This code is designed to interact with the WOS API using "premium" or "lite" access protocols.
wos.pycontains code for interacting with the API directly.woscalls.pyincludes calls made to theWOSclassmetawos.pycan be used to extract metadata from search results.buildsearch.pybuilds search strings from data supplied in a certain style of tsv file.
It's possible to work with the API class, WOS, directly using syntax like the following:
from wos import Wos
wos = Wos(client="Lite")
This will initiate the search client but not yet run any API calls. There are two options for the "client" keyword argument: "Search" and "Lite". The Lite API should be available to all institutions that subscribe to at least the Web of Knowledge Core Collection. Contact your representative to have access opened to your IP address, if it isn't already. More details here. The Lite API provides basic search and basic metadata retrieval functionality, but not citation retrieval.
The "Search" client is an available at an additional cost, and does provide access to both forward and backward citation data. Access is available on a project basis, or as a yearly subscription. I would recommend first requesting a trial period to work with the API.
wos.authorize()
The authorize function attempts to authenticate based on IP address. If successful, an authorization token will be attached to all future requests in the session. Output:
Search client authorized.
Next a new search can be established:
wos.query_parameters('AU=(Peiretti AND Palmegiano) AND SO=(Animal Feed Science and Technology) AND PY=2004', database_id="WOK")
To do so establish the query_parameters object, which should include a query string along with, optionally, a set of parameters. Details on the structure of the query can be found in the API documentation (which can be requested from ThomsonReutuers, but which I'm also making available here (in possibly an outdated version).
Parameters can be provided as the following keyword arguments:
- time_begin (str) -- date in YYYY-MM-DD format.
- time_end (str) -- date in YYYY-MM-DD format.
- database_id (str) -- from the WOS set of database abbreviations. "WOS" correpsonds to the WOS core collection.
- query_language (str) -- "en" the only currently allowed value.
- symbolic_timespan (str) -- a human-readable timespan, e.g. "4weeks", must be null if time_begin and time_end used.
- editions (list) -- TODO list of sub-components of the selected database to use.
Output:
(queryParameters){
databaseId = "WOK"
userQuery = "AU=(Peiretti AND Palmegiano) AND SO=(Animal Feed Science and Technology) AND PY=2004"
editions[] = <empty>
symbolicTimeSpan = None
timeSpan =
(timeSpan){
begin = "1900-01-01"
end = "2015-10-01"
}
queryLanguage = "en"
}
Next, establish what results to return with the retrieve_parameters object.
wos.retrieve_parameters()
The retrieve_parameters allow for some control of the data that is returned.
- first_record (int) -- The number of the first record to return in the search.
- count (int) -- Number of records to return (maximum 100).
- sort_field (list) -- TODO Field to sort by (should be WOS field abbreviation).
Output:
(retrieveParameters){
firstRecord = 1
count = 100
sortField[] = <empty>
}
You can then carry out the search, using the query and retrieval parameters you just created like this:
wos.search(wos.qp, wos.rp)
Output:
Found 1 Results for AU=(Peiretti AND Palmegiano) AND SO=(Animal Feed Science and Technology) AND PY=2004
Results can be found in the wos.search_results.records object, if any results were returned. More generally, wos.search_results can be used to find info about the response, including number of results. Output:
(searchResults){
queryId = "1"
recordsFound = 1
recordsSearched = 198371943
records[] =
(liteRecord){
uid = "WOS:000224567700011"
title[] =
(labelValuesPair){
label = "Title"
value[] =
"Chemical composition, organic matter digestibility and fatty acid content of evening primrose (Oenothera paradoxa) during its growth cycle",
},
source[] =
(labelValuesPair){
label = "Issue"
value[] =
"3-4",
},
(labelValuesPair){
label = "Pages"
value[] =
"293-299",
},
(labelValuesPair){
label = "Published.BiblioDate"
value[] =
"OCT 15",
},
(labelValuesPair){
label = "Published.BiblioYear"
value[] =
"2004",
},
(labelValuesPair){
label = "SourceTitle"
value[] =
"ANIMAL FEED SCIENCE AND TECHNOLOGY",
},
(labelValuesPair){
label = "Volume"
value[] =
"116",
},
authors[] =
(labelValuesPair){
label = "Authors"
value[] =
"Peiretti, PG",
"Palmegiano, GB",
"Masoero, G",
},
keywords[] =
(labelValuesPair){
label = "Keywords"
value[] =
"lipids",
"forage",
"foodstuffs",
"ruminant",
"nutrition",
},
other[] =
(labelValuesPair){
label = "Contributor.ResearcherID.Names"
value[] =
"Peiretti, Pier Giorgio",
"Peiretti, Pier Giorgio",
},
(labelValuesPair){
label = "Contributor.ResearcherID.ResearcherIDs"
value[] =
"B-6871-2013",
None,
},
(labelValuesPair){
label = "Identifier.Doi"
value[] =
"10.1016/j.anifeedsci.2004.07.001",
},
(labelValuesPair){
label = "Identifier.Ids"
value[] =
"863ND",
},
(labelValuesPair){
label = "Identifier.Issn"
value[] =
"0377-8401",
},
(labelValuesPair){
label = "Identifier.Xref_Doi"
value[] =
"10.1016/j.anifeedsci.2004.07.001",
},
(labelValuesPair){
label = "ResearcherID.Disclaimer"
value[] =
"ResearcherID data provided by Thomson Reuters",
},
},
}
Additional methods can be used to get cited references as well as citing articles if the "Search" client is used.
Code from above:
from wos import Wos
wos = Wos(client="Lite")
wos.authorize()
wos.query_parameters('AU=(Peiretti AND Palmegiano) AND SO=(Animal Feed Science and Technology) AND PY=2004', database_id="WOK")
wos.retrieve_parameters(view_field=["title", "name"])
wos.search(wos.qp, wos.rp)
The above code should return 1 result and can be used as a test to ensure search is working properly.
The WosCalls class provides a means of interacting with the API at a higher level. That is, lists of search strings or sets of search parameters can be provided to run in batch. This functionality is still very much work in progress.
from woscalls import WosCalls
wosc = WosCalls(search_queries=bs.searches, database_id="WOK")
wosc.get_all_search_results()
In this case the bs.searches object contains a Python list of 10 search strings, each of which is being carried out in sequence. Output:
Search client authorized.
Found 1 Results for AU=(Lambertsen) AND PY=1966 AND SO=(Acta Agric* Scand*)
Found 1 Results for AU=(Bentes) AND PY=1986 AND SO=(Acta Amazonica)
Found 1 Results for AU=(Maia) AND PY=1978 AND SO=(Acta Amazonica)
Found 0 Results for AU=(Loth) AND PY=1991 AND SO=(Agrochimica)
Found 1 Results for AU=(Jellum AND Powell) AND PY=1971 AND SO=(Agron* J*)
Found 5 Results for AU=(Bertoni) AND PY=1994 AND SO=(An* Asoc* Quim* Argent*)
Found 0 Results for AU=(Balnchini) AND PY=1981 AND SO=(Anal* Chem*)
Found 1 Results for AU=(Peiretti AND Palmegiano AND Masoero) AND PY=2004 AND SO=(Animal Feed Science and Technology)
Found 0 Results for AU=(Adhikari) AND PY=1991 AND SO=(Bangladesh J* Sci* Ind* Res*)
Found 0 Results for AU=(Serrano AND Guzm?n) AND PY=1994 AND SO=(Biochem* Systemat* Ecology)
Process complete.
Returned 10 records
The WosCalls class is additionally a place to house content-specific methods that build on the Wos clas. See run_phylo_process method as it develops.
The BuildSearch class is currently quite content specific but could in principle be broadened to allow for automatically generating searches from data in other formats, such as CSV or JSON. Currently the algorithm below assumes a very specific data structure to work.
from buildsearch import BuildSearch
bs = BuildSearch("data/ohlrogge/ohlrogge_test_10.txt")
bs.make_search_list() # from here the object bs.searches can be used in WosCalls()
Output:
File loaded.
Check bs.searches to view objects:
bs.searches
Output:
[u'AU=(Lambertsen) AND PY=1966 AND SO=(Acta Agric* Scand*)',
u'AU=(Bentes) AND PY=1986 AND SO=(Acta Amazonica)',
u'AU=(Maia) AND PY=1978 AND SO=(Acta Amazonica)',
u'AU=(Loth) AND PY=1991 AND SO=(Agrochimica)',
u'AU=(Jellum AND Powell) AND PY=1971 AND SO=(Agron* J*)',
u'AU=(Bertoni) AND PY=1994 AND SO=(An* Asoc* Quim* Argent*)',
u'AU=(Balnchini) AND PY=1981 AND SO=(Anal* Chem*)',
u'AU=(Peiretti AND Palmegiano AND Masoero) AND PY=2004 AND SO=(Animal Feed Science and Technology)',
u'AU=(Adhikari) AND PY=1991 AND SO=(Bangladesh J* Sci* Ind* Res*)',
u'AU=(Serrano AND Guzm?n) AND PY=1994 AND SO=(Biochem* Systemat* Ecology)']
This list of searches, suitable to pass as an argument in the WosCalls class.
If I can be of help in using this code, or if you have suggestions for improvement, please do contact me.