@ -2,27 +2,21 @@ package io.legado.app.model.analyzeRule
//通用的规则切分处理
class RuleAnalyzer ( data : String , code : Boolean = false ) {
class RuleAnalyzer ( data : String , code : Boolean = false ) {
private var queue : String = data //被处理字符串
private var pos = 0 //处理到的位置
private var rule = arrayOf < String > ( )
private var pos = 0 //当前处理到的位置
private var start = 0 //当前处理字段的开始
private var startX = 0 //当前规则的开始
private var start = 0 //每次处理字段的开始
private var startX = 0 //规则的开始
private var step : Int = 0 //分割字符的长度
private var rule = arrayOf < String > ( ) //分割出的规则列表
private var step : Int = 0 //分割字符的长度
var elementsType = " " //当前分割字符串
val chompBalanced = if ( code ) :: chompCodeBalanced else :: chompRuleBalanced //设置平衡组函数，json或JavaScript时设置成chompCodeBalanced，否则为chompRuleBalanced
//设置平衡组函数，json或JavaScript时设置成chompCodeBalanced，否则为chompRuleBalanced
val chompBalanced = if ( code ) :: chompCodeBalanced else :: chompRuleBalanced
var elementsType = " "
//当前平衡字段
fun currBalancedString ( stepStart : Int = 1 , stepEnd : Int = 1 ) : String { //stepStart平衡字符的起始分隔字串长度，stepEnd平衡字符的结束分隔字串长度
return queue . substring ( startX + stepStart , pos - stepEnd ) //当前平衡字段
}
fun trim ( ) { // 修剪当前规则之前的"@"或者空白符
fun trim ( ) { // 修剪当前规则之前的"@"或者空白符
while ( queue [ pos ] == '@' || queue [ pos ] < '!' ) pos ++
}
@ -33,14 +27,7 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
//当前拉取字段
fun currString ( ) : String {
return queue . substring ( start , pos ) //当前拉取到的字段
}
//剩余字串
private fun remainingString ( ) : String {
start = pos
pos = queue . length
return queue . substring ( start )
return queue . substring ( start , pos ) //当前拉取到的字段
}
/ * *
@ -48,7 +35,7 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
* @return 若剩余字串中已无字符则返回true
* /
val isEmpty : Boolean
get ( ) = queue . length - pos == 0 //是否处理到最后
get ( ) = queue . length - pos == 0 //是否处理到最后
/ * *
* 消耗剩余字串中一个字符 。
@ -72,15 +59,13 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
* @param seq 分隔字符 * * 区分大小写 * *
* @return 是否找到相应字段 。
* /
fun consumeTo ( seq : String , setStartPos : Boolean = true ) : Boolean {
if ( setStartPos ) start = pos //将处理到的位置设置为规则起点
fun consumeTo ( seq : String ) : Boolean {
start = pos //将处理到的位置设置为规则起点
val offset = queue . indexOf ( seq , pos )
return if ( offset != - 1 ) {
pos = offset
true
} else false
}
/ * *
@ -88,7 +73,7 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
* @param seq 匹配字符串序列
* @return 成功返回true并设置间隔 ， 失败则直接返回fasle
* /
fun consumeToAny ( vararg seq : String ) : Boolean {
fun consumeToAny ( vararg seq : String ) : Boolean {
var pos = pos //声明新变量记录匹配位置，不更改类本身的位置
@ -113,13 +98,13 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
* @param seq 匹配字符序列
* @return 返回匹配位置
* /
private fun findToAny ( vararg seq : Char ) : Int {
private fun findToAny ( vararg seq : Char ) : Int {
var pos = pos //声明新变量记录匹配位置，不更改类本身的位置
while ( pos != queue . length ) {
for ( s in seq ) if ( queue [ pos ] == s ) return pos //匹配则返回位置
for ( s in seq ) if ( queue [ pos ] == s ) return pos //匹配则返回位置
pos ++ //逐个试探
@ -129,13 +114,15 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
}
//其中js只要符合语法，就不用避开任何阅读关键字，自由发挥
fun chompJsBalanced ( f : ( ( Char ) -> Boolean ? ) = {
when ( it ) {
'{' -> true //开始嵌套一层
'}' -> false //闭合一层嵌套
else -> null
}
} ) : Boolean {
fun chompJsBalanced (
f : ( ( Char ) -> Boolean ? ) = {
when ( it ) {
'{' -> true //开始嵌套一层
'}' -> false //闭合一层嵌套
else -> null
}
}
) : Boolean {
var pos = pos //声明变量记录临时处理位置
var depth = 0 //嵌套深度
var bracketsDepth = 0 //[]嵌套深度
@ -151,22 +138,23 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
if ( pos == queue . length ) break
var c = queue [ pos ++ ]
if ( c != '\\' ) { //非转义字符
if ( c == '\'' && ! commits && ! commit && ! regex && !in DoubleQuote && !in OtherQuote ) inSingleQuote = !in SingleQuote //匹配具有语法功能的单引号
else if ( c == '"' && ! commits && ! commit && ! regex && !in SingleQuote && !in OtherQuote ) inDoubleQuote = !in DoubleQuote //匹配具有语法功能的双引号
else if ( c == '`' && ! commits && ! commit && ! regex && !in SingleQuote && !in DoubleQuote ) inOtherQuote = !in OtherQuote //匹配具有语法功能的'`'
if ( c == '\'' && ! commits && ! commit && ! regex && !in DoubleQuote && !in OtherQuote ) inSingleQuote =
!in SingleQuote //匹配具有语法功能的单引号
else if ( c == '"' && ! commits && ! commit && ! regex && !in SingleQuote && !in OtherQuote ) inDoubleQuote =
!in DoubleQuote //匹配具有语法功能的双引号
else if ( c == '`' && ! commits && ! commit && ! regex && !in SingleQuote && !in DoubleQuote ) inOtherQuote =
!in OtherQuote //匹配具有语法功能的'`'
else if ( c == '/' && ! commits && ! commit && ! regex && !in SingleQuote && !in DoubleQuote && !in OtherQuote ) { //匹配注释或正则起点
c = queue [ pos ++ ]
when ( c ) {
'/' -> commit = true //匹配单行注释起点
'*' -> commits = true //匹配多行注释起点
else -> regex = true //匹配正则起点
when ( c ) {
'/' -> commit = true //匹配单行注释起点
'*' -> commits = true //匹配多行注释起点
else -> regex = true //匹配正则起点
}
}
else if ( commits && c == '*' ) { //匹配多行注释终点
} else if ( commits && c == '*' ) { //匹配多行注释终点
c = queue [ pos ++ ]
if ( c == '/' ) commits = false
}
else if ( regex && c == '/' ) { //正则的终点或[]平衡
if ( c == '/' ) commits = false
} else if ( regex && c == '/' ) { //正则的终点或[]平衡
when ( c ) {
'/' -> regex = false //匹配正则终点
@ -177,25 +165,25 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
']' -> bracketsDepth -- //闭合一层嵌套[]
}
} else if ( c == '\n' ) commit = false
} else if ( c == '\n' ) commit = false
if ( commits || commit || regex || inSingleQuote || inDoubleQuote || inOtherQuote ) continue //语法单元未匹配结束，直接进入下个循环
if ( commits || commit || regex || inSingleQuote || inDoubleQuote || inOtherQuote ) continue //语法单元未匹配结束，直接进入下个循环
val fn = f ( c ) ?: continue
if ( fn ) depth ++ else depth -- //嵌套或者闭合
} else pos ++
} else pos ++
} while ( depth > 0 || bracketsDepth > 0 ) //拉出全部符合js语法的字段
} while ( depth > 0 || bracketsDepth > 0 ) //拉出全部符合js语法的字段
return if ( depth > 0 || bracketsDepth > 0 ) false else {
return if ( depth > 0 || bracketsDepth > 0 ) false else {
this . pos = pos //同步位置
true
}
}
/ * *
* 拉出一个代码平衡组 ， 存在转义文本 ， 没有实体字符 ， 通常以 { } 作为模块
* 拉出一个非内嵌 代码平衡组 ， 存在转义文本
* /
fun chompCodeBalanced ( open : Char , close : Char ) : Boolean {
@ -214,22 +202,21 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
if ( c == '\'' && !in DoubleQuote ) inSingleQuote = !in SingleQuote //匹配具有语法功能的单引号
else if ( c == '"' && !in SingleQuote ) inDoubleQuote = !in DoubleQuote //匹配具有语法功能的双引号
if ( inSingleQuote || inDoubleQuote ) continue //语法单元未匹配结束，直接进入下个循环
if ( inSingleQuote || inDoubleQuote ) continue //语法单元未匹配结束，直接进入下个循环
if ( c == '[' ) depth ++ //开始嵌套一层
else if ( c == ']' ) depth -- //闭合一层嵌套
else if ( depth == 0 ) {
if ( c == '[' ) depth ++ //开始嵌套一层
else if ( c == ']' ) depth -- //闭合一层嵌套
else if ( depth == 0 ) {
//处于默认嵌套中的非默认字符不需要平衡，仅depth为0时默认嵌套全部闭合，此字符才进行嵌套
if ( c == open ) otherDepth ++
else if ( c == close ) otherDepth --
if ( c == open ) otherDepth ++
else if ( c == close ) otherDepth --
}
} else pos ++
} else pos ++
} while ( depth > 0 || otherDepth > 0 ) //拉出一个平衡字串
return if ( depth > 0 || otherDepth > 0 ) false else {
startX = this . pos //内嵌规则起始
return if ( depth > 0 || otherDepth > 0 ) false else {
this . pos = pos //同步位置
true
}
@ -251,18 +238,18 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
if ( c == '\'' && !in DoubleQuote ) inSingleQuote = !in SingleQuote //匹配具有语法功能的单引号
else if ( c == '"' && !in SingleQuote ) inDoubleQuote = !in DoubleQuote //匹配具有语法功能的双引号
if ( inSingleQuote || inDoubleQuote ) continue //语法单元未匹配结束，直接进入下个循环
else if ( c == '\\' ) { //不在引号中的转义字符才将下个字符转义
if ( inSingleQuote || inDoubleQuote ) continue //语法单元未匹配结束，直接进入下个循环
else if ( c == '\\' ) { //不在引号中的转义字符才将下个字符转义
pos ++
continue
}
if ( c == open ) depth ++ //开始嵌套一层
else if ( c == close ) depth -- //闭合一层嵌套
if ( c == open ) depth ++ //开始嵌套一层
else if ( c == close ) depth -- //闭合一层嵌套
} while ( depth > 0 ) //拉出一个平衡字串
} while ( depth > 0 ) //拉出一个平衡字串
return if ( depth > 0 ) false else {
return if ( depth > 0 ) false else {
this . pos = pos //同步位置
true
}
@ -272,18 +259,18 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
* 不用正则 , 不到最后不切片也不用中间变量存储 , 只在序列中标记当前查找字段的开头结尾 , 到返回时才切片 , 高效快速准确切割规则
* 解决jsonPath自带的 " && " 和 " || " 与阅读的规则冲突 , 以及规则正则或字符串中包含 " && " 、 " || " 、 " %% " 、 " @ " 导致的冲突
* /
tailrec fun splitRule ( vararg split : String ) : Array < String > { //首段匹配,elementsType为空
tailrec fun splitRule ( vararg split : String ) : Array < String > { //首段匹配,elementsType为空
if ( split . size == 1 ) {
if ( split . size == 1 ) {
elementsType = split [ 0 ] //设置分割字串
return if ( ! consumeTo ( elementsType ) ) {
return if ( ! consumeTo ( elementsType ) ) {
rule += queue . substring ( startX )
rule
} else {
} else {
step = elementsType . length //设置分隔符长度
splitRule ( )
} //递归匹配
} else if ( ! consumeToAny ( * split ) ) { //未找到分隔符
} else if ( ! consumeToAny ( * split ) ) { //未找到分隔符
rule += queue . substring ( startX )
return rule
}
@ -291,7 +278,7 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
val end = pos //记录分隔位置
pos = start //重回开始，启动另一种查找
do {
do {
val st = findToAny ( '[' , '(' ) //查找筛选器位置
if ( st == - 1 ) {
@ -323,11 +310,10 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
pos += step //跳过分隔符
}
return if ( pos > st ) {
return if ( pos > st ) {
startX = start
splitRule ( ) //首段已匹配,但当前段匹配未完成,调用二段匹配
}
else { //执行到此，证明后面再无分隔字符
} else { //执行到此，证明后面再无分隔字符
rule += queue . substring ( pos ) //将剩余字段压入数组末尾
rule
}
@ -337,13 +323,13 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
val next = if ( queue [ pos ] == '[' ) ']' else ')' //平衡组末尾字符
if ( ! chompBalanced ( queue [ pos ] , next ) ) throw Error (
queue . substring (
0 ,
start
) + " 后未平衡 "
queue . substring (
0 ,
start
) + " 后未平衡 "
) //拉出一个筛选器,不平衡则报错
} while ( end > pos )
} while ( end > pos )
start = pos //设置开始查找筛选器位置的起始位置
@ -351,12 +337,12 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
}
@JvmName ( " splitRuleNext " )
private tailrec fun splitRule ( ) : Array < String > { //二段匹配被调用,elementsType非空(已在首段赋值),直接按elementsType查找,比首段采用的方式更快
private tailrec fun splitRule ( ) : Array < String > { //二段匹配被调用,elementsType非空(已在首段赋值),直接按elementsType查找,比首段采用的方式更快
val end = pos //记录分隔位置
pos = start //重回开始，启动另一种查找
do {
do {
val st = findToAny ( '[' , '(' ) //查找筛选器位置
if ( st == - 1 ) {
@ -384,11 +370,10 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
pos += step //跳过分隔符
}
return if ( pos > st ) {
return if ( pos > st ) {
startX = start
splitRule ( ) //首段已匹配,但当前段匹配未完成,调用二段匹配
}
else { //执行到此，证明后面再无分隔字符
} else { //执行到此，证明后面再无分隔字符
rule += queue . substring ( pos ) //将剩余字段压入数组末尾
rule
}
@ -398,24 +383,23 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
val next = if ( queue [ pos ] == '[' ) ']' else ')' //平衡组末尾字符
if ( ! chompBalanced ( queue [ pos ] , next ) ) throw Error (
queue . substring (
0 ,
start
) + " 后未平衡 "
queue . substring (
0 ,
start
) + " 后未平衡 "
) //拉出一个筛选器,不平衡则报错
} while ( end > pos )
} while ( end > pos )
start = pos //设置开始查找筛选器位置的起始位置
return if ( ! consumeTo ( elementsType ) ) {
return if ( ! consumeTo ( elementsType ) ) {
rule += queue . substring ( startX )
rule
} else splitRule ( ) //递归匹配
} else splitRule ( ) //递归匹配
}
/ * *
* 替换内嵌规则
* @param inner 起始标志 , 如 { $ . 或 { {
@ -424,67 +408,32 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
* @param fr 查找到内嵌规则时 ， 用于解析的函数
*
* * /
fun innerRule ( inner : String , startStep : Int = 1 , endStep : Int = 1 , fr : ( String ) -> String ? ) : String {
val start0 = pos //规则匹配前起点
fun innerRule (
inner : String ,
startStep : Int = 1 ,
endStep : Int = 1 ,
fr : ( String ) -> String ?
) : String {
startX = pos //设置规则起点
val st = StringBuilder ( )
while ( !is Empty && consumeTo ( inner ) ) { //拉取成功返回true，ruleAnalyzes里的字符序列索引变量pos后移相应位置，否则返回false,且isEmpty为true
val posPre = pos //记录上次结束位置
if ( chompBalanced ( '{' , '}' ) ) {
val frv = fr ( currBalancedString ( startStep , endStep ) )
if ( frv != null ) {
st . append ( queue . substring ( start , posPre ) + frv ) //压入内嵌规则前的内容，及内嵌规则解析得到的字符串
while ( consumeTo ( inner ) ) { //拉取成功返回true，ruleAnalyzes里的字符序列索引变量pos后移相应位置，否则返回false,且isEmpty为true
if ( chompCodeBalanced ( '{' , '}' ) ) {
val frv = fr ( queue . substring ( start + startStep , pos - endStep ) )
if ( frv != null ) {
st . append ( queue . substring ( startX , start ) + frv ) //压入内嵌规则前的内容，及内嵌规则解析得到的字符串
startX = pos //记录下次规则起点
continue //获取内容成功，继续选择下个内嵌规则
}
}
pos += inner . length //拉出字段不平衡，inner只是个普通字串，跳到此inner后继续匹配
}
//匹配前起点与当前规则起点相同，证明无替换成功的内嵌规则,返回空字符串。否则返回替换后的字符串
return if ( start0 == start ) " " else {
st . append ( remainingString ( ) ) //压入剩余字符串
st . toString ( )
}
st . append ( queue . substring ( startX ) ) //压入剩余字符串
return st . toString ( )
}
// /**
// * 匹配并返回标签中的属性键字串（字母、数字、-、_、:）
// * @return 属性键字串
// */
// fun consumeAttributeKey(start:Int = pos): String {
// while (!isEmpty && (Character.isLetterOrDigit(queue[pos]) || matchesAny('-', '_', ':'))) pos++
// return queue.substring(start, pos)
// }
// fun splitRule(query:String,item:String = "other",listItem:String = "allInOne"):String{
//
// val cuurItem = item //当前项类型，list->列表项 mulu->章节列表项 url->链接项 search->搜索链接项 find发现链接列表项 other->其他项
// val cuurList = listItem//当前界面总列表项类型，allInOne，json，xml，kotin，java
// var Reverse = false //是否反转列表
//
// consumeWhitespace() //消耗开头空白
// var fisrt = consume() //拉出并消费首字符
//
// when(item){
// "search" ->
// "find" ->
// "mulu" -> if(fisrt == '-'){
// Reverse=true //开启反转
// consumeWhitespace() //拉出所有空白符
// fisrt = consume() //首字符后移
// }
// else ->
//
// }
//
// return query
// }
companion object {
/ * *
* 转义字符
@ -495,7 +444,7 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
* 阅读共有分隔字串起始部分
* " ## " , " @@ " , " {{ " , " {[ " , " <js> " , " @js: "
* /
val splitList = arrayOf ( " ## " , " @@ " , " {{ " , " {[ " , " <js> " , " @js: " )
val splitList = arrayOf ( " ## " , " @@ " , " {{ " , " {[ " , " <js> " , " @js: " )
/ * *
* 发现 ‘ 名称 - 链接 ’ 分隔字串
@ -519,7 +468,7 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
* 结果为元素列表的项的同规则组合结构
* " || " , " && " , " %% "
* /
val splitListReSplit = arrayOf ( " || " , " && " , " %% " )
val splitListReSplit = arrayOf ( " || " , " && " , " %% " )
/ * *
* js脚本结束字串
@ -542,27 +491,27 @@ class RuleAnalyzer(data: String, code:Boolean = false) {
/ * *
* '[' , ']' , '(' , ')' , '{' , '}'
* /
val splitListPublic = charArrayOf ( '[' , ']' , '(' , ')' , '{' , '}' )
val splitListPublic = charArrayOf ( '[' , ']' , '(' , ')' , '{' , '}' )
/ * *
* '*' , " / " , " // " , " : " , " :: " , " @ " , " | " , " @xpath: "
* /
val splitListXpath = arrayOf ( " * " , " / " , " // " , " : " , " :: " , " @ " , " | " , " @xpath: " )
val splitListXpath = arrayOf ( " * " , " / " , " // " , " : " , " :: " , " @ " , " | " , " @xpath: " )
/ * *
* '*' , '$' , " . " , " .. " , " @json: "
* /
val splitListJson = arrayOf ( '*' , '$' , " . " , " .. " , " @json: " )
val splitListJson = arrayOf ( '*' , '$' , " . " , " .. " , " @json: " )
/ * *
* '*' , " + " , " ~ " , " . " , " , " , " | " , " @ " , " @css: " , " : "
* /
val splitListCss = arrayOf ( '*' , " + " , " ~ " , " . " , " , " , " | " , " @ " , " @css: " , " : " )
val splitListCss = arrayOf ( '*' , " + " , " ~ " , " . " , " , " , " | " , " @ " , " @css: " , " : " )
/ * *
* " - " , " . " , " ! " , " @ " , " @@ "
* /
val splitListDefault = arrayOf ( " - " , " . " , " ! " , " @ " , " @@ " )
val splitListDefault = arrayOf ( " - " , " . " , " ! " , " @ " , " @@ " )
}
}
bushixuanqi
3 years ago