[web] safety belt to avoid overriding pageid with loadxhtml()
HtmlHeaders.define_var() now accepts a third optional 'override' paramater.
If set to False, the variable won't be evaluated if it's already defined
in the page.
from unittest import TestCase
import os.path as osp
from xml.etree.cElementTree import ElementTree, fromstring, tostring, dump
from tempfile import NamedTemporaryFile
from subprocess import Popen as sub
from cubicweb.utils import can_do_pdf_conversion
from cubicweb.ext.xhtml2fo import ReportTransformer
DATADIR = osp.join(osp.dirname(__file__), 'data')
class PDFTC(TestCase):
def test_xhtml_to_fop_to_pdf(self):
if not can_do_pdf_conversion():
self.skip('dependencies not available : check pysixt and fop')
xmltree = ElementTree()
xmltree.parse(osp.join(DATADIR, 'sample1.xml'))
foptree = ReportTransformer(u'contentmain').transform(xmltree)
# next
foptmp = NamedTemporaryFile()
foptree.write(foptmp)
foptmp.flush()
pdftmp = NamedTemporaryFile()
fopproc = sub(['/usr/bin/fop', foptmp.name, pdftmp.name])
fopproc.wait()
del foptmp
pdftmp.seek(0) # a bit superstitious
reference = open(osp.join(DATADIR, 'sample1.pdf'), 'r').read()
output = pdftmp.read()
# XXX almost equals due to ID, creation date, so it seems to fail
self.assertEquals( len(output), len(reference) )
# cut begin & end 'cause they contain variyng data
self.assertTextEquals(output[150:1500], reference[150:1500])