def format(self, article, subscriber, codes=None):
"""
Constructs a dictionary that represents the parameters passed to the SMS InsertAlerts stored procedure
:return: returns the sequence number of the subscriber and the constructed parameter dictionary
"""
try:
pub_seq_num = superdesk.get_resource_service('subscribers').generate_sequence_number(subscriber)
sms_message = article.get('sms_message', article.get('abstract', ''))
# category = 1 is used to indicate a test message
category = '1' if superdesk.app.config.get('TEST_SMS_OUTPUT', True) is True \
else article.get('anpa_category', [{}])[0].get('qcode').upper()
odbc_item = {'Sequence': pub_seq_num, 'Category': category,
'Headline': to_ascii(get_text(sms_message, content='html')).replace('\'', '\'\''),
'Priority': map_priority(article.get('priority'))}
body = self.append_body_footer(article)
if article[ITEM_TYPE] == CONTENT_TYPE.TEXT:
body = get_text(body, content='html')
odbc_item['StoryText'] = to_ascii(body).replace('\'', '\'\'') # @article_text
odbc_item['ident'] = '0'
return [(pub_seq_num, json.dumps(odbc_item))]
except Exception as ex:
raise FormatterError.AAPSMSFormatterError(ex, subscriber)
def format(self, article, subscriber, codes=None):
try:
formatted_doc = {}
formatted_doc['headline'] = get_text(article.get('headline', ''),
content='html')
formatted_doc['headline'] = formatted_doc['headline'].replace(
'\'', '\'\'').replace('\xA0', ' ')
formatted_doc['keyword'] = article.get('slugline',
'').replace('\'', '\'\'')
# body formatting
if article.get(FORMAT) == FORMATS.PRESERVED:
body = get_text(self.append_body_footer(article),
content='html')
formatted_doc['article_text'] = body.replace('\'', '\'\'')
elif article.get(FORMAT, FORMATS.HTML) == FORMATS.HTML:
body = self.get_wrapped_text_content(
to_ascii(self.append_body_footer(article))).replace(
'\'', '\'\'')
formatted_doc['article_text'] = body
self.refine_article_body(formatted_doc, article)
# Frame the text output according to AAP requirement
formatted_output = 'KEYWORD: ' + formatted_doc.get('keyword',
'') + '\r\n'
formatted_output += 'HEADLINE: ' + formatted_doc.get(
'headline', '') + '\r\n'
formatted_output += ' ' + formatted_doc.get('article_text', '')
return [(0, json.dumps({'article_text': formatted_output}))]
except Exception as ex:
raise FormatterError.AAPTextFormatterError(ex, subscriber)
def _transform_to_ninjs(self, article, subscriber, recursive=True):
return {
"uuid":
article["guid"],
"createdTimestamp":
format_datetime(article["firstcreated"]),
"latestVersionTimestamp":
format_datetime(article["versioncreated"]),
"publicationTimestamp":
format_datetime(article["firstpublished"]),
"authors":
[author["sub_label"] for author in article.get("authors") or []],
"language":
article["language"],
"pubStatus":
True,
"concepts":
self._format_concepts(article),
"headline":
get_text(article["headline"]),
"preamble":
get_text(article["abstract"], lf_on_block=True).strip()
if article.get("abstract") else "",
"dateline":
article["dateline"]["text"] if article.get("dateline")
and article["dateline"].get("text") else "",
"body": [
line.strip() for line in get_text(article["body_html"],
lf_on_block=True).split("\n")
if line
],
}
def _format_body_content(self, article, body_content):
nitf_body = []
if article.get('ednote'):
nitf_body.append(to_ascii(self._format_line(article.get('ednote'))))
if article.get(BYLINE):
nitf_body.append(to_ascii(self._format_line(get_text(article.get(BYLINE)))))
if article.get(FORMAT) == FORMATS.PRESERVED:
nitf_body.append(to_ascii(get_text(self.append_body_footer(article), content='html')))
else:
body = article.get('body_html', '')
# we need to inject the dateline
if article.get('dateline', {}).get('text') and not article.get('auto_publish', False):
body_html_elem = parse_html(article.get('body_html'))
ptag = body_html_elem.find('.//p')
if ptag is not None:
ptag.text = article['dateline']['text'] + ' ' + (ptag.text or '')
body = to_string(body_html_elem)
nitf_body.append(self.get_text_content(body))
if article.get('body_footer'):
nitf_body.append(self.get_text_content(article.get('body_footer', '')))
sign_off = '{} {}'.format(article.get('source') or '', (article.get('sign_off') or '')).strip()
if sign_off:
nitf_body.append(to_ascii(self._format_line(sign_off)))
SubElement(body_content, 'pre').text = ''.join(nitf_body)
def get_odbc_item(self,
article,
subscriber,
category,
codes,
pass_through=False):
"""
Construct an odbc_item with the common key value pairs populated, if pass_through is true then the headline
original headline is maintained.
:param article:
:param subscriber:
:param category:
:param codes:
:param pass_through:
:return:
"""
article['headline'] = get_text(article.get('headline', ''),
content='html')
pub_seq_num = superdesk.get_resource_service(
'subscribers').generate_sequence_number(subscriber)
odbc_item = dict(
originator=article.get('source', None),
sequence=pub_seq_num,
category=category.get('qcode').lower(),
author=get_text(article.get('byline', '') or '',
content='html').replace('\'', '\'\''),
keyword=SluglineMapper().map(
article=article,
category=category.get('qcode').upper(),
truncate=True).replace('\'', '\'\'') if not pass_through else
(article.get('slugline', '') or '').replace('\'', '\'\''),
subject_reference=set_subject(category, article),
take_key=(article.get('anpa_take_key', '')
or '').replace('\'', '\'\''))
if 'genre' in article and len(article['genre']) >= 1:
odbc_item['genre'] = article['genre'][0].get('name', None)
else:
odbc_item['genre'] = 'Current' # @genre
odbc_item['news_item_type'] = 'News'
odbc_item['fullStory'] = 1
odbc_item['ident'] = '0' # @ident
odbc_item['selector_codes'] = ' '.join(codes) if codes else ' '
headline = to_ascii(LocatorMapper().get_formatted_headline(
article,
category.get('qcode').upper()))
odbc_item['headline'] = headline.replace('\'',
'\'\'').replace('\xA0', ' ')
self.expand_subject_codes(odbc_item)
self.set_usn(odbc_item, article)
return pub_seq_num, odbc_item
def append_body_footer(self, article):
"""
Checks if the article has any Public Service Announcements and if available appends each of them to the body.
:return: body with public service announcements.
"""
try:
article['body_html'] = article['body_html'].replace('<br>', '<br/>')
except KeyError:
pass
body = ''
if article[ITEM_TYPE] in [CONTENT_TYPE.TEXT, CONTENT_TYPE.PREFORMATTED]:
body = article.get('body_html', '')
elif article[ITEM_TYPE] in [CONTENT_TYPE.AUDIO, CONTENT_TYPE.PICTURE, CONTENT_TYPE.VIDEO]:
body = article.get('description', '')
if body and article.get(FORMAT, '') == FORMATS.PRESERVED:
body = body.replace('\n', '\r\n').replace('\r\r', '\r')
parsed = parse_html(body, content='html')
for br in parsed.xpath('//br'):
br.tail = '\r\n' + br.tail if br.tail else '\r\n'
etree.strip_elements(parsed, 'br', with_tail=False)
body = etree.tostring(parsed, encoding="unicode")
if body and article.get('body_footer'):
footer = article.get('body_footer')
if article.get(FORMAT, '') == FORMATS.PRESERVED:
body = '{}\r\n{}'.format(body, get_text(footer))
else:
body = '{}{}'.format(body, footer)
return body
def _format_body_content(self, article, body_content):
if article.get(FORMAT) == FORMATS.PRESERVED:
pre = get_text(self.append_body_footer(article))
SubElement(body_content, 'pre').text = pre
else:
self.map_html_to_xml(body_content,
self.append_body_footer(article))
def _set_headline(self, item, value):
if not value:
# if there is no headline, we use first 100 chars of body
# cf. SDNTB-481
value = text_utils.get_text(item.get('body_html', ''),
'html')[:100]
item['headline'] = value
def append_body_footer(self, article):
"""
Checks if the article has any Public Service Announcements and if available appends each of them to the body.
:return: body with public service announcements.
"""
try:
article["body_html"] = article["body_html"].replace("<br>", "<br/>")
except KeyError:
pass
body = ""
if article[ITEM_TYPE] in [CONTENT_TYPE.TEXT, CONTENT_TYPE.PREFORMATTED]:
body = article.get("body_html", "")
elif article[ITEM_TYPE] in [CONTENT_TYPE.AUDIO, CONTENT_TYPE.PICTURE, CONTENT_TYPE.VIDEO]:
body = article.get("description", "")
if body and article.get(FORMAT, "") == FORMATS.PRESERVED:
body = body.replace("\n", "\r\n").replace("\r\r", "\r")
parsed = parse_html(body, content="html")
for br in parsed.xpath("//br"):
br.tail = "\r\n" + br.tail if br.tail else "\r\n"
etree.strip_elements(parsed, "br", with_tail=False)
body = etree.tostring(parsed, encoding="unicode")
if body and article.get("body_footer"):
footer = article.get("body_footer")
if article.get(FORMAT, "") == FORMATS.PRESERVED:
body = "{}\r\n{}".format(body, get_text(footer))
else:
body = "{}{}".format(body, footer)
return body
def append_body_footer(self, article):
"""
Checks if the article has any Public Service Announcements and if available appends each of them to the body.
:return: body with public service announcements.
"""
try:
article['body_html'] = article['body_html'].replace('<br>', '<br/>')
except KeyError:
pass
body = ''
if article[ITEM_TYPE] in [CONTENT_TYPE.TEXT, CONTENT_TYPE.PREFORMATTED]:
body = article.get('body_html', '')
elif article[ITEM_TYPE] in [CONTENT_TYPE.AUDIO, CONTENT_TYPE.PICTURE, CONTENT_TYPE.VIDEO]:
body = article.get('description', '')
if body and article.get(FORMAT, '') == FORMATS.PRESERVED:
body = body.replace('\n', '\r\n').replace('\r\r', '\r')
parsed = parse_html(body, content='html')
for br in parsed.xpath('//br'):
br.tail = '\r\n' + br.tail if br.tail else '\r\n'
etree.strip_elements(parsed, 'br', with_tail=False)
body = etree.tostring(parsed, encoding="unicode")
if body and article.get('body_footer'):
footer = article.get('body_footer')
if article.get(FORMAT, '') == FORMATS.PRESERVED:
body = '{}\r\n{}'.format(body, get_text(footer))
else:
body = '{}{}'.format(body, footer)
return body
def _set_headline(self, item, value):
if not value:
# if there is no headline, we use first 100 chars of body
# cf. SDNTB-481
value = text_utils.get_text(item.get("body_html", ""),
"html")[:100]
item["headline"] = value
def _set_revision_history(self, article):
"""Get revision history of published article
:param dict article:
"""
query = {
'query': {
'filtered': {
'filter': {
'bool': {
'must': {
'term': {
'item_id': article.get('item_id')
}
}
}
}
}
},
'sort': [{
'versioncreated': {
'order': 'asc'
}
}]
}
req = ParsedRequest()
repos = 'published,archived'
req.args = {
'source': json.dumps(query),
'repo': repos,
'aggregations': 0
}
revisions = list(
get_resource_service('search').get(req=req, lookup=None))
revisions_tag = []
for rev in revisions:
local_date = utc_to_local(
config.DEFAULT_TIMEZONE,
rev.get('firstpublished') if rev.get(ITEM_STATE)
== CONTENT_STATE.PUBLISHED else rev.get('versioncreated'))
date_string = datetime.strftime(local_date,
'%b XXX, %Y %H:%M %Z').replace(
'XXX', str(local_date.day))
if rev.get(ITEM_STATE) == CONTENT_STATE.PUBLISHED:
revisions_tag.append('<li>{} {}</li>'.format(
'First published', date_string))
else:
revision_markup = '{} {}'.format('Revision published',
date_string)
ednote = get_text(rev.get('ednote') or '',
content='html').strip()
if rev.get(ITEM_STATE) == CONTENT_STATE.CORRECTED and ednote:
revision_markup += '<br><i>{}</i>'.format(ednote)
revisions_tag.append('<li>{}</li>'.format(revision_markup))
article['_revision_history'] = '<ul>{}</ul>'.format(
''.join(revisions_tag)) if revisions_tag else ''
def _fill_definition_short(self, document, item):
content = document.find('content')
if content is not None:
item['definition_short'] = text_utils.get_text(
content.text,
content='html',
lf_on_block=True,
space_on_elements=True).strip()
def get_item_body(item):
body = []
for field in ("body_html", "abstract"):
try:
body.extend([p.strip() for p in get_text(item[field], "html", True).split("\n") if p.strip()])
except KeyError:
pass
return body
def parse_item(self, tree):
item = super().parse_item(tree)
meta = tree.find(self.qname('contentMeta'))
organisation = meta.xpath('./iptc:subject[@type="cpnat:organisation"][@literal]', namespaces=NS)
if organisation:
item['abstract'] = format_maxlength('FOR: {}. {}'.format(
organisation[0].get('literal').upper().rstrip('.'),
get_text(item['body_html']).replace(' ', ' '),
), 200)
return item
def format(self, article, subscriber, codes=None):
"""
Constructs a dictionary that represents the parameters passed to the SMS InsertAlerts stored procedure
:return: returns the sequence number of the subscriber and the constructed parameter dictionary
"""
try:
pub_seq_num = superdesk.get_resource_service(
'subscribers').generate_sequence_number(subscriber)
sms_message = article.get('sms_message',
article.get('abstract', ''))
# category = 1 is used to indicate a test message
category = '1' if superdesk.app.config.get('TEST_SMS_OUTPUT', True) is True \
else article.get('anpa_category', [{}])[0].get('qcode').upper()
odbc_item = {
'Sequence':
pub_seq_num,
'Category':
category,
'Headline':
to_ascii(get_text(sms_message,
content='html')).replace('\'', '\'\''),
'Priority':
map_priority(article.get('priority'))
}
body = self.append_body_footer(article)
if article[ITEM_TYPE] == CONTENT_TYPE.TEXT:
body = get_text(body, content='html')
odbc_item['StoryText'] = to_ascii(body).replace(
'\'', '\'\'') # @article_text
odbc_item['ident'] = '0'
return [(pub_seq_num, json.dumps(odbc_item))]
except Exception as ex:
raise FormatterError.AAPSMSFormatterError(ex, subscriber)
def get_permalink(item):
code = item["_id"][-6:]
try:
title = item["extra"][PERMALINK] or ""
slug = slugify(get_text(title, 'html'))
except (KeyError, AttributeError):
slug = ""
return urljoin(
BASE_URL,
"/{lang}/{code}/".format(
lang=item.get("language", "en"), code="-".join(filter(bool, [slug, code])),
),
)
def add_byline(self, odbc_item, article):
"""
Add the byline to the article text
:param odbc_item:
:param article:
:return:
"""
if article.get('byline') and article.get('byline') != '':
byline = get_text(article.get('byline', ''), content='html')
if len(byline) >= 3 and byline[:2].upper() != 'BY':
byline = 'By ' + byline
byline = ' {}\r\n\r\n'.format(byline).replace('\'', '\'\'')
odbc_item['article_text'] = byline + odbc_item['article_text']
def add_byline(self, odbc_item, article):
"""
Add the byline to the article text
:param odbc_item:
:param article:
:return:
"""
if article.get('byline') and article.get('byline') != '':
byline = get_text(article.get('byline', ''), content='html')
if len(byline) >= 3 and byline[:2].upper() != 'BY':
byline = 'By ' + byline
byline = '\x19 {}\x19\r\n'.format(byline).replace('\'', '\'\'')
odbc_item['article_text'] = byline + odbc_item['article_text']
def get_odbc_item(self, article, subscriber, category, codes, pass_through=False):
"""
Construct an odbc_item with the common key value pairs populated, if pass_through is true then the headline
original headline is maintained.
:param article:
:param subscriber:
:param category:
:param codes:
:param pass_through:
:return:
"""
article['headline'] = get_text(article.get('headline', ''), content='html')
pub_seq_num = superdesk.get_resource_service('subscribers').generate_sequence_number(subscriber)
odbc_item = dict(originator=article.get('source', None), sequence=pub_seq_num,
category=category.get('qcode').lower(),
author=get_text(article.get('byline', '') or '', content='html').replace('\'', '\'\''),
keyword=SluglineMapper().map(article=article,
category=category.get('qcode').upper(),
truncate=True).replace('\'', '\'\'') if not pass_through else
(article.get('slugline', '') or '').replace('\'', '\'\''),
subject_reference=set_subject(category, article),
take_key=(article.get('anpa_take_key', '') or '').replace('\'', '\'\''))
if 'genre' in article and len(article['genre']) >= 1:
odbc_item['genre'] = article['genre'][0].get('name', None)
else:
odbc_item['genre'] = 'Current' # @genre
odbc_item['news_item_type'] = 'News'
odbc_item['fullStory'] = 1
odbc_item['ident'] = '0' # @ident
odbc_item['selector_codes'] = ' '.join(codes) if codes else ' '
headline = to_ascii(LocatorMapper().get_formatted_headline(article, category.get('qcode').upper()))
odbc_item['headline'] = headline.replace('\'', '\'\'').replace('\xA0', ' ')
self.expand_subject_codes(odbc_item)
self.set_usn(odbc_item, article)
return pub_seq_num, odbc_item
def populate(item, **kwargs):
"""Populate the abstract field with the first sentence of the body"""
# get the list of sentences of the body
if not item.get("body_html", None):
item["abstract"] = "No body found to use for abstract..."
else:
sentences = p.split(item["body_html"])
# chop the first sentence to size for abstract (64)
if sentences and len(sentences) > 0:
item["abstract"] = get_text(sentences[0][:64]).strip()
return item
def populate(item, **kwargs):
"""Populate the abstract field with the first sentence of the body"""
# get the list of sentences of the body
if not item.get('body_html', None):
item['abstract'] = 'No body found to use for abstract...'
else:
sentences = p.split(item['body_html'])
# chop the first sentence to size for abstract (64)
if sentences and len(sentences) > 0:
item['abstract'] = get_text(sentences[0][:64]).strip()
return item
def _process_headline(self, anpa, article, category):
# prepend the locator to the headline if required
article['headline'] = get_text(article.get('headline', ''))
headline = to_ascii(LocatorMapper().get_formatted_headline(article, category.decode('UTF-8').upper()))
# Set the maximum size to 64 including the sequence number if any
if len(headline) > 64:
if article.get('sequence'):
digits = len(str(article['sequence'])) + 1
shortened_headline = '{}={}'.format(headline[:-digits][:(64 - digits)], article['sequence'])
anpa.append(shortened_headline.encode('ascii', 'replace'))
else:
anpa.append(headline[:64].encode('ascii', 'replace'))
else:
anpa.append(headline.encode('ascii', 'replace'))
anpa.append(b'\x0D\x0A')
def _format_content(self, article, news_item, nitf):
"""Adds the content set to the xml
:param dict article:
:param Element newsItem:
:param Element nitf:
"""
content_set = SubElement(news_item, 'contentSet')
if article.get(FORMAT) == FORMATS.PRESERVED:
inline_data = text_utils.get_text(self.append_body_footer(article))
SubElement(content_set, 'inlineData',
attrib={'contenttype': 'text/plain'}).text = inline_data
elif article[ITEM_TYPE] in [CONTENT_TYPE.TEXT, CONTENT_TYPE.COMPOSITE]:
inline = SubElement(content_set, 'inlineXML',
attrib={'contenttype': 'application/nitf+xml'})
inline.append(nitf)
def _process_headline(self, anpa, article, category):
# prepend the locator to the headline if required
article['headline'] = get_text(article.get('headline', ''))
headline = to_ascii(LocatorMapper().get_formatted_headline(article, category.decode('UTF-8').upper()))
# Set the maximum size to 64 including the sequence number if any
if len(headline) > 64:
if article.get('sequence'):
digits = len(str(article['sequence'])) + 1
shortened_headline = '{}={}'.format(headline[:-digits][:(64 - digits)], article['sequence'])
anpa.append(shortened_headline.encode('ascii', 'replace'))
else:
anpa.append(headline[:64].encode('ascii', 'replace'))
else:
anpa.append(headline.encode('ascii', 'replace'))
anpa.append(b'\x0D\x0A')
def truncate_article_body(items, monitoring_profile, full_text=False):
# To make sure PDF creator and RTF creator does truncate for linked_text settings
# Manually truncate it
for i in items:
i['body_str'] = get_text(i.get('body_html', ''), content='html', lf_on_block=True)
if monitoring_profile['alert_type'] == 'linked_text':
if not full_text and len(i['body_str']) > 160:
i['body_str'] = i['body_str'][:159] + '...'
if monitoring_profile.get('format_type') == 'monitoring_pdf':
body_lines = i.get('body_str', '').split('\n')
altered_html = ''
for line in body_lines:
altered_html = '{}<div class="line">{}</div>'.format(altered_html, line)
i['body_str'] = altered_html
def _format_content(self, article, news_item, nitf):
"""Adds the content set to the xml
:param dict article:
:param Element newsItem:
:param Element nitf:
"""
content_set = SubElement(news_item, 'contentSet')
if article.get(FORMAT) == FORMATS.PRESERVED:
inline_data = text_utils.get_text(self.append_body_footer(article))
SubElement(content_set, 'inlineData',
attrib={'contenttype': 'text/plain'}).text = inline_data
elif article[ITEM_TYPE] in [CONTENT_TYPE.TEXT, CONTENT_TYPE.COMPOSITE]:
inline = SubElement(content_set, 'inlineXML',
attrib={'contenttype': 'application/nitf+xml'})
inline.append(nitf)
def format_for_source(self, article, subscriber, source, codes=None):
"""Constructs a dictionary that represents the parameters passed to the IPNews InsertNews stored procedure
:type article: object
:return: returns the sequence number of the subscriber and the constructed parameter dictionary
"""
pass_through = article.get('auto_publish', False)
try:
docs = []
for category in self._get_category_list(article.get('anpa_category')):
# All NZN sourced content is AAP content for the AAP output formatted
article['source'] = source
pub_seq_num, odbc_item = self.get_odbc_item(article, subscriber, category, codes, pass_through)
if article.get(FORMAT) == FORMATS.PRESERVED: # @article_text
body = get_text(self.append_body_footer(article))
odbc_item['article_text'] = body.replace('\'', '\'\'')
odbc_item['texttab'] = 't'
elif article.get(FORMAT, FORMATS.HTML) == FORMATS.HTML:
body = self.get_wrapped_text_content(
to_ascii(self.append_body_footer(article))).replace('\'', '\'\'')
# if we have a dateline inject it
if 'dateline' in article and 'text' in article.get('dateline', {}) and not pass_through:
if body.startswith(' '):
body = ' {} {}'.format(article.get('dateline')
.get('text').replace('\'', '\'\''),
body[3:])
odbc_item['article_text'] = body
odbc_item['texttab'] = 'x'
if not pass_through:
self.add_ednote(odbc_item, article)
self.add_byline(odbc_item, article)
odbc_item['article_text'] += '\r\n' + article.get('source', '')
sign_off = article.get('sign_off', '') or ''
if len(sign_off) > 0:
odbc_item['article_text'] += ' ' + sign_off
odbc_item['service_level'] = get_service_level(category, article) # @service_level
odbc_item['wordcount'] = article.get('word_count') or 0 # @wordcount
odbc_item['priority'] = map_priority(article.get('priority')) # @priority
docs.append((pub_seq_num, json.dumps(odbc_item)))
return docs
except Exception as ex:
raise FormatterError.AAPIpNewsFormatterError(ex, subscriber)
def parse(self, xml, provider=None):
self.root = xml
try:
item = self.parse_item(xml)
if not item.get('headline'):
item['headline'] = text_utils.get_text(item.get('body_html', ''), 'html')[:100]
try:
abstract = xml.xpath("//iptc:description[@role='drol:summary']", namespaces={'iptc': IPTC_NS})[0].text
except IndexError:
pass
else:
if abstract:
item['abstract'] = abstract
return [item]
except Exception as ex:
raise ParserError.newsmlTwoParserError(ex, provider)
def map_html_to_xml(self, element, html):
"""
Map the html text tags to xml
:param element: The xml element to populate
:param html: the html to parse the text from
:return:
"""
html = html.replace('<br>', '<br/>').replace('</br>', '')
html = re.sub('[\x00-\x09\x0b\x0c\x0e-\x1f]', '', html)
html = html.replace('\n', ' ')
html = re.sub(r'\s\s+', ' ', html)
parsed = parse_html(html, content='html')
for tag in parsed.xpath('/html/div/child::*'):
p = etree.Element('p')
p.text = to_ascii(
get_text(to_string(tag, method='html'), content='html'))
element.append(p)
def _set_revision_history(self, article):
"""Get revision history of published article
:param dict article:
"""
query = {
'query': {
'filtered': {
'filter': {
'bool': {
'must': {
'term': {'item_id': article.get('item_id')}
}
}
}
}
},
'sort': [
{'versioncreated': {'order': 'asc'}}
]
}
req = ParsedRequest()
repos = 'published,archived'
req.args = {'source': json.dumps(query), 'repo': repos, 'aggregations': 0}
revisions = list(get_resource_service('search').get(req=req, lookup=None))
revisions_tag = []
for rev in revisions:
local_date = utc_to_local(
config.DEFAULT_TIMEZONE,
rev.get('firstpublished') if rev.get(ITEM_STATE) == CONTENT_STATE.PUBLISHED
else rev.get('versioncreated')
)
date_string = datetime.strftime(local_date, '%b XXX, %Y %H:%M %Z').replace('XXX', str(local_date.day))
if rev.get(ITEM_STATE) == CONTENT_STATE.PUBLISHED:
revisions_tag.append('<li>{} {}</li>'.format('First published', date_string))
else:
revision_markup = '{} {}'.format('Revision published', date_string)
ednote = get_text(rev.get('ednote') or '', content='html').strip()
if rev.get(ITEM_STATE) == CONTENT_STATE.CORRECTED and ednote:
revision_markup += '<br><i>{}</i>'.format(ednote)
revisions_tag.append('<li>{}</li>'.format(revision_markup))
article['_revision_history'] = '<ul>{}</ul>' .format(''.join(revisions_tag)) if revisions_tag else ''
def callback(item, **kwargs):
diff = {}
if not item.get('body_html'):
return diff
rate = get_rate()
text = get_text(item['body_html'], 'html', True)
def repl(m, is_fr=False):
if m.group('currency') and m.group('currency') != 'US':
return
if is_fr:
num = m.group('num').replace(',', '.').replace(' ', '')
else:
num = m.group('num').replace(',', '')
converted = decimal.Decimal(num) * rate
if m.group('decimal'):
_format = '{:.3f}' # convert 55.21 to 73.73 - round to 3 decimals and strip last one
if is_fr and ' ' in m.group('num') or not is_fr and ',' in m.group(
'num'):
_format = '{:,.3f}'
fixed = _format.format(converted)[:-1]
else:
_format = '{:.1f}0' # convert 55 to 73.70 - round to 1 decimal and add 0
if is_fr and ' ' in m.group('num') or not is_fr and ',' in m.group(
'num'):
_format = '{:,.1f}0'
fixed = _format.format(converted).replace('.00', '')
# keep leeding whitespace so on client it won't
# replace $500 in C$500
diff[m.group(0).rstrip(
)] = '{whitespace} ({en_currency}{value}{mil}{fr_currency})'.format(
whitespace=m.group(0).rstrip(),
en_currency='' if is_fr else 'C$',
value=fixed if not is_fr else fixed.replace(',', ' ').replace(
'.', ','),
mil=m.group('mil') or '',
fr_currency=' $ CAN' if is_fr else '',
).rstrip()
re.sub(CURRENCY_REGEX, repl, text)
re.sub(CURRENCY_REGEX_FR, functools.partial(repl, is_fr=True), text)
return (item, diff)
def callback(item, **kwargs):
diff = {}
if not item.get("body_html"):
return diff
rate = get_rate()
text = get_text(item["body_html"], "html", True)
def repl(m, is_fr=False):
if m.group("currency") and m.group("currency") != "US":
return
if is_fr:
num = m.group("num").replace(",", ".").replace(" ", "")
else:
num = m.group("num").replace(",", "")
converted = decimal.Decimal(num) * rate
if m.group("decimal"):
_format = "{:.3f}" # convert 55.21 to 73.73 - round to 3 decimals and strip last one
if is_fr and " " in m.group("num") or not is_fr and "," in m.group(
"num"):
_format = "{:,.3f}"
fixed = _format.format(converted)[:-1]
else:
_format = "{:.1f}0" # convert 55 to 73.70 - round to 1 decimal and add 0
if is_fr and " " in m.group("num") or not is_fr and "," in m.group(
"num"):
_format = "{:,.1f}0"
fixed = _format.format(converted).replace(".00", "")
# keep leeding whitespace so on client it won't
# replace $500 in C$500
diff[m.group(0).rstrip(
)] = "{whitespace} ({en_currency}{value}{mil}{fr_currency})".format(
whitespace=m.group(0).rstrip(),
en_currency="" if is_fr else "C$",
value=fixed if not is_fr else fixed.replace(",", " ").replace(
".", ","),
mil=m.group("mil") or "",
fr_currency=" $ CAN" if is_fr else "",
).rstrip()
re.sub(CURRENCY_REGEX, repl, text)
re.sub(CURRENCY_REGEX_FR, functools.partial(repl, is_fr=True), text)
return (item, diff)
def format_for_source(self, article, subscriber, source, codes=None):
try:
pass_through = article.get('auto_publish', False)
docs = []
for category in self._get_category_list(
article.get('anpa_category')):
article['source'] = source
pub_seq_num, odbc_item = self.get_odbc_item(
article, subscriber, category, codes, pass_through)
if article.get(FORMAT) == FORMATS.PRESERVED: # @article_text
body = get_text(self.append_body_footer(article),
content='html')
odbc_item['article_text'] = body.replace('\'', '\'\'')
else:
body = self.get_text_content(
to_ascii(self.append_body_footer(article)))
if 'dateline' in article \
and 'text' in article.get('dateline', {}) and not pass_through:
if body.startswith(' '):
body = ' {} {}'.format(
article.get('dateline').get('text'), body[3:])
odbc_item['article_text'] = body.replace('\'', '\'\'')
if not pass_through:
self.add_ednote(odbc_item, article)
self.add_byline(odbc_item, article)
odbc_item['article_text'] += '\r\n' + source
sign_off = article.get('sign_off', '') or ''
if len(sign_off) > 0:
odbc_item['article_text'] += ' ' + sign_off
odbc_item['category'] = odbc_item.get('category', '').upper()
odbc_item['selector_codes'] = odbc_item.get(
'selector_codes', '').upper()
docs.append((pub_seq_num, json.dumps(odbc_item)))
return docs
except Exception as ex:
raise FormatterError.AAPNewscentreFormatterError(ex, subscriber)
def parse(self, xml, provider=None):
self.root = xml
try:
item = self.parse_item(xml)
if not item.get('headline'):
item['headline'] = text_utils.get_text(
item.get('body_html', ''), 'html')[:100]
try:
abstract = xml.xpath(
"//iptc:description[@role='drol:summary']",
namespaces={'iptc': IPTC_NS})[0].text
except IndexError:
pass
else:
if abstract:
item['abstract'] = abstract
return [item]
except Exception as ex:
raise ParserError.newsmlTwoParserError(ex, provider)
def _sanitize_fields(self, doc, validator):
"""If maxlength or minlength is specified in the validator then remove any markups from that field
:param doc: Article to be validated
:param validator: Validation rule
:return: updated article
"""
fields_to_check = ['minlength', 'maxlength']
item_schema = validator.get('schema', {})
extra_schema = item_schema.get('extra', {}).get('schema', {})
schemes_docs = [(item_schema, doc), (extra_schema, doc.get('extra', {}))]
for schema, content in schemes_docs:
for field in schema:
if content.get(field) and schema.get(field) and type(content[field]) is str and \
any(k in schema[field] for k in fields_to_check):
try:
content[field] = get_text(content[field])
except (ValueError, TypeError):
# fails for json fields like subject, genre
pass
def parse_newsitem(self, item, newsitem_el):
super().parse_newsitem(item, newsitem_el)
# mapping services-products from category, and have only one product
matching = False
for category in item.get('anpa_category', []):
qcode = self.MAPPING_CATEGORY.get(category.get('qcode'),
'NEWS/GENERAL')
item.setdefault('subject', []).append({
'name': qcode,
'qcode': qcode,
'parent': 'NEWS',
'scheme': 'services-products'
})
matching = True
if not matching:
item.setdefault('subject', []).append({
'name': 'NEWS/GENERAL',
'qcode': 'NEWS/GENERAL',
'parent': 'NEWS',
'scheme': 'services-products'
})
# add content for headline when it is empty
if item.get('urgency') in (1, 2) and not item.get('headline'):
for line in get_text(item.get('body_html', ''),
lf_on_block=True).split('\n'):
if line.strip():
item['headline'] = 'URGENT: ' + line.strip()
break
# Label must be empty
item['subject'] = [
i for i in item['subject'] if i.get('scheme') != 'label'
]
# Source is AFP
credit = {"name": 'AFP', "qcode": 'AFP', "scheme": "sources"}
item.setdefault('subject', []).append(credit)
if item.get('urgency') == 4:
item['urgency'] = 3
return item
def _sanitize_fields(self, doc, validator):
"""If maxlength or minlength is specified in the validator then remove any markups from that field
:param doc: Article to be validated
:param validator: Validation rule
:return: updated article
"""
fields_to_check = ['minlength', 'maxlength']
item_schema = validator.get('schema', {})
extra_schema = item_schema.get('extra', {}).get('schema', {})
schemes_docs = [(item_schema, doc),
(extra_schema, doc.get('extra', {}))]
for schema, content in schemes_docs:
for field in schema:
if content.get(field) and schema.get(field) and type(content[field]) is str and \
any(k in schema[field] for k in fields_to_check):
try:
content[field] = get_text(content[field])
except (ValueError, TypeError):
# fails for json fields like subject, genre
pass
def parse_item(self, tree):
item = super().parse_item(tree)
meta = tree.find(self.qname("contentMeta"))
organisation = meta.xpath(
'./iptc:subject[@type="cpnat:organisation"][@literal]',
namespaces=NS)
if organisation:
org_name = organisation[0].get("literal")
item["abstract"] = format_maxlength(
"FOR: {}. {}".format(
org_name.upper().rstrip("."),
get_text(item["body_html"]).replace(" ", " "),
),
200,
)
item.setdefault("subject", []).append({
"name": org_name,
"qcode": org_name,
"scheme": cp.ORGANISATION,
})
return item
def upload_document(item):
item_name = item.get("headline") or item.get("slugline")
if not item_name or not item.get("body_html"):
return
payload = {
"lang": {
"fromLang": "en",
"toLang": "fr",
},
"name": item_name,
"state": "new",
"text": {
"original": get_text(item["body_html"]),
},
}
resp = sess.post(
ULTRAD_URL, json=payload, headers=get_headers(), timeout=ULTRAD_TIMEOUT
)
raise_for_resp_error(resp)
data = get_json(resp)
return data["_id"]
def get_value(self, article):
try:
return get_text(article[self.field.name]).replace('\n', ' ')
except (etree.XMLSyntaxError, ValueError):
return article[self.field.name]
def format(self, article, subscriber, codes=None):
try:
docs = []
formatted_article = deepcopy(article)
for category in self._get_category_list(formatted_article.get('anpa_category')):
mapped_source = self._get_mapped_source(formatted_article)
formatted_article[config.ID_FIELD] = formatted_article.get('item_id',
formatted_article.get(config.ID_FIELD))
pub_seq_num = superdesk.get_resource_service('subscribers').generate_sequence_number(subscriber)
anpa = []
if codes:
anpa.append(b'\x05')
anpa.append(' '.join(codes).encode('ascii'))
anpa.append(b'\x0D\x0A')
# start of message header (syn syn soh)
anpa.append(b'\x16\x16\x01')
anpa.append(get_service_level(category, formatted_article).encode('ascii'))
# story number
anpa.append(str(pub_seq_num).zfill(4).encode('ascii'))
# field seperator
anpa.append(b'\x0A') # -LF
anpa.append(map_priority(formatted_article.get('priority')).encode('ascii'))
anpa.append(b'\x20')
anpa.append(category['qcode'].lower().encode('ascii'))
anpa.append(b'\x13')
# format identifier
if formatted_article.get(FORMAT, FORMATS.HTML) == FORMATS.PRESERVED:
anpa.append(b'\x12')
else:
anpa.append(b'\x11')
anpa.append(b'\x20')
# keyword
keyword = 'bc-{}'.format(self.append_legal(article=formatted_article, truncate=True)).replace(' ', '-')
keyword = keyword[:24] if len(keyword) > 24 else keyword
anpa.append(keyword.encode('ascii'))
anpa.append(b'\x20')
# version field
anpa.append(b'\x20')
# reference field
anpa.append(b'\x20')
# filing date
anpa.append('{}-{}'.format(formatted_article['_updated'].strftime('%m'),
formatted_article['_updated'].strftime('%d')).encode('ascii'))
anpa.append(b'\x20')
# add the word count
anpa.append(str(formatted_article.get('word_count', '0000')).zfill(4).encode('ascii'))
anpa.append(b'\x0D\x0A')
anpa.append(b'\x02') # STX
self._process_headline(anpa, formatted_article, category['qcode'].encode('ascii'))
keyword = SluglineMapper().map(article=formatted_article, category=category['qcode'].upper(),
truncate=True).encode('ascii', 'ignore')
anpa.append(keyword)
take_key = (formatted_article.get('anpa_take_key', '') or '').encode('ascii', 'ignore')
anpa.append((b'\x20' + take_key) if len(take_key) > 0 else b'')
anpa.append(b'\x0D\x0A')
if formatted_article.get('ednote', '') != '':
ednote = '{}\r\n'.format(to_ascii(formatted_article.get('ednote')))
anpa.append(ednote.encode('ascii', 'replace'))
if formatted_article.get(BYLINE):
anpa.append(get_text(formatted_article.get(BYLINE)).encode('ascii', 'replace'))
anpa.append(b'\x0D\x0A')
if formatted_article.get(FORMAT) == FORMATS.PRESERVED:
anpa.append(get_text(self.append_body_footer(formatted_article),
content='html').encode('ascii', 'replace'))
else:
body = to_ascii(formatted_article.get('body_html', ''))
# we need to inject the dateline
if formatted_article.get('dateline', {}).get('text') and not article.get('auto_publish', False):
body_html_elem = parse_html(formatted_article.get('body_html'))
ptag = body_html_elem.find('.//p')
if ptag is not None:
ptag.text = formatted_article['dateline']['text'] + ' ' + (ptag.text or '')
body = to_string(body_html_elem)
anpa.append(self.get_text_content(body))
if formatted_article.get('body_footer'):
anpa.append(self.get_text_content(to_ascii(formatted_article.get('body_footer', ''))))
anpa.append(b'\x0D\x0A')
anpa.append(mapped_source.encode('ascii'))
sign_off = (formatted_article.get('sign_off', '') or '').encode('ascii')
anpa.append((b'\x20' + sign_off) if len(sign_off) > 0 else b'')
anpa.append(b'\x0D\x0A')
anpa.append(b'\x03') # ETX
# time and date
anpa.append(datetime.datetime.now().strftime('%d-%m-%y %H-%M-%S').encode('ascii'))
anpa.append(b'\x04') # EOT
anpa.append(b'\x0D\x0A\x0D\x0A\x0D\x0A\x0D\x0A\x0D\x0A\x0D\x0A\x0D\x0A\x0D\x0A')
docs.append({'published_seq_num': pub_seq_num, 'encoded_item': b''.join(anpa),
'formatted_item': b''.join(anpa).decode('ascii')})
return docs
except Exception as ex:
raise FormatterError.AnpaFormatterError(ex, subscriber)
def _parse_content(self, article):
"""Parse body_html and mapping to fields required for apple news format
:param article:
"""
statement_regex = re.compile(r'^The Statement$', re.IGNORECASE)
analysis_regex = re.compile(r'^The Analysis$', re.IGNORECASE)
verdict_regex = re.compile(r'^The Verdict$', re.IGNORECASE)
references_regex = re.compile(r'^The References$', re.IGNORECASE)
url_regex = re.compile(r'(?:(?:https|http)://)[\w/\-?=%.]+\.[\w/\-?=%.]+', re.IGNORECASE)
abstract = get_text(article.get('abstract'), content='html').strip()
article['_title'] = abstract
body_html = article.get('body_html')
article['_analysis_first_line'] = ''
article['_analysis'] = ''
article['_statement'] = ''
article['_statement_attribution'] = ''
article['_verdict1'] = ''
article['_verdict2'] = ''
article['_references'] = ''
article['_revision_history'] = ''
if article.get(ITEM_STATE) == CONTENT_STATE.KILLED or article.get(ITEM_STATE) == CONTENT_STATE.RECALLED:
article['_title'] = 'This article has been removed.'
article['_analysis_first_line'] = 'This article has been removed.'
article['_analysis'] = 'This article has been removed.'
article['_statement'] = 'This article has been removed.'
article['_statement_attribution'] = 'This article has been removed.'
article['_verdict1'] = 'This article has been removed.'
article['_verdict2'] = 'This article has been removed.'
article['_references'] = 'This article has been removed.'
self._set_revision_history(article)
return
parsed_content = parse_html(body_html, content='html')
statement_found = False
analysis_found = False
analysis_first_line = False
verdict1_found = False
verdict2_found = False
references_found = False
statement_elements = []
for top_level_tag in parsed_content.xpath('/html/div/child::*'):
tag_text = format_text_content(top_level_tag).strip()
if not tag_text:
continue
if not verdict1_found:
if not statement_found:
match = statement_regex.search(tag_text)
if match:
statement_found = True
continue
else:
# statement found
match = verdict_regex.search(tag_text)
if match:
verdict1_found = True
if len(statement_elements) > 1:
statement_length = len(statement_elements) - 1
for i in range(statement_length):
article['_statement'] += get_text(
to_string(statement_elements[i], remove_root_div=False),
content='html'
).strip()
if statement_length > 1 and i != statement_length - 1:
article['_statement'] += '\r\n'
article['_statement_attribution'] = get_text(
to_string(statement_elements[-1:][0], remove_root_div=False),
content='html'
).strip()
elif len(statement_elements) == 1:
article['_statement'] = to_string(
statement_elements[0],
remove_root_div=False
)
continue
statement_elements.append(top_level_tag)
continue
if verdict1_found and not analysis_found:
match = analysis_regex.search(tag_text)
if match:
analysis_found = True
else:
article['_verdict1'] += to_string(top_level_tag, remove_root_div=False)
continue
if analysis_found and not verdict2_found:
if not analysis_first_line:
article['_analysis_first_line'] = tag_text
analysis_first_line = True
match = verdict_regex.search(tag_text)
if match:
verdict2_found = True
else:
article['_analysis'] += to_string(top_level_tag, remove_root_div=False)
continue
if verdict2_found and not references_found:
match = references_regex.search(tag_text)
if match:
references_found = True
else:
article['_verdict2'] += to_string(top_level_tag, remove_root_div=False)
continue
if references_found:
def replacement(match_object):
value = match_object.group(0)
if value:
return '<a href="{0}">{0}</a>'.format(value)
return ''
tag_text = re.sub(r'^\d*\s*[.):]?', '', tag_text).strip()
article['_references'] += '<li>{}</li>'.format(
re.sub(url_regex, replacement, tag_text)
)
if len(article['_references']):
article['_references'] = '<ol>{}</ol>'.format(article['_references'])
if not article.get('_statement') and article.get('_statement_attribution'):
# if statement is not as per the format
article['_statement'] = article.get('_statement_attribution')
article['_statement_attribution'] = ''
self._set_revision_history(article)
def plaintext_filter(value):
"""Filter out html from value."""
return get_text(value).replace('\n', ' ').strip()
def parse(self, xml, provider=None):
self.root = xml
try:
item = self.parse_item(xml)
if not item.get('headline'):
item['headline'] = text_utils.get_text(item.get('body_html', ''), 'html')[:100]
# abstract
try:
abstract = xml.xpath("//iptc:description[@role='drol:summary']", namespaces={'iptc': IPTC_NS})[0].text
except IndexError:
pass
else:
if abstract:
item['abstract'] = abstract
# genre
for genre_elt in xml.xpath("//iptc:genre", namespaces={'iptc': IPTC_NS}):
qcode = genre_elt.get('qcode')
if qcode is None:
continue
elif qcode.startswith('sttgenre:'):
qcode = qcode[9:]
genre_data = {'qcode': qcode}
name_elt = genre_elt.find(self.qname('name'))
name = name_elt.text if name_elt is not None and name_elt.text else ""
try:
name = self.getVocabulary("genre", qcode, name)
except ValueError:
continue
else:
genre_data['name'] = name
item.setdefault('genre', []).append(genre_data)
elif qcode.startswith('sttversion:'):
qcode = qcode[11:]
version_data = {'qcode': qcode, 'scheme': 'sttversion'}
name_elt = genre_elt.find(self.qname('name'))
name = name_elt.text if name_elt is not None and name_elt.text else ""
try:
name = self.getVocabulary("sttgenre", qcode, name)
except ValueError:
continue
else:
version_data['name'] = name
item.setdefault('subject', []).append(version_data)
# location
for location_elt in xml.xpath("//iptc:assert", namespaces={'iptc': IPTC_NS}):
qcode = location_elt.get("qcode")
if not qcode or not qcode.startswith("sttlocmeta:default:"):
continue
qcode = qcode[19:]
location_data = {"scheme": "sttlocmeta:default", "qcode": qcode}
for broader_elt in location_elt.xpath(".//iptc:broader[@type='cpnat:geoArea']",
namespaces={'iptc': IPTC_NS}):
qcode = broader_elt.get('qcode')
if not qcode:
continue
for key, mapping in STT_LOCATION_MAP.items():
if qcode.startswith(key + ":"):
if "qcode" in mapping:
qcode = qcode[len(key) + 1:]
try:
name = broader_elt.find(self.qname('name')).text
except AttributeError:
name = ""
try:
name = self.getVocabulary(key, qcode, name)
except ValueError:
continue
else:
location_data[mapping["qcode"]] = qcode
if "name" in mapping:
location_data[mapping["name"]] = name
item.setdefault('place', []).append(location_data)
# public editorial note
if 'ednote' in item:
# stt has specific roles for public and private editorial notes
# so we remove ednote found by parent parser, as it takes first one
# as a public note
del item['ednote']
try:
ednote = xml.xpath("//iptc:edNote[@role='sttnote:public']", namespaces={'iptc': IPTC_NS})[0].text
except IndexError:
pass
else:
if ednote:
item['ednote'] = ednote
# private editorial note
try:
private_note = xml.xpath("//iptc:edNote[@role='sttnote:private']", namespaces={'iptc': IPTC_NS})[0].text
except IndexError:
pass
else:
if private_note:
item.setdefault('extra', {})['sttnote_private'] = private_note
return [item]
except Exception as ex:
raise ParserError.newsmlTwoParserError(ex, provider)
def _transform_to_ninjs(self, article, subscriber, recursive=True):
ninjs = {
'guid': article.get(GUID_FIELD, article.get('uri')),
'version': str(article.get(config.VERSION, 1)),
'type': self._get_type(article)
}
if article.get('byline'):
ninjs['byline'] = article['byline']
located = article.get('dateline', {}).get('located', {})
if located:
ninjs['located'] = located.get('city', '')
for copy_property in self.direct_copy_properties:
if article.get(copy_property) is not None:
ninjs[copy_property] = article[copy_property]
if 'body_text' not in article and 'alt_text' in article:
ninjs['body_text'] = article['alt_text']
if 'title' in article:
ninjs['headline'] = article['title']
if article.get('body_html'):
ninjs['body_html'] = self.append_body_footer(article)
if article.get('description'):
ninjs['description_html'] = self.append_body_footer(article)
if article.get('place'):
ninjs['place'] = self._format_place(article)
if article.get('profile'):
ninjs['profile'] = self._format_profile(article['profile'])
extra_items = None
if recursive:
if article[ITEM_TYPE] == CONTENT_TYPE.COMPOSITE:
ninjs[ASSOCIATIONS] = self._get_associations(article, subscriber)
if article.get(ASSOCIATIONS):
associations, extra_items = self._format_related(article, subscriber)
ninjs[ASSOCIATIONS].update(associations)
elif article.get(ASSOCIATIONS):
ninjs[ASSOCIATIONS], extra_items = self._format_related(article, subscriber)
elif article.get(ASSOCIATIONS):
ninjs[ASSOCIATIONS], extra_items = self._format_related(article, subscriber)
if extra_items:
ninjs.setdefault(EXTRA_ITEMS, {}).update(extra_items)
if article.get(EMBARGO):
ninjs['embargoed'] = get_utc_schedule(article, EMBARGO).isoformat()
if article.get('priority'):
ninjs['priority'] = article['priority']
else:
ninjs['priority'] = 5
if article.get('subject'):
ninjs['subject'] = self._get_subject(article)
if article.get('anpa_category'):
ninjs['service'] = self._get_service(article)
if article.get('renditions'):
ninjs['renditions'] = self._get_renditions(article)
elif 'url' in article:
ninjs['renditions'] = self._generate_renditions(article)
# SDPA-317
if 'abstract' in article:
abstract = article.get('abstract', '')
ninjs['description_html'] = abstract
ninjs['description_text'] = text_utils.get_text(abstract)
elif article.get('description_text'):
ninjs['description_text'] = article.get('description_text')
if article.get('company_codes'):
ninjs['organisation'] = [{'name': c.get('name', ''), 'rel': 'Securities Identifier',
'symbols': [{'ticker': c.get('qcode', ''),
'exchange': c.get('security_exchange', '')}]}
for c in article['company_codes']]
elif 'company' in article:
ninjs['organisation'] = [{'name': article['company']}]
if article.get('rewrite_of'):
ninjs['evolvedfrom'] = article['rewrite_of']
if not ninjs.get('copyrightholder') and not ninjs.get('copyrightnotice') and not ninjs.get('usageterms'):
ninjs.update(superdesk.get_resource_service('vocabularies').get_rightsinfo(article))
if 'genre' in article:
ninjs['genre'] = self._get_genre(article)
if article.get('flags', {}).get('marked_for_legal'):
ninjs['signal'] = self._format_signal_cwarn()
if article.get('attachments'):
ninjs['attachments'] = self._format_attachments(article)
if ninjs['type'] == CONTENT_TYPE.TEXT and ('body_html' in ninjs or 'body_text' in ninjs):
if 'body_html' in ninjs:
body_html = ninjs['body_html']
word_count = text_utils.get_word_count(body_html)
char_count = text_utils.get_char_count(body_html)
readtime = text_utils.get_reading_time(body_html, word_count, article.get('language'))
else:
body_text = ninjs['body_text']
word_count = text_utils.get_text_word_count(body_text)
char_count = len(body_text)
readtime = text_utils.get_reading_time(body_text, word_count, article.get('language'))
ninjs['charcount'] = char_count
ninjs['wordcount'] = word_count
ninjs['readtime'] = readtime
if article.get('authors'):
ninjs['authors'] = self._format_authors(article)
return ninjs
def _format_body_content(self, article, body_content):
if article.get(FORMAT) == FORMATS.PRESERVED:
pre = get_text(self.append_body_footer(article))
SubElement(body_content, 'pre').text = pre
else:
self.map_html_to_xml(body_content, self.append_body_footer(article))
def _set_headline(self, item, value):
if not value:
# if there is no headline, we use first 100 chars of body
# cf. SDNTB-481
value = text_utils.get_text(item.get('body_html', ''), 'html')[:100]
item['headline'] = value
def _ednote_filter(self, ednote):
return text_utils.get_text(ednote, lf_on_block=True).strip()
