Merge branch 'master' of github.com:DocNow/twarc

DocNow · Aug 28, 2017 · ece85a2 · ece85a2
2 parents ce6e1be + 8ea9569
commit ece85a2
Show file tree

Hide file tree

Showing 2 changed files with 27 additions and 6 deletions.
diff --git a/utils/json2csv.py b/utils/json2csv.py
@@ -5,18 +5,36 @@
 CSV columns. If you'd like it adjusted send a pull request!
 """
 
-import csv
 import sys
 import json
+import codecs
+import argparse
 import fileinput
 
 if sys.version_info[0] < 3:
-    sys.exit("Sorry, json2csv.py requires python v3")
+    try:
+        import unicodecsv as csv
+    except ImportError:
+        sys.exit("unicodecsv is required for python 2")
+else:
+    import csv
 
 def main():
-    sheet = csv.writer(sys.stdout)
+    parser = argparse.ArgumentParser()
+    parser.add_argument('--output', '-o', help='write output to file instead of stdout')
+    parser.add_argument('files', metavar='FILE', nargs='*', help='files to read, if empty, stdin is used')
+    args = parser.parse_args()
+
+    if args.output:
+        sheet = csv.writer(codecs.open(args.output, 'wb', 'utf-8'))
+    else:
+        sheet = csv.writer(sys.stdout)
+
     sheet.writerow(get_headings())
-    for line in fileinput.input():
+
+    files = args.files if len(args.files) > 0 else ('-',)
+    for line in fileinput.input(files, openhook=fileinput.hook_encoded("utf-8")):
+
         tweet = json.loads(line)
         sheet.writerow(get_row(tweet))
 

diff --git a/utils/wordcloud.py b/utils/wordcloud.py
@@ -2,6 +2,7 @@
 
 from __future__ import print_function
 import re
+import sys
 import json
 import fileinput
 
@@ -48,7 +49,7 @@
 
 wordcloud_js = urlopen('https://raw.githubusercontent.com/jasondavies/d3-cloud/master/build/d3.layout.cloud.js').read()
 
-print(("""<!DOCTYPE html>
+output = """<!DOCTYPE html>
 <html>
 <head>
 <meta charset="utf-8">
@@ -93,4 +94,6 @@
 </script>
 </body>
 </html>
-""" % (wordcloud_js.decode('utf8'), json.dumps(words, indent=2))).encode('utf8'))
+""" % (wordcloud_js.decode('utf8'), json.dumps(words, indent=2))
+
+sys.stdout.write(output)