Change address recognizer (not final)

.gitignore

@@ -2,4 +2,5 @@
 .venv
 __pycache__
 .env
-data*.csv
+data*.csv
+.idea/

parser/__main__.py

@@ -8,7 +8,7 @@ from . import pipeline
 
 def job():
     parser = pipeline()
-    parser.save_df(f'./data_{parser.today.strftime("%d-%m-%y_%H:%M")}.csv')
+    parser.save_df(f'./data_{parser.today.strftime("%d-%m-%y_%H.%M")}.csv')
 
 
 if len(sys.argv) == 2:

parser/address.py

@@ -7,20 +7,30 @@ import pandas as pd
 
 T = TypeVar("T")
 
-STREET_PREFIXES = ("ул.", "бул.", "пр.", "ул", "бул", "пр", "ш.", "ш", "пер.", "пер")
-HOUSES_PREFIXES = ("д.", "д")
+SETTLEMENTS_PREFIXES=("г","мо","р-н","п","д","гп","c","хутор","массив","тер","СНТ","СТ","ДСК","ДНП","ДПК","НП","садоводство")
+STREET_PREFIXES = ("ул", "бул", "пр", "ш", "пер", "дор", "маг", "наб", "пл", "просп", "туп", "аллея", "мост", "парк", "кольцо","проезд")
+HOUSES_PREFIXES = ("д.", "д","уч","участок")
+BUILDING_PREFIXES=("к", "корп")
+LETTER=("лит", "литера")
 
-
-def unfold_house_ranges(token: str) -> str:
+def unfold_house_ranges(address:str, token: str) -> List[str]:
+    adresses=[]
     pairs_strings = re.findall(r"([\d]+-[\d]+)", token)
     for pair_string in pairs_strings:
         a, b = pair_string.split("-")
         a, b = int(a), int(b)
 
         if b > a:
-            token = token.replace(pair_string, ", ".join(map(str, range(a, b + 1))))
+            token = token.replace(pair_string,"")
+            adresses += [address + " "+ token + number for number in map(str, range(a, b + 1))]
 
-    return token
+
+        else:
+            token = token.replace("-", "/")
+            adresses += address + " " + token
+    if not adresses:
+        adresses.append(address + " " + token)
+    return adresses
 
 
 def unfold_houses_list(token: str) -> List[str]:
@@ -42,28 +52,54 @@ def flatten(arr: Iterable[List[T]]) -> List[T]:
     return sum(arr, [])
 
 
+# TODO: переработать систему из if в нормальный вид и классификация чисел/букв
 def split_address(address: str) -> List[str]:
     if ";" in address:
-        return flatten(map(unfold_houses_list, address.split(";")))
-    elif "," in address:
-        tokens = re.split(r"(,)", address)
+        address = address.replace(";", ",")
+    if "," in address:
+        tokens = address.split(",")
 
         tokens = list(map(str.strip, filter(lambda token: token != "", tokens)))
-
         res = []
-        accumulator = ""
+        accumulator = []
 
         for i in range(len(tokens)):
-            if any_of_in(STREET_PREFIXES, tokens[i].lower()) and any_of_in(
-                STREET_PREFIXES, accumulator.lower()
-            ):
-                res += unfold_houses_list(accumulator)
-                accumulator = ""
 
-            accumulator += tokens[i]
+            # TODO: напселённые пункты
+            # if any_of_in(SETTLEMENTS_PREFIXES, tokens[i].lower())
+            #     accumulator += tokens[i]
 
-        res += unfold_houses_list(accumulator)
+            # улицы
+            if any_of_in(STREET_PREFIXES, tokens[i].lower()):
+                if accumulator and any_of_in(STREET_PREFIXES, "".join(accumulator).lower() ):
+                    res.append( " ".join(accumulator))
+                    accumulator=[]
+                accumulator.append(tokens[i])
 
+            # дома
+            elif any_of_in(HOUSES_PREFIXES, tokens[i].lower()):
+                if accumulator and any_of_in(HOUSES_PREFIXES, accumulator[-1].lower()):
+                    res.append(" ".join(accumulator))
+                    accumulator.pop()
+                res.append(unfold_house_ranges(" ".join(accumulator),tokens[i]))
+                accumulator=res[-1]
+                res.pop()
+            # корпус
+            elif any_of_in(BUILDING_PREFIXES, tokens[i].lower()):
+                if accumulator and any_of_in(BUILDING_PREFIXES, accumulator[-1].lower() ):
+                    res.append( " ".join(accumulator))
+                    accumulator.pop()
+                accumulator.append(tokens[i])
+            # литера
+            elif any_of_in(LETTER, tokens[i].lower()):
+                if accumulator and any_of_in(LETTER, accumulator[-1].lower() ):
+                    res.append(" ".join(accumulator))
+                    accumulator.pop()
+                accumulator.append (tokens[i])
+            else:
+                accumulator.append(tokens[i])
+
+        res.append(" ".join(accumulator))
         return res
 
     return [address]