annotate scripts/search/tab2json @ 14:be7787c36e58 default tip

new: nofity LGSercies for deleted files
author Zoe Hong <zhong@mpiwg-berlin.mpg.de>
date Mon, 02 Nov 2015 16:41:23 +0100
parents a50cf11e5178
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
10
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
1 #!/usr/bin/env python
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
2 import sys
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
3 from optparse import OptionParser
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
4 import csv
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
5 try:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
6 import json
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
7 except ImportError:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
8 import simplejson as json
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
9
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
10 parser = OptionParser()
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
11 options, args = parser.parse_args()
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
12
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
13 if args:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
14 csv_file = open(args[0])
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
15 else:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
16 csv_file = sys.stdin
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
17
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
18 reader = csv.DictReader(csv_file, delimiter="\t")
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
19 rows = [row for row in reader]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
20 for row in rows:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
21 if "name" in row and "alias" in row and row["id"] == "1":
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
22 del row["id"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
23 del row["owner"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
24 row["permissionRoot"] = "false"
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
25 elif "title" in row:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
26 row["@type"] = "dataset"
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
27 row["files"] = []
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
28 row["versions"] = []
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
29 del row["id"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
30 del row["owner"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
31 del row["title"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
32 del row["author"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
33 del row["citationDate"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
34 del row["distributor"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
35 elif "contentType" in row:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
36 del row["id"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
37 row["permissionRoot"] = "false"
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
38 #print "must be a file..."
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
39 dataset_id = row["dataset"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
40 #row["dataset"] = {"id": dataset_id}
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
41 else:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
42 del row["id"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
43 row["permissionRoot"] = "false"
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
44 del row["owner"]
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
45 #if row["id"] == "1":
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
46 print json.dumps(row)
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
47 csv_file.close()
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
48 # sample dataverse file:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
49 #id name alias owner contactEmail description affiliation
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
50 #2 Birds birds 1 birds@birds.com A birds dataverse Birds Inc.
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
51 # sample dataset file:
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
52 #id title author owner description citationDate distributor
a50cf11e5178 Rewrite LGDataverse completely upgrading to dataverse4.0
Zoe Hong <zhong@mpiwg-berlin.mpg.de>
parents:
diff changeset
53 #1 birdstudy1 Dr. Finch 1 About birds 2013-12-11 For the Birds