X-Git-Url: https://code.communitydata.science/covid19.git/blobdiff_plain/36167295ecfe297780be2ec42ef1213e26d014e0..207b1f8b95096f2dcde913f7556003ad59e41123:/transliterations/src/wikidata_search.py

diff --git a/transliterations/src/wikidata_search.py b/transliterations/src/wikidata_search.py
index 792d61f..a3abbc0 100644
--- a/transliterations/src/wikidata_search.py
+++ b/transliterations/src/wikidata_search.py
@@ -3,25 +3,30 @@ from os import path
 from sys import stdout
 from wikidata_api_calls import search_wikidata, get_wikidata_api
 import csv
+from itertools import chain
 
 class Wikidata_ResultSet:
     def __init__(self):
         self.results = []
 
     def extend(self, term, results):
-        self.results.extend([Wikidata_Result(term, result, i)
-                                    for i, result in enumerate(results)])
+        self.results.append(
+            (Wikidata_Result(term, result, i)
+             for i, result in enumerate(results))
+        )
 
-    def to_csv(self, outfile=None):
+    def to_csv(self, outfile=None, mode='w'):
         if outfile is None:
             of = stdout
 
         else:
-            of = open(outfile,'w',newline='')
-
+            if path.exists(outfile) and mode != 'w':
+                of = open(outfile,'a',newline='')
+            else:
+                of = open(outfile,'w',newline='')
         writer = csv.writer(of)
         writer.writerow(Wikidata_Result.__slots__)
-        writer.writerows(map(Wikidata_Result.to_list, self.results))
+        writer.writerows(map(Wikidata_Result.to_list, chain(* self.results)))
 
 
 class Wikidata_Result:
@@ -46,20 +51,45 @@ class Wikidata_Result:
                 self.search_position,
                 self.timestamp]
     
-def run_wikidata_searches(terms_file = '../data/input/base_terms.txt', outfile="../data/output/wikidata_search_results.csv"):
-
+def run_wikidata_searches(terms):
+    api = get_wikidata_api()
     resultset = Wikidata_ResultSet()
-    for term in open(terms_file,'r'):
-        api = get_wikidata_api()
+    for term in terms:
         search_results = search_wikidata(api, term)
         resultset.extend(term, search_results)
+    return resultset
+
+def read_google_trends_files(terms_files):
+    def _read_file(infile):
+        return csv.DictReader(open(infile,'r',newline=''))
+
+    for row in chain(* [_read_file(terms_file) for terms_file in terms_files]):
+        yield row['query']
+
 
-    resultset.to_csv(outfile)
+def trawl_google_trends(terms_files, outfile = None, mode='w'):
+    terms = list(read_google_trends_files(terms_files))
+    resultset = run_wikidata_searches(terms)
+    resultset.to_csv(outfile, mode)
 
+def trawl_base_terms(infiles, outfile = None, mode='w'):
+    terms = list(chain(* (open(infile,'r') for infile in infiles)))
+    resultset = run_wikidata_searches(terms)
+    resultset.to_csv(outfile, mode)
 
     ## search each of the base terms in wikidata
 
     # store unique entities found in the search results, the position in the search result, and the date
 
 if __name__ == "__main__":
-    run_wikidata_searches()
+    import argparse
+    parser = argparse.ArgumentParser("Search wikidata for entities related to a set of terms.")
+    parser.add_argument('inputs', type=str, nargs='+', help='one or more files to read')
+    parser.add_argument('--use-gtrends', action='store_true', help = 'toggle whether the input is the output from google trends')
+    parser.add_argument('--output', type=str, help='an output file. defaults to stdout')
+    parser.add_argument('--overwrite', action='store_true', help = 'overwrite existing output files instead of appending')
+    args = parser.parse_args()
+    if args.use_gtrends:
+        trawl_google_trends(args.inputs, args.output)
+    else:
+        trawl_base_terms(args.inputs, args.output)