after][district]['growth']
+   total_target = total_after - growth_after
+   data[day_target][district]['total'] = total_target
+   total_before = data[day_before][district]['total']
+   data[day_target][district]['growth'] = total_target - total_before
+
 dates_sorted = list(data.keys())
 dates_sorted.sort()
 dates_sorted.reverse()