This microservice exports ADS records with various formats including BibTex, AASTex, and multiple fielded and xml options.
$ virtualenv python
$ source python/bin/activate
$ pip install -r requirements.txt
$ pip install -r dev-requirements.txt
$ vim local_config.py # edit, edit
On your desktop run:
$ py.test
To get e.g. BibTeX for a set of records you do a POST request to the endpoint
https://api.adsabs.harvard.edu/v1/export/bibtex
within the POST header payload is defined as
{"bibcode":["<bibcode1>","<bibcode2>", ...]}
For example to get the BibTeX for the record with bibcode 2015ApJS..219...21Z, you would do
curl -H "Authorization: Bearer <your API token>" -H "Content-Type: application/json" -X POST -d '{"bibcode":["2015ApJS..219...21Z"]}' https://api.adsabs.harvard.edu/v1/export/bibtex
and the API then responds in JSON with
{u'msg': u'Retrieved 1 abstracts, starting with number 1. Total number selected: 1.', u'export': u'@ARTICLE{2015ApJS..219...21Z,\n author = {{Zhang}, M. and {Fang}, M. and {Wang}, H. and {Sun}, J. and \n\t{Wang}, M. and {Jiang}, Z. and {Anathipindika}, S.},\n title = "{A Deep Near-infrared Survey toward the Aquila Molecular Cloud. I. Molecular Hydrogen Outflows}",\n journal = {\\apjs},\narchivePrefix = "arXiv",\n eprint = {1506.08372},\n primaryClass = "astro-ph.SR",\n keywords = {infrared: ISM, ISM: jets and outflows, shock waves, stars: formation, stars: winds, outflows},\n year = 2015,\n month = aug,\n volume = 219,\n eid = {21},\n pages = {21},\n doi = {10.1088/0067-0049/219/2/21},\n adsurl = {http://adsabs.harvard.edu/abs/2015ApJS..219...21Z},\n adsnote = {Provided by the SAO/NASA Astrophysics Data System}\n}\n\n'}
Note that there is an optional parameter sort
for sorting the records from solr. The default sort option (if omitted) is by date desc, bibcode desc
. For example
curl -H "Authorization: Bearer <your API token>" -H "Content-Type: application/json" -X POST -d '{"bibcode":["2015ApJS..219...21Z", "2018AAS...23221409A"], "sort":"bibcode desc"}' https://api.adsabs.harvard.edu/v1/export/bibtex
To return the records in the exact order of input bibcode list, set sort
to no sort
.
Note that for endpoints /bibtex
and /bibtexabs
optional parameters maxauthor
, authorcutoff
, keyformat
and journalformat
can be passed in.
maxauthor
is maxinum number of authors displayed. The default values for maxauthor
for /bibtex
and /bibtexabs
respectivley are 10 and 0, where 0 means all.authorcutoff
is the threshold for truncating number of authors. If the number of authors is larger than authorcutoff
, author list is truncated and maxauthor
number of authors will be returned followed by et al.
. If authorcutoff
is not specified, the default of 200 is used.keyformat
allows user to customize bibtex key and could contain some combination of authors' last name(s), publication year, journal, and bibcode. User is now able to pick the key generation algorithm by specifying a custom format for it. To provide a specific example, this is our default format for 2019AAS...23338108A:
@INPROCEEDINGS{2019AAS...23338108A,
author = {{Accomazzi}, Alberto and {Kurtz}, Michael J. and {Henneken}, Edwin and
...
In addition keyformat
accepts specifier %zm
to enumerate keys (one character alphabet) should duplicate keys get created. Note that if enumeration specifier is not included, even if duplicate key are found, service does not enumerate the keys.
Now user can define one of the following:
Accomazzi:2019 -- %1H:%Y
Accomazzi:2019:AAS -- %1H:%Y:%q
Accomazzi2019 -- %1H%Y
Accomazzi2019AAS -- %1H%Y%q
AccomazziKurtz2019 -- %2H%Y
Note that for endpoint /rss
an optional parameter link
can be passed in. link
is the query url that generated the bibcodes.
For example querying ADS Bumblebee for
author:"accomazzi" and year:2018
at the end of March 2018, returns the following 4 bibcodes, and generates the url as specified below:
{"bibcode":["2018arXiv180303598K","2018arXiv180101021FB","2018AAS...23136217A","2018AAS...23130709A"], "link":"https://ui.adsabs.harvard.edu/#search/q=author%3A%22accomazzi%22%20and%20year%3A2018&sort=date%20desc%2C%20bibcode%20desc"}
which can be passed as payload to /rss
endpoint.
Note that for the endpoints /aastex
there is an optional parameter journalformat
which allows user to decide on the format of journal name. 1
indicates to use AASTeX macros if there are any (default), otherwise full journal name is exported. 2
means use journal abbreviations and 3
means use full journal name.
The /csl endpoint allows various citation style language (CSL) and output formats to be defined in the payload as follows:
{"bibcode":["1980ApJS...44..137K","1980ApJS...44..489B"], "style":"", "format":"", "sort":"date desc, bibcode, desc"}
The style parameter can be any of the CSL formats listed in sections 3 and 4 (e.g., aastex, icarus, mnras, soph, aspc, apsj, ieee, agu, gsa, ams). The format specifies the output type: 1 (Unicode), 2 (HTML), or 3 (LaTeX). Additionally, there is an optional journalformat parameter for certain styles (aastex, aspc, and aasj) to specify the journal name format, with values 1 (AASTeX macros, default), 2 (journal abbreviations), or 3 (full journal name).
Note that, as mentioned above, all the CSL formats can also be accessed through their dedicated endpoints, as listed earlier, using a payload like:
{"bibcode":["1980ApJS...44..137K","1980ApJS...44..489B"], "sort":"date desc, bibcode, desc"}
{"bibcode":["1980ApJS...44..137K","1980ApJS...44..489B"], "format":"%ZEncoding:latex%ZLinelength:0\bibitem[%4m(%Y)]{%R} %5.3l\ %Y, %j, %V, %p.\n"}
GET endpoints are similar to the POSTS endpoints. The curl
command has the following syntax:
curl -H "Authorization: Bearer <your API token>" <endpoint>/<bibcode>`
curl -H "Authorization: Bearer <your API token>" -H "Content-Type: application/json" -X POST -d <payload> <endpoint>
where <endpoint>
is /convert
and <payload>
is formatted as:
{"format":
For example:
curl -H "Authorization: Bearer <your API token>" -H "Content-Type: application/json" -X POST -d '{"format":"\\\\bibitem[%\\2m%(y)]\\{%za1%y} %\\8l %\\Y,%\\j,%\\V,%\\p"}' https://api.adsabs.harvard.edu/v1/export/convert
A new optional parameter, authorlimit
, has been introduced to limit the number of authors and their affiliations returned in exported records. This parameter addresses the growing issue of excessively large author lists, which are becoming increasingly common, with some papers featuring thousands of authors.
The authorlimit
parameter specifies the maximum number of authors and affiliations to retrieve from Solr and format in the service.
This parameter will be utilized in the ADS UI. However, when using the API, it is optional. If omitted, the service will return all authors and affiliations.
{"bibcode": ["1980ApJS...44..137K", "1980ApJS...44..489B"], "authorlimit": 500}
To retrieve all records without this limit using the API, you can use the provided script: https://github.com/adsabs/export_service/blob/master/export_with_api.py.
Golnaz, Edwin