-- Applescript for BibDesk -- Extract the DOI information from the PDF -- Last revision by L. Lim, 2012 tell application "BibDesk" -- without document, there is no selection, so nothing to do if (count of documents) = 0 then beep display dialog "No documents found." buttons {"¥"} default button 1 giving up after 3 end if set theDoc to document 1 tell theDoc set theSel to selection set thePub to item 1 of theSel get value of field "DOI" of thePub set pasteItem to result as string set theExistingDOI to pasteItem if theExistingDOI = "" then set theDOIString to "" tell thePub set theCount to count of (get linked files) --return theCount set theAlreadyLinkedFiles to " " if theCount = 0 then error if theCount > 0 then --set theAlreadyLinkedFiles to the POSIX path of (get linked files) set theAlreadyLinkedFiles to (get linked files) set theFirstLinkedFile to item 1 of theAlreadyLinkedFiles set theFirstFile to the POSIX path of theFirstLinkedFile -- Extract the DOI from the first linked file using pdftotext and sed set theDOIString to my findDOI(theFirstFile) --return theDOIString end if end tell else set theDOIString to theExistingDOI end if -- Now that we have a DOI, construct an ADS URL and a DOI URL from it; add to linked URL and linked files if theDOIString is not "" then -- If it's not already in there, copy to remote URL list set theDOIString2 to my removebrackets(theDOIString) -- Add the DOI URL to the list of URLs if it's not already there set urlName to "http://dx.doi.org/" & theDOIString2 set theExistingURLs to linked URLs of thePub if (theDOIString is not "") and (urlName is not in (theExistingURLs)) then add urlName to end of linked URLs of thePub -- Add the ADS URL to its own field (Adsurl) if it's not already there set theExistingAdsurl to value of field "Adsurl" of thePub if "adsabs" is not in theExistingAdsurl then set adsurlName to "http://adsabs.harvard.edu/doi/" & theDOIString2 set value of field "Adsurl" of thePub to adsurlName end if end if -- Now use the ADS URL to fill out the Bibdesk record for this publication. -- First save the existing values of every field so they won't be overwritten by the new ADS information: --- 1. construct "newPub" with the ADS values --- 2. copy thePub's linked URL and linked file lists to newPub, along with the abstract, if any --- 3. overwrite all fields of "thePub" from newPub --- 4. delete the now-redundant newPub. get value of field "Adsurl" of thePub set theAdsurl to result as string if theAdsurl is "" then beep set theMessage to "No Adsurl" display dialog theMessage buttons {"¥"} default button 1 giving up after 3 else -- Download the BibTeX information from ADS set theBibtexUrl to my getADSBibtexUrl(theAdsurl) --return theBibtexUrl --set TheCommand to "/sw/bin/wget -O - '" & urlName & "' 2>&1" --set TheCommand to "curl '" & theBibtexUrl & "'" set TheCommand to "curl " & the quoted form of theBibtexUrl do shell script TheCommand set theNewBibtexString to the result --return TheCommand --return theTitle if theNewBibtexString = "" then beep set theMessage to "No ADS BibTeX downloaded." & TheCommand display dialog theMessage buttons {"¥"} default button 1 giving up after 4 --return theNewBibtexString else --return theNewBibtexString -- Temporarily stash the ADS results in "newPub" set newPub to make new publication with properties {BibTeX string:theNewBibtexString} at end of publications -- Preserve the original entry's linked file list set theAlreadyLinkedFiles to linked files of thePub --return theAlreadyLinkedFiles tell newPub make new linked file with data theAlreadyLinkedFiles at end of linked files end tell -- Preserve the original entry's abstract, if it has one -- If the original entry doesn't already have an abstract, fill it from ADS set theExistingAbstract to the abstract of thePub if theExistingAbstract is not "" then set the abstract of newPub to theExistingAbstract else --set TheCommand to "curl '" & theAdsurl & "'" set theGoodAdsurl to my cleanADSUrl(theAdsurl) set theAbstract to my getADSAbstract(theGoodAdsurl) --return theAbstract if theAbstract is not " " then get theAbstract set pasteItem to result as string --return pasteItem set abstract of newPub to pasteItem else set theMessage to "No abstract found" display dialog theMessage buttons {"¥"} default button 1 giving up after 4 end if end if -- Now that the linked file list and abstract have been copied into the "newPub", we can overwrite every field in "ThePub" -- Copy all the fields in "newPub" into the original entry "ThePub" set theNewFields to every field of newPub repeat with theField in every field of newPub set fieldName to name of theField --return fieldName set value of field fieldName of thePub to (get value of field fieldName of newPub) end repeat get the cite key of newPub set newCiteKey to result --return newCiteKey set cite key of thePub to newCiteKey -- Copy the ADS URL from its own special field to the "normal" list of linked URLs get value of field "Adsurl" of thePub set pasteItem to result as string set Adsurl to pasteItem tell thePub -- generate a new value for the field, we use the "Url" field as a dummy if (theAdsurl is not "") and (theAdsurl is not in (get linked URLs)) then  add theAdsurl to end of linked URLs end tell remove newPub --return theAbstract --return theNewFields end if -- end of "if a non-empty ADS Bibtex entry has been downloaded" end if -- end of "if a valid ADS Url exists" end tell end tell -- end of the main program ------------------------- on findDOI(theFilename) -- Given a PDF file, extract the DOI. -- Any number of digits + a period + any number of alphanumeric characters + a solidus + at least one of (alphanumeric characters, parentheses, periods, or dashes) -- Also, artifically add a "doi" to the front of the resulting string to use later as a delimiter -- For SsRev, allow a space ([:blank:]) to be the delimiter between "DOI" and the DOI -- For ApJ and AAP, try adding another solidus -- Do not allow consecutive solidi? -- Do not allow the DOI to end with a solidus, period, or comma -- -- Try allowing ONE whitespace character after "10.1111" for the PNAS format (the dot can be a slash or -- whitespace or any other non-EOL character) --- the problem with PNAS is that it looks like a slash, but pdftotext reads it as a space. --- the extracted DOI has a space and does not resolve properly in ADS --- could substitute a slash, but that would be awfully specific to PNAS. For now just --- handle PNAS manually. set theSedslashString to "'s_.*[Dd][Oo][Ii][:)/[:blank:]] *\\([[:digit:]][[:digit:]]*\\.[[:alnum:]]*.[[:alnum:]]*[[:alnum:]/)(.-]*[[:alnum:]]\\).*_doi\\1_p'" set theSedSlashCommand to "sed -n -e " & theSedslashString set TheCommand to "/sw/bin/pdftotext -l 1 " & quoted form of theFilename & " - | " & theSedSlashCommand --return TheCommand --display dialog TheCommand buttons {"¥"} default button 1 giving up after 5 do shell script TheCommand set theDOIString to result --return theDOIString --display dialog theDOIString buttons {"¥"} default button 1 giving up after 3 -- Fall back to a simpler DOI -- Any number of digits + a period + any number of alphanumeric characters + a solidus + at least one of (alphanumeric characters, parentheses, periods, or dashes) -- Also, artifically add a "doi" to the front of the resulting string to use later as a delimiter if theDOIString = "" then display dialog "No DOI found. Retrying" buttons {"¥"} default button 1 giving up after 3 set theSedString to "'s_.*[Dd][Oo][Ii][:)] *\\([[:digit:]][[:digit:]]*.[[:alnum:])(.-]*/[[:alnum:])(.-][[:alnum:]):(.-]*[[:alnum:]-]\\).*_doi\\1_p'" --return theSedString set theSedCommand to "sed -n -e " & theSedString set theGrepCommand to " grep -i doi " set TheCommand to "/sw/bin/pdftotext -l 1 " & quoted form of theFilename & " - | " & theSedCommand --return TheCommand do shell script TheCommand set theDOIString to result --display dialog theDOIString buttons {"¥"} default button 1 giving up after 3 end if -- If the DOI isn't found on the first page, try searching the rest of the document if theDOIString = "" then --set TheCommand to "/sw/bin/pdftotext -f 2 " & quoted form of theFilename & " - | " & theSedCommand set TheCommand to "/sw/bin/pdftotext -f 2 " & quoted form of theFilename & " - | " & theSedSlashCommand do shell script TheCommand set theDOIString to result end if -- If there's no labeled DOI, try again with just the form if theDOIString = "" then set theSedString to "'s_.* *\\(10\\.[[:alnum:])(.-]*/[[:alnum:]][[:alnum:]):(.-]*[[:alnum:]):(.-/]*[[:alnum:])(]\\).*_doi\\1_p'" set theSedCommand to "sed -n -e " & theSedString set TheCommand to "/sw/bin/pdftotext -l 1 " & quoted form of theFilename & " - | " & theSedCommand --return TheCommand do shell script TheCommand set theDOIString to result display dialog theDOIString buttons {"¥"} default button 1 giving up after 3 end if if theDOIString = "" then beep display dialog "No DOI found." buttons {"¥"} default button 1 giving up after 3 else -- In case more than one DOI has been returned, select just the first one --display dialog theDOIString buttons {"¥"} default button 1 giving up after 3 set savedTextItemDelimiters to AppleScript's text item delimiters set AppleScript's text item delimiters to "doi" set thefirstDOIString to the second text item of theDOIString --set thefirstDOIString to AppleScript's text item delimiters & partofthefirstDOIString -- Eliminate trailing carriage returns set AppleScript's text item delimiters to ASCII character 13 set thefirstDOIString to the first text item of thefirstDOIString set AppleScript's text item delimiters to savedTextItemDelimiters set theDOIString to thefirstDOIString end if return theDOIString end findDOI ------------------------- on removebrackets(theDOIString) -- If the DOI is in brackets (ScienceDirect will do this) then remove the brackets if the first item of theDOIString is "{" then set savedTextItemDelimiters to AppleScript's text item delimiters -- replace theBadString with theGoodString in theBibtexUrl set AppleScript's text item delimiters to "{" set theDOIString2 to the second text item of theDOIString set AppleScript's text item delimiters to "}" set theDOIString2 to the first text item of theDOIString2 set AppleScript's text item delimiters to savedTextItemDelimiters else set theDOIString2 to theDOIString end if return theDOIString2 end removebrackets ------------------------- on getADSBibtexUrl(theAdsurl) set theBibtexUrl to my cleanADSUrl(theAdsurl) -- Eliminate carriage returns if (theBibtexUrl contains (ASCII character 13)) then set savedTextItemDelimiters to AppleScript's text item delimiters -- replace theBadString with theGoodString in theBibtexUrl set AppleScript's text item delimiters to ASCII character 13 set the text_item_list to every text item of theBibtexUrl set AppleScript's text item delimiters to "" set theBibtexUrl to the text_item_list as string set AppleScript's text item delimiters to savedTextItemDelimiters end if --return theBibtexUrl set theBibtexUrl to theBibtexUrl & "&data_type=BIBTEX" return theBibtexUrl end getADSBibtexUrl ------------------------- on cleanADSUrl(theAdsurl) set theBadString to "cgi-bin/nph-bib_query?" if (theAdsurl contains theBadString) then set theGoodString to "abs/" set savedTextItemDelimiters to AppleScript's text item delimiters -- replace theBadString with theGoodString in theAdsurl set AppleScript's text item delimiters to theBadString set the text_item_list to every text item of theAdsurl set AppleScript's text item delimiters to theGoodString set theGoodAdsurl to the text_item_list as string set AppleScript's text item delimiters to savedTextItemDelimiters else set theGoodAdsurl to theAdsurl end if set theBadString to "cgi-bin/nph-data_query?" if (theGoodAdsurl contains theBadString) then set theGoodString to "abs/" set savedTextItemDelimiters to AppleScript's text item delimiters -- replace theBadString with theGoodString in theBibtexUrl set AppleScript's text item delimiters to theBadString set the text_item_list to every text item of theGoodAdsurl set AppleScript's text item delimiters to theGoodString set theGoodAdsurl to the text_item_list as string set AppleScript's text item delimiters to savedTextItemDelimiters end if return theGoodAdsurl end cleanADSUrl ------------------------- on getADSAbstract(theAdsurl) -- Download an ADS record (from the supplied URL) and extract the abstract text. set TheCommand to "curl '" & theAdsurl & "'" do shell script TheCommand set theAbstract to the result --if theAbstract = "" then error if theAbstract contains "Abstract" then try set savedTextItemDelimiters to AppleScript's text item delimiters set AppleScript's text item delimiters to {"Abstract"} get last text item of theAbstract set theAbstract to the result set AppleScript's text item delimiters to {"
"} get first text item of theAbstract set theAbstract to the result set AppleScript's text item delimiters to savedTextItemDelimiters end try else set theAbstract to " " end if return theAbstract end getADSAbstract -------------------------