def test_create_and_delete_document(self):
url = self.gemini_example.url_for(0)
content = self.gemini_example.get_from_url(url)
document = HarvestedDocument(url=url, content=content)
document.save()
document_id = document.id
self.assert_equal(document.url, url)
self.assert_equal(document.content, content)
self.delete_commit(document)
self.assert_raises(Exception, HarvestedDocument.get, document_id)
def assert_read_values(self, example_index, expect_values):
url = self.gemini_example.url_for(file_index=example_index)
content = self.gemini_example.get_from_url(url)
document = HarvestedDocument(url=url, content=content)
values = document.read_values()
self.assert_gemini_values(values, expect_values)
def write_package_from_gemini_string(self, content):
"""Create or update a Package based on some content that has
come from a URL.
Also store the raw content as a HarvestedDocument (with
references to its source and its package)
"""
# Look for previously harvested document matching Gemini GUID
gemini_document = GeminiDocument(content)
gemini_values = gemini_document.read_values()
gemini_guid = gemini_values["guid"]
harvested_documents = HarvestedDocument.filter(guid=gemini_guid).all()
if len(harvested_documents) > 1:
# A programming error; should never happen
raise Exception("More than one harvested document GUID %s" % gemini_guid)
elif len(harvested_documents) == 1:
# we've previously harvested this (i.e. it's an update)
harvested_doc = harvested_documents[0]
if harvested_doc.source.id != self.job.source.id:
# A 'user' error: there are two or more sources
# pointing to the same harvested document
raise HarvesterError("Another source is using metadata GUID %s" % self.job.source.id)
# XXX Not strictly true - we need to check the title, package resources etc
if harvested_doc.read_values() == gemini_values:
# nothing's changed
return None
package = harvested_doc.package
else:
harvested_doc = None
package = None
extras = {"publisher": int(self.job.source.publisher_ref or 0), "INSPIRE": "True"}
# Just add some of the metadata as extras, not the whole lot
for name in ["bbox-east-long", "bbox-north-lat", "bbox-south-lat", "bbox-west-long", "abstract", "guid"]:
extras[name] = gemini_values[name]
package_data = {"name": str(gemini_guid), "title": gemini_values["title"], "extras": extras}
resource_locator = (
gemini_values.get("resource-locator", []) and gemini_values["resource-locator"][0].get("url") or ""
)
if resource_locator:
package_data["resources"] = [
{"url": resource_locator, "description": "Resource locator", "format": "Unverified"},
{
"url": "%s/api/2/rest/harvesteddocument/%s/xml/%s.xml"
% (config.get("ckan.api_url", "/").rstrip("/"), gemini_guid, gemini_guid),
"description": "Source GEMINI 2 document",
"format": "XML",
},
{
"url": "%s/api/2/rest/harvesteddocument/%s/html/%s.html"
% (config.get("ckan.api_url", "/").rstrip("/"), gemini_guid, gemini_guid),
"description": "Formatted GEMINI 2 document",
"format": "HTML",
},
]
if package == None:
# Create new package from data.
package = self._create_package_from_data(package_data)
else:
package = self._update_package_from_data(package, package_data)
harvested_doc = HarvestedDocument(content=content, guid=gemini_guid, package=package, source=self.job.source)
harvested_doc.save()
return package
