From 9a00fb82daa75ad74354a2625621ebf19f3bd2f4 Mon Sep 17 00:00:00 2001 From: bushixuanqi <57338301+bushixuanqi@users.noreply.github.com> Date: Sun, 4 Jul 2021 14:30:13 +0800 Subject: [PATCH 1/5] Update RuleAnalyzer.kt MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 简化部分规则，调整压入切分出的规则时的判断顺序，避免重复压入 --- .../app/model/analyzeRule/RuleAnalyzer.kt | 144 +++++++++--------- 1 file changed, 75 insertions(+), 69 deletions(-) diff --git a/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt b/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt index 73a861acf..09129de7a 100644 --- a/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt +++ b/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt @@ -8,7 +8,6 @@ class RuleAnalyzer(data: String) { private var pos = 0 //处理到的位置 private var start = 0 //每次处理字段的开始 - private var end:Int = queue.length //每次处理字段的终点 private var step:Int = 0 //分割字符的长度 var elementsType = "" @@ -29,7 +28,7 @@ class RuleAnalyzer(data: String) { } //剩余字串 - fun remainingString(): String { + private fun remainingString(): String { start = pos pos = queue.length return queue.substring(start) @@ -82,69 +81,73 @@ class RuleAnalyzer(data: String) { return queue.regionMatches(pos, seq, 0, seq.length, ignoreCase = true) } - /** - * 测试下个字符是否与序列中相应位置的字符相等。 - * @param seq ：被检查的字符列表 - * @return 相等就为 true ，不相等则为 false - */ - fun matchesAny(vararg seq: Char): Boolean { - if (isEmpty) return false - for (c in seq) { - if (queue[pos] == c) { - return true - } - } - return false - } - - /** - * 测试下个字符（串）是否与参数列表里的序列存在匹配。 不区分大小写。 - * @param seq ：被不区分大小写检查的字符串列表 - * @return 只要匹配就为 true ，没有匹配则为 false - */ - fun matchesAny(vararg seq: String): Boolean { - for (s in seq) { - if (matches(s)) { - step = s.length - return true - } - } - return false - } - /** * 从剩余字串中拉出一个字符串，直到但不包括匹配序列，或剩余字串用完。 * @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 val offset = queue.indexOf(seq, pos) return if (offset != -1) { pos = offset true } else false + + } + + /** + * 从剩余字串中拉出一个字符串，直到且包括匹配序列（匹配参数列表中一项即为匹配），或剩余字串用完。 + * @param seq 匹配字符串序列 + * @return 成功返回true并推移pos，失败返回fasle而不推移pos + */ + fun consumeToAny(vararg seq:String): Boolean { + + start = pos + + while (!isEmpty) { + + for (s in seq) { + if (matches(s)) { + step = s.length //间隔数 + return true //匹配就返回 true + } + } + + pos++ //逐个试探 + } + + pos = start //匹配失败，位置回退 + + return false } /** * 从剩余字串中拉出一个字符串，直到但不包括匹配序列（匹配参数列表中一项即为匹配），或剩余字串用完。 - * @param f 消费函数，返回true表示消费，fasle表示不消费 - * @param setStartPos 设置开始消费位置 - * @return 消耗的字符串 + * @param seq 匹配字符序列 + * @return 匹配到的位置 */ - fun consumeToAny(setStartPos:Boolean = true, f:()->Boolean,): Boolean { - if(setStartPos)start = pos //将处理到的位置设置为规则起点 - while (!isEmpty && !f()) { - pos++ + private fun findToAny(vararg seq:Char): Int { + + for (s in seq){ + + val offset = queue.indexOf(s, pos) + + if (offset != -1) return offset //只要匹配到，就返回相应位置 + } - return !isEmpty + + return -1 } //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?) = { - if ( it == '{' )true //开始嵌套一层 - else if ( it == '}') false //闭合一层嵌套 - else null + when (it) { + '{' -> true //开始嵌套一层 + '}' -> false //闭合一层嵌套 + else -> null + } } ): Boolean { start = pos var depth = 0 //嵌套深度 @@ -178,20 +181,20 @@ class RuleAnalyzer(data: String) { } else if(regex && c == '/') { //正则的终点或[]平衡 - if(c == '/')regex = false//匹配正则终点 - - //为了保证当open为（ 且 close 为 ）时，正则中[(]或[)]的合法性。故对[]这对在任何规则中都平衡的成对符号做匹配。 - // 注：正则里[(]、[)]、[{]、[}]都是合法的，所以只有[]必须平衡。 + when (c) { + '/' -> regex = false//匹配正则终点 - else if ( c == '[' )bracketsDepth++ //开始嵌套一层[] - else if ( c== ']') bracketsDepth-- //闭合一层嵌套[] + //为了保证当open为（ 且 close 为 ）时，正则中[(]或[)]的合法性。故对[]这对在任何规则中都平衡的成对符号做匹配。 + // 注：正则里[(]、[)]、[{]、[}]都是合法的，所以只有[]必须平衡。 + '[' -> bracketsDepth++ //开始嵌套一层[] + ']' -> bracketsDepth-- //闭合一层嵌套[] + } } 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) - if (fn == null) continue + val fn = f(c) ?: continue if (fn) depth++ else depth-- //嵌套或者闭合 }else { //转义字符 @@ -233,8 +236,7 @@ class RuleAnalyzer(data: String) { if ( c == open )depth++ //开始嵌套一层 else if ( c== close) depth-- //闭合一层嵌套 else if(depth == 0 && f != null) { //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 - val fn = f(c) - if (fn == null) continue + val fn = f(c) ?: continue if (fn) otherDepth++ else otherDepth-- } @@ -257,24 +259,28 @@ class RuleAnalyzer(data: String) { */ tailrec fun splitRule(vararg split: String): Array{ //首段匹配,elementsType为空 - if (!consumeToAny { matchesAny(* split) }) return arrayOf(queue) //未找到分隔符 + if(split.size == 1) { + elementsType = split[0] //设置分割字串 + step = elementsType.length //设置分隔符长度 + return splitRule(arrayOf()) //仅一个分隔字串时，直接二段解析更快 + }else if (!consumeToAny(* split)) return arrayOf(queue) //未找到分隔符 - end = pos - val st = if( consumeToAny(false){ matchesAny( '(','[' ) } )pos else -1 //查找筛选器 - pos = end + val st = findToAny( '(','[' ) //查找筛选器 if(st == -1) { var rule = arrayOf(queue.substring(0, pos)) //压入分隔的首段规则到数组 + elementsType = queue.substring(pos, pos + step) //设置组合类型 pos += step //跳过分隔符 - elementsType = queue.substring(pos - step, pos) //设置组合类型 - while (consumeToAny { matchesAny(* split) }) { //循环切分规则压入数组 + while (consumeToAny(* split)) { //循环切分规则压入数组 rule += queue.substring(start, pos) pos += step //跳过分隔符 } - rule+= queue.substring(start) //将剩余字段压入数组末尾 + + rule += queue.substring(pos) //将剩余字段压入数组末尾 + return rule } @@ -282,14 +288,16 @@ class RuleAnalyzer(data: String) { var rule = arrayOf(queue.substring(0, pos)) //压入分隔的首段规则到数组 + elementsType = queue.substring(pos, pos + step) //设置组合类型 pos += step //跳过分隔符 - elementsType = queue.substring(pos - step, pos) //设置组合类型 - while (pos < st && consumeToAny { matchesAny( * split ) }) { - rule += queue.substring(start, pos) //循环切分规则压入数组 + while (consumeToAny( * split ) && pos < st ) { //循环切分规则压入数组 + rule += queue.substring(start, pos) pos += step //跳过分隔符 } + rule + }else null pos = st //位置推移到筛选器处 @@ -316,9 +324,7 @@ class RuleAnalyzer(data: String) { if (!consumeTo(elementsType,false)) return rules + queue.substring(start) //此处consumeTo(...)开始位置不是规则的开始位置,start沿用上次设置 - end = pos - val st = if( consumeToAny(false){ matchesAny( '(','[' ) } )pos else -1 //查找筛选器 - pos = end + val st = findToAny( '(','[' ) //查找筛选器 if(st == -1) { var rule = rules + queue.substring(start, pos) //压入本次分隔的首段规则到数组 @@ -327,14 +333,14 @@ class RuleAnalyzer(data: String) { rule += queue.substring(start, pos) pos += step //跳过分隔符 } - rule += queue.substring(start) //将剩余字段压入数组末尾 + rule += queue.substring(pos) //将剩余字段压入数组末尾 return rule } val rule = if(st > pos ){//先匹配到st1pos，表明"&&","||"不在选择器中，将选择器前"&&","||"分隔的字段依次压入数组 var rule = rules + queue.substring(start, pos) //压入本次分隔的首段规则到数组 pos += step //跳过分隔符 - while (pos < st && consumeTo(elementsType)) { //循环切分规则压入数组 + while (consumeTo(elementsType) && pos < st) { //循环切分规则压入数组 rule += queue.substring(start, pos) pos += step //跳过分隔符 } @@ -512,4 +518,4 @@ class RuleAnalyzer(data: String) { val splitListDefault = arrayOf("-",".","!","@","@@") } -} \ No newline at end of file +} From d8a83f9e156a9a238d644d6df9a1abfd5050209f Mon Sep 17 00:00:00 2001 From: bushixuanqi <57338301+bushixuanqi@users.noreply.github.com> Date: Sun, 4 Jul 2021 14:35:07 +0800 Subject: [PATCH 2/5] Update RuleAnalyzer.kt MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 优化注释描述 --- .../java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt b/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt index 09129de7a..2bcf3555f 100644 --- a/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt +++ b/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt @@ -255,7 +255,7 @@ class RuleAnalyzer(data: String) { /** * 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"&&"或"||"或"%%"而导致的冲突 + * 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{ //首段匹配,elementsType为空 @@ -284,7 +284,7 @@ class RuleAnalyzer(data: String) { return rule } - val rule = if(st >pos ){ //先匹配到st1pos，表明"&&","||"不在选择器中，将选择器前"&&","||"分隔的字段依次压入数组 + val rule = if(st >pos ){ //先匹配到st1pos，表明分隔字串不在选择器中，将选择器前分隔字串分隔的字段依次压入数组 var rule = arrayOf(queue.substring(0, pos)) //压入分隔的首段规则到数组 @@ -337,7 +337,7 @@ class RuleAnalyzer(data: String) { return rule } - val rule = if(st > pos ){//先匹配到st1pos，表明"&&","||"不在选择器中，将选择器前"&&","||"分隔的字段依次压入数组 + val rule = if(st > pos ){//先匹配到st1pos，表明分隔字串不在选择器中，将选择器前分隔字串分隔的字段依次压入数组 var rule = rules + queue.substring(start, pos) //压入本次分隔的首段规则到数组 pos += step //跳过分隔符 while (consumeTo(elementsType) && pos < st) { //循环切分规则压入数组 From c9f2071e004c41dc1345995102dc573adf9a0cfa Mon Sep 17 00:00:00 2001 From: bushixuanqi <57338301+bushixuanqi@users.noreply.github.com> Date: Sun, 4 Jul 2021 14:37:37 +0800 Subject: [PATCH 3/5] Update AnalyzeByJSoup.kt MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 允许索引作为首规则，此时相对于children。 head@[1]@text 与 head@children[1]@text 等价 --- .../io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/app/src/main/java/io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt b/app/src/main/java/io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt index 4ce3be1bb..be1615a46 100644 --- a/app/src/main/java/io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt +++ b/app/src/main/java/io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt @@ -213,7 +213,7 @@ class AnalyzeByJSoup(doc: Any) { val curList = mutableListOf() //当前数字区间 var l = "" //暂存数字字符串 - val head = rus[rus.length-1] == ']' //是否为常规索引写法 + val head = rus.last() == ']' //是否为常规索引写法 if(head){ //常规索引写法[index...] @@ -236,7 +236,7 @@ class AnalyzeByJSoup(doc: Any) { else -> { - //为保证查找顺序，区间和单个索引都添加到同一集合 + //为保证查找顺序，区间和单个索引都添加到同一集合 if(curList.isEmpty())indexSet.indexs.add(curInt!!) else{ @@ -265,7 +265,7 @@ class AnalyzeByJSoup(doc: Any) { curMinus = false //重置 } } - } else while (len --> 1) { //阅读原本写法，逆向遍历,至少两位前置字符,如 p. + } else while (len --> 0) { //阅读原本写法，逆向遍历,至少两位前置字符,如 . val rl = rus[len] if (rl == ' ') continue //跳过空格 @@ -314,8 +314,9 @@ class AnalyzeByJSoup(doc: Any) { val rules = ruleStr.split(".") elements.addAll( - when (rules[0]) { - "children" -> temp.children() + if(ruleStr.isEmpty()) temp.children() //允许索引直接作为根元素，此时前置规则为空，效果与children相同 + else when (rules[0]) { + "children" -> temp.children() //允许索引直接作为根元素，此时前置规则为空，效果与children相同 "class" -> temp.getElementsByClass(rules[1]) "tag" -> temp.getElementsByTag(rules[1]) "id" -> Collector.collect(Evaluator.Id(rules[1]), temp) From 0978ff83fd4c059c1d6b2bbebd91794f18536b6a Mon Sep 17 00:00:00 2001 From: bushixuanqi <57338301+bushixuanqi@users.noreply.github.com> Date: Sun, 4 Jul 2021 15:00:03 +0800 Subject: [PATCH 4/5] Update RuleAnalyzer.kt MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 修正注释描述 --- .../java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt b/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt index 2bcf3555f..50d91ba15 100644 --- a/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt +++ b/app/src/main/java/io/legado/app/model/analyzeRule/RuleAnalyzer.kt @@ -98,9 +98,9 @@ class RuleAnalyzer(data: String) { } /** - * 从剩余字串中拉出一个字符串，直到且包括匹配序列（匹配参数列表中一项即为匹配），或剩余字串用完。 + * 从剩余字串中拉出一个字符串，直到但不包括匹配序列（匹配参数列表中一项即为匹配），或剩余字串用完。 * @param seq 匹配字符串序列 - * @return 成功返回true并推移pos，失败返回fasle而不推移pos + * @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 { @@ -126,7 +126,7 @@ class RuleAnalyzer(data: String) { /** * 从剩余字串中拉出一个字符串，直到但不包括匹配序列（匹配参数列表中一项即为匹配），或剩余字串用完。 * @param seq 匹配字符序列 - * @return 匹配到的位置 + * @return 返回匹配位置 */ private fun findToAny(vararg seq:Char): Int { From d01d77fc4ffda227221a9a4919cb4bc196230d5d Mon Sep 17 00:00:00 2001 From: bushixuanqi <57338301+bushixuanqi@users.noreply.github.com> Date: Sun, 4 Jul 2021 15:22:38 +0800 Subject: [PATCH 5/5] Update AnalyzeByJSoup.kt MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 扩大索引向前的匹配范围，索引前无规则时相当于前面的规则是children --- .../java/io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/app/src/main/java/io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt b/app/src/main/java/io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt index be1615a46..3bb2e4e33 100644 --- a/app/src/main/java/io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt +++ b/app/src/main/java/io/legado/app/model/analyzeRule/AnalyzeByJSoup.kt @@ -219,7 +219,7 @@ class AnalyzeByJSoup(doc: Any) { len-- //跳过尾部']' - while (len-- > 0) { //逆向遍历,至少有一位前置字符,如 [ + while (len-- >= 0) { //逆向遍历,可以无前置规则 var rl = rus[len] if (rl == ' ') continue //跳过空格 @@ -265,7 +265,7 @@ class AnalyzeByJSoup(doc: Any) { curMinus = false //重置 } } - } else while (len --> 0) { //阅读原本写法，逆向遍历,至少两位前置字符,如 . + } else while (len-- >= 0) { //阅读原本写法，逆向遍历,可以无前置规则 val rl = rus[len] if (rl == ' ') continue //跳过空格