def work():
''' Fetch and process a job by running the McXtrace simulation '''
# fetch job
run = one_or_none(fetch(SimRun, status="waiting").order_by('created'))
if run is None:
return
run.status = "running"
run.save()
print "Running job: ", run.ref
# create output folder; works as a lock
workdir = "%s/%%s" % (WORK_PATH % run.ref)
try:
os.mkdir(workdir % "")
except OSError:
# Someone else beat us to it, bail out
print "Skipping: already running."
return
# try to process job
try:
processJob(run, workdir)
except:
exc = traceback.format_exc()
file(workdir % 'err.txt', 'a').write('\n' + STACKTRACE % exc)
print exc
# mark job as completed
run.status = "done"
run.save()
print "Done."
def work():
''' Fetch and process a job by running the McXtrace simulation '''
# fetch job
run = one_or_none(fetch(SimRun, status="waiting").order_by('created'))
if run is None:
return
run.status = "running"
run.save()
print "Running job: ", run.ref
# create output folder; works as a lock
workdir = "%s/%%s" % (WORK_PATH % run.ref)
try:
os.mkdir(workdir % "")
except OSError:
# Someone else beat us to it, bail out
print "Skipping: already running."
return
# try to process job
try:
processJob(run, workdir)
except:
exc = traceback.format_exc()
file(workdir % 'err.txt', 'a').write('\n' + STACKTRACE % exc)
print exc
# mark job as completed
run.status = "done"
run.save()
print "Done."
def work():
""" Fetch and process a job by running the McStas simulation """
# fetch job
run = one_or_none(fetch(SimRun, status="waiting").order_by("created"))
if run is None:
return
run.status = "running"
run.save()
print "Running job: ", run.ref
# create output folder; works as a lock
workdir = "%s/%%s" % (WORK_PATH % run.ref)
try:
os.mkdir(workdir % "")
except OSError:
# Someone else beat us to it, bail out
print "Skipping: already running."
return
# try to process job
try:
processJob(run, workdir)
except:
exc = traceback.format_exc()
file(workdir % "err.txt", "a").write("\n" + STACKTRACE % exc)
print exc
# mark job as completed
run.status = "done"
with transaction.commit_on_success():
run.save()
print "Done."
def lambda_handler(event, context=None):
platforms = fetch('CURRENT_TARGETS', event['version']).strip().split('\n')
if event['buildInput']['platforms']:
platforms = [
p for p in platforms if p in event['buildInput']['platforms']
]
return platforms
def download_file(reqs_ses, url, filepath):
res = common.fetch(
requests_session=reqs_ses,
url=url,
method='get',
expect_status=200,
)
common.write_file( # Store page to disk
file_path=filepath, data=res.content)
logging.debug('Saved {0} to {1}'.format(url, filepath))
return
def load_archive(req_ses, blog_name):
base_url = 'http://{0}.tumblr.com/archive'.format(blog_name)
all_history = ''
# Load first block
first_response = common.fetch(
requests_session=req_ses,
url=base_url,
method='get',
)
common.write_file(file_path=os.path.join('debug', 'first_response.html'),
data=first_response.content)
all_history += first_response.content
# Find last post date
last_post_date = find_last_post_date(html=first_response.content)
# Load subsequent history
while True:
# Load next block
scroll_url = 'http://{0}.tumblr.com/archive?before_time={1}'.format(
blog_name, last_post_date)
scroll_response = common.fetch(
requests_session=req_ses,
url=base_url,
method='get',
)
common.write_file(file_path=os.path.join('debug',
'scroll_response.html'),
data=scroll_response.content)
all_history += scroll_response.content
# Find last post date
last_post_date = find_last_post_date(html=scroll_response.content)
# Stop if no more posts
if not last_post_date:
break
# Store combined page
common.write_file(file_path=os.path.join('debug', 'all_history.html'),
data=all_history)
return all_history
def save_followed_file(req_ses, followed_list_path):
logging.info('Saving list of followed blogs')
followed_list_path = os.path.join('dl', 'followed.opml')
# Get followed list file
# www.tumblr.com/following.opml
logging.debug('Loading followed list')
followed_res = common.fetch(
requests_session=req_ses,
url='https://www.tumblr.com/following.opml',
method='get',
)
common.write_file( # Save to file
file_path=followed_list_path,
data=followed_res.content)
logging.info('Saved list of followed blogs')
return
def find_blog_name_thorough(req_ses, blog_url): # TODO
logging.debug(
'Using slower, more thorough name-finding on {0!r}'.format(blog_url))
# Extract domain
# 'http://nsfw.kevinsano.com'
# 'nsfw.kevinsano.com'
domain_search = re.search(r'(?:https?://)?([^\\/]+\.\w+)/?', blog_url)
if domain_search:
domain = domain_search.group(1)
logging.debug('domain={0!r}'.format(domain))
else:
logging.error('Could not identify domain! Failing.')
return None
# Genreate archive page URL
blog_rss_url = 'http://{0}/rss'.format(domain)
logging.debug('blog_rss_url={0!r}'.format(blog_rss_url))
rss_path = os.path.join('debug',
'run_grab_site.find_blog_name_thorough.rss.rss')
# Load archive page
rss_res = common.fetch(
requests_session=req_ses,
url=blog_rss_url,
method='get',
)
common.write_file( # Save to file for debugging
file_path=rss_path,
data=rss_res.content)
# Extract blog name from page
# '<generator>Tumblr (3.0; @nsfwkevinsano)</generator>'
# 'nsfwkevinsano'
name_search = re.search('<generator>[^<]{0,25}@([^)<]+)\)</generator>',
rss_res.content)
if name_search:
blog_name = name_search.group(1)
logging.debug('blog_name={0!r}'.format(blog_name))
return blog_name
def main():
parser = argparse.ArgumentParser()
parser.add_argument('--debug', action='store_true')
parser.add_argument('--arch', default=common.arch)
parser.add_argument('--skia-dir')
parser.add_argument('--skia-release', default='m93-87e8842e8c')
parser.add_argument('--skija-version')
(args, _) = parser.parse_known_args()
# Fetch Skia
build_type = 'Debug' if args.debug else 'Release'
if args.skia_dir:
skia_dir = os.path.abspath(args.skia_dir)
os.chdir(common.root + '/platform')
else:
os.chdir(common.root + '/platform')
skia_dir = "Skia-" + args.skia_release + "-" + common.system + "-" + build_type + '-' + common.arch
if not os.path.exists(skia_dir):
zip = skia_dir + '.zip'
common.fetch('https://github.com/JetBrains/skia-build/releases/download/' + args.skia_release + '/' + zip, zip)
with zipfile.ZipFile(zip, 'r') as f:
print("Extracting", zip)
f.extractall(skia_dir)
os.remove(zip)
skia_dir = os.path.abspath(skia_dir)
print("Using Skia from", skia_dir)
# CMake
os.makedirs("build", exist_ok = True)
common.check_call([
"cmake",
"-G", "Ninja",
"-DCMAKE_BUILD_TYPE=" + build_type,
"-DSKIA_DIR=" + skia_dir,
"-DSKIA_ARCH=" + common.arch]
+ (["-DCMAKE_OSX_ARCHITECTURES=" + {"x64": "x86_64", "arm64": "arm64"}[common.arch]] if common.system == "macos" else [])
+ [".."],
cwd=os.path.abspath('build'))
# Ninja
common.check_call(["ninja"], cwd=os.path.abspath('build'))
# Codesign
if common.system == "macos" and os.getenv("APPLE_CODESIGN_IDENTITY"):
subprocess.call(["codesign",
# "--force",
# "-vvvvvv",
"--deep",
"--sign",
os.getenv("APPLE_CODESIGN_IDENTITY"),
"build/libskija_" + common.arch + ".dylib"])
# javac
modulepath = []
if args.skija_version:
modulepath += [
common.fetch_maven('org.jetbrains.skija', 'skija-shared', args.skija_version, repo='https://packages.jetbrains.team/maven/p/skija/maven')
]
else:
build_shared.main()
modulepath += ['../shared/target/classes']
os.chdir(common.root + '/platform')
sources = common.glob('java-' + common.classifier, '*.java')
common.javac(sources, 'target/classes', modulepath = modulepath, release = '9')
# Copy files
target = 'target/classes/org/jetbrains/skija'
if common.classifier == 'macos-x64':
common.copy_newer('build/libskija_x64.dylib', target + '/macos/x64/libskija_x64.dylib')
elif common.classifier == 'macos-arm64':
common.copy_newer('build/libskija_arm64.dylib', target + '/macos/arm64/libskija_arm64.dylib')
elif common.classifier == 'linux':
common.copy_newer('build/libskija.so', target + '/linux/libskija.so')
elif common.classifier == 'windows':
common.copy_newer('build/skija.dll', target + '/windows/skija.dll')
common.copy_newer(skia_dir + '/out/' + build_type + '-' + common.arch + '/icudtl.dat',
target + '/windows/icudtl.dat')
return 0
#!/usr/bin/env python
import time
from common import fetch, pluralize, logger, db
period_sleep = 1
# polling loop
while (True):
time.sleep(period_sleep)
# fetch last known ledger index
last_seq = int(db.ledgers.find_one(sort=[('seq', -1)])['seq'])
# get latest closed ledger
response = fetch({"method": "ledger_closed", "params": [{}]})
seq = response['ledger_index']
logger.info('FETCHED ledger {}'.format(seq))
# how many did we miss?
delta = seq - last_seq
if (delta == 0):
logger.info('local database is up to date')
continue
else:
logger.info('LAGGING by {} ledger{}'.format(delta, pluralize(delta)))
# parse missing ledgers
for idx in range(last_seq + 1, seq + 1):
# ledgerEntryType
def tumblr_login(req_ses, email, username, password):
logging.info('Logging in as {0!r}'.format(username))
# Load front page to look normal
logging.debug('Loading front page to prepare login attempt')
response_1 = common.fetch(
requests_session=req_ses,
url='https://www.tumblr.com/login',
method='get',
)
common.write_file(
file_path=os.path.join('debug', 'login.response_1.html'),
data=response_1.content
)
# Load login page
logging.debug('Loading login page')
response_2 = common.fetch(
requests_session=req_ses,
url='https://www.tumblr.com/login',
method='get',
)
common.write_file(
file_path=os.path.join('debug', 'login.response_2.html'),
data=response_2.content
)
# Get key from login page
#'<meta name="tumblr-form-key" id="tumblr_form_key" content="!1231544361914|zzONO1XougbCpvRupb561N630">'
token_search = re.search('<meta name="tumblr-form-key" id="tumblr_form_key" content="([a-zA-Z0-9!|]+)">', response_2.content, re.IGNORECASE)
token = token_search.group(1)
# Perform login
logging.debug('Sending login request')
response_3 = common.fetch(
requests_session=req_ses,
url='https://www.tumblr.com/login',
method='post',
data={
'determine_email': email,
'user[email]': email,
'user[password]': password,
'form_key': token,
},
expect_status=200,
)
common.write_file(
file_path=os.path.join('debug', 'login.response_3.html'),
data=response_3.content
)
# Validate login worked
logging.debug('Checking if login worked')
response_4 = common.fetch(
requests_session=req_ses,
url='https://www.tumblr.com/dashboard',
method='get',
)
common.write_file(
file_path=os.path.join('debug', 'login.response_4.html'),
data=response_4.content
)
# TODO
logging.warning('Login validation still TODO')
logging.info('Logged in as {0!r}'.format(username))
return
def fetch_test_data(name):
common.fetch(URL_PREFIX + name,
os.path.join(IDNA_TEST_DATA_DIR, os.path.basename(name)))
def save_post_media(Thread, Image, db_ses, req_ses, media_base_path, board_name, post):
"""Save media from a post, adding it to the DB.
Expects py8chan type Post object.
Commits changes."""
if (not config.media_enable_saving):
# If media saving disabled completely
return
# Save any new media (from new posts).
for image in post.all_files():# For each image, if any:
# Lookup image hash in DB.
image_find_query = db_ses.query(Image)\
.filter(Image.hash_md5 == image.file_md5_hex)
existing_image_row = image_find_query.first()
if existing_image_row:
# Check if image needs (re)downloading
do_redownload = decide_if_media_redownload(
media_base_path=media_base_path,
board_name=board_name,
image_row=existing_image_row
)
if (not do_redownload):
return None# Be f*****g explicit about what we're doing.
# If image not in DB, download image then add it to DB.
if (config.media_download_enable_full):# If disabled fullsized media files will not be downloaded.
# Download image:
# Generate path to save image to
image_filename = image.filename
if (image_filename == 'deleted'):
logging.info('File was deleted and cannot be downloaded: {0!r}')
assert2( (type(image_filename) is str), value=image_filename)# Should be text. (Sanity check remote value)
assert2( (8 <= len(image_filename) <= 128), value=image_filename)# is image hash so about 64 chars. (Sanity check remote value)
board_path = os.path.join(media_base_path, board_name)
image_filepath = generate_media_filepath(
media_base_path=media_base_path,
media_type='image',# TODO: Validate that 'image' is what we're using for this value
filename=image.filename
)
file_extension = image.file_extension
assert2( (type(file_extension) is str), value=file_extension)# Should be text. (Sanity check remote value)
assert2( (0 <= len(file_extension) <= 16), value=file_extension)# Short text. (Sanity check remote value)
image_url = image.file_url
assert2( (type(image_url) is str), image_url)# Should be text. (Sanity check remote value)
assert2( (16 <= len(image_url) <= 256), image_url)# Short text. (Sanity check remote value)
# Load image from server
try:
logging.debug('save_post_media() image_url={0!r}; image_filepath={1!r};'.format(image_url, image_filepath))
image_resp = common.fetch(
requests_session=req_ses,
url = image_url,
)
except common.FetchGot404 as err:
## logging.exception(err)
logging.warning('Could not fetch primary remote file, skipping this image.')
continue# Skip handling of this image download.
# Save image file
common.write_file(
file_path=image_filepath,
data=image_resp.content
)
# Calculate hashes. (We use more than one because hash collisions are a thing.)
with open(image_filepath, 'rb') as image_f:
# https://www.pythoncentral.io/hashing-files-with-python/
# Filesize in bytes
size_bytes = os.path.getsize(image_filepath)# https://stackoverflow.com/questions/2104080/how-to-check-file-size-in-python
hash_md5 = common.hash_file_md5(filepath=image_filepath)
hash_sha1 = common.hash_file_sha1(filepath=image_filepath)
hash_sha256 = common.hash_file_sha256(filepath=image_filepath)
hash_sha512 = common.hash_file_sha512(filepath=image_filepath)
print('BREAKPOINT before md5 check')
# Sanitycheck recieved file's MD5 against what the server told us. (Do we really need to do this? Does having this check make us less reliable?)
md5_algorithm_works = (hash_md5 == image.file_md5)
if (not md5_algorithm_works):
logging.warning('MD5 IMPLIMENTATION IN USE IS NOT CONSISTENT WITH EXPECTED DATA! DO NOT USE IN PRODUCTION!')# TODO: Crash on this happening once everything else is working.
logging.debug('hash_md5={0!r}'.format(hash_md5))
logging.debug('image.file_md5={0!r}'.format(image.file_md5))
logging.debug('image.file_md5_hex={0!r}'.format(image.file_md5_hex))
print('BREAKPOINT in md5 check failed')# F**K WHY IS ENCODING SUCH A PAIN IN THE ASS?!
else:
logging.warning('MD5 hashes matched this time')
print('BREAKPOINT md5 check passed')
print('BREAKPOINT after md5 check')
else:
# If image download disabled:
# NULL values for unfetched data
filename_full = None
size_bytes = None
hash_md5 = None
hash_sha1 = None
hash_sha256 = None
hash_sha512 = None
if (config.media_download_enable_thumb):# If disabled thumbnails will not be downloaded.
# Download thumbnail:
thumbnail_url = image.thumbnail_url
# Genreate thumbnail path
filename_thumbnail = os.path.basename(thumbnail_url)# TODO: Less shitty handling of this. (Feels wrong to use filsystem code for a URL)
thumbnail_filepath = generate_media_filepath(
media_base_path=media_base_path,
media_type='thumb',
filename=filename_thumbnail
)
try:
# Load thumbnail from server
thumbnail_resp = common.fetch(
requests_session=req_ses,
url = thumbnail_url,
)
except common.FetchGot404 as err:
## logging.exception(err)
logging.warning('Could not fetch thumbnail remote file, skipping this image.')# TODO: Decide if missing thumbnail should or should not be enough to skip image add.
continue# Skip handling of this image download.
# Save thumbnail file
common.write_file(
file_path=thumbnail_filepath,
data=image_resp.content
)
else:
# If no thumbnail download:
# NULL values for unfetched data
filename_thumbnail = None
# Create DB record for image
new_image_row = Image(
# Identification of image characteristics
size_bytes = size_bytes, # Size of the fullsized image in bytes.
hash_md5 = hash_md5, # MD5 hash of full file.
hash_sha1 = hash_sha1, # SHA1 hash of full file.
hash_sha256 = hash_sha256, # SHA256 hash of full file.
hash_sha512 = hash_sha512, # SHA512 hash of full file.
# Files on disk
file_extension = file_extension, # File extention of the fullview file.
filename_full = image_filename, # Fullsized media file's filename.
filename_thumbnail = filename_thumbnail, # Thumbnail's filename. Does not care if OP or reply.
)
# Stage new image entry to DB.
db_ses.add(new_image_row)
# Commit new image entry.
db_ses.commit()
logging.info('Added image to DB: {0!r}'.format(image_filepath))
time.sleep(config.media_download_delay)# Ratelimiting
continue# Done saving this image.
return None# Can we return a list of DB IDs for the media?
