escorts: extract from upstream instead of konect

entries/escorts/__init__.py

@@ -17,16 +17,17 @@
 # along with this program.  If not, see <http://www.gnu.org/licenses/>.
 
 from .. import *
+import csv
 
 title = "Brazilian prostitution network (2010)"
 description = """A bipartite network of escort and individuals who buy sex from them in Brazil, extracted from a Brazilian online community for such ratings. Edges represent a purchase of sexual intercourse with an escort, and the edge weight gives the buyer's rating of the escort: -1 (bad), 0 (neutral), or +1 (good). Other edge attributes are available. Edges are timestamped.[^icon]
 [^icon]: Description obtained from the [ICON](https://icon.colorado.edu) project."""
 tags = ['Social', 'Public Health', 'Weighted', 'Timestamps']
-url = 'http://konect.cc/networks/escorts'
+url = 'https://doi.org/10.1371/journal.pcbi.1001109'
 citation = [('L. Rocha, F. Liljeros, & P. Holme. "Simulated epidemics in an empirical spatiotemporal network of 50,185 sexual contacts." PLoS Comput Biol, 7(3). (2011).', 'http://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1001109')]
 icon_hash = '5832214b8cf714f319c427a0'
-upstream_prefix = 'http://konect.cc/files'
-files = [('download.tsv.escorts.tar.bz2', None, 'konect')]
+upstream_prefix = 'https://doi.org/10.1371'
+files = [('journal.pcbi.1001109.s001', None, 'csv')]
 
 def fetch_upstream(force=False):
     return fetch_upstream_files(__name__.split(".")[-1], upstream_prefix, files,
@@ -44,6 +45,20 @@ def parse(alts=None):
         if isinstance(fnames, str):
             fnames = [fnames]
         with ExitStack() as stack:
-            fs = [stack.enter_context(open_upstream_file(name, fn, "rb")) for fn in fnames]
-            g = parse_graph(fs, fmt, directed=True)
+            fs = [stack.enter_context(open_upstream_file(name, fn, "r")) for fn in fnames]
+            g = Graph(directed=True)
+            reader = csv.reader(fs[0], delimiter=";")
+            def edges():
+                for row in reader:
+                    if len(row) == 1:
+                        continue
+                    yield [int(x) for x in row]
+            eprops = ["time", "rating", "anal", "oral", "kiss"]
+            for p in eprops:
+                g.ep[p] = g.new_ep("int")
+            g.add_edge_list(edges(), hashed=False,
+                            eprops=[g.ep[p] for p in eprops])
+            g.vp.male = g.new_vp("bool")
+            g.vp.male.a = g.degree_property_map("out").a > 0
+            g.set_directed(False)
         yield alt, g