uk_bin_collection 0.91.1__py3-none-any.whl → 0.92.0__py3-none-any.whl

uk_bin_collection/tests/input.json

@@ -247,6 +247,16 @@
         "wiki_name": "Chorley Council",
         "wiki_note": "Chorley needs to be passed both a Postcode & UPRN in the format of UPRNXXXXXX to work. Find this on [FindMyAddress](https://www.findmyaddress.co.uk/search)"
     },
+    "ColchesterCityCouncil": {
+        "house_number": "29",
+        "paon": "29",
+        "postcode": "CO2 8UN",
+        "skip_get_url": false,
+        "url": "https://www.colchester.gov.uk/your-recycling-calendar",
+        "web_driver": "http://selenium:4444",
+        "wiki_name": "Colchester City Council",
+        "wiki_note": "Pass the house name/number in the house number parameter, wrapped in double quotes"
+    },
     "ConwyCountyBorough": {
         "postcode": "LL30 2DF",
         "uprn": "100100429249",
@@ -1220,10 +1230,10 @@
         "wiki_name": "Wiltshire Council"
     },
     "WindsorAndMaidenheadCouncil": {
-        "house_number": "11",
-        "postcode": "TW19 5NF",
+        "web_driver": "http://selenium:4444",
+        "uprn": "100080371082",
         "skip_get_url": true,
-        "url": "https://my.rbwm.gov.uk/special/find-your-collection-dates",
+        "url": "https://forms.rbwm.gov.uk/bincollections?uprn=",
         "wiki_name": "Windsor and Maidenhead Council"
     },
     "WokingBoroughCouncil": {

uk_bin_collection/uk_bin_collection/councils/ColchesterCityCouncil.py

@@ -0,0 +1,103 @@
+from bs4 import BeautifulSoup
+from selenium.webdriver.common.by import By
+from selenium.webdriver.support import expected_conditions as EC
+from selenium.webdriver.support.wait import WebDriverWait
+
+from uk_bin_collection.uk_bin_collection.common import *
+from uk_bin_collection.uk_bin_collection.get_bin_data import AbstractGetBinDataClass
+
+
+# import the wonderful Beautiful Soup and the URL grabber
+class CouncilClass(AbstractGetBinDataClass):
+    """
+    Concrete classes have to implement all abstract operations of the
+    base class. They can also override some operations with a default
+    implementation.
+    """
+
+    def parse_data(self, page: str, **kwargs) -> dict:
+        driver = None
+        try:
+            data = {"bins": []}
+            user_paon = kwargs.get("paon")
+            user_postcode = kwargs.get("postcode")
+            web_driver = kwargs.get("web_driver")
+            headless = kwargs.get("headless")
+            check_paon(user_paon)
+            check_postcode(user_postcode)
+
+            # Create Selenium webdriver
+            driver = create_webdriver(web_driver, headless, None, __name__)
+            driver.get(
+                "https://www.colchester.gov.uk/your-recycling-calendar/?start=true"
+            )
+
+            accept_button = WebDriverWait(driver, timeout=30).until(
+                EC.element_to_be_clickable((By.ID, "ccc-notify-accept"))
+            )
+            accept_button.click()
+
+            # Wait for the postcode field to appear then populate it
+            inputElement_postcode = WebDriverWait(driver, 10).until(
+                EC.presence_of_element_located((By.CLASS_NAME, "input-text"))
+            )
+            inputElement_postcode.send_keys(user_postcode)
+
+            # Click search button
+            findAddress = WebDriverWait(driver, 10).until(
+                EC.presence_of_element_located((By.CLASS_NAME, "button-small"))
+            )
+            findAddress.click()
+
+            # Wait for the 'Select address' dropdown to appear and select option matching the house name/number
+            WebDriverWait(driver, 10).until(
+                EC.element_to_be_clickable(
+                    (
+                        By.XPATH,
+                        "//select[@class='input-select']//option[contains(., '"
+                        + user_paon
+                        + "')]",
+                    )
+                )
+            ).click()
+
+            WebDriverWait(driver, 10).until(
+                EC.presence_of_element_located((By.CLASS_NAME, "button-small"))
+            ).click()
+
+            # Wait for the collections table to appear
+            WebDriverWait(driver, 10).until(
+                EC.presence_of_element_located((By.CLASS_NAME, "recycling-calendar"))
+            )
+
+            soup = BeautifulSoup(driver.page_source, features="html.parser")
+
+            recyclingcalendar = soup.find("div", {"class": "recycling-calendar"})
+
+            rows = recyclingcalendar.find_all(
+                "div", {"class": "recycling-calendar-row"}
+            )
+
+            for row in rows:
+                collectiondate = datetime.strptime(
+                    row.find("strong").get_text(),
+                    "%d %B %Y",
+                )
+                listobj = row.find("ul")
+                for li in listobj.find_all("li"):
+                    dict_data = {
+                        "type": li.get_text().strip(),
+                        "collectionDate": collectiondate.strftime("%d/%m/%Y"),
+                    }
+                    data["bins"].append(dict_data)
+
+        except Exception as e:
+            # Here you can log the exception if needed
+            print(f"An error occurred: {e}")
+            # Optionally, re-raise the exception if you want it to propagate
+            raise
+        finally:
+            # This block ensures that the driver is closed regardless of an exception
+            if driver:
+                driver.quit()
+        return data

uk_bin_collection/uk_bin_collection/councils/WindsorAndMaidenheadCouncil.py

@@ -1,7 +1,6 @@
 from datetime import datetime
-from urllib.parse import urlparse
 
-import requests
+import dateutil.parser
 from bs4 import BeautifulSoup
 
 from uk_bin_collection.uk_bin_collection.common import *
@@ -17,119 +16,35 @@ class CouncilClass(AbstractGetBinDataClass):
     """
 
     def parse_data(self, page: str, **kwargs) -> dict:
-        root_url = "https://my.rbwm.gov.uk"
-        href_url = ""
-        api_url = "https://my.rbwm.gov.uk/block_refresh/block/47/node/136968?"
-        user_postcode = kwargs.get("postcode")
-        user_paon = kwargs.get("paon")
-
         data = {"bins": []}
+        user_uprn = kwargs.get("uprn")
+        web_driver = kwargs.get("web_driver")
+        headless = kwargs.get("headless")
+        check_uprn(user_uprn)
 
-        requests.packages.urllib3.disable_warnings()
-        s = requests.Session()
-        # Form start
-        headers = {
-            "authority": "my.rbwm.gov.uk",
-            "accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8",
-            "accept-language": "en-GB,en;q=0.8",
-            "cache-control": "max-age=0",
-            "referer": "https://my.rbwm.gov.uk/special/your-collection-dates?uprn=100080371082&subdate=2022-08-19&addr=11%20Douglas%20Lane%20Wraysbury%20Staines%20TW19%205NF",
-            "sec-fetch-dest": "document",
-            "sec-fetch-mode": "navigate",
-            "sec-fetch-site": "same-origin",
-            "sec-fetch-user": "?1",
-            "sec-gpc": "1",
-            "upgrade-insecure-requests": "1",
-            "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.5112.102 Safari/537.36",
-        }
-        s.get(
-            "https://my.rbwm.gov.uk/special/find-your-collection-dates", headers=headers
-        )
+        root_url = "https://forms.rbwm.gov.uk/bincollections?uprn="
+        api_url = root_url + user_uprn
 
-        # Select address
-        headers = {
-            "authority": "my.rbwm.gov.uk",
-            "accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8",
-            "accept-language": "en-GB,en;q=0.8",
-            "cache-control": "max-age=0",
-            "origin": "https://my.rbwm.gov.uk",
-            "referer": "https://my.rbwm.gov.uk/special/find-your-collection-dates",
-            "sec-fetch-dest": "document",
-            "sec-fetch-mode": "navigate",
-            "sec-fetch-site": "same-origin",
-            "sec-fetch-user": "?1",
-            "sec-gpc": "1",
-            "upgrade-insecure-requests": "1",
-            "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.5112.102 Safari/537.36",
-        }
-        request_data = {
-            "atTxtStreet": user_postcode,
-            "nodeid": "x",
-            "formname": "x",
-            "pg": "20",
-            "start": "1",
-            "selectaddress": "Select this address",
-            "selectheading": "The following addresses match the address you entered - choose your address",
-            "arg": "",
-        }
-        response = s.post(
-            "https://my.rbwm.gov.uk/special/address-selector-collection-dates",
-            headers=headers,
-            data=request_data,
-        )
+        # Create Selenium webdriver
+        driver = create_webdriver(web_driver, headless, None, __name__)
+        driver.get(api_url)
 
-        soup = BeautifulSoup(response.content, features="html.parser")
+        soup = BeautifulSoup(driver.page_source, features="html.parser")
         soup.prettify()
 
-        table = soup.find("table")
-        if table:
-            table_rows = table.find_all("tr")
-            for tr in table_rows:
-                td = tr.find_all("td")
-                # row = [i.text for i in td]
-                for item in td:
-                    if user_paon in item.text and user_postcode in item.text:
-                        href_url = td[1].find("a").get("href")
-                        continue
-
-            # Getting to bin data
-            headers = {
-                "authority": "my.rbwm.gov.uk",
-                "accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8",
-                "accept-language": "en-GB,en;q=0.8",
-                "referer": "https://my.rbwm.gov.uk/special/address-selector-collection-dates",
-                "sec-fetch-dest": "document",
-                "sec-fetch-mode": "navigate",
-                "sec-fetch-site": "same-origin",
-                "sec-fetch-user": "?1",
-                "sec-gpc": "1",
-                "upgrade-insecure-requests": "1",
-                "user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/104.0.5112.102 Safari/537.36",
-            }
-            params = {}
-            parsed_params = urlparse(href_url).query.split("&")
-            for item in parsed_params:
-                values = item.split("=")
-                params.update({values[0]: values[1]})
+        # Get collections div
+        next_collection_div = soup.find("div", {"class": "widget-bin-collections"})
 
-            s.get(root_url + href_url, params=params, headers=headers)
-            response = s.get(
-                api_url + href_url.split("?")[1], params=params, headers=headers
-            )
-
-            soup = BeautifulSoup(response.content, features="html.parser")
-            soup.prettify()
-
-            table_rows = soup.find_all("tr")
-            for tr in table_rows:
+        for tbody in next_collection_div.find_all("tbody"):
+            for tr in tbody.find_all("tr"):
                 td = tr.find_all("td")
-                row = [i.text for i in td]
-
-                if len(row) > 0:
-                    dict_data = {
-                        "type": row[0],
-                        "collectionDate": row[1],
-                    }
-                    data["bins"].append(dict_data)
+                next_collection_type = td[0].get_text()
+                next_collection_date = dateutil.parser.parse(td[1].get_text())
+                print(next_collection_date)
+                dict_data = {
+                    "type": next_collection_type,
+                    "collectionDate": next_collection_date.strftime("%d/%m/%Y"),
+                }
+                data["bins"].append(dict_data)
 
         return data

{uk_bin_collection-0.91.1.dist-info → uk_bin_collection-0.92.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: uk_bin_collection
-Version: 0.91.1
+Version: 0.92.0
 Summary: Python Lib to collect UK Bin Data
 Author: Robert Bradley
 Author-email: robbrad182@gmail.com

{uk_bin_collection-0.91.1.dist-info → uk_bin_collection-0.92.0.dist-info}/RECORD

@@ -2,7 +2,7 @@ uk_bin_collection/README.rst,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,
 uk_bin_collection/tests/council_feature_input_parity.py,sha256=DO6Mk4ImYgM5ZCZ-cutwz5RoYYWZRLYx2tr6zIs_9Rc,3843
 uk_bin_collection/tests/features/environment.py,sha256=VQZjJdJI_kZn08M0j5cUgvKT4k3iTw8icJge1DGOkoA,127
 uk_bin_collection/tests/features/validate_council_outputs.feature,sha256=SJK-Vc737hrf03tssxxbeg_JIvAH-ddB8f6gU1LTbuQ,251
-uk_bin_collection/tests/input.json,sha256=wvkZlpCUMF9dJl3nl0Uy1G8OwwInlIBr5s45YHa6HhY,61255
+uk_bin_collection/tests/input.json,sha256=Q3l16fwk4hmvIG6tY6SgP2aZCn-_ghBbMnGjuZCtS_c,61684
 uk_bin_collection/tests/output.schema,sha256=ZwKQBwYyTDEM4G2hJwfLUVM-5v1vKRvRK9W9SS1sd18,1086
 uk_bin_collection/tests/step_defs/step_helpers/file_handler.py,sha256=Ygzi4V0S1MIHqbdstUlIqtRIwnynvhu4UtpweJ6-5N8,1474
 uk_bin_collection/tests/step_defs/test_validate_council.py,sha256=LrOSt_loA1Mw3vTqaO2LpaDMu7rYJy6k5Kr-EOBln7s,3424
@@ -45,6 +45,7 @@ uk_bin_collection/uk_bin_collection/councils/CheshireEastCouncil.py,sha256=VZQsi
 uk_bin_collection/uk_bin_collection/councils/CheshireWestAndChesterCouncil.py,sha256=M58kIHCAZu1K5qBp6_M4sw8nMvHz38YkoxY_sNuhvOI,4780
 uk_bin_collection/uk_bin_collection/councils/ChichesterDistrictCouncil.py,sha256=HxrLcJves7ZsE8FbooymeecTUmScY4R7Oi71vwCePPo,4118
 uk_bin_collection/uk_bin_collection/councils/ChorleyCouncil.py,sha256=M7HjuUaFq8aSnOf_9m1QS4MmPPMmPhF3mLHSrfDPtV0,5194
+uk_bin_collection/uk_bin_collection/councils/ColchesterCityCouncil.py,sha256=Mny-q2rQkWe2Tj1gINwEM1L4AkqQl1EDMAaKY0-deD4,3968
 uk_bin_collection/uk_bin_collection/councils/ConwyCountyBorough.py,sha256=el75qv2QyfWZBU09tJLvD8vLQZ9pCg73u1NBFs6ybo8,1034
 uk_bin_collection/uk_bin_collection/councils/CornwallCouncil.py,sha256=W7_wWHLbE0cyE90GkNaZHmR7Lbd2Aq48A1hKMUNO-vg,2806
 uk_bin_collection/uk_bin_collection/councils/CrawleyBoroughCouncil.py,sha256=_BEKZAjlS5Ad5DjyxqAEFSLn8F-KYox0zmn4BXaAD6A,2367
@@ -183,7 +184,7 @@ uk_bin_collection/uk_bin_collection/councils/WestNorthamptonshireCouncil.py,sha2
 uk_bin_collection/uk_bin_collection/councils/WestSuffolkCouncil.py,sha256=HMFWxM7VMhBuC7iubNGbZYEoCVWi--gRHDJMVdPPFOM,2633
 uk_bin_collection/uk_bin_collection/councils/WiganBoroughCouncil.py,sha256=3gqFA4-BVx_In6QOu3KUNqPN4Fkn9iMlZTeopMK9p6A,3746
 uk_bin_collection/uk_bin_collection/councils/WiltshireCouncil.py,sha256=it2Oh5Kmq3lD30gAZgk2bzZPNCtJcFHyQO1NgOQtfvU,5653
-uk_bin_collection/uk_bin_collection/councils/WindsorAndMaidenheadCouncil.py,sha256=yG5iu294q138Bvk-0IPd4pJDFsOQPDmtU5XXytxnAnA,5554
+uk_bin_collection/uk_bin_collection/councils/WindsorAndMaidenheadCouncil.py,sha256=7Qhznj95ktAQjpWm5C8pbD5UcvfXm7Mwb7_DQxwjGSM,1777
 uk_bin_collection/uk_bin_collection/councils/WokingBoroughCouncil.py,sha256=37igH9g0xe4XIhRhcJ-ZJBU8MxTp5yzgpadWbdE33Yg,5205
 uk_bin_collection/uk_bin_collection/councils/WychavonDistrictCouncil.py,sha256=YuZdzEW0CZLwusm1VQcGRIKXAab_UDFLaCnN60itt_E,5776
 uk_bin_collection/uk_bin_collection/councils/WyreCouncil.py,sha256=zDDa7n4K_zm5PgDL08A26gD9yOOsOhuexI3x2seaBF4,3511
@@ -191,8 +192,8 @@ uk_bin_collection/uk_bin_collection/councils/YorkCouncil.py,sha256=I2kBYMlsD4bId
 uk_bin_collection/uk_bin_collection/councils/council_class_template/councilclasstemplate.py,sha256=4s9ODGPAwPqwXc8SrTX5Wlfmizs3_58iXUtHc4Ir86o,1162
 uk_bin_collection/uk_bin_collection/create_new_council.py,sha256=m-IhmWmeWQlFsTZC4OxuFvtw5ZtB8EAJHxJTH4O59lQ,1536
 uk_bin_collection/uk_bin_collection/get_bin_data.py,sha256=YvmHfZqanwrJ8ToGch34x-L-7yPe31nB_x77_Mgl_vo,4545
-uk_bin_collection-0.91.1.dist-info/LICENSE,sha256=vABBUOzcrgfaTKpzeo-si9YVEun6juDkndqA8RKdKGs,1071
-uk_bin_collection-0.91.1.dist-info/METADATA,sha256=lc_NSdYQBqKFcC13H1v1XV2sTVRLLK2Tiqtg-WVwvew,16642
-uk_bin_collection-0.91.1.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
-uk_bin_collection-0.91.1.dist-info/entry_points.txt,sha256=36WCSGMWSc916S3Hi1ZkazzDKHaJ6CD-4fCEFm5MIao,90
-uk_bin_collection-0.91.1.dist-info/RECORD,,
+uk_bin_collection-0.92.0.dist-info/LICENSE,sha256=vABBUOzcrgfaTKpzeo-si9YVEun6juDkndqA8RKdKGs,1071
+uk_bin_collection-0.92.0.dist-info/METADATA,sha256=I8N4Q2TwfwCtMksjuMddP2UDKku-BH3LrMasukfm4Z0,16642
+uk_bin_collection-0.92.0.dist-info/WHEEL,sha256=sP946D7jFCHeNz5Iq4fL4Lu-PrWrFsgfLXbbkciIZwg,88
+uk_bin_collection-0.92.0.dist-info/entry_points.txt,sha256=36WCSGMWSc916S3Hi1ZkazzDKHaJ6CD-4fCEFm5MIao,90
+uk_bin_collection-0.92.0.dist-info/RECORD,,