# gscript.run_command('g.region', flags='p')

input = options['input']

# input = '{}_net'.format(oinput)

layer = options['layer']

output = options['output']

node_layer = options['node_layer']

table = '{}_{}'.format(output, 1)

# # a tempfile would be needed if graph could be read from an edgelist

# tmpFile = grass.tempfile()

gscript.verbose(_("Reading network data..."))

## Read Network data from vector map

gscript.run_command('v.net', flags='c', input=input,

output=output, operation='nodes',

node_layer=node_layer, quiet=True)

#gscript.run_command('v.db.addtable', map=output, layer=2, key='cat')

#gscript.run_command('v.to.db', map=output, layer=node_layer, option='cat', columns='cat', quiet=True)

# Data has to be parsed or written to file as StringIO objects are not supported by igraph

# https://github.com/igraph/python-igraph/issues/8

net = gscript.read_command('v.net', input=output, points=output, node_layer=node_layer,

operation='report', quiet=True).split('\n')

# Parse network data and extract vertices, edges and edge names

edges = []

vertices = []

edge_cat = []

for l in net:

if l != '':

# Names for edges and vertices have to be of type string

# Names (cat) for edges

edge_cat.append(l.split(' ')[0])

# From- and to-vertices for edges

edges.append((l.split(' ')[1], l.split(' ')[2]))

# Names (cat) for from-vertices

vertices.append(l.split(' ')[1])

# Names (cat) for to-vertices

vertices.append(l.split(' ')[2])

# Create Graph object

g = Graph().as_directed()

# Add vertices with names

vertices.sort()

vertices = set(vertices)

g.add_vertices(list(vertices))

# Add vertices with names

g.add_edges(edges)

gscript.verbose(_("Computing neighborhood..."))

# Compute number of vertices that can be reached from each vertex

# Indicates upstream or downstream position of a node

g.vs['nbh'] = g.neighborhood_size(mode='out', order=g.diameter())

g.vs['cl'] = g.as_undirected().clusters().membership

# Compute incoming degree centrality

# sources have incoming degree centrality of 0

g.vs['indegree'] = g.degree(type="in")

# Compute outgoing degree centrality

# outlets have outgoing degree centrality of 0

g.vs['outdegree'] = g.degree(type="out")

gscript.verbose(_("Writing result to table..."))

# Get Attributes

attrs = []

for n in g.vs:

attrs.append((int(n['name']), int(n['nbh']), int(n['cl']),

int(n['indegree']),int(n['outdegree'])))

# Write results back to attribute table

# Note: Backend depenent! For a more general solution this has to be handled

path = '$GISDBASE/$LOCATION_NAME/$MAPSET/sqlite/sqlite.db'

conn = sqlite3.connect(get_path(path))

c = conn.cursor()

c.execute('DROP TABLE IF EXISTS {}'.format(table))

# Create temporary table

c.execute('''CREATE TABLE {}

(cat integer, neighborhood integer,

cluster integer, indegree integer,

outdegree integer)'''.format(table))

conn.commit()

# Insert data into temporary table

c.executemany('INSERT INTO {} VALUES (?,?,?,?,?)'.format(table), attrs)

# Save (commit) the changes

conn.commit()

# Connect table to output node layer

gscript.run_command('v.db.connect', map=output, table=table, layer=node_layer, flags='o')

# Join temporary table to output

#gscript.run_command('v.db.join', map=output, layer=node_layer,

# column='cat', other_table=tmpTable,

# other_column='cat', quiet=True)

# Remove temporary table

#c = conn.cursor()

#c.execute('DROP TABLE IF EXISTS {}'.format(tmpTable))

#conn.commit()

# We can also close the connection if we are done with it.

# Just be sure any changes have been committed or they will be lost.