added script to register DOIs for SOMAR

somar/somar_register_dois.py

@@ -0,0 +1,187 @@
+import sys
+import pandas as pd
+import requests
+import json
+import base64
+import os
+from datetime import datetime
+
+# set variables.
+working_dir = os.path.dirname(os.path.abspath(__file__))
+draft_dois = os.path.join(working_dir, "draft_dois.csv")
+minted_dois = os.path.join(working_dir, "minted_dois.csv")
+xml_dir = os.path.join(working_dir, "XML")
+
+try:
+    if sys.argv[1].lower() == 'test':
+        credentials = os.path.join(working_dir, "test_credentials.json")
+    elif sys.argv[1].lower() == 'prod':
+        credentials = os.path.join(working_dir, "prod_credentials.json")
+    else:
+        print('\n\nWARNING: script only accepts "prod" or "test" as first argument.')
+        sys.exit(1)
+
+except IndexError:
+    print('\n\nUsage: python somar_dois.py <prod OR test>')
+    sys.exit(1)
+
+#make sure we have our draft_dois CSV in our working dir
+for fi in [credentials, draft_dois, xml_dir]:
+    if not os.path.exists(fi):
+        print(f'\n\nWARNING: "{fi}" does not exist. Please add and then re-run script.')
+        sys.exit(1)
+
+#read in our credential info
+with open(credentials, "r") as file:
+    credential_info = json.load(file)
+
+#set auth variables
+try:
+    user = credential_info['user']
+    password = credential_info['password']
+    prefix = credential_info['prefix']
+    datacite_url = credential_info['url']
+except KeyError:
+    print(f'\n\nWARNING: make sure that {credentials} includes "user", "password", "prefix", and "url" values.')
+    sys.exit(1)
+
+#Read in data.
+somar = pd.read_csv(draft_dois)
+
+# Clean up dataframe by removing 'Unnamed' columns
+somar = somar.loc[:, ~somar.columns.str.contains('^Unnamed')]
+
+# Set up additional columns, if needed
+if 'DOI' in somar.columns and somar['DOI'].dtype != 'object':
+    somar['DOI'] = somar['DOI'].astype('object')
+elif 'DOI' not in somar.columns:
+    somar['DOI'] = pd.Series(dtype='object')
+
+if 'Result' not in somar.columns:
+    somar['Result'] = pd.Series(dtype='object')
+
+#Additional vars: an empty list for suffixes and an index counter, set to 0 
+new_doi_suffixes = []
+
+# loop through the dataframe
+print('Gathering info and minting DOIs...')
+
+for index, row in somar.iterrows():
+
+    print(f'\n - Working on {row["StudyTitle"]}')
+
+    # skip item if it has already been added
+    if pd.notna(row['Result']) and 'Success' in str(row['Result']):
+        print('\tAlready added to DataCite; moving on to next item!')
+        continue
+
+    #extract the item ID from its SOMAR URL
+    item_url = row['URL']
+    item_id =item_url.rsplit('/', 1)[1]
+
+    #set up our JSON data to mint the DOI
+    print('\n\tMinting DOI...')
+    data_content = {
+        "data": {
+            "type": "dois",
+            "attributes": {
+                "prefix": prefix,
+                "titles": [
+                    {
+                        "title": row['StudyTitle']
+                    }
+                ],
+                "publisher": "ICPSR - Interuniversity Consortium for Political and Social Research",
+                "publicationYear": str(datetime.now().year),
+                "url": item_url,
+                "schemaVersion": "http://datacite.org/schema/kernel-4"
+            }
+        }
+    }
+
+    data = json.dumps(data_content).replace('\n', '').replace('\r', '').encode()
+    headers = {
+        'content-type': 'application/vnd.api+json',
+    }
+
+    response = requests.post(
+        datacite_url,
+        headers=headers,
+        data=data,
+        auth=(user, password)
+    )
+
+    #Check to make sure we got a successful ('201') response code.
+    if str(response.status_code) == '201':
+        print('\t  - DOI succeeded!')
+
+        # assign variables: locally and in our dataframe
+        item_doi = response.json()['data']['attributes']['doi']
+        somar.loc[index, 'DOI'] = item_doi
+
+    else:
+        print(f'\t  - DOI failed; code {response.status_code}')
+        print(f'\t  - {response.json()}')
+        somar.loc[index, 'Result'] = 'Failure'
+        continue
+
+    # proceed to update the metadata for our new DataCite object
+    print('\n\tUpdating metadata...')
+
+    # First, make sure the XML file exists; if not, provide error message and move on to next item
+    xml_file = os.path.join(xml_dir, row['XMLFile'])
+    if not os.path.exists(xml_file):
+        print(f'\t  - WARNING: {xml_file} does not exist; verify file and associated name (and update draft_dois.csv, if necessary).')
+        somar.loc[index, 'Result'] = 'Failure'
+        continue
+
+    # base64 encode the XML so it can be included in our API call
+    with open(xml_file, 'rb') as file:
+        xml_encoded = base64.b64encode(file.read()).decode('utf-8')
+
+    #create update json. Key things here: we have to pass the doi, we have to have a publish event, we have to pass the now-encoded xml.
+    payload = {
+        'data': {
+            'type': 'dois', 
+            'attributes': {
+                'event': 'publish', 
+                'identifiers': {
+                    'identifier': item_doi, 
+                    'identifierType': 'DOI'
+                }, 
+                'doi': item_doi, 
+                'xml': xml_encoded
+            }
+        }
+    }
+
+    #set up vars
+    data = json.dumps(payload).replace('\n', '').replace('\r', '').encode()
+    headers = {
+        'Content-Type': 'application/vnd.api+json',
+    }
+
+    # submit API request
+    response = requests.put(
+        f'{datacite_url}/{item_doi}',
+        headers=headers,
+        data=data,
+        auth=(user, password),
+    )
+
+    #Response code should be 200; document outcome in dataframe
+    if str(response.status_code) == "200":
+        print('\t  - Metadata update succeeded!')
+        somar.loc[index, 'Result'] = "Success"
+    else:
+        print(f'\t  - Metadata update failed; code {response.status_code}')
+        print(f'\t  - {response.json()}')
+        somar.loc[index, 'Result'] = 'Failure'
+        continue
+
+#Save the dataframe as a csv, for easiest copy/pasting into the google sheet. MAKE SURE TO KEEP THIS, YOU WILL NEED IT WHEN YOU UPDATE.
+somar.to_csv(minted_dois, index=False)
+try:
+    os.replace(minted_dois, draft_dois)
+except Exception as e:
+    print(f"Error replacing file: {e}")