aleph_soruces/ directory cointains few xml files for test purposes, which are
read, parsed and saved to:

    parsed_outputs/$(basename).txt -- outputs from EPublication namedtuple
                                      conversion to string

    shelve_files/$(basename).shelve -- for EPublication named tuples saved to
                                    shelve under "data" key

    xml_outputs/$(basename) -- outputs from MARCXMLRecord class converted back
                               to XML


$(basename) is taken from original xml file using os.path.basename(). See
to_shelve.py for details.

Files from parsed_outputs/ should be on XML level (not byte level) same as files
from aleph_soruces/.

Files from shelve_files/ can be used to test to test AMQP communication, as
they are serialized structures from __init__.py.

parsed_outputs/ are there for humans, just to check, that everything works as
expected.


Shelve example:

>>> import shelve
>>> f = shelve.open("shelve_files/example.xml.shelve")
>>> f["data"]
EPublication(nazev='D\xc3\xadt\xc4\x9b, \xc5\xa1kola a matematika :', podnazev='konstruktivistick\xc3\xa9 p\xc5\x99\xc3\xadstupy k vyu\xc4\x8dov\xc3\xa1n\xc3\xad /', vazba='(bro\xc5\xbe.)', cena='', castDil='', nazevCasti='', nakladatelVydavatel='Port\xc3\xa1l,', datumVydani='2009', poradiVydani='2., aktual. vyd.', zpracovatelZaznamu='ABA001', kategorieProRIV='', mistoDistribuce='', distributor='', datumDistribuce='', datumProCopyright='', format='23 cm', url='', mistoVydani='Praha :', ISBNSouboruPublikaci=['978-80-7367-397-0'], autori=[Author(firstName='Milan,', lastName='Hejn\xc3\xbd,', title=''), Author(firstName='Franti\xc5\xa1ek,', lastName='Ku\xc5\x99ina', title='')], originaly=[])