-;;; doi-utils.el --- DOI utilities for making bibtex entries and downloading PDFs -*- lexical-binding: t; -*-
+;;; doi-utils.el --- DOI utilities for making bibtex entries
;; Copyright (C) 2015 John Kitchin
;; Author: John Kitchin <jkitchin@andrew.cmu.edu>
;; Keywords: convenience
+;; Version: 0.1
+;; Package-Requires: ((org-ref))
;; This program is free software; you can redistribute it and/or modify
;; it under the terms of the GNU General Public License as published by
;;; Commentary:
-;; This package provides functionality to download PDFs and bibtex entries from a DOI, as well as to update a bibtex entry from a DOI. It depends slightly on org-ref, to determine where to save pdf files too, and where to insert bibtex entries in the default bibliography.
+;; This package provides functionality to download PDFs and bibtex entries from a DOI, as well as to update a bibtex entry from a DOI. It depends slightly on org-ref, to determine where to save pdf files too, and where to insert bibtex entries in the default bibliography.
;; The principle commands you will use from here are:
;; - doi-utils-add-bibtex-entry-from-region to add an entry from a highlighed doi to your default bibliography.
;; - doi-utils-update-bibtex-entry-from-doi with cursor in an entry to update its fields.
-;; Package-Requires: ((org-ref))
-
(require 'json)
;;; Code:
;; There are some subtleties in doing this that are described here. To get the redirect, we have to use url-retrieve, and a callback function. The callback does not return anything, so we communicate through global variables. url-retrieve is asynchronous, so we have to make sure to wait for it to finish.
(defvar *doi-utils-waiting* t
- "stores waiting state for url retrieval.")
+ "Stores waiting state for url retrieval.")
(defvar *doi-utils-redirect* nil
- "stores redirect url from a callback function")
+ "Stores redirect url from a callback function.")
(defun doi-utils-redirect-callback (&optional status)
- "callback for url-retrieve to set the redirect"
+ "Callback for `url-retrieve' to set the redirect.
+Optional argument STATUS Unknown why this is optional."
(when (plist-get status :error)
(signal (car (plist-get status :error)) (cdr(plist-get status :error))))
(when (plist-get status :redirect) ; is nil if there none
;; To actually get the redirect we use url-retrieve like this.
(defun doi-utils-get-redirect (doi)
- "get redirect url from dx.doi.org/doi"
+ "Get redirect url from dx.DOI.org/doi."
;; we are going to wait until the url-retrieve is done
(setq *doi-utils-waiting* t)
;; start with no redirect. it will be set in the callback.
;; Once we have a redirect for a particular doi, we need to compute the url to the pdf. We do this with a series of functions. Each function takes a single argument, the redirect url. If it knows how to compute the pdf url it does, and returns it. We store the functions in a variable:
(defvar doi-utils-pdf-url-functions nil
- "list of functions that return a url to a pdf from a redirect url. Each function takes one argument, the redirect url. The function must return a pdf-url, or nil.")
+ "List of functions that return a url to a pdf from a redirect url. Each function takes one argument, the redirect url. The function must return a pdf-url, or nil.")
;; ** APS journals
(defun aps-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://journals.aps.org" *doi-utils-redirect*)
(replace-regexp-in-string "/abstract/" "/pdf/" *doi-utils-redirect*)))
;; ** Science
(defun science-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://www.sciencemag.org" *doi-utils-redirect*)
(concat *doi-utils-redirect* ".full.pdf")))
;; ** Nature
(defun nature-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://www.nature.com" *doi-utils-redirect*)
(let ((result *doi-utils-redirect*))
(setq result (replace-regexp-in-string "/full/" "/pdf/" result))
(defun doi-utils-get-wiley-pdf-url (redirect-url)
- "wileyscience direct hides the pdf url in html. we get it out here"
+ "Wileyscience direct hides the pdf url in html.
+We get it out here by parsing the html.
+Argument REDIRECT-URL URL you are redirected to."
(setq *doi-utils-waiting* t)
(url-retrieve redirect-url
(lambda (status)
- (beginning-of-buffer)
+ (goto-char (point-min))
(re-search-forward "<iframe id=\"pdfDocument\" src=\"\\([^\"]*\\)\"" nil)
(setq *doi-utils-pdf-url* (match-string 1)
- ,*doi-utils-waiting* nil)))
+ *doi-utils-waiting* nil)))
(while *doi-utils-waiting* (sleep-for 0.1))
*doi-utils-pdf-url*)
(defun wiley-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://onlinelibrary.wiley.com" *doi-utils-redirect*)
(doi-utils-get-wiley-pdf-url
(replace-regexp-in-string "/abstract" "/pdf" *doi-utils-redirect*))
;; ** Springer
(defun springer-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://link.springer.com" *doi-utils-redirect*)
(replace-regexp-in-string "/article/" "/content/pdf/" (concat *doi-utils-redirect* ".pdf"))))
;; we just change /abs/ to /pdf/.
(defun acs-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://pubs.acs.org" *doi-utils-redirect*)
(replace-regexp-in-string "/abs/" "/pdf/" *doi-utils-redirect*)))
;; ** IOP
(defun iop-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://iopscience.iop.org" *doi-utils-redirect*)
(let ((tail (replace-regexp-in-string
"^http://iopscience.iop.org" "" *doi-utils-redirect*)))
;; ** JSTOR
(defun jstor-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://www.jstor.org" *doi-utils-redirect*)
(concat (replace-regexp-in-string "/stable/" "/stable/pdfplus/" *doi-utils-redirect*) ".pdf")))
;; ** AIP
(defun aip-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://scitation.aip.org" *doi-utils-redirect*)
;; get stuff after content
(let (p1 p2 s p3)
- (setq p2 (replace-regexp-in-string "^http://scitation.aip.org/" "" *doi-utils-redirect*))
+ (setq p2 (replace-regexp-in-string
+ "^http://scitation.aip.org/" "" *doi-utils-redirect*))
(setq s (split-string p2 "/"))
(setq p1 (mapconcat 'identity (-remove-at-indices '(0 6) s) "/"))
(setq p3 (concat "/" (nth 0 s) (nth 1 s) "/" (nth 2 s) "/" (nth 3 s)))
;; ** Taylor and Francis
(defun tandfonline-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://www.tandfonline.com" *doi-utils-redirect*)
(replace-regexp-in-string "/abs/\\|/full/" "/pdf/" *doi-utils-redirect*)))
;; ** ECS
(defun ecs-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://jes.ecsdl.org" *doi-utils-redirect*)
(replace-regexp-in-string "\.abstract$" ".full.pdf" *doi-utils-redirect*)))
(defun ecst-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://ecst.ecsdl.org" *doi-utils-redirect*)
(concat *doi-utils-redirect* ".full.pdf")))
;; ** RSC
(defun rsc-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://pubs.rsc.org" *doi-utils-redirect*)
(let ((url (downcase *doi-utils-redirect*)))
(setq url (replace-regexp-in-string "articlelanding" "articlepdf" url))
;; You cannot compute these pdf links; they are embedded in the redirected pages.
(defvar *doi-utils-pdf-url* nil
- "stores url to pdf download from a callback function")
+ "Stores url to pdf download from a callback function.")
(defun doi-utils-get-science-direct-pdf-url (redirect-url)
- "science direct hides the pdf url in html. we get it out here"
+ "Science direct hides the pdf url in html. W get it out here.
+REDIRECT-URL is where the pdf url will be in."
(setq *doi-utils-waiting* t)
(url-retrieve redirect-url
(lambda (status)
(beginning-of-buffer)
(re-search-forward "pdfurl=\"\\([^\"]*\\)\"" nil t)
(setq *doi-utils-pdf-url* (match-string 1)
- ,*doi-utils-waiting* nil)))
+ *doi-utils-waiting* nil)))
(while *doi-utils-waiting* (sleep-for 0.1))
*doi-utils-pdf-url*)
(defun science-direct-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://www.sciencedirect.com" *doi-utils-redirect*)
(doi-utils-get-science-direct-pdf-url *doi-utils-redirect*)
*doi-utils-pdf-url*))
;; which actually redirect to
;; http://www.sciencedirect.com/science/article/pii/S0927025609004558
(defun linkinghub-elsevier-pdf-url (*doi-utils-redirect*)
- (when (string-match "^http://linkinghub.elsevier.com/retrieve" *doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
+ (when (string-match
+ "^http://linkinghub.elsevier.com/retrieve" *doi-utils-redirect*)
(let ((second-redirect (replace-regexp-in-string
"http://linkinghub.elsevier.com/retrieve"
"http://www.sciencedirect.com/science/article"
- ,*doi-utils-redirect*)))
+ *doi-utils-redirect*)))
(message "getting pdf url from %s" second-redirect)
*doi-utils-pdf-url*)))
;; http://www.pnas.org/content/early/2014/05/08/1319030111.full.pdf+html?with-ds=yes
(defun pnas-pdf-url (*doi-utils-redirect*)
+ "Get url to the pdf from *DOI-UTILS-REDIRECT*."
(when (string-match "^http://www.pnas.org" *doi-utils-redirect*)
(concat *doi-utils-redirect* ".full.pdf?with-ds=yes")))
;; ** Get the pdf url for a doi
(defun doi-utils-get-pdf-url (doi)
- "returns a url to a pdf for the doi if one can be
-calculated. Loops through the functions in `doi-utils-pdf-url-functions'
-until one is found"
+ "Return a url to a pdf for the DOI if one can be calculated.
+Loops through the functions in `doi-utils-pdf-url-functions'
+until one is found."
(doi-utils-get-redirect doi)
(unless *doi-utils-redirect*
;; ** Finally, download the pdf
(defun doi-utils-get-bibtex-entry-pdf ()
- "download pdf for entry at point if the pdf does not already
-exist locally. The entry must have a doi. The pdf will be saved
+ "Download pdf for entry at point if the pdf does not already exist locally.
+The entry must have a doi. The pdf will be saved
to `org-ref-pdf-directory', by the name %s.pdf where %s is the
-bibtex label. Files will not be overwritten. The pdf will be
+bibtex label. Files will not be overwritten. The pdf will be
checked to make sure it is a pdf, and not some html failure
-page. you must have permission to access the pdf. We open the pdf
+page. you must have permission to access the pdf. We open the pdf
at the end."
(interactive)
(save-excursion
;; I [[http://homepages.see.leeds.ac.uk/~eeaol/notes/2013/02/doi-metadata/][found]] you can download metadata about a DOI from http://dx.doi.org. You just have to construct the right http request to get it. Here is a function that gets the metadata as a plist in emacs.
(defun doi-utils-get-json-metadata (doi)
- "Try to get json metadata for DOI. Open the DOI in a browser if we do not get it."
+ "Try to get json metadata for DOI. Open the DOI in a browser if we do not get it."
(let ((url-request-method "GET")
(url-mime-accept-string "application/citeproc+json")
(json-object-type 'plist)
(if (string-match "Resource not found" json-data)
(progn
(browse-url (concat "http://dx.doi.org/" doi))
- (error "Resource not found. Opening website."))
+ (error "Resource not found. Opening website"))
(json-read-from-string json-data)))))
;; We can use that data to construct a bibtex entry. We do that by defining a template, and filling it in. I wrote this template expansion code which makes it easy to substitute values like %{} in emacs lisp.
(defun doi-utils-expand-template (s)
- "expand a template containing %{} with the eval of its contents"
+ "Expand a string template S containing %{} with the eval of its contents."
(replace-regexp-in-string "%{\\([^}]+\\)}"
(lambda (arg)
(let ((sexp (substring arg 2 -1)))
(setq doi-utils-bibtex-type-generators nil)
(defun doi-utils-concat-prepare (lst &optional acc)
- "Given a list `lst' of strings and other expressions, which are
-intented to passed to `concat', concat any subsequent strings,
+ "Minimize the number of args passed to `concat' from LST.
+Given a list LST of strings and other expressions, which are
+intended to be passed to `concat', concat any subsequent strings,
minimising the number of arguments being passed to `concat'
-without changing the results."
+without changing the results. ACC is the list of additional
+expressions."
(cond ((null lst) (nreverse acc))
((and (stringp (car lst))
(stringp (car acc)))
(t (doi-utils-concat-prepare (cdr lst) (cons (car lst) acc)))))
(defmacro doi-utils-def-bibtex-type (name matching-types &rest fields)
- "Define a BibTeX type identified by (symbol) `name' with
-`fields' (given as symbols), matching to retrieval expressions in
-`doi-utils-json-metadata-extract'. This type will only be used
+ "Define a BibTeX type identified by (symbol) NAME with
+FIELDS (given as symbols), matching to retrieval expressions in
+`doi-utils-json-metadata-extract'. This type will only be used
when the `:type' parameter in the JSON metadata is contained in
-`matching-types' - a list of strings."
+MATCHING-TYPES - a list of strings."
`(push (lambda (type results)
- (when (or ,@(mapcar (lambda (match-type) `(string= type ,match-type)) matching-types))
+ (when
+ (or ,@(mapcar
+ (lambda (match-type)
+ `(string= type ,match-type)) matching-types))
(let ,(mapcar (lambda (field)
- (let ((field-expr (assoc field doi-utils-json-metadata-extract)))
+ (let ((field-expr
+ (assoc field doi-utils-json-metadata-extract)))
(if field-expr
;; need to convert to string first
`(,(car field-expr) (format "%s" ,(cadr field-expr)))
- (error "unknown bibtex field type %s" field))))
+ (error "Unknown bibtex field type %s" field))))
fields)
(concat
,@(doi-utils-concat-prepare
;; With the code generating the bibtex entry in place, we can glue it to the json retrieval code.
(defun doi-utils-doi-to-bibtex-string (doi)
- "return a bibtex entry as a string for the doi. Not all types are supported yet."
+ "Return a bibtex entry as a string for the DOI. Not all types are supported yet."
(let* ((results (doi-utils-get-json-metadata doi))
(type (plist-get results :type)))
;(format "%s" results) ; json-data
(defun doi-utils-insert-bibtex-entry-from-doi (doi)
- "insert bibtex entry from a doi. Also cleans entry using
-org-ref, and tries to download the corresponding pdf."
+ "Insert bibtex entry from a DOI.
+Also cleans entry using org-ref, and tries to download the corresponding pdf."
(interactive "sDOI :")
(insert (doi-utils-doi-to-bibtex-string doi))
(backward-char)
kill-ring starts like a DOI, then that is the intial
prompt. Otherwise, you have to type or pste in a DOI."
(interactive
- (list (read-input "DOI: "
+ (list (read-string "DOI: "
;; now set initial input
(cond
;; If region is active and it starts like a doi we want it.
(defun bibtex-set-field (field value &optional nodelim)
- "set field to value in bibtex file. create field if it does not exist"
+ "Set FIELD to VALUE in bibtex file. create field if it does not exist.
+Optional argument NODELIM see `bibtex-make-field'."
(interactive "sfield: \nsvalue: ")
(bibtex-beginning-of-entry)
(let ((found))
(defun plist-get-keys (plist)
- "return keys in a plist"
- (loop
+ "Return keys in a PLIST."
+ (cl-loop
for key in results by #'cddr collect key))
(defun doi-utils-update-bibtex-entry-from-doi (doi)
- "update fields in a bibtex entry from the doi. Every field will be updated, so previous changes will be lost."
+ "Update fields in a bibtex entry from the DOI. Every field will be updated, so previous change will be lost."
(interactive (list
(or (replace-regexp-in-string "http://dx.doi.org/" "" (bibtex-autokey-get-field "doi"))
(read-string "DOI: "))))
(defun doi-utils-update-field ()
+ "Update the field at point in the bibtex entry.
+Data is retrieved from the doi in the entry."
(interactive)
(let* ((doi (bibtex-autokey-get-field "doi"))
(results (doi-utils-get-json-metadata doi))
;; These are pretty easy to construct, so we can write functions that will create them and open the url in our browser. There are some other options that could be considered, but since we usually have a doi, it seems like the best way to go for creating the links. Here are the functions.
(defun doi-utils-wos (doi)
- "Open Web of Science entry for DOI"
+ "Open Web of Science entry for DOI."
(interactive "sDOI: ")
(browse-url
(format
"http://ws.isiknowledge.com/cps/openurl/service?url_ver=Z39.88-2004&rft_id=info:doi/%s" doi)))
(defun doi-utils-wos-citing (doi)
- "Open Web of Science citing articles entry. May be empty if none are found"
+ "Open Web of Science citing articles entry for DOI.
+May be empty if none are found."
(interactive "sDOI: ")
(browse-url
(concat
"&svc_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Asch_svc&svc.citing=yes")))
(defun doi-utils-wos-related (doi)
- "Open Web of Science related articles page."
+ "Open Web of Science related articles page for DOI."
(interactive "sDOI: ")
(browse-url
(concat "http://ws.isiknowledge.com/cps/openurl/service?url_ver=Z39.88-2004&rft_id=info%3Adoi%2F"
(defun doi-utils-open (doi)
+ "Open DOI in browser."
(interactive "sDOI: ")
(browse-url (concat "http://dx.doi.org/" doi)))
(defun doi-utils-open-bibtex (doi)
- "Search through `reftex-default-bibliography' for DOI."
+ "Search through variable `reftex-default-bibliography' for DOI."
(interactive "sDOI: ")
(catch 'file
(dolist (f reftex-default-bibliography)
(defun doi-utils-google-scholar (doi)
- "Google scholar the word at point or selection."
+ "Google scholar the DOI."
(interactive "sDOI: ")
(browse-url
(format
(defun doi-utils-pubmed (doi)
- "Pubmed the word at point or selection."
+ "Search Pubmed for the DOI."
(interactive "sDOI: ")
(browse-url
(format
(defvar doi-link-menu-funcs '()
- "Functions to run in doi menu. Each entry is a list of (key menu-name function).
-The function must take one argument, the doi.")
+ "Functions to run in doi menu.
+Each entry is a list of (key menu-name function). The function
+must take one argument, the doi.")
(setq doi-link-menu-funcs
'(("o" "pen" doi-utils-open)
(defun doi-link-menu (link-string)
"Generate the link menu message, get choice and execute it.
-Options are stored in `doi-link-menu-funcs'."
+Options are stored in `doi-link-menu-funcs'.
+Argument LINK-STRING Passed in on link click."
(interactive)
(message
(concat
;; The idea here is to perform a query on Crossref, get a helm buffer of candidates, and select the entry(ies) you want to add to your bibtex file. You can select a region, e.g. a free form citation, or set of words, or you can type the query in by hand.
(defun doi-utils-add-entry-from-crossref-query (query bibtex-file)
+ "Search Crossref with QUERY and use helm to select an entry to add to BIBTEX-FILE."
(interactive (list
- (read-input
+ (read-string
"Query: "
;; now set initial input
(cond
projects / org-ref.git / blobdiff