proposed regex function extracted

mini-kep · Dec 9, 2017 · c87c6ba · c87c6ba
1 parent 16f9074
commit c87c6ba
Showing 1 changed file with 90 additions and 58 deletions.
diff --git a/issues/namelist.py b/issues/namelist.py
@@ -1,53 +1,55 @@
-names = [
-  "BRENT", 
-  "CPI_ALCOHOL_rog", 
-  "CPI_FOOD_rog", 
-  "CPI_NONFOOD_rog", 
-  "CPI_rog", 
-  "CPI_SERVICES_rog", 
-  "EXPORT_GOODS_bln_usd", 
-  "GDP_bln_rub", 
-  "GDP_yoy", 
-  "GOV_EXPENSE_CONSOLIDATED_bln_rub", 
-  "GOV_EXPENSE_FEDERAL_bln_rub", 
-  "GOV_EXPENSE_SUBFEDERAL_bln_rub", 
-  "GOV_REVENUE_CONSOLIDATED_bln_rub", 
-  "GOV_REVENUE_FEDERAL_bln_rub", 
-  "GOV_REVENUE_SUBFEDERAL_bln_rub", 
-  "GOV_SURPLUS_FEDERAL_bln_rub", 
-  "GOV_SURPLUS_SUBFEDERAL_bln_rub", 
-  "IMPORT_GOODS_bln_usd", 
-  "INDPRO_rog", 
-  "INDPRO_yoy", 
-  "INVESTMENT_bln_rub", 
-  "INVESTMENT_rog", 
-  "INVESTMENT_yoy", 
-  "RETAIL_SALES_bln_rub", 
-  "RETAIL_SALES_FOOD_bln_rub", 
-  "RETAIL_SALES_FOOD_rog", 
-  "RETAIL_SALES_FOOD_yoy", 
-  "RETAIL_SALES_NONFOOD_bln_rub", 
-  "RETAIL_SALES_NONFOOD_rog", 
-  "RETAIL_SALES_NONFOOD_yoy", 
-  "RETAIL_SALES_rog", 
-  "RETAIL_SALES_yoy", 
-  "TRANSPORT_FREIGHT_bln_tkm", 
-  "UNEMPL_pct", 
-  "USDRUR_CB", 
-  "UST_10YEAR", 
-  "UST_1MONTH", 
-  "UST_1YEAR", 
-  "UST_20YEAR", 
-  "UST_2YEAR", 
-  "UST_30YEAR", 
-  "UST_3MONTH", 
-  "UST_3YEAR", 
-  "UST_5YEAR", 
-  "UST_6MONTH", 
-  "UST_7YEAR", 
-  "WAGE_NOMINAL_rub", 
-  "WAGE_REAL_rog", 
-  "WAGE_REAL_yoy"
+import re
+
+NAMES = [
+    "BRENT",
+    "CPI_ALCOHOL_rog",
+    "CPI_FOOD_rog",
+    "CPI_NONFOOD_rog",
+    "CPI_rog",
+    "CPI_SERVICES_rog",
+    "EXPORT_GOODS_bln_usd",
+    "GDP_bln_rub",
+    "GDP_yoy",
+    "GOV_EXPENSE_CONSOLIDATED_bln_rub",
+    "GOV_EXPENSE_FEDERAL_bln_rub",
+    "GOV_EXPENSE_SUBFEDERAL_bln_rub",
+    "GOV_REVENUE_CONSOLIDATED_bln_rub",
+    "GOV_REVENUE_FEDERAL_bln_rub",
+    "GOV_REVENUE_SUBFEDERAL_bln_rub",
+    "GOV_SURPLUS_FEDERAL_bln_rub",
+    "GOV_SURPLUS_SUBFEDERAL_bln_rub",
+    "IMPORT_GOODS_bln_usd",
+    "INDPRO_rog",
+    "INDPRO_yoy",
+    "INVESTMENT_bln_rub",
+    "INVESTMENT_rog",
+    "INVESTMENT_yoy",
+    "RETAIL_SALES_bln_rub",
+    "RETAIL_SALES_FOOD_bln_rub",
+    "RETAIL_SALES_FOOD_rog",
+    "RETAIL_SALES_FOOD_yoy",
+    "RETAIL_SALES_NONFOOD_bln_rub",
+    "RETAIL_SALES_NONFOOD_rog",
+    "RETAIL_SALES_NONFOOD_yoy",
+    "RETAIL_SALES_rog",
+    "RETAIL_SALES_yoy",
+    "TRANSPORT_FREIGHT_bln_tkm",
+    "UNEMPL_pct",
+    "USDRUR_CB",
+    "UST_10YEAR",
+    "UST_1MONTH",
+    "UST_1YEAR",
+    "UST_20YEAR",
+    "UST_2YEAR",
+    "UST_30YEAR",
+    "UST_3MONTH",
+    "UST_3YEAR",
+    "UST_5YEAR",
+    "UST_6MONTH",
+    "UST_7YEAR",
+    "WAGE_NOMINAL_rub",
+    "WAGE_REAL_rog",
+    "WAGE_REAL_yoy"
 ]
 
 # part of issue https://github.com/mini-kep/db/blob/master/doc/listing.md
@@ -58,21 +60,51 @@
 concepts = dict(labor=['WAGE_*', 'UNEMPL'],
                 output=['IND*', 'TRANSPORT_FREIGHT'])
 
-# TODO 1: make a selection function that produces a list of 
+
+# TODO 1: make a selection function that produces a list of
 #       variable names based on regex or variable head
 
 #       use <https://github.com/mini-kep/parser-rosstat-kep/blob/master/src/csv2df/util_label.py>
 #       to split labels
 
 # TODO 2: write tests for the fucntion using the guidelines
 #         <https://github.com/mini-kep/guidelines/blob/master/testing.md>
+
+
+def make_regex(patterns):
+    matched = []
+    base_regex = re.compile('([A-Z]+_?)+')
+    for p in patterns:
+        match = base_regex.match(p)
+        matched.append(match.group())
+    return '|'.join(matched)
+    # 1. get_re() - это 'не вынести регекс в отдельную функцию'
+    #    make_regex() делает что-то полезное с законченной зоно ответственности
+    #    + ее можно тестировать
+    # 2. небольшой костыль был: 
+    #    search = search[:-1]
+    #    если что-то надо с хвоста отрезать - непрвильно объединение сделали
+    # 3. теперь галвный вопрос - зачем это делалось?
+
+assert make_regex(['WAGE_*', 'UNEMPL']) == 'WAGE_|UNEMPL'
 
-def make_namelist(patterns):
-    pass
+
+
+def make_namelist(patterns, names=NAMES):
+    namelist = []
+    regex = re.compile(make_regex(patterns))
+    for name in names:
+        hit = regex.search(name)
+        if hit:
+            namelist.append(name)
+    return namelist
+
+
+
 
 labor = make_namelist(concepts['labor'])
-assert set(labor) == set(["WAGE_NOMINAL_rub", 
-  "WAGE_REAL_rog", 
-  "WAGE_REAL_yoy",
-  "UNEMPL_pct"
-  ])
+assert set(labor) == set(["WAGE_NOMINAL_rub",
+                          "WAGE_REAL_rog",
+                          "WAGE_REAL_yoy",
+                          "UNEMPL_pct"
+                          ])