优化
This commit is contained in:
gedoor
@@ -51,10 +51,12 @@ class TextFile(private val book: Book) {
|
|||||||
} else {
|
} else {
|
||||||
book.tocUrl.toPattern(Pattern.MULTILINE)
|
book.tocUrl.toPattern(Pattern.MULTILINE)
|
||||||
}
|
}
|
||||||
return analyze(rulePattern)
|
return rulePattern?.let {
|
||||||
|
analyze(rulePattern)
|
||||||
|
} ?: analyze()
|
||||||
}
|
}
|
||||||
|
|
||||||
private fun analyze(pattern: Pattern?): ArrayList<BookChapter> {
|
private fun analyze(pattern: Pattern): ArrayList<BookChapter> {
|
||||||
val toc = arrayListOf<BookChapter>()
|
val toc = arrayListOf<BookChapter>()
|
||||||
LocalBook.getBookInputStream(book).use { bis ->
|
LocalBook.getBookInputStream(book).use { bis ->
|
||||||
var blockContent: String
|
var blockContent: String
|
||||||
@@ -76,142 +78,103 @@ class TextFile(private val book: Book) {
|
|||||||
.also { length = it } > 0
|
.also { length = it } > 0
|
||||||
) {
|
) {
|
||||||
blockPos++
|
blockPos++
|
||||||
//如果存在Chapter
|
var end = bufferStart + length
|
||||||
if (pattern != null) {
|
for (i in bufferStart + length - 1 downTo 0) {
|
||||||
var end = bufferStart + length
|
if (buffer[i] == blank) {
|
||||||
for (i in bufferStart + length - 1 downTo 0) {
|
end = i
|
||||||
if (buffer[i] == blank) {
|
break
|
||||||
end = i
|
|
||||||
break
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
//将数据转换成String, 不能超过length
|
}
|
||||||
blockContent = String(buffer, 0, end, charset)
|
//将数据转换成String, 不能超过length
|
||||||
buffer.copyInto(buffer, 0, end, bufferStart + length)
|
blockContent = String(buffer, 0, end, charset)
|
||||||
bufferStart = bufferStart + length - end
|
buffer.copyInto(buffer, 0, end, bufferStart + length)
|
||||||
length = end
|
bufferStart = bufferStart + length - end
|
||||||
//当前Block下使过的String的指针
|
length = end
|
||||||
var seekPos = 0
|
//当前Block下使过的String的指针
|
||||||
//进行正则匹配
|
var seekPos = 0
|
||||||
val matcher: Matcher = pattern.matcher(blockContent)
|
//进行正则匹配
|
||||||
//如果存在相应章节
|
val matcher: Matcher = pattern.matcher(blockContent)
|
||||||
while (matcher.find()) { //获取匹配到的字符在字符串中的起始位置
|
//如果存在相应章节
|
||||||
val chapterStart = matcher.start()
|
while (matcher.find()) { //获取匹配到的字符在字符串中的起始位置
|
||||||
//获取章节内容
|
val chapterStart = matcher.start()
|
||||||
val chapterContent = blockContent.substring(seekPos, chapterStart)
|
//获取章节内容
|
||||||
val chapterLength = chapterContent.toByteArray(charset).size
|
val chapterContent = blockContent.substring(seekPos, chapterStart)
|
||||||
val lastStart = toc.lastOrNull()?.start ?: 0
|
val chapterLength = chapterContent.toByteArray(charset).size
|
||||||
if (curOffset + chapterLength - lastStart > 50000) {
|
val lastStart = toc.lastOrNull()?.start ?: 0
|
||||||
bis.close()
|
if (curOffset + chapterLength - lastStart > 50000) {
|
||||||
//移除不匹配的规则
|
|
||||||
tocRules.removeFirstOrNull()
|
|
||||||
return analyze(tocRules.firstOrNull()?.rule?.toPattern(Pattern.MULTILINE))
|
|
||||||
}
|
|
||||||
//如果 seekPos == 0 && nextChapterPos != 0 表示当前block处前面有一段内容
|
|
||||||
//第一种情况一定是序章 第二种情况是上一个章节的内容
|
|
||||||
if (seekPos == 0 && chapterStart != 0) { //获取当前章节的内容
|
|
||||||
if (toc.isEmpty()) { //如果当前没有章节，那么就是序章
|
|
||||||
//加入简介
|
|
||||||
if (StringUtils.trim(chapterContent).isNotEmpty()) {
|
|
||||||
val qyChapter = BookChapter()
|
|
||||||
qyChapter.title = "前言"
|
|
||||||
qyChapter.start = 0
|
|
||||||
qyChapter.end = chapterLength.toLong()
|
|
||||||
toc.add(qyChapter)
|
|
||||||
}
|
|
||||||
//创建当前章节
|
|
||||||
val curChapter = BookChapter()
|
|
||||||
curChapter.title = matcher.group()
|
|
||||||
curChapter.start = chapterLength.toLong()
|
|
||||||
toc.add(curChapter)
|
|
||||||
} else { //否则就block分割之后，上一个章节的剩余内容
|
|
||||||
//获取上一章节
|
|
||||||
val lastChapter = toc.last()
|
|
||||||
//将当前段落添加上一章去
|
|
||||||
lastChapter.end =
|
|
||||||
lastChapter.end!! + chapterLength.toLong()
|
|
||||||
//创建当前章节
|
|
||||||
val curChapter = BookChapter()
|
|
||||||
curChapter.title = matcher.group()
|
|
||||||
curChapter.start = lastChapter.end
|
|
||||||
toc.add(curChapter)
|
|
||||||
}
|
|
||||||
} else {
|
|
||||||
if (toc.isNotEmpty()) { //获取章节内容
|
|
||||||
//获取上一章节
|
|
||||||
val lastChapter = toc.last()
|
|
||||||
lastChapter.end =
|
|
||||||
lastChapter.start!! + chapterContent.toByteArray(charset).size.toLong()
|
|
||||||
//创建当前章节
|
|
||||||
val curChapter = BookChapter()
|
|
||||||
curChapter.title = matcher.group()
|
|
||||||
curChapter.start = lastChapter.end
|
|
||||||
toc.add(curChapter)
|
|
||||||
} else { //如果章节不存在则创建章节
|
|
||||||
val curChapter = BookChapter()
|
|
||||||
curChapter.title = matcher.group()
|
|
||||||
curChapter.start = 0
|
|
||||||
curChapter.end = 0
|
|
||||||
toc.add(curChapter)
|
|
||||||
}
|
|
||||||
}
|
|
||||||
//设置指针偏移
|
|
||||||
seekPos += chapterContent.length
|
|
||||||
}
|
|
||||||
if (seekPos == 0 && length > 50000) {
|
|
||||||
bis.close()
|
bis.close()
|
||||||
//移除不匹配的规则
|
//移除不匹配的规则
|
||||||
tocRules.remove(tocRules.removeFirstOrNull())
|
tocRules.removeFirstOrNull()
|
||||||
return analyze(tocRules.firstOrNull()?.rule?.toPattern(Pattern.MULTILINE))
|
return tocRules.firstOrNull()?.let {
|
||||||
|
analyze(it.rule.toPattern(Pattern.MULTILINE))
|
||||||
|
} ?: analyze()
|
||||||
}
|
}
|
||||||
} else { //进行本地虚拟分章
|
//如果 seekPos == 0 && nextChapterPos != 0 表示当前block处前面有一段内容
|
||||||
bufferStart = 0
|
//第一种情况一定是序章 第二种情况是上一个章节的内容
|
||||||
//章节在buffer的偏移量
|
if (seekPos == 0 && chapterStart != 0) { //获取当前章节的内容
|
||||||
var chapterOffset = 0
|
if (toc.isEmpty()) { //如果当前没有章节，那么就是序章
|
||||||
//当前剩余可分配的长度
|
//加入简介
|
||||||
var strLength = length
|
if (StringUtils.trim(chapterContent).isNotEmpty()) {
|
||||||
//分章的位置
|
val qyChapter = BookChapter()
|
||||||
var chapterPos = 0
|
qyChapter.title = "前言"
|
||||||
while (strLength > 0) {
|
qyChapter.start = 0
|
||||||
++chapterPos
|
qyChapter.end = chapterLength.toLong()
|
||||||
//是否长度超过一章
|
toc.add(qyChapter)
|
||||||
if (strLength > maxLengthWithNoToc) { //在buffer中一章的终止点
|
|
||||||
var end = length
|
|
||||||
//寻找换行符作为终止点
|
|
||||||
for (i in chapterOffset + maxLengthWithNoToc until length) {
|
|
||||||
if (buffer[i] == blank) {
|
|
||||||
end = i
|
|
||||||
break
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
val chapter = BookChapter()
|
//创建当前章节
|
||||||
chapter.title = "第${blockPos}章($chapterPos)"
|
val curChapter = BookChapter()
|
||||||
chapter.start = curOffset + chapterOffset + 1
|
curChapter.title = matcher.group()
|
||||||
chapter.end = curOffset + end
|
curChapter.start = chapterLength.toLong()
|
||||||
toc.add(chapter)
|
toc.add(curChapter)
|
||||||
//减去已经被分配的长度
|
} else { //否则就block分割之后，上一个章节的剩余内容
|
||||||
strLength -= (end - chapterOffset)
|
//获取上一章节
|
||||||
//设置偏移的位置
|
val lastChapter = toc.last()
|
||||||
chapterOffset = end
|
//将当前段落添加上一章去
|
||||||
} else {
|
lastChapter.end =
|
||||||
val chapter = BookChapter()
|
lastChapter.end!! + chapterLength.toLong()
|
||||||
chapter.title = "第" + blockPos + "章" + "(" + chapterPos + ")"
|
//创建当前章节
|
||||||
chapter.start = curOffset + chapterOffset + 1
|
val curChapter = BookChapter()
|
||||||
chapter.end = curOffset + length
|
curChapter.title = matcher.group()
|
||||||
toc.add(chapter)
|
curChapter.start = lastChapter.end
|
||||||
strLength = 0
|
toc.add(curChapter)
|
||||||
|
}
|
||||||
|
} else {
|
||||||
|
if (toc.isNotEmpty()) { //获取章节内容
|
||||||
|
//获取上一章节
|
||||||
|
val lastChapter = toc.last()
|
||||||
|
lastChapter.end =
|
||||||
|
lastChapter.start!! + chapterContent.toByteArray(charset).size.toLong()
|
||||||
|
//创建当前章节
|
||||||
|
val curChapter = BookChapter()
|
||||||
|
curChapter.title = matcher.group()
|
||||||
|
curChapter.start = lastChapter.end
|
||||||
|
toc.add(curChapter)
|
||||||
|
} else { //如果章节不存在则创建章节
|
||||||
|
val curChapter = BookChapter()
|
||||||
|
curChapter.title = matcher.group()
|
||||||
|
curChapter.start = 0
|
||||||
|
curChapter.end = 0
|
||||||
|
toc.add(curChapter)
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
//设置指针偏移
|
||||||
|
seekPos += chapterContent.length
|
||||||
|
}
|
||||||
|
if (seekPos == 0 && length > 50000) {
|
||||||
|
bis.close()
|
||||||
|
//移除不匹配的规则
|
||||||
|
tocRules.remove(tocRules.removeFirstOrNull())
|
||||||
|
return tocRules.firstOrNull()?.let {
|
||||||
|
analyze(it.rule.toPattern(Pattern.MULTILINE))
|
||||||
|
} ?: analyze()
|
||||||
}
|
}
|
||||||
|
|
||||||
//block的偏移点
|
//block的偏移点
|
||||||
curOffset += length.toLong()
|
curOffset += length.toLong()
|
||||||
|
|
||||||
if (pattern != null) {
|
//设置上一章的结尾
|
||||||
//设置上一章的结尾
|
val lastChapter = toc.last()
|
||||||
val lastChapter = toc.last()
|
lastChapter.end = curOffset
|
||||||
lastChapter.end = curOffset
|
|
||||||
}
|
|
||||||
|
|
||||||
//当添加的block太多的时候，执行GC
|
//当添加的block太多的时候，执行GC
|
||||||
if (blockPos % 15 == 0) {
|
if (blockPos % 15 == 0) {
|
||||||
@@ -231,7 +194,92 @@ class TextFile(private val book: Book) {
|
|||||||
|
|
||||||
System.gc()
|
System.gc()
|
||||||
System.runFinalization()
|
System.runFinalization()
|
||||||
book.tocUrl = pattern?.pattern() ?: ""
|
book.tocUrl = pattern.pattern()
|
||||||
|
book.save()
|
||||||
|
return toc
|
||||||
|
}
|
||||||
|
|
||||||
|
private fun analyze(): ArrayList<BookChapter> {
|
||||||
|
val toc = arrayListOf<BookChapter>()
|
||||||
|
LocalBook.getBookInputStream(book).use { bis ->
|
||||||
|
//加载章节
|
||||||
|
var curOffset: Long = 0
|
||||||
|
//block的个数
|
||||||
|
var blockPos = 0
|
||||||
|
//读取的长度
|
||||||
|
var length: Int
|
||||||
|
val buffer = ByteArray(bufferSize)
|
||||||
|
var bufferStart = 3
|
||||||
|
bis.read(buffer, 0, 3)
|
||||||
|
if (Utf8BomUtils.hasBom(buffer)) {
|
||||||
|
bufferStart = 0
|
||||||
|
curOffset = 3
|
||||||
|
}
|
||||||
|
//获取文件中的数据到buffer，直到没有数据为止
|
||||||
|
while (bis.read(buffer, bufferStart, bufferSize - bufferStart)
|
||||||
|
.also { length = it } > 0
|
||||||
|
) {
|
||||||
|
bufferStart = 0
|
||||||
|
blockPos++
|
||||||
|
//章节在buffer的偏移量
|
||||||
|
var chapterOffset = 0
|
||||||
|
//当前剩余可分配的长度
|
||||||
|
var strLength = length
|
||||||
|
//分章的位置
|
||||||
|
var chapterPos = 0
|
||||||
|
while (strLength > 0) {
|
||||||
|
++chapterPos
|
||||||
|
//是否长度超过一章
|
||||||
|
if (strLength > maxLengthWithNoToc) { //在buffer中一章的终止点
|
||||||
|
var end = length
|
||||||
|
//寻找换行符作为终止点
|
||||||
|
for (i in chapterOffset + maxLengthWithNoToc until length) {
|
||||||
|
if (buffer[i] == blank) {
|
||||||
|
end = i
|
||||||
|
break
|
||||||
|
}
|
||||||
|
}
|
||||||
|
val chapter = BookChapter()
|
||||||
|
chapter.title = "第${blockPos}章($chapterPos)"
|
||||||
|
chapter.start = curOffset + chapterOffset + 1
|
||||||
|
chapter.end = curOffset + end
|
||||||
|
toc.add(chapter)
|
||||||
|
//减去已经被分配的长度
|
||||||
|
strLength -= (end - chapterOffset)
|
||||||
|
//设置偏移的位置
|
||||||
|
chapterOffset = end
|
||||||
|
} else {
|
||||||
|
val chapter = BookChapter()
|
||||||
|
chapter.title = "第" + blockPos + "章" + "(" + chapterPos + ")"
|
||||||
|
chapter.start = curOffset + chapterOffset + 1
|
||||||
|
chapter.end = curOffset + length
|
||||||
|
toc.add(chapter)
|
||||||
|
strLength = 0
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
//block的偏移点
|
||||||
|
curOffset += length.toLong()
|
||||||
|
|
||||||
|
//当添加的block太多的时候，执行GC
|
||||||
|
if (blockPos % 15 == 0) {
|
||||||
|
System.gc()
|
||||||
|
System.runFinalization()
|
||||||
|
}
|
||||||
|
}
|
||||||
|
}
|
||||||
|
for (i in toc.indices) {
|
||||||
|
val bean = toc[i]
|
||||||
|
bean.index = i
|
||||||
|
bean.bookUrl = book.bookUrl
|
||||||
|
bean.url = (MD5Utils.md5Encode16(book.originName + i + bean.title))
|
||||||
|
}
|
||||||
|
book.latestChapterTitle = toc.last().title
|
||||||
|
book.totalChapterNum = toc.size
|
||||||
|
|
||||||
|
System.gc()
|
||||||
|
System.runFinalization()
|
||||||
|
book.tocUrl = ""
|
||||||
book.save()
|
book.save()
|
||||||
return toc
|
return toc
|
||||||
}
|
}
|
||||||
|
|||||||
Reference in New Issue
Block a user