def generate_sitemap(blog_posts):
data = DataSource(populate=False)
# write sitemap to _site (to be used as index for static site search)
with open("_site/sitemap.json", "w") as output:
json.dump(site_to_json(data_source=data, blog_posts=blog_posts),
output)
print_progress(text='Generate sitemap index')
def build_website_api_batch(batch):
with DataSource(populate=False) as data:
for website in batch:
stats = data.sites.get_datapoint(website)
with open(f'_site/data/sites/global/{website}.json',
'w') as output:
json.dump(stats._asdict(), output)
def build_website_pages_batch(batch):
with DataSource(populate=False) as data:
template = get_template(data, "website-page.html", path_to_root='..')
for rank, site in batch:
website_page(template, data.sites.get_datapoint(site), rank + 1,
data)
def build_explorer():
data = DataSource(populate=False)
build_packed_data(data)
temp_folder = Path("temp")
if not temp_folder.exists():
temp_folder.mkdir()
table_to_csv(data.trackers, "temp/trackers.csv")
table_to_csv(data.sites, "temp/sites.csv")
table_to_csv(data.companies, "temp/companies.csv")
table_to_csv(data.sites_trackers, "temp/sites_trackers.csv")
month = data.trackers.last_month
shutil.make_archive(
f"_site/data/wtm-data-{month}", "zip", "temp"
)
shutil.rmtree(temp_folder.as_posix(), ignore_errors=True)
with open(f"_site/explorer.html", "w") as output:
output.write(render_template(
template=get_template(data, name="explorer.html"),
download_link=f"data/wtm-data-{month}.zip"
))
print_progress(text="Generated Exporable Dataset")
def build_tracker_api_batch(batch):
with DataSource(populate=False) as data:
gh_data_dir = Path('_site/data/trackers/ghostery')
if not gh_data_dir.exists():
gh_data_dir.mkdir(parents=True)
for tracker_id in batch:
build_tracker_json(tracker_id, data)
def build_tracker_page_batch(batch):
data = DataSource(populate=False)
template = get_template(data, name='tracker-page.html', path_to_root='..')
for tracker_id in batch:
page_data = tracker_page_data(tracker_id,
data.trackers.get_datapoint(tracker_id),
data)
tracker_page(template, page_data)
def build_blogpost_pages(blog_posts):
data = DataSource(populate=False)
for blog_post in blog_posts:
# TODO: Move template out after footnotes markdown extension does
# not save global state
template = get_template(data,
"blog-page.html",
render_markdown=True,
path_to_root="..")
with open(f'_site/blog/{blog_post.get("filename")}.html',
"w") as output:
output.write(
render_template(path_to_root="..",
template=template,
blog_post=blog_post))
print_progress(text="Generate blog posts")
def feed_event(self, event):
futures = []
with concurrent.futures.ProcessPoolExecutor(max_workers=8) as executor:
###################################################################
# This needs to be first, as other tasks will need to write in #
# the resulting folders. #
###################################################################
# Depends on folder: 'static/'
if event & STATIC_FOLDER:
create_site_structure(static_path=STATIC_PATH)
print_progress(text='Create _site')
###################################################################
# We then reload data in memory, before generating the site #
###################################################################
# Depends on folder: 'data/'
if self.data_source is None or event & DATA_FOLDER:
# class where all data can be accessed from
data_source = DataSource()
print_progress(text='Load data sources')
# Depends on: 'blog/'
if self.blog_posts is None or event & BLOG_FOLDER:
self.blog_posts = load_blog_posts()
print_progress(text='Load blog posts')
###################################################################
# Once site structure has been created and data is refreshed, we #
# can build all parts of the site in parallel, since there is no #
# dependencies between them. #
###################################################################
# Depends on: 'templates/', 'data/'
if event & DATA_FOLDER or event & TEMPLATES_FOLDER:
print_progress(text='Generate error pages')
copy_custom_error_pages(data=data_source)
def batched_job(inp, batch_fn, batch_size, message):
batches = []
input_size = len(inp)
for batch in [
inp[i:i + batch_size]
for i in range(0, input_size, batch_size)
]:
submission = executor.submit(batch_fn, batch=batch)
batches.append(submission)
futures.append(submission)
for i, f in enumerate(
concurrent.futures.as_completed(batches)):
print_progress(
text=
f"{message} {min((i+1) * batch_size, input_size)}/{input_size}"
)
return batches
# Explorer: depends on 'data/'
if event & DATA_FOLDER or event & STATIC_FOLDER:
futures.append(executor.submit(build_explorer, ))
# Depends on: 'data/', 'blog/', 'templates/'
if event & DATA_FOLDER or event & BLOG_FOLDER or event & TEMPLATES_FOLDER:
futures.append(
executor.submit(generate_sitemap,
blog_posts=self.blog_posts))
# Depends on: 'data/', 'templates/'
if event & DATA_FOLDER or event & TEMPLATES_FOLDER:
# Home
build_home(data=data_source)
build_privacy_policy(data=data_source)
# Trackers
trackers = [id for id, _ in data_source.trackers.iter()]
batched_job(trackers, build_tracker_page_batch, 150,
"Generate tracker pages")
build_trackers_list(data=data_source)
# Websites
websites = list(
enumerate([id for id, _ in data_source.sites.iter()]))
batched_job(websites, build_website_pages_batch, 400,
"Generate website pages")
build_website_list(data=data_source)
# Companies
build_company_reach_chart_page(data=data_source)
# Depends on: 'data/', 'blog/', 'templates/'
if event & DATA_FOLDER or event & BLOG_FOLDER or event & TEMPLATES_FOLDER:
futures.append(
executor.submit(build_blogpost_pages,
blog_posts=self.blog_posts))
futures.append(
executor.submit(build_rss_feeds,
blog_posts=self.blog_posts))
build_blogpost_list(data=data_source,
blog_posts=self.blog_posts)
if event & DATA_FOLDER:
build_tracker_db()
trackers = [id for id, _ in data_source.trackers.iter()]
data_dir = Path('_site/data/trackers/global')
if not data_dir.exists():
data_dir.mkdir(parents=True)
batched_job(trackers, build_tracker_api_batch, 150,
"Generate Tracker API pages")
site_data_dir = Path('_site/data/sites/global')
if not site_data_dir.exists():
site_data_dir.mkdir(parents=True)
sites = [id for id, _ in data_source.sites.iter()]
batched_job(sites, build_website_api_batch, 400,
"Generate Website API pages")
# TODO: uncomment when company profiles are ready
# if args['site'] or args['companies']:
# company_process = Process(target=build_company_pages, args=(data_source,))
# company_process.start()
# Wait for all jobs to finish
concurrent.futures.wait(futures)
# Getting the `result` of each promise (although none is expected)
# allows to re-raise exception happening in children processes. If
# we don't do it, exceptions will be silently ignored.
for future in futures:
future.result()
print('Done')
def feed_event(self, event):
futures = []
with concurrent.futures.ThreadPoolExecutor() as executor:
###################################################################
# This needs to be first, as other tasks will need to write in #
# the resulting folders. #
###################################################################
# Depends on folder: 'static/'
if event & STATIC_FOLDER:
create_site_structure(static_path=STATIC_PATH)
print_progress(text='Create _site')
###################################################################
# We then reload data in memory, before generating the site #
###################################################################
# Depends on folder: 'data/'
if self.data_source is None or event & DATA_FOLDER:
# class where all data can be accessed from
data_source = DataSource()
print_progress(text='Load data sources')
# Depends on: 'blog/'
if self.blog_posts is None or event & BLOG_FOLDER:
self.blog_posts = load_blog_posts()
print_progress(text='Load blog posts')
###################################################################
# Once site structure has been created and data is refreshed, we #
# can build all parts of the site in parallel, since there is no #
# dependencies between them. #
###################################################################
# Depends on: 'templates/', 'data/'
if event & DATA_FOLDER or event & TEMPLATES_FOLDER:
print_progress(text='Generate error pages')
copy_custom_error_pages(data=data_source)
# Depends on: 'data/', 'templates/'
if event & DATA_FOLDER or event & TEMPLATES_FOLDER:
# Home
futures.append(executor.submit(build_home, data=data_source))
# Trackers
futures.append(
executor.submit(build_trackers_list, data=data_source))
futures.append(
executor.submit(build_tracker_pages, data=data_source))
# Websites
futures.append(
executor.submit(build_website_list, data=data_source))
futures.append(
executor.submit(build_website_pages, data=data_source))
# Depends on: 'data/', 'blog/', 'templates/'
if event & DATA_FOLDER or event & BLOG_FOLDER or event & TEMPLATES_FOLDER:
futures.append(
executor.submit(build_blogpost_list,
data=data_source,
blog_posts=self.blog_posts))
futures.append(
executor.submit(build_blogpost_pages,
data=data_source,
blog_posts=self.blog_posts))
# Depends on: 'data/', 'blog/', 'templates/'
if event & DATA_FOLDER or event & BLOG_FOLDER or event & TEMPLATES_FOLDER:
futures.append(
executor.submit(generate_sitemap,
data=data_source,
blog_posts=self.blog_posts))
# TODO: uncomment when company profiles are ready
# if args['site'] or args['companies']:
# company_process = Process(target=build_company_pages, args=(data_source,))
# company_process.start()
# Wait for all jobs to finish
concurrent.futures.wait(futures)
# Getting the `result` of each promise (although none is expected)
# allows to re-raise exception happening in children processes. If
# we don't do it, exceptions will be silently ignored.
for future in futures:
future.result()
print('Done')
def build_api_batch(batch):
data = DataSource(populate=False)
for tracker_id in batch:
build_tracker_json(tracker_id, data)