def unpack(self):
"""
Unpack the downloads into the root directory for this map
"""
global VERBOSE
if VERBOSE:
print "Unpacking data files to disk..."
# need to check what file type we've got now...
file_types = {
'.csv': lambda x: None, # don't need to unpack uncompressed files
'.sql': lambda x: None,
'.xls': lambda x: None,
'.xlsx': lambda x: None,
'.html': lambda x: None,
'.pdf': lambda x: None,
'.tar': unpack_tar,
'.gz': unpack_gzip,
'.tgz': unpack_tar,
'.tar.gz': unpack_tar,
'.zip': unpack_zip,
}
# get all files in working directory of this map
files = os.listdir(TMP_DIRECTORY + '%s/' % self.__name__)
# iterate through files
for f in files:
file_name = os.path.basename(f)
# separate out the file extension
root, ext = guess_extension(file_name)
# using file type, extract this file!
file_types[ext](os.path.basename(f))
def install(self):
"""
Does installation of the files into user's chosen database
This is a primarily internal method, but if base it should just get called.
NOTES:
- Does installation have to assume that it can just install from each of the files available? Do we
have to re-write the installer for something complex like the US Census? And is that an acceptable level
of configuration for a Map?
TODO:
- Need to fix how headers work -- can specify whether headers are present, whether all data should be installed
into the same database?
"""
# check if we need a separate db for each url or whether one is enough
# one is enough if specified here
if self.db_name:
db_name = self.db_name
self.db.create_db(self.__name__)
# for every file url
#files = os.listdir(TMP_DIRECTORY + '%s/' % self.__name__)
for k, v in self.data.iteritems():
root, ext = guess_extension(v['url'])
file_name = os.path.basename(root + ext)
# If we don't have a db name, we should find it in the URLs
if self.db_name:
db_name = self.db_name
else:
db_name = v['database']
self.db.create_db(db_name=db_name)
if ext == ".sql":
# if we have a SQL file, we should run that
# TODO: THIS DOESN'T ACTUALLY WORK, BUT WE NEED TO DO SOMETHING LIKE THIS
self.db.query(f)
elif ext in (".csv", ".pdf", ".xls", ".xlsx", ".html"):
# create messy2sql instance
m2s = Messy2SQL(file_name, DATABASES['sql']['type'])
# if we have PDF, HTML, CSV, or Excel files, we should use messy2sql
# get a table query, run it!
fh = open((TMP_DIRECTORY + self.__name__ + '/' + file_name), 'rb')
# use messytables to build a MessyTables RowSet with file type
rows = {
'.csv': CSVTableSet(fh).tables[0],
# '.pdf': PDFTableSet(file_name),
# '.xlsx': XLSTableSet(file_name),
# '.xls': XLSTableSet(file_name),
# '.html': HTMLTableSet(file_name),
}[ext]
# use the rowset here to create a sql table query and execute
self.db.create_table(query = m2s.create_sql_table(rows), db_name=db_name)
# get insert statements
self.db.insert(query = m2s.create_sql_insert(rows), db_name=db_name, table_name=root)
else:
pass
