def write_clusters(xmlcollection, clusters, base_clust_dir, type_='soft'):
"""
@param xmlcollection: collection of XML documents
@param clusters: Structure containing the clustered document
indices and the stems used for doing so.
@param base_clust_dir: Where the base directory for storing
clusters reside.
@param type_: Default arg is 'soft' indicating our clusters
params reflects soft clusters. Other possible
argument: 'hard' for hard clustering.
"""
clust_dir = base_clust_dir + type_ + sep
"""
try:
makedirs(clust_dir)
except OSError, e:
print(e)
"""
# XXX: Check
clust_no = 1
for docs, clust_stems in clusters:
specific_clust_dir = clust_dir + str(clust_no) + sep
try:
makedirs(specific_clust_dir)
except OSError, e:
print(e)
for doc_id in docs:
xmldoc = xmlcollection.get_doc(doc_id)
f = open(specific_clust_dir + xmldoc.get_id(), "w", get_def_enc())
f.write(xmldoc.get_rawcontent())
f.close()
print "clust_dir: ", clust_dir, "clust_no: ", clust_no, " [written]"
f = open(specific_clust_dir + "clust_stems", "w", get_def_enc())
g = open("/home/hernani/clust_stems", "a", get_def_enc()) # DEBUG line
for clust_stem in clust_stems:
f.write(str(clust_stem) + "\n") # stem should come here directly
g.write(str(clust_stem) + "\n") # DEBUG line
f.close()
g.close()
clust_no += 1
def export_d3_js(G,
files_dir=get_web_output_dir(),
graphname=get_def_graph_name(),
group=None,
width=get_webgraph_res()[0],
height=get_webgraph_res()[1],
node_labels=False,
encoding=get_def_enc()):
"""
A function that exports a NetworkX graph as an interavtice D3.js object.
The function builds a folder, containing the graph's formatted JSON, the D3.js
JavaScript, and an HTML page to load the graph in a browser.
Parameters
----------
G : graph
a NetworkX graph
files_dir : string, optional
name of directory to save files
graphname : string, optional
the name of the graph being save as JSON, will appears in directory as 'graphname.json'
group : string, optional
The name of the 'group' key for each node in the graph. This is used to
assign nodes to exclusive partitions, and for node coloring if visualizing.
width : int, optional
width (px) of display frame for graph object in browser window
height : int, optional
height (px) of display frame for graph object in browser window
node_labels : bool, optional
If true, nodes are displayed with labels in browser
encoding: string, optional
Specify which encoding to use when writing file.
Examples
--------
>>> from scipy import random
>>> from networkx.readwrite import d3_js
>>> G = nx.random_lobster(20, .8, .8)
>>> low = 0
>>> high = 5
>>> G.add_nodes_from(map(lambda i: (i, {'group': random.random_integers(low, high, 1)[0]}), G.nodes()))
>>> G.add_edges_from(map(lambda e: (e[0], e[1], {'weight': random.random_integers(low+1, high, 1)[0]}), G.edges()))
>>> d3_js.export_d3_js(G, files_dir="random_lobster", graphname="random_lobster_graph", node_labels=False)
"""
if not os.path.exists(files_dir):
os.makedirs(files_dir)
# Begin by creating the necessary JS and HTML files
write_d3_js(G, path=files_dir+"/"+graphname+".json", group=group, encoding=encoding)
graph_force_html = open(files_dir+'/'+graphname+'.html', 'w')
for line in d3_html.split("\n"):
if line.find('"../../d3.js"') > 0:
line = line.replace('"../../d3.js"', '"d3/d3.js"')
if line.find('"../../d3.geom.js"') > 0:
line = line.replace('"../../d3.geom.js"', '"d3/d3.geom.js"')
if line.find('"../../d3.layout.js"') > 0:
line = line.replace('"../../d3.layout.js"', '"d3/d3.layout.js"')
if line.find('"force.css"') > 0:
line = line.replace('"force.css"', '"d3/force.css"')
if line.find('"force.js"') > 0:
line = line.replace('"force.js"', '"'+graphname+'.js"')
graph_force_html.write(line+'\n'.encode(encoding))
graph_force_html.close()
argv.append(get_mailfolder() + "*")
for arg in argv[1:]:
for filename in glob(arg):
# If works, document is well-formed
try:
parseFile(filename)
# If exception occurs, document is not well-formed; add to
# collection of invalid docs.
except Exception, e:
add_invalid_docs(filename, str(e))
print filename
continue
# Prepare XML file to write invalid input XML files of the
# collection into.
invalid_xml_filehandler = open(get_invalid_xml_filename(), "w", get_def_enc())
invalid_xmldoc = Doc()
invalid_xmldocs = invalid_xmldoc.createElement("invalid_xmldocsection")
invalid_xmldoc.appendChild(invalid_xmldocs)
invalidstat = defaultdict(int)
# Check collection of invalid docs and effectively write XML
# invalid file.
for err, no in invalidstat.items():
print err + " : " + str(no)
invalid_xml_filehandler.write(invalid_xmldoc.toprettyxml())
invalid_xml_filehandler.close()
if len(invalidstat.values()) == 0:
print "No XML errors found in " + get_mailfolder()
else:
print "XML file with detailed error info written to " \
