def ingest_expired_text(text_id):
from texts.models import Text, TextMeta, HtmlVisualization
from annis.models import AnnisServer
# Set up an instance of the logger
logger = logging.getLogger(__name__)
# Define HTML Formats and the ANNIS server to query
annis_server = AnnisServer.objects.all()[:1]
vdisplay = Xvfb()
vdisplay.start()
driver = webdriver.Firefox()
if len(annis_server) > 0:
annis_server = annis_server[0]
# Ensure trailing slash in the annis server base domain
if not annis_server.base_domain.endswith("/"):
annis_server.base_domain += "/"
else:
logger.error("Error with single text re-ingest, no ANNIS server found")
return False
# Get the text
try:
text = Text.objects.get(id=text_id)
except Text.DoesNotExist:
logger.error("Error with single text re-ingest, no text found for text id", text_id)
return False
# Log the import information
logger.info(" -- Importing " + text.corpus.title + " " + text.title + " " + str(text.id))
# Query ANNIS for the metadata for the document
meta_query_url = annis_server.base_domain + annis_server.document_metadata_url.replace(":corpus_name", text.corpus.annis_corpus_name).replace(":document_name", text.title)
res = request.urlopen(meta_query_url)
xml = res.read()
soup = BeautifulSoup(xml)
meta_items = soup.find_all("annotation")
# Remove old textmeta items
# And save each meta item for the text document
text.text_meta.all().delete()
for meta_item in meta_items:
text_meta = TextMeta()
text_meta.name = meta_item.find("name").text
text_meta.value = meta_item.find("value").text
text_meta.pre = meta_item.find("pre").text
meta_corpus_name = meta_item.find("corpusName")
if meta_corpus_name:
text_meta.corpus_name = meta_corpus_name.text
text_meta.save()
text.text_meta.add(text_meta)
# Remove the html visualizations from the text
# Query ANNIS for each HTML format of the documents
text.html_visualizations.all().delete()
for html_format in text.corpus.html_visualization_formats.all():
# Add the corpus corpus name to the URL
corpora_url = annis_server.base_domain + annis_server.html_visualization_url.replace(":corpus_name", text.corpus.annis_corpus_name).replace(":document_name", text.title).replace(":html_visualization_format", html_format.slug)
# Fetch the HTML for the corpus/document/html_format from ANNIS
driver.get(corpora_url)
# Wait for visualization to load in browser
try:
element = WebDriverWait(driver, 20).until(
EC.presence_of_element_located((By.CLASS_NAME, "htmlvis"))
)
driver.delete_all_cookies()
body = driver.find_element_by_xpath("/html/body")
text_html = body.get_attribute("innerHTML")
styles = driver.find_elements_by_xpath("/html/head/style")
except:
text_html = ""
styles = []
driver.quit()
driver = webdriver.Firefox()
# Check to ensure there's html returned
# if "Could not query document" in text_html or "error" in text_html:
if "Client response status: 403" in text_html:
logger.error(" -- Error fetching " + corpora_url)
text_html = ""
# Remove Javascript from the body content
if len(text_html):
# Add the styles
for style_elem in styles:
style_css = style_elem.get_attribute("innerHTML")
text_html = text_html + "<style>" + style_css + "</style>"
# For script element in the html, remove it
script_elems = re.findall(r'<script.*script>', text_html, re.DOTALL)
for script_elem in script_elems:
text_html = text_html.replace(script_elem, "")
# Create the new html_visualization
html_visualization = HtmlVisualization()
html_visualization.visualization_format = html_format
html_visualization.html = text_html
html_visualization.save()
# Add the html visualization to the text
text.html_visualizations.add(html_visualization)
# Save the updated text
text.is_expired = False
text.save()
driver.quit()
vdisplay.stop()
return text.id
def collect(corpus, text, annis_server):
corpus_name = corpus.annis_corpus_name
formats = corpus.html_visualization_formats.all()
logger.info('Fetching %d visualizations' % len(formats))
driver = None
for html_format in formats:
html_vis_url = annis_server.url_html_visualization(corpus_name, text.title, html_format.slug)
vis_tries_left = MAX_VIS_TRIES
text_html = False
while not text_html and vis_tries_left:
try:
if not driver:
logger.debug("Starting browser")
try:
driver = webdriver.Chrome(os.environ.get('CHROMEDRIVER', '/usr/lib/chromium-browser/chromedriver'))
except Exception as e:
logger.error('Unable to start browser: %s' % e)
return
logger.debug(driver)
logger.info(html_format.title)
retries_left = 5
connection_accepted = False
vis_fetch_start_time = time()
while not connection_accepted and retries_left:
try:
driver.get(html_vis_url)
connection_accepted = True
except ConnectionRefusedError as cre:
logger.warning(cre)
retries_left -= 1
sleep(15)
if retries_left == 0:
raise VisServerRefusingConn()
logger.info('Calling WebDriverWait')
WebDriverWait(driver, 20).until(EC.presence_of_element_located((By.CLASS_NAME, "htmlvis")))
text_html = driver.find_element_by_xpath("/html/body").get_attribute("innerHTML")
logger.info('WebDriverWait returned\t%s\t%s\t%s\t%d\t%d\t%f' % (
corpus_name, text.title, html_format.slug, len(text_html),
MAX_VIS_TRIES - vis_tries_left, time() - vis_fetch_start_time))
driver.delete_all_cookies()
except Exception as e:
vis_tries_left -= 1
logger.error('Error getting %s: %s' % (html_vis_url, e))
logger.error('Page source: ' + driver.page_source)
driver.quit()
driver = None
if not text_html:
logger.error('Unable to get %s in %d tries.' % (html_vis_url, MAX_VIS_TRIES))
else:
# Add the styles
for style_elem in driver.find_elements_by_xpath("/html/head/style"):
text_html += "<style>" + style_elem.get_attribute("innerHTML") + "</style>"
# Remove JavaScript elements
for script_elem in re.findall(r'<script.*script>', text_html, re.DOTALL):
text_html = text_html.replace(script_elem, "")
vis = HtmlVisualization()
vis.visualization_format = html_format
vis.html = text_html
vis.save()
text.html_visualizations.add(vis)
self_max_mem, child_max_mem = [resource.getrusage(who).ru_maxrss for who in (resource.RUSAGE_SELF, resource.RUSAGE_CHILDREN)]
logger.info('Max mem, self: {:,}, children: {:,}'.format(self_max_mem, child_max_mem))
if driver:
driver.quit()
def collect(corpus, text, annis_server, driver):
corpus_name = corpus.annis_corpus_name
formats = corpus.html_visualization_formats.all()
logger.info('Fetching %d visualizations' % len(formats))
for html_format in formats:
html_vis_url = annis_server.url_html_visualization(corpus_name, text.title, html_format.slug)
try:
logger.info(html_format.title)
retries_left = 5
connection_accepted = False
while not connection_accepted and retries_left:
try:
driver.get(html_vis_url)
connection_accepted = True
except ConnectionRefusedError as cre:
logger.warning(cre)
driver.close()
retries_left -= 1
sleep(15)
if retries_left == 0:
raise VisServerRefusingConn()
WebDriverWait(driver, 60 * 2).until(EC.presence_of_element_located((By.CLASS_NAME, "htmlvis")))
driver.delete_all_cookies()
body = driver.find_element_by_xpath("/html/body")
text_html = body.get_attribute("innerHTML")
styles = driver.find_elements_by_xpath("/html/head/style")
except Exception as e:
logger.error('Error getting %s: %s' % (html_vis_url, e))
logger.error('Page source: ' + driver.page_source)
text_html = ""
styles = []
if "Client response status: 403" in text_html:
logger.error(" -- Error fetching " + html_vis_url)
text_html = ""
if text_html:
# Add the styles
for style_elem in styles:
style_css = style_elem.get_attribute("innerHTML")
text_html = text_html + "<style>" + style_css + "</style>"
# For script element in the html, remove it
script_elems = re.findall(r'<script.*script>', text_html, re.DOTALL)
for script_elem in script_elems:
text_html = text_html.replace( script_elem, "" )
vis = HtmlVisualization()
vis.visualization_format = html_format
vis.html = text_html
vis.save()
text.html_visualizations.add(vis)
self_max_mem, child_max_mem = [resource.getrusage(who).ru_maxrss for who in (resource.RUSAGE_SELF, resource.RUSAGE_CHILDREN)]
logger.info('Max mem, self: {:,}, children: {:,}'.format(self_max_mem, child_max_mem))
