def test_extref(self):
self.content.node = etree.fromstring(self.EXTREF)
fmt = format_ead(self.content)
self.assert_('<a href="http://pid.emory.edu/ark:/25593/8zgst">Irish Literary Miscellany</a>'
in fmt, 'extref tag converted to a href')
self.content.node = etree.fromstring(self.EXTREF_NOLINK)
fmt = format_ead(self.content)
self.assert_('<a>Irish Literary Miscellany</a>'
in fmt, 'formatter should not fail when extref has no href')
def test_title_emph(self):
self.content.node = etree.fromstring(self.TITLE_EMPH)
fmt = format_ead(self.content)
self.assert_('<em>Biographical source:</em> "Shaw, George' in fmt,
"emph tag rendered correctly in section with title")
self.assert_('<span class="ead-title">Contemporary Authors Online</span>, Gale' in fmt,
"title rendered correctly in sectino with emph tag")
def test_title(self):
self.content.node = etree.fromstring(self.TITLE)
fmt = format_ead(self.content)
self.assert_('magazine <span class="ead-title">The Smart Set</span> from' in fmt,
"title tag converted correctly to span class ead-title")
# title variants
# - doublequotes
self.content.node = etree.fromstring(self.TITLE_QUOT)
fmt = format_ead(self.content)
self.assertEqual('"Terminus"', fmt)
# - multiple
self.content.node = etree.fromstring(self.TITLE_MULTI)
fmt = format_ead(self.content)
self.assertEqual('Some Author: "Terminus", "Saturday"', fmt)
# - multiple titles + RDFa
fmt = format_ead(self.content, rdfa=True)
self.assertEqual('<span rel="dc:creator"><span typeof="schema:Person"><span property="schema:name">Some Author</span></span></span>: "<span inlist="inlist" property="dc:title">Terminus</span>", "<span inlist="inlist" property="dc:title">Saturday</span>"',
fmt)
def test_exist_match(self):
self.content.node = etree.fromstring(self.EXIST_MATCH)
fmt = format_ead(self.content)
self.assert_('Pitts v. <span class="exist-match">Freeman</span>'
in fmt, 'exist:match tag converted to span for highlighting')
def test_notrans(self):
self.content.node = etree.fromstring(self.NOTRANS)
fmt = format_ead(self.content)
self.assert_('magazine <span class="ead-title">The Smart Set</span>...' in fmt,
"nested format rendered correctly")
def test_bold(self):
self.content.node = etree.fromstring(self.BOLD)
fmt = format_ead(self.content)
self.assert_('<span class="ead-bold">Pitts v. Freeman</span> school desegregation' in fmt,
"render bold converted correctly to span class ead-bold")
def test_italics(self):
self.content.node = etree.fromstring(self.ITALICS)
fmt = format_ead(self.content)
self.assert_('<span class="ead-italic">Pitts v. Freeman</span> school desegregation' in fmt,
"render italic converted correctly to span class ead-italic")
def check_eadxml(ead):
"""Sanity checks specific to the EAD xml, independent of file or eXist.
Checks the following:
- series and index ids are present
- fields used for search/browse title match code expectations:
- at most one top-level origination
- no leading whitespace in list-title (origination or unittitle)
- alphabetical first letter (for first-letter browse)
- eadid matches site URL regex
:param ead: :class:`~findingaids.fa.models.FindingAid` ead instance to be checked
:returns: list of all errors found
:rtype: list
"""
# NOTE: throughout, be sure to use unicode instead of string
errors = []
# check that series ids are set
if ead.dsc and ead.dsc.hasSeries():
for series in ead.dsc.c:
errors.extend(check_series_ids(series))
# check that any index ids are set
for index in ead.archdesc.index:
if not index.id:
errors.append("%(node)s id attribute is not set for %(label)s"
% {'node': local_name(index.node), 'label': unicode(index.head)})
# eadid matches appropriate site URL regex
if not re.match('^%s$' % EADID_URL_REGEX, ead.eadid.value): # entire eadid should match regex
errors.append("eadid '%s' does not match site URL regular expression" \
% ead.eadid.value)
# multiple tests to ensure xml used for search/browse list-title matches what code expects
# -- since list title is pulled from multiple places, give enough context so it can be found & corrected
list_title_path = "%s/%s" % (local_name(ead.list_title.node.getparent()),
local_name(ead.list_title.node))
# - check for at most one top-level origination
origination_count = ead.node.xpath('count(e:archdesc/e:did/e:origination)',
namespaces={'e': EAD_NAMESPACE})
if int(origination_count) > 1:
errors.append("Site expects only one archdesc/did/origination; found %d" \
% origination_count)
# container list formatting (based on encoding practice) expects only 2 containers per did
# - dids with more than 2 containers
containers = ead.node.xpath('//e:did[count(e:container) > 2]',
namespaces={'e': EAD_NAMESPACE})
if len(containers):
errors.append("Site expects maximum of 2 containers per did; found %d did(s) with more than 2" \
% len(containers))
errors.append(['Line %d: %s' % (c.sourceline, tostring(c)) for c in containers])
# - dids with only one container
containers = ead.node.xpath('//e:did[count(e:container) = 1]',
namespaces={'e': EAD_NAMESPACE})
if len(containers):
errors.append("Site expects 2 containers per did; found %d did(s) with only 1" \
% len(containers))
errors.append(['Line %d: %s' % (c.sourceline, tostring(c)) for c in containers])
# - no leading whitespace in list title
title_node = ead.node.xpath("%s/text()" % ead.list_title_xpath,
namespaces={'e': EAD_NAMESPACE})
if hasattr(title_node[0], 'text'):
title_text = title_node[0].text
else:
title_text = unicode(title_node)
if title_text is None:
errors.append("List title seems to be empty")
elif re.match(r'\s+', title_text):
# using node.text because unicode() normalizes, which obscures whitespace problems
errors.append("Found leading whitespace in list title field (%s): '%s'" %
(list_title_path, ead.list_title.node.text))
# report with enough context that they can find the appropriate element to fix
# - first letter of title matches regex -- only check if whitespace test fails
elif not re.match(TITLE_LETTERS, ead.first_letter):
errors.append("First letter ('%s') of list title field %s does not match browse letter URL regex '%s'" % \
(ead.first_letter, list_title_path, TITLE_LETTERS))
# leading space in unit title (could be list title but might not be)
# NOTE: title can contain and even start with subtags such as <title>
# or <emph>, which is hard to account for with lxml or text() xpath.
# Using format_ead to generate html that would be displayed, and then
# stripping tags to check for any leading whitespace within a leading tag
title = striptags(format_ead(ead.unittitle))
if re.match(r'\s+', title):
errors.append("Found leading whitespace in unittitle: '%s'" %
title)
# leading whitespace in control access fields (if any)
if ead.archdesc.controlaccess and ead.archdesc.controlaccess.controlaccess:
for ca in ead.archdesc.controlaccess.controlaccess:
for term in ca.terms:
# NOTE: using node text because term.value is now normalized
if re.match(r'\s+', unicode(term.node.text)):
errors.append("Found leading whitespace in controlaccess term '%s' (%s)" \
% (term.node.text, local_name(term.node)))
# eadid url should contain resolvable ARK
if ead.eadid.url is None or not is_ark(ead.eadid.url):
errors.append("eadid url is either not set or not an ARK. " +
"To correct, run the prep process again.")
# eadid identifier should contain short-form ARK
if ead.eadid.identifier is None or not is_ark(ead.eadid.identifier):
errors.append("eadid identifier is either not set or not an ARK" +
"To correct, run the prep process again.")
# short- and long-form ARKs should match each other
if ead.eadid.url is not None and ead.eadid.identifier is not None and \
not ead.eadid.url.endswith(ead.eadid.identifier):
errors.append("eadid url and identifier do not match: url '%s' should end with identifier '%s'" \
% (ead.eadid.url, ead.eadid.identifier))
return errors
