def updateDataBase(token):
repository = Repository(token)
print("Get Java Repositories")
repositoriesJava = repository.get_repositories("java", 100)
print("Get Python Repositories")
repositoriesPython = repository.get_repositories("python", 100)
repositories = repositoriesJava + repositoriesPython
RepositoryData().update(repositories)
def insert_repo_and_user(gituser, repolist):
Session = sessionmaker(bind=engine)
session = Session()
raw_repos = session.query(Repository).filter(
Repository.username == gituser).all()
aquired_repos = []
print('Repositorios ja armazenados deste usuario:')
for repo in raw_repos:
print(f'{repo.username}:{repo.reponame}')
aquired_repos.append(repo.reponame)
print(f'\nRepositorios do usuario adiquiridos no git: {repolist}\n')
delete_list = select_delete_candidates(aquired_repos, repolist)
insert_list = select_insert_candidates(aquired_repos, repolist)
#print(delete_list)
#print(insert_list)
if delete_list is not None:
for delete_this in delete_list:
print(
f'Usuario nao possui mais {delete_this} em sua conta, deletando...\n'
)
session.query(Repository).filter(
Repository.reponame == delete_this).delete()
if insert_list is not None:
for insert_this in insert_list:
print(
f'Novo repositorio pelo nome de {insert_this} encontrado, adicionando...\n'
)
newadd = Repository(gituser, insert_this)
session.add(newadd)
session.commit()
def show_all_forums():
print("Forums at DB:" + database_config.HOST)
print("-------------------------------------")
print("forum_id | forum_link")
repository = Repository.Repository()
forums = repository.get_all_forums()
for f in forums:
print(str(f.forum_id) + " | " + f.link)
def get_texts_and_prepare_data_frame(date_from, date_to, forum_id):
"""
Perform SQL Query to database and turns the result into DataFrame
"""
repository = Repository.Repository()
data = repository.get_posts(date_from, date_to, forum_id)
data_frame = pd.DataFrame(
data, columns=['post', 'post_date', 'topic_title', 'category'])
return data_frame
def getRepo(ctx, repository):
if repository not in ctx.obj['repositories']:
print('[ERROR] The repository does not exist, %s' % repository,
file=sys.stderr)
sys.exit(1)
try:
mirror_path = os.path.realpath(
os.path.join(ctx.obj['current_path'],
ctx.obj['repositories'][repository]['mirror-path']))
except KeyError as err:
print >> sys.stderr, '[ERROR] Incorrect repositories configuration file, repository: %s [%s]' % (
repository, err)
sys.exit(1)
repo = Repository(repository, mirror_path,
**ctx.obj['repositories'][repository])
return repo
def __init__(self, repo_client=Repository(adapter=UserRepository)):
self.repo_client = repo_client
languages = ", ".join(languages)
await repository.get_or_create(
external_id=repo["id"],
name=repo["full_name"],
languages=languages,
owner_id=owner.id,
)
# Waiting a bit to avoid too much requests
print("Waiting 1 second before next request")
await asyncio.sleep(1)
await scrap_github(repository, user)
async def fetch_data(url, session):
async with session.get(url) as response:
if response.status == 403 or response.status == 401:
raise ValueError("API limit exceeded")
return await response.json()
if __name__ == "__main__":
engine = create_engine(config.DB_URL)
session = sessionmaker(bind=engine)()
repository = Repository(session)
user = User(session)
asyncio.run(scrap_github(repository, user))
def __init__(self, repo_client=Repository(adapter=ReportRepository)):
self.repo_client = repo_client
self.platforms = ['osx', 'linux', 'win', 'win7']
def export_posts(forum_id, date_from, date_to, filename):
repository = Repository.Repository()
data = repository.get_posts(date_from, date_to, forum_id)
df = pd.DataFrame(data,
columns=['post', 'post_date', 'topic_title', 'category'])
df.to_csv(filename, sep=';', escapechar='\\', encoding='utf8')
def __init__(self):
self.repository = r.Repository()
self.forum = None
def __init__(self):
self.repository = r.Repository()
self.strategy_initialized = False
self.forum = None
def finish_strategy(self):
repository = r.Repository()
all_categories = repository.get_all_categories(self.forum)
categories_data_frame = pd.DataFrame.from_records(
[category.to_dict() for category in all_categories])
categories_data_frame.to_csv("config/categories.csv", sep=";")
def __init__(self, repo_client=Repository(adapter=SessionRepository)):
self.repo_client = repo_client
def repository(dbsession):
return Repository(dbsession)
class CategoriesSpider(scrapy.Spider):
name = 'categories'
repository = Repository.Repository()
def __init__(self, start_url, scrap_mode, **kwargs):
super().__init__(**kwargs)
self.start_urls = [start_url]
self.base_domain = start_url
self.scrap_mode = scrap_mode
self.logger_dbg = logging_util.get_logger("logs/debug")
self.rule_provider = rp.RuleProvider()
self.rule_provider.prepare_model()
self.scraping_strategy = scraping_strategy_builder.get_strategy(
scrap_mode)
self.forum = self.scraping_strategy.init_strategy(self.base_domain)
def parse(self, response):
soup = BeautifulSoup(response.body, features="lxml")
parent = response.meta.get('parent')
for tag in self.rule_provider.possible_tags:
elements_with_tag = soup.body.findAll(tag)
for html_element in elements_with_tag:
if html_util.element_has_css_class(html_element):
predicted = self.rule_provider.predict(
tag, html_element["class"])
yield from self.scraping_strategy.execute_strategy(
html_element, parent, predicted, tag,
self.rule_provider, self)
def parse_categories(self, html_element, predicted, parent):
"""
Executes action of parsing the categories
"""
category = None
""" Title found"""
if predicted == self.rule_provider.get_mapping(m.category_title):
link = html_element['href']
title = str(html_element.contents[0])
category = self.repository.save_category(title, link, parent,
self.forum)
self.logger_dbg.info(title + " " + self.base_domain + link)
""" Unwrapping needed """
if predicted == self.rule_provider.get_mapping(m.category_whole):
try:
first_a_html_element_inside_whole = html_element.findAll(
"a")[0]
link = first_a_html_element_inside_whole['href']
title = str(first_a_html_element_inside_whole.contents[0])
category = self.repository.save_category(
title, link, parent, self.forum)
self.logger_dbg.info(title + " " + self.base_domain + link)
except BaseException as e:
self.logger_dbg.error(str(e))
self.logger_dbg.error("Can't find category inside: " +
str(html_element))
if category is not None and html_util.url_not_from_other_domain(
category.link, self.base_domain):
yield scrapy.Request(url=build_link(self.base_domain,
category.link),
callback=self.parse,
meta={'parent': category})
def parse_topics(self, html_element, parent):
author = None
date = None
link = None
title = None
for tag in self.rule_provider.possible_tags_topics:
elements_inside_tag = html_element.findAll(tag)
for elem in elements_inside_tag:
if html_util.element_has_css_class(elem):
predicted = self.rule_provider.predict(tag, elem["class"])
if predicted == self.rule_provider.get_mapping(
m.topic_title):
title = elem.contents[0]
link = elem['href']
self.logger_dbg.info(title + " " + link)
if predicted == self.rule_provider.get_mapping(m.author):
author = elem.contents[0]
if predicted == self.rule_provider.get_mapping(
m.topic_date):
date = dpt.parse_date(elem.contents)
""" Additional check english speaking invision """
time_tags = html_element.findAll("time")
if len(time_tags) > 0:
date = dpt.parse_english_date(time_tags[0].contents)
link = html_element.findAll('a')[0]['href']
title = html_element.findAll('a')[0]['title']
if title is None or link is None:
self.logger_dbg.info("Can't find topic inside: " +
str(html_element))
return
if not filtering.topic_meets_criterion(title, author, date):
return
topic = self.repository.save_topic(author, date, link, parent, title)
self.logger_dbg.info("Scrapped topic: " + str(topic))
yield scrapy.Request(dont_filter=True,
url=build_link(self.base_domain, topic.link),
callback=self.parse,
meta={'parent': topic})
def parse_posts(self, html_element, parent):
"""
Executes action of parsing the post
"""
self.logger_dbg.info("Parsing post of topic: " + parent.title)
author = None
date = None
content = None
"""Go through all the possible tags that occur in posts and determine their function """
for tag in self.rule_provider.possible_tags_posts:
elements_with_tag = html_element.findAll(tag)
for elem in elements_with_tag:
if html_util.element_has_css_class(elem):
predicted = self.rule_provider.predict(tag, elem["class"])
if predicted == self.rule_provider.get_mapping(
m.post_body):
content = filtering.assign_new_value_if_changed_and_not_null(
content, ppt.contents_to_plain_text(elem.contents))
if predicted == self.rule_provider.get_mapping(m.author):
author = elem.contents[0]
if predicted == self.rule_provider.get_mapping(
m.post_date):
date = dpt.parse_date(elem.contents)
"""Perform additional check for english format of date """
time_tags = html_element.findAll("time")
if len(time_tags) > 0 and date is None:
date = dpt.parse_english_date(time_tags[0].contents)
if content is not None and filtering.post_meets_criterions(
content, author, date):
self.repository.save_post(author, content, date, parent)
def go_to_next_page(self, html_element, parent, predicted):
"""
Executes action of going to the next page
"""
if predicted == self.rule_provider.get_mapping(m.next_page):
try:
first_a_html_element_inside_whole = html_element.findAll(
"a")[0]
link = first_a_html_element_inside_whole['href']
self.logger_dbg.info("Going to next page: " + str(parent) +
" unwrapped url: " + link)
yield scrapy.Request(url=build_link(self.base_domain, link),
callback=self.parse,
meta={'parent': parent})
except BaseException as e:
self.logger_dbg.error("Couldn't go to next page of: " +
str(parent) + " due to: " + str(e))
self.logger_dbg.error("Element that caused the problem: " +
str(html_element))
elif predicted == self.rule_provider.get_mapping(m.next_page_link):
self.logger_dbg.info("Going to next page: " + str(parent) +
" url: " + html_element['href'])
yield scrapy.Request(url=build_link(self.base_domain,
html_element['href']),
callback=self.parse,
meta={'parent': parent})
def closed(self, reason):
self.scraping_strategy.finish_strategy()
