使用AutoHotKey(AHK)自动根据文献名查找文献DOI,并查找pdf

引子

google学术上的文章有些没有pdf链接,其中大部分可以在scihub上下载到。如果要在scihub上下载文献pdf,需要文献的DOI号,通常需要再在crossref上查找DOI,之后再去scihub上查找pdf。这样一套下来效率太低,因此有了这篇文章。
如果你认为此脚本可以有更多的可能性或者有更好的实现方式,欢迎在评论区讨论;如果你觉得脚本对你有用,还请顺手点赞收藏,你的支持是对我最大的鼓励😊

脚本功能

该脚本可以实现的功能如下:

  1. 自动查找与鼠标选中内容相同标题的文献DOI,如果多篇文献标题相同(概率极低),则只会查找crossref中第一篇文章的DOI,并保存到系统剪贴板;
  2. 根据第1步中的DOI信息,自动在默认浏览器中打开这篇文献的scihub链接,方便查看和下载;

脚本使用方法

单击F3键,实现功能2;
双击F3键,实现功能1
长按F3键,F3键的原始功能。

脚本

$F3::
    KeyWait, F3, T0.6 ;wait 0.6s
    if (ErrorLevel)
        sendinput {F3} ;long press
    else {
        KeyWait, F3, D T0.3 ;wait 0.3s
        if (ErrorLevel)
            GetPdfFromSciHub(false) ;single press
        else
            GetPdfFromSciHub(true) ;double press
    }
    KeyWait, F3
return

; get pdf according to title
GetPdfFromSciHub(onlyDOI := false)
{
    transMultiToSingle()
    ti:=RegExReplace(clipboard,"[ ]+"," ") ; keep only one space
    title:=StrReplace(ti," ","+") ; replace space by +
    url := "https://api.crossref.org/works?query=" . title ; create url
    sJson := GetTxtFromUrl(url) ; get json
    parsed := JSON.Load(sJson) ; parse json
    for item in parsed.message.items
    {
        StringLower, oti, % parsed.message.items[item].title[1] 
        StringLower, iti, ti
        if(oti == iti)
        {
            doi:=parsed.message.items[item].DOI
            if onlyDOI
            {
                msgbox,0x40000 ,,% doi,1
                clipboard:=doi
            }
            else
                run, https://sci-hub.tw/%doi%
            return
        }
    }
    MsgBox,,,No article, 3
}

; transform multiline clipboard string to single line
transMultiToSingle()
{
    Clipboard := ""
    SendInput, ^c
    ClipWait, 1
    if ErrorLevel
        Return
    
    ;Later two lines transform multiline string to single line string, and save results in clipboard. This is very useful when copying multiline string from pdf.
    Clipboard:=RegExReplace(Clipboard,"\r\n[ #;%]*([\x{3002}\x{ff1b}\x{ff0c}\x{ff1a}\x{201c}\x{201d}\x{ff08}\x{ff09}\x{3001}\x{ff1f}\x{300a}\x{300b}\x{4e00}-\x{9fa5}])","$1")
    Clipboard:=RegExReplace(Clipboard,"\r\n[ #;%]*"," ")
    ;Clipboard:=RegExReplace(Clipboard,"\r\n[ #;%]*([A-z])"," $1")
    ;Clipboard:=RegExReplace(Clipboard,"\r\n[ #;%]*([^A-z])","$1")
}

; get text content from url and return
GetTxtFromUrl(url) 
{
    whr := ComObjCreate("WinHttp.WinHttpRequest.5.1") 
    whr.Open("GET", url, true)
    whr.Send()
    ; Using 'true' above and the call below allows the script to remain responsive.
    whr.WaitForResponse()
return whr.ResponseText
}

/**
 * Lib: JSON.ahk
 *     JSON lib for AutoHotkey.
 * Version:
 *     v2.1.3 [updated 04/18/2016 (MM/DD/YYYY)]
 * License:
 *     WTFPL [http://wtfpl.net/]
 * Requirements:
 *     Latest version of AutoHotkey (v1.1+ or v2.0-a+)
 * Installation:
 *     Use #Include JSON.ahk or copy into a function library folder and then
 *     use #Include <JSON>
 * Links:
 *     GitHub:     - https://github.com/cocobelgica/AutoHotkey-JSON
 *     Forum Topic - http://goo.gl/r0zI8t
 *     Email:      - cocobelgica <at> gmail <dot> com
 */


/**
 * Class: JSON
 *     The JSON object contains methods for parsing JSON and converting values
 *     to JSON. Callable - NO; Instantiable - YES; Subclassable - YES;
 *     Nestable(via #Include) - NO.
 * Methods:
 *     Load() - see relevant documentation before method definition header
 *     Dump() - see relevant documentation before method definition header
 */
class JSON
{
	/**
	 * Method: Load
	 *     Parses a JSON string into an AHK value
	 * Syntax:
	 *     value := JSON.Load( text [, reviver ] )
	 * Parameter(s):
	 *     value      [retval] - parsed value
	 *     text    [in, ByRef] - JSON formatted string
	 *     reviver   [in, opt] - function object, similar to JavaScript's
	 *                           JSON.parse() 'reviver' parameter
	 */
	class Load extends JSON.Functor
	{
		Call(self, ByRef text, reviver:="")
		{
			this.rev := IsObject(reviver) ? reviver : false
		; Object keys(and array indices) are temporarily stored in arrays so that
		; we can enumerate them in the order they appear in the document/text instead
		; of alphabetically. Skip if no reviver function is specified.
			this.keys := this.rev ? {} : false

			static quot := Chr(34), bashq := "\" . quot
			     , json_value := quot . "{[01234567890-tfn"
			     , json_value_or_array_closing := quot . "{[]01234567890-tfn"
			     , object_key_or_object_closing := quot . "}"

			key := ""
			is_key := false
			root := {}
			stack := [root]
			next := json_value
			pos := 0

			while ((ch := SubStr(text, ++pos, 1)) != "") {
				if InStr(" `t`r`n", ch)
					continue
				if !InStr(next, ch, 1)
					this.ParseError(next, text, pos)

				holder := stack[1]
				is_array := holder.IsArray

				if InStr(",:", ch) {
					next := (is_key := !is_array && ch == ",") ? quot : json_value

				} else if InStr("}]", ch) {
					ObjRemoveAt(stack, 1)
					next := stack[1]==root ? "" : stack[1].IsArray ? ",]" : ",}"

				} else {
					if InStr("{[", ch) {
					; Check if Array() is overridden and if its return value has
					; the 'IsArray' property. If so, Array() will be called normally,
					; otherwise, use a custom base object for arrays
						static json_array := Func("Array").IsBuiltIn || ![].IsArray ? {IsArray: true} : 0
					
					; sacrifice readability for minor(actually negligible) performance gain
						(ch == "{")
							? ( is_key := true
							  , value := {}
							  , next := object_key_or_object_closing )
						; ch == "["
							: ( value := json_array ? new json_array : []
							  , next := json_value_or_array_closing )
						
						ObjInsertAt(stack, 1, value)

						if (this.keys)
							this.keys[value] := []
					
					} else {
						if (ch == quot) {
							i := pos
							while (i := InStr(text, quot,, i+1)) {
								value := StrReplace(SubStr(text, pos+1, i-pos-1), "\\", "\u005c")

								static tail := A_AhkVersion<"2" ? 0 : -1
								if (SubStr(value, tail) != "\")
									break
							}

							if (!i)
								this.ParseError("'", text, pos)

							  value := StrReplace(value,  "\/",  "/")
							, value := StrReplace(value, bashq, quot)
							, value := StrReplace(value,  "\b", "`b")
							, value := StrReplace(value,  "\f", "`f")
							, value := StrReplace(value,  "\n", "`n")
							, value := StrReplace(value,  "\r", "`r")
							, value := StrReplace(value,  "\t", "`t")

							pos := i ; update pos
							
							i := 0
							while (i := InStr(value, "\",, i+1)) {
								if !(SubStr(value, i+1, 1) == "u")
									this.ParseError("\", text, pos - StrLen(SubStr(value, i+1)))

								uffff := Abs("0x" . SubStr(value, i+2, 4))
								if (A_IsUnicode || uffff < 0x100)
									value := SubStr(value, 1, i-1) . Chr(uffff) . SubStr(value, i+6)
							}

							if (is_key) {
								key := value, next := ":"
								continue
							}
						
						} else {
							value := SubStr(text, pos, i := RegExMatch(text, "[\]\},\s]|$",, pos)-pos)

							static number := "number", integer :="integer"
							if value is %number%
							{
								if value is %integer%
									value += 0
							}
							else if (value == "true" || value == "false")
								value := %value% + 0
							else if (value == "null")
								value := ""
							else
							; we can do more here to pinpoint the actual culprit
							; but that's just too much extra work.
								this.ParseError(next, text, pos, i)

							pos += i-1
						}

						next := holder==root ? "" : is_array ? ",]" : ",}"
					} ; If InStr("{[", ch) { ... } else

					is_array? key := ObjPush(holder, value) : holder[key] := value

					if (this.keys && this.keys.HasKey(holder))
						this.keys[holder].Push(key)
				}
			
			} ; while ( ... )

			return this.rev ? this.Walk(root, "") : root[""]
		}

		ParseError(expect, ByRef text, pos, len:=1)
		{
			static quot := Chr(34), qurly := quot . "}"
			
			line := StrSplit(SubStr(text, 1, pos), "`n", "`r").Length()
			col := pos - InStr(text, "`n",, -(StrLen(text)-pos+1))
			msg := Format("{1}`n`nLine:`t{2}`nCol:`t{3}`nChar:`t{4}"
			,     (expect == "")     ? "Extra data"
			    : (expect == "'")    ? "Unterminated string starting at"
			    : (expect == "\")    ? "Invalid \escape"
			    : (expect == ":")    ? "Expecting ':' delimiter"
			    : (expect == quot)   ? "Expecting object key enclosed in double quotes"
			    : (expect == qurly)  ? "Expecting object key enclosed in double quotes or object closing '}'"
			    : (expect == ",}")   ? "Expecting ',' delimiter or object closing '}'"
			    : (expect == ",]")   ? "Expecting ',' delimiter or array closing ']'"
			    : InStr(expect, "]") ? "Expecting JSON value or array closing ']'"
			    :                      "Expecting JSON value(string, number, true, false, null, object or array)"
			, line, col, pos)

			static offset := A_AhkVersion<"2" ? -3 : -4
			throw Exception(msg, offset, SubStr(text, pos, len))
		}

		Walk(holder, key)
		{
			value := holder[key]
			if IsObject(value) {
				for i, k in this.keys[value] {
					; check if ObjHasKey(value, k) ??
					v := this.Walk(value, k)
					if (v != JSON.Undefined)
						value[k] := v
					else
						ObjDelete(value, k)
				}
			}
			
			return this.rev.Call(holder, key, value)
		}
	}

	/**
	 * Method: Dump
	 *     Converts an AHK value into a JSON string
	 * Syntax:
	 *     str := JSON.Dump( value [, replacer, space ] )
	 * Parameter(s):
	 *     str        [retval] - JSON representation of an AHK value
	 *     value          [in] - any value(object, string, number)
	 *     replacer  [in, opt] - function object, similar to JavaScript's
	 *                           JSON.stringify() 'replacer' parameter
	 *     space     [in, opt] - similar to JavaScript's JSON.stringify()
	 *                           'space' parameter
	 */
	class Dump extends JSON.Functor
	{
		Call(self, value, replacer:="", space:="")
		{
			this.rep := IsObject(replacer) ? replacer : ""

			this.gap := ""
			if (space) {
				static integer := "integer"
				if space is %integer%
					Loop, % ((n := Abs(space))>10 ? 10 : n)
						this.gap .= " "
				else
					this.gap := SubStr(space, 1, 10)

				this.indent := "`n"
			}

			return this.Str({"": value}, "")
		}

		Str(holder, key)
		{
			value := holder[key]

			if (this.rep)
				value := this.rep.Call(holder, key, ObjHasKey(holder, key) ? value : JSON.Undefined)

			if IsObject(value) {
			; Check object type, skip serialization for other object types such as
			; ComObject, Func, BoundFunc, FileObject, RegExMatchObject, Property, etc.
				static type := A_AhkVersion<"2" ? "" : Func("Type")
				if (type ? type.Call(value) == "Object" : ObjGetCapacity(value) != "") {
					if (this.gap) {
						stepback := this.indent
						this.indent .= this.gap
					}

					is_array := value.IsArray
				; Array() is not overridden, rollback to old method of
				; identifying array-like objects. Due to the use of a for-loop
				; sparse arrays such as '[1,,3]' are detected as objects({}). 
					if (!is_array) {
						for i in value
							is_array := i == A_Index
						until !is_array
					}

					str := ""
					if (is_array) {
						Loop, % value.Length() {
							if (this.gap)
								str .= this.indent
							
							v := this.Str(value, A_Index)
							str .= (v != "") ? v . "," : "null,"
						}
					} else {
						colon := this.gap ? ": " : ":"
						for k in value {
							v := this.Str(value, k)
							if (v != "") {
								if (this.gap)
									str .= this.indent

								str .= this.Quote(k) . colon . v . ","
							}
						}
					}

					if (str != "") {
						str := RTrim(str, ",")
						if (this.gap)
							str .= stepback
					}

					if (this.gap)
						this.indent := stepback

					return is_array ? "[" . str . "]" : "{" . str . "}"
				}
			
			} else ; is_number ? value : "value"
				return ObjGetCapacity([value], 1)=="" ? value : this.Quote(value)
		}

		Quote(string)
		{
			static quot := Chr(34), bashq := "\" . quot

			if (string != "") {
				  string := StrReplace(string,  "\",  "\\")
				; , string := StrReplace(string,  "/",  "\/") ; optional in ECMAScript
				, string := StrReplace(string, quot, bashq)
				, string := StrReplace(string, "`b",  "\b")
				, string := StrReplace(string, "`f",  "\f")
				, string := StrReplace(string, "`n",  "\n")
				, string := StrReplace(string, "`r",  "\r")
				, string := StrReplace(string, "`t",  "\t")

				static rx_escapable := A_AhkVersion<"2" ? "O)[^\x20-\x7e]" : "[^\x20-\x7e]"
				while RegExMatch(string, rx_escapable, m)
					string := StrReplace(string, m.Value, Format("\u{1:04x}", Ord(m.Value)))
			}

			return quot . string . quot
		}
	}

	/**
	 * Property: Undefined
	 *     Proxy for 'undefined' type
	 * Syntax:
	 *     undefined := JSON.Undefined
	 * Remarks:
	 *     For use with reviver and replacer functions since AutoHotkey does not
	 *     have an 'undefined' type. Returning blank("") or 0 won't work since these
	 *     can't be distnguished from actual JSON values. This leaves us with objects.
	 *     Replacer() - the caller may return a non-serializable AHK objects such as
	 *     ComObject, Func, BoundFunc, FileObject, RegExMatchObject, and Property to
	 *     mimic the behavior of returning 'undefined' in JavaScript but for the sake
	 *     of code readability and convenience, it's better to do 'return JSON.Undefined'.
	 *     Internally, the property returns a ComObject with the variant type of VT_EMPTY.
	 */
	Undefined[]
	{
		get {
			static empty := {}, vt_empty := ComObject(0, &empty, 1)
			return vt_empty
		}
	}

	class Functor
	{
		__Call(method, ByRef arg, args*)
		{
		; When casting to Call(), use a new instance of the "function object"
		; so as to avoid directly storing the properties(used across sub-methods)
		; into the "function object" itself.
			if IsObject(method)
				return (new this).Call(method, arg, args*)
			else if (method == "")
				return (new this).Call(arg, args*)
		}
	}
}

参考资料

脚本中的JSON解析部分代码是从Github上copy的,在此附上参考地址AutoHotkey-JSON

评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值