curl working but not request

[joelczinn]

Hello,

Thanks so much for this service. I would love to take advantage of it, but I'm running into an authorization issue, it seems.

I would expect the following code to work:

import requests
bibcodes="bibcode\n1907AN....174...59.\n1908PA.....16..445.\n1989LNP...334..242S"
r = requests.post('https://api.adsabs.harvard.edu/v1/search/bigquery',
    params={'q':'*:*', 'wt':'json', 'fq':'{!bitset}', 'fl':'bibcode'},
    headers={'Authorization': 'Bearer:TOKEN'},
    data=bibcodes)
print(r.text)

However, the result is:

'{\n  "error": "Unauthorized"\n}\n'

However, the equivalent request via curl:

curl 'https://api.adsabs.harvard.edu/v1/search/bigquery?q=*:*&fl=bibcode'  -X 'POST' -H 'Authorization: Bearer TOKEN' -H 'Content-Type: big-query/csv' -d $'bibcode\n1907AN....174...59.\n1908PA.....16..445.\n1989LNP...334..242S'

works and returns:

{
  "responseHeader":{
    "status":0,
    "QTime":31,
    "params":{
      "q":"*:*",
      "fl":"bibcode",
      "start":"0",
      "internal_logging_params":"X-Amzn-Trace-Id=Root=1-64fbb58b-52d35276622d99767e24944b",
      "fq":"{!bitset}",
      "rows":"10",
      "wt":"json"}},
  "response":{"numFound":3,"start":0,"docs":[
      {
        "bibcode":"1989LNP...334..242S"},
      {
        "bibcode":"1907AN....174...59."},
      {
        "bibcode":"1908PA.....16..445."}]
  }}

requests is v2.31.0 and Python is v3.8.6.

Thanks in advance for the help!

[tjacovich]

Hello,

Thank you for reaching out to ADS. Looking at the code you provided, I think the issue may be a subtle difference in formatting in this line here:

    headers={'Authorization': 'Bearer:TOKEN'},

The authentication methods are very sensitive to syntax, so this line should really be

    headers={'Authorization': 'Bearer TOKEN'},

like it is in the cURL.

    -H 'Authorization: Bearer TOKEN'

I hope this helps,

-Taylor

[joelczinn]

Hello,

I'm coming back to this to clarify that changing Bearer:TOKEN to Bearer TOKEN did not work... Are there any other ideas of what might be going on?

Thanks again!

[tjacovich]

Hello,

I took the python code you provided and ran it against python-3.8 and requests-v2.31.0 but could not reproduce the issue with my personal token. I modified the code slightly and confirmed this is working.

import requests
TOKEN=REPLACE_TOKEN
bibcodes="bibcode\n1907AN....174...59.\n1908PA.....16..445.\n1989LNP...334..242S"
r = requests.post('https://api.adsabs.harvard.edu/v1/search/bigquery',
    params={'q':'*:*', 'wt':'json', 'fq':'{!bitset}', 'fl':'bibcode'},
    headers={'Authorization': 'Bearer '+str(TOKEN)},
    data=bibcodes)
print(r.text)

For more detailed troubleshooting if this doesn't solve the issue, can you send an email with a detailed description of your current issue, as well as the goal of your queries, to adshelp[at]cfa.harvard.edu?

Thanks,

Taylor