Update crawl_id.py

Adding some sources.
google · Nov 9, 2019 · 8ccc08b · 8ccc08b
1 parent 4c6cb38
commit 8ccc08b
Showing 1 changed file with 7 additions and 1 deletion.
diff --git a/Lib/corpuscrawler/crawl_id.py b/Lib/corpuscrawler/crawl_id.py
@@ -19,5 +19,11 @@
 
 def crawl(crawler):
     out = crawler.get_output(language='id')
-    crawl_udhr(crawler, out, filename='udhr_ind.txt')
+    crawler.crawl_abc_net_au(out, program_id='indonesian')
+    crawler.crawl_voice_of_america(out, host='voaindonesia.com')
+    crawl_bbc_news(crawler, out, urlprefix='/indonesia/')
     crawl_deutsche_welle(crawler, out, prefix='/id/')
+    crawl_udhr(crawler, out, filename='udhr_ind.txt')
+    crawl_bibleis(crawler, out, bible='INDASV')   
+    crawl_bibleis(crawler, out, bible='INDWBT')
+    crawl_bibleis(crawler, out, bible='INDSHV')