Sorry to bring this back up, but although the underlying problems seem resolved and the HD and HB formats are fixed, the "not integrated" message still appears for the Files tab for some records. For example: https://documents.classe.cornell.edu/record/2078?ln=en
Fortunately, exporting the xml of our entire collection and then running "bibupload -r" seems to fix this. Please just let me know if there's anything else I should look into first, or if there are any other fixes you'd recommend. Thanks again, Devin > On Jan 11, 2017, at 4:23 PM, Devin A. Bougie <[email protected]> wrote: > > Hi Tibor, > > I believe everything has now been fixed. Thank you so much for all of your > time and help. > > On Jan 11, 2017, at 10:59 AM, Tibor Simko <[email protected]> wrote: >> ... >> You should have the same version of `lxml` installed via >> `requirements.txt`. Perhaps you also have a native OS-packaged lxml >> version which causes some conflicts? > > That was exactly the problem. I've now fixed my python setup and the pip > commands to install and use the versions specified in requirements.txt > regardless of what's installed at the system level. > > inveniocfg --run-unit-tests now only shows two failures: > > ------ > ... > ====================================================================== > FAIL: search engine - parsing structured regexp refersto query > ---------------------------------------------------------------------- > Traceback (most recent call last): > File > "/mnt/documents/invenio/lib/python/invenio/search_engine_unit_tests.py", line > 236, in test_parsing_structured_regexp_refersto_query > [['+', '(one|two)', 'refersto', 'r']]) > File > "/mnt/documents/invenio/lib/python/invenio/search_engine_unit_tests.py", line > 85, in _check > repr(result_wanted)) > AssertionError: obtained [['+', 'refersto:/(one|two)/', '', 'w']] instead of > [['+', '(one|two)', 'refersto', 'r']] > > ====================================================================== > FAIL: SPIRES and invenio search syntax - find a ellis and citedby:hawking > ---------------------------------------------------------------------- > Traceback (most recent call last): > File > "/mnt/documents/invenio/lib/python/invenio/search_engine_query_parser_unit_tests.py", > line 990, in test_simple_syntax_mixing > self._compare_searches(inv_search, combo_search) > File > "/mnt/documents/invenio/lib/python/invenio/search_engine_query_parser_unit_tests.py", > line 372, in _compare_searches > (repr(result_obtained), repr(result_wanted)) > AssertionError: SPIRES parsed as [['+', 'ellis', 'author', 'w'], ['+', > 'citedby:"hawking"', '', 'w']] instead of [['+', 'ellis', 'author', 'w'], > ['+', 'citedby:hawking', '', 'w']] > > ---------------------------------------------------------------------- > Ran 852 tests in 129.849s > > FAILED (failures=2) > ------ > >> ... >> You can try installing the rxp parser, which I don't see in your output. >> RXP is the fastest parser available, and it works well on my box: >> >> $ sudo -u www-data python >> /opt/invenio/lib/python/invenio/bibrecord_unit_tests.py 2>&1 | grep >> create_record >> bibrecord - create_record() with lxml ... ok >> bibrecord - create_record() with minidom ... ok >> bibrecord - create_record() with pyRXP ... ok > > This also succeeds now that I've fixed my python setup and pip command. > > Thanks again! > > Devin >
