Merge branch 'master' of github.com:KUlishevgeniy/c22712

This commit is contained in:
pErfEcto2
2023-03-26 22:09:04 +03:00
12 changed files with 312 additions and 49 deletions
+35 -43
View File
@@ -4,9 +4,8 @@
<option name="autoReloadType" value="SELECTIVE" /> <option name="autoReloadType" value="SELECTIVE" />
</component> </component>
<component name="ChangeListManager"> <component name="ChangeListManager">
<list default="true" id="e5615680-c090-497d-a104-1337ed85f850" name="Changes" comment="ганьг или баньг&#10;СХС"> <list default="true" id="e5615680-c090-497d-a104-1337ed85f850" name="Changes" comment="parser">
<change beforePath="$PROJECT_DIR$/.idea/workspace.xml" beforeDir="false" afterPath="$PROJECT_DIR$/.idea/workspace.xml" afterDir="false" /> <change beforePath="$PROJECT_DIR$/.idea/workspace.xml" beforeDir="false" afterPath="$PROJECT_DIR$/.idea/workspace.xml" afterDir="false" />
<change beforePath="$PROJECT_DIR$/main.py" beforeDir="false" afterPath="$PROJECT_DIR$/main.py" afterDir="false" />
</list> </list>
<option name="SHOW_DIALOG" value="false" /> <option name="SHOW_DIALOG" value="false" />
<option name="HIGHLIGHT_CONFLICTS" value="true" /> <option name="HIGHLIGHT_CONFLICTS" value="true" />
@@ -36,39 +35,12 @@
<option name="hideEmptyMiddlePackages" value="true" /> <option name="hideEmptyMiddlePackages" value="true" />
<option name="showLibraryContents" value="true" /> <option name="showLibraryContents" value="true" />
</component> </component>
<component name="PropertiesComponent">{ <component name="PropertiesComponent"><![CDATA[{
&quot;keyToString&quot;: { "keyToString": {
&quot;last_opened_file_path&quot;: &quot;C:/Users/rusta/c22712&quot;, "last_opened_file_path": "C:/Users/71332/PycharmProjects/c22712"
&quot;settings.editor.selected.configurable&quot;: &quot;com.jetbrains.python.configuration.PyActiveSdkModuleConfigurable&quot;
} }
}</component> }]]></component>
<component name="RecentsManager"> <component name="RunManager" selected="Python.zaparsil">
<key name="CopyFile.RECENT_KEYS">
<recent name="C:\Users\rusta\c22712" />
</key>
</component>
<component name="RunManager" selected="Python.main">
<configuration name="1page1win" type="PythonConfigurationType" factoryName="Python" temporary="true" nameIsGenerated="true">
<module name="22712" />
<option name="INTERPRETER_OPTIONS" value="" />
<option name="PARENT_ENVS" value="true" />
<envs>
<env name="PYTHONUNBUFFERED" value="1" />
</envs>
<option name="SDK_HOME" value="" />
<option name="WORKING_DIRECTORY" value="$PROJECT_DIR$/../OneDrive/Рабочий стол/парсинг" />
<option name="IS_MODULE_SDK" value="false" />
<option name="ADD_CONTENT_ROOTS" value="true" />
<option name="ADD_SOURCE_ROOTS" value="true" />
<option name="SCRIPT_NAME" value="$PROJECT_DIR$/../OneDrive/Рабочий стол/парсинг/1page1win.py" />
<option name="PARAMETERS" value="" />
<option name="SHOW_COMMAND_LINE" value="false" />
<option name="EMULATE_TERMINAL" value="false" />
<option name="MODULE_MODE" value="false" />
<option name="REDIRECT_INPUT" value="false" />
<option name="INPUT_FILE" value="" />
<method v="2" />
</configuration>
<configuration name="main" type="PythonConfigurationType" factoryName="Python" nameIsGenerated="true"> <configuration name="main" type="PythonConfigurationType" factoryName="Python" nameIsGenerated="true">
<module name="22712" /> <module name="22712" />
<option name="INTERPRETER_OPTIONS" value="" /> <option name="INTERPRETER_OPTIONS" value="" />
@@ -76,10 +48,9 @@
<envs> <envs>
<env name="PYTHONUNBUFFERED" value="1" /> <env name="PYTHONUNBUFFERED" value="1" />
</envs> </envs>
<option name="SDK_HOME" value="C:\Users\rusta\PycharmProjects\pythonProject\venv\Scripts\python.exe" /> <option name="SDK_HOME" value="" />
<option name="SDK_NAME" value="Python 3.11 (pythonProject)" />
<option name="WORKING_DIRECTORY" value="$PROJECT_DIR$" /> <option name="WORKING_DIRECTORY" value="$PROJECT_DIR$" />
<option name="IS_MODULE_SDK" value="false" /> <option name="IS_MODULE_SDK" value="true" />
<option name="ADD_CONTENT_ROOTS" value="true" /> <option name="ADD_CONTENT_ROOTS" value="true" />
<option name="ADD_SOURCE_ROOTS" value="true" /> <option name="ADD_SOURCE_ROOTS" value="true" />
<EXTENSION ID="PythonCoverageRunConfigurationExtension" runner="coverage.py" /> <EXTENSION ID="PythonCoverageRunConfigurationExtension" runner="coverage.py" />
@@ -92,9 +63,30 @@
<option name="INPUT_FILE" value="" /> <option name="INPUT_FILE" value="" />
<method v="2" /> <method v="2" />
</configuration> </configuration>
<configuration name="zaparsil" type="PythonConfigurationType" factoryName="Python" temporary="true" nameIsGenerated="true">
<module name="22712" />
<option name="INTERPRETER_OPTIONS" value="" />
<option name="PARENT_ENVS" value="true" />
<envs>
<env name="PYTHONUNBUFFERED" value="1" />
</envs>
<option name="SDK_HOME" value="" />
<option name="WORKING_DIRECTORY" value="$PROJECT_DIR$/Задания/task1/Sukhanov" />
<option name="IS_MODULE_SDK" value="true" />
<option name="ADD_CONTENT_ROOTS" value="true" />
<option name="ADD_SOURCE_ROOTS" value="true" />
<option name="SCRIPT_NAME" value="$PROJECT_DIR$/Задания/task1/Sukhanov/zaparsil.py" />
<option name="PARAMETERS" value="" />
<option name="SHOW_COMMAND_LINE" value="false" />
<option name="EMULATE_TERMINAL" value="false" />
<option name="MODULE_MODE" value="false" />
<option name="REDIRECT_INPUT" value="false" />
<option name="INPUT_FILE" value="" />
<method v="2" />
</configuration>
<recent_temporary> <recent_temporary>
<list> <list>
<item itemvalue="Python.1page1win" /> <item itemvalue="Python.zaparsil" />
</list> </list>
</recent_temporary> </recent_temporary>
</component> </component>
@@ -143,12 +135,12 @@
<option name="project" value="LOCAL" /> <option name="project" value="LOCAL" />
<updated>1676934066947</updated> <updated>1676934066947</updated>
</task> </task>
<task id="LOCAL-00006" summary="ганьг или баньг&#10;СХС"> <task id="LOCAL-00006" summary="parser">
<created>1678906921509</created> <created>1679839777822</created>
<option name="number" value="00006" /> <option name="number" value="00006" />
<option name="presentableId" value="LOCAL-00006" /> <option name="presentableId" value="LOCAL-00006" />
<option name="project" value="LOCAL" /> <option name="project" value="LOCAL" />
<updated>1678906921509</updated> <updated>1679839777822</updated>
</task> </task>
<option name="localTasksCounter" value="7" /> <option name="localTasksCounter" value="7" />
<servers /> <servers />
@@ -170,14 +162,14 @@
<component name="VcsManagerConfiguration"> <component name="VcsManagerConfiguration">
<MESSAGE value="Первый коммит" /> <MESSAGE value="Первый коммит" />
<MESSAGE value="parser" /> <MESSAGE value="parser" />
<MESSAGE value="ганьг или баньг&#10;СХС" /> <option name="LAST_COMMIT_MESSAGE" value="parser" />
<option name="LAST_COMMIT_MESSAGE" value="ганьг или баньг&#10;СХС" />
</component> </component>
<component name="XDebuggerManager"> <component name="XDebuggerManager">
<breakpoint-manager> <breakpoint-manager>
<breakpoints> <breakpoints>
<line-breakpoint enabled="true" suspend="THREAD" type="python-line"> <line-breakpoint enabled="true" suspend="THREAD" type="python-line">
<url>file://$PROJECT_DIR$/main.py</url> <url>file://$PROJECT_DIR$/main.py</url>
<line>8</line>
<option name="timeStamp" value="1" /> <option name="timeStamp" value="1" />
</line-breakpoint> </line-breakpoint>
</breakpoints> </breakpoints>
+1 -2
View File
@@ -28,7 +28,6 @@ pictures = b_soup.find_all('div', class_="catalog-list-item__img-wrapper")
for i in range(15): for i in range(15):
url = 'https://amwine.ru'+pictures[i].find('a').find('img').attrs['data-src'] url = 'https://amwine.ru'+pictures[i].find('a').find('img').attrs['data-src']
filename = f"Programming\\23.03\img\{i}.jpg" filename = f"Programming\\23.03\img\{i}.jpg"
print(filename)
wget.download(url, filename) wget.download(url, filename)
ins_qwery = f"""insert into public.Parser(page_name, price, priceDis, mark, scr) values ('{name[i].text}', '{price[i].text}', '{priceDis[i].text}', '{mark[i].text}', '{filename}')""" ins_qwery = f"""insert into public.Parser(page_name, price, priceDis, mark, scr) values ('{name[i].text}', '{price[i].text}', '{priceDis[i].text}', '{mark[i].text}', '{filename}')"""
cursor.execute(ins_qwery) cursor.execute(ins_qwery)
@@ -39,4 +38,4 @@ for i in range(15):
cursor.close() cursor.close()
connection.close() connection.close()
+32
View File
@@ -0,0 +1,32 @@
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
import psycopg2
import wget
browser = webdriver.Chrome(service=Service('C:\Desktop\exe\chromedriver.exe'))
browser.get('https://flawery.ru/moscow/bouquets/event-yanvary25/?min=0&max=max&sorting=discount&filter_24h=on')
soup = BeautifulSoup(browser.page_source, "lxml")
Name = soup.find_all(attrs={"class": "catalog_title"})
Price = soup.find_all(attrs={"class": "catalog_price_now"})
Delivery_Time = soup.find_all(attrs={"class": "catalog_express"})
Discount = soup.find_all(attrs={"class": "catalog_sale"})
Image = soup.find_all(attrs={"class": "catalog_item catalog_item_popup"})
connection = psycopg2.connect(host='localhost', dbname='FHWDB', user='postgres', password='Q1w2e3r4')
cursor = connection.cursor()
create_q = '''CREATE TABLE Parse
(ID serial primary key, Name varchar(100), Price varchar(9), Delivery_Time varchar(25), Discount varchar(10), src varchar(110))'''
cursor.execute(create_q)
connection.commit()
for j in range(60):
url = 'https://flawery.ru'+Image[j].find('a').find('picture').find('img').attrs['src']
tempf = f"C:\\Users\\user\\Desktop\\FHWDB{j}.jpg"
wget.download(url, tempf)
insert_query = f'''INSERT into public.Parse(Name, Price, Delivery_Time, Discount, src) values ('{Name[j].text}', '{Price[j].text}', '{Delivery_Time[j].text}', '{Discount[j].text}', '{tempf}') '''
cursor.execute(insert_query)
connection.commit()
cursor.close()
connection.close()
+1 -1
View File
@@ -1,4 +1,4 @@
#made by dialuna (or just Diana Gasanova) <3 # Diana Gasanova
import requests import requests
from bs4 import BeautifulSoup as bs from bs4 import BeautifulSoup as bs
URL = "https://www.kinopoisk.ru/lists/movies/top-250-2020/" URL = "https://www.kinopoisk.ru/lists/movies/top-250-2020/"
@@ -0,0 +1,26 @@
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
from wget import download
import psycopg2
connection = psycopg2.connect(host='localhost', dbname='dbdata',
user='postgres', password='Q1w2e3r4')
cursor=connection.cursor()
s = Service('C:\\Users\\Dasha\\Desktop\\Progect\\chromedriver.exe')
browser = webdriver.Chrome(service=s)
url = 'https://www.tretyakovgallery.ru/'
browser.get(url)
html_text=browser.page_source
soup = BeautifulSoup(html_text, 'lxml')
products = soup.find_all(attrs={"class": "building-item"})
for i, product in enumerate(products):
link = product.find("img").attrs.get("src")
tag = product.find("strong")
if tag is not None:
download(url + link, f"img\\{i}.jpg")
cursor.execute(f"""INSERT INTO public.parsing(name, link)
VALUES
( '{tag.text}', '{url + link}')""")
connection.commit()
cursor.close()
connection.close()
+71
View File
@@ -0,0 +1,71 @@
import psycopg2
import wget
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
import time
i = 1
s = Service("С:\DATA\ChromeDriver\chromedriver.exe")
browser = webdriver.Chrome(service=s)
browser.get("https://www.mvideo.ru/noutbuki-planshety-komputery-8/noutbuki-118?from=under_search")
height = browser.execute_script("return document.body.scrollHeight")
time.sleep(4)
while 900*i < height:
height = browser.execute_script("return document.body.scrollHeight")
browser.execute_script(f"window.scrollTo(0,{900*i})")
time.sleep(1)
i+=1
html_text = browser.page_source
soup = BeautifulSoup(html_text, 'lxml')
productst = soup.find_all('a', class_="product-title__text")
pricest = soup.find_all('span', class_="price__main-value")
features = soup.find_all('ul', class_="product-feature-list product-feature-list--undefined")
pictures = soup.find_all('img', class_="product-picture__img product-picture__img--list")
products, prices, diagonals, resolutions, CPUs, RAMs, Graphics_controllers, Volumes = [],[],[],[],[],[],[],[]
for i in range(len(productst)):
products.append(productst[i].text.strip())
prices.append(pricest[i].text.replace("\xa0", "").strip())
for feature in features:
u = feature.text
diagonals.append(u[20:u.find("\"",20)+1])
resolutions.append(u[u.find("\"",20)+2:u.find(".", 23)+1])
CPUs.append(u[u.find("Процессор",20)+9:u.find("ГГц",20)+3].replace("\xa0",""))
RAMs.append(" ".join(u[u.find("(RAM)",20)+5:u.find("Графический контроллер")].split()))
Graphics_controllers.append(u[u.find("Графический контроллер")+22:u.rfind("Объем")-2])
Volumes.append(u[u.rfind("Объем")+6:].replace("\xa0", "").strip())
for i in range(len(Volumes)):
u = Volumes[i]
if u.rfind("D")!=-1: Volumes[i] = u[:u.rfind("D")+1]+ " " + u[u.rfind("D")+1:]
else: Volumes[i] = u[:u.rfind("С")+1]+ " " + u[u.rfind("С")+1:]
connection=psycopg2.connect(host='localhost', dbname='dbdata', user='postgres', password='Q1w2e3r4')
cursor = connection.cursor()
insert = """CREATE TABLE public.laptops(
id serial primary key,
product varchar(100),
price varchar(15),
diagonal varchar(5),
resolution varchar(20),
CPU varchar(50),
RAM varchar(10),
Graphics_Controller varchar(40),
Volume varchar(25),
src varchar(100)
);
"""
cursor.execute(insert)
connection.commit()
for i in range(len(products)):
url = "https://"+pictures[i]['src']
filename = f"C:\DATA\Images\{i+1}.jpg"
wget.download(url, filename)
insert = f"""INSERT INTO public.laptops(
product, price, diagonal, resolution, CPU, RAM, Graphics_Controller, Volume, src)
VALUES
('{products[i]}', '{prices[i]}', '{diagonals[i]}', '{resolutions[i]}', '{CPUs[i]}', '{RAMs[i]}', '{Graphics_controllers[i]}', '{Volumes[i]}', '{filename}');"""
cursor.execute(insert)
connection.commit()
cursor.execute("select * from laptops")
print("Результат", cursor.fetchall())
cursor.close()
connection.close()
@@ -0,0 +1,42 @@
import psycopg2
import wget
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
s = Service("С:\DATA\ChromeDriver\chromedriver.exe")
browser = webdriver.Chrome(service=s)
browser.get("https://online.metro-cc.ru/virtual/assortiment_rioba-4887?from=under_search&page=2")
html_text = browser.page_source
soup = BeautifulSoup(html_text, 'lxml')
productst = soup.find_all('span', class_="product-card-name__text")
pricest = soup.find_all('span', class_="product-price__sum-rubles")
pictures = soup.find_all('img', class_="product-card-photo__image")
connection=psycopg2.connect(host='localhost', dbname='dbdata', user='postgres', password='Q1w2e3r4')
cursor = connection.cursor()
products, prices = [], []
for i in range(len(productst)):
products.append(productst[i].text.replace("\n", "").strip())
prices.append(pricest[i].text+" Р/шт")
insert = """CREATE TABLE public.Chocolate(
id serial primary key,
product varchar(100),
price varchar(25),
src varchar(50)
);
"""
cursor.execute(insert)
connection.commit()
for i in range(len(products)):
url = pictures[i]['src']
filename = f"C:\DATA\Images2\{i+1}.jpg"
wget.download(url, filename)
insert = f"""INSERT INTO public.Chocolate(
product, price, src)
VALUES
('{products[i]}', '{prices[i]}', '{filename}');"""
cursor.execute(insert)
connection.commit()
cursor.execute("select * from Chocolate")
print("Результат", cursor.fetchall())
cursor.close()
connection.close()
@@ -0,0 +1,41 @@
import psycopg2
import wget
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
connection = psycopg2.connect(host='localhost', dbname='dbdata', user='postgres', password='Q1w2e3r4')
cursor = connection.cursor()
cr_qwery = """ create table Parser
(id serial primary key, name varchar(100), sale varchar(30), price varchar(10), rate varchar(10), picture varchar(100))"""
#cursor.execute(cr_qwery)
connection.commit()
driver = Service('D:\GOGdriver\chromedriver.exe')
browser = webdriver.Chrome(service=driver)
browser.get('https://www.perekrestok.ru/cat/promo/10')
html_code = browser.page_source
soup = BeautifulSoup(html_code, 'lxml')
name = soup.find_all('div', class_="product-card__title")
sale = soup.find_all('div', class_="sc-kYrlTI ciSrDA product-card__badge")
price = soup.find_all('div', class_="price-new")
rate = soup.find_all('div', class_="rating-value")
picture = soup.find_all('div', class_="product-card__image-wrapper")
for i in range(15):
url = picture[i].find('img').attrs['src']
filename = f"C:\\Users\\71332\PycharmProjects\pythonProject1\\venv\Pictures\{i}.jpg"
wget.download(url, filename)
ins_qwery = f"""insert into public.Parser(name, sale, price, rate, picture) values ('{name[i].text}', '{sale[i].text}', '{price[i].text}', '{rate[i].text}', '{filename}')"""
cursor.execute(ins_qwery)
connection.commit()
cursor.close()
connection.close()
+16
View File
@@ -0,0 +1,16 @@
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
driver = Service('D:\GOGdriver\chromedriver.exe')
browser = webdriver.Chrome(service=driver)
browser.get('https://www.kinopoisk.ru/lists/movies/popular-films/')
html_code = browser.page_source
soup = BeautifulSoup(html_code, 'lxml')
name = soup.find_all('div', class_="desktop-list-main-info_secondaryTitleSlot__mc0mI")
print(soup)
print(name)
for i in name:
print(i.text)
+32 -2
View File
@@ -1,12 +1,42 @@
from bs4 import BeautifulSoup from bs4 import BeautifulSoup
from selenium import webdriver from selenium import webdriver
from selenium.webdriver.chrome.service import Service from selenium.webdriver.chrome.service import Service
import psycopg2
import wget
connection = psycopg2.connect(host='localhost', dbname='dbdata', user='postgres', password='Q1w2e3r4')
cursor = connection.cursor()
s = Service('D:\Games\data\chromedriver.exe') s = Service('D:\Games\data\chromedriver.exe')
browser = webdriver.Chrome(service=s) browser = webdriver.Chrome(service=s)
browser.get('https://www.volkswagen.ru/polo/') browser.get('https://www.volkswagen.ru/polo/')
html_text = browser.page_source html_text = browser.page_source
soup = BeautifulSoup(html_text, 'lxml') soup = BeautifulSoup(html_text, 'lxml')
creat_table = """ create table Cars_volks
(id serial primary key, car_name varchar(20),
price varchar(15), adress varchar(300),
scr varchar(150)) """
#cursor.execute(creat_table)
connection.commit()
car_names = soup.find_all('div', class_='avn001-2_name') car_names = soup.find_all('div', class_='avn001-2_name')
prices = soup.find_all('div', class_='avn001-2_price-container') prices = soup.find_all('div', class_='avn001-2_price-container')
for car_name, price in zip(car_names, prices): adresses = soup.find_all('div', class_='avn001-2_dealer-link__text')
print(f"Название машины:{car_name.text} | Цена: {price.text} рублей") pictures = soup.find_all('div', class_='avn001-2_image image__container')
#for car_name, adress, price in zip(car_names, adresses, prices):
# print(f"Название машины:{car_name.text} | Адрес диллера: {adress.text} | Цена: {price.text} рублей")
for i in range(len(car_names)):
url = pictures[i].find('img').attrs['src']
filename = f"c22712\\Задания\\task1\\nikitin\\img\\{i}.jpg"
# wget.download(url, filename) представим что картинки скачались:)
insert_qwery = f"""INSERT INTO public.Cars_volks(car_name, price, adress, scr)
VALUES ('{car_names[i].text}', '{prices[i].text}', '{adresses[i].text}', '{filename}')""";
cursor.execute(insert_qwery)
connection.commit()
cursor.close()
connection.close()
+1 -1
View File
@@ -1 +1 @@
def print_hi(' hello') абоба
@@ -0,0 +1,14 @@
from bs4 import BeautifulSoup
from selenium import webdriver
from selenium.webdriver.chrome.service import Service
s = Service('C:\Desktop\exe\chromedriver.exe')
browser = webdriver.Chrome(service=s)
browser.get('https://www.hellride.ru/catalog/zapchasti-dlya-tryukovyh-samokatov/deki')
html_text = browser.page_source
soup = BeautifulSoup(html_text, 'lxml')
prices = soup.find_all(attrs={"class": "product-card__price"})
titles = soup.find_all(attrs={"class": "product-card__title"})
for price, title in zip(prices, titles):
print(f" название {title.text} ; цена: {price.text}")