Javascript 如何基于单词作为键分隔符来读取文本文件的块?
我有一个.txt文件,格式如下:Javascript 如何基于单词作为键分隔符来读取文本文件的块?,javascript,arrays,node.js,file,lodash,Javascript,Arrays,Node.js,File,Lodash,我有一个.txt文件,格式如下: Part #368 - XXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX XXXXXXXXXXXXXXXXX
Part #368 - XXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
Part #369 - XXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
Part #370 - XXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
我是这样读文件的:
var lines = fs.readFileSync('file.txt', 'utf-8')
.split('\n')
.filter(Boolean);
所以它返回一个文件行数组。如何获取以“Part”字符串开头的文件块
类似于这样,但不是获取以“Part”开头的字符串,而是要从“Part”字符串到下一个“Part”字符串的所有行。这将创建一个行数组
var parts = _.reduce(lines, function( result, line ) {
if (line.indexOf('Part') !== -1) result.push([]);
_.last(result).push(line);
return result;
}, []);
JSON流
根据,下面是一个从扩展而来的类,该类将文件分隔为JSON数组流:
const { Transform } = require('stream');
class Delimited extends Transform {
constructor({ delimiter = /\r?\n/g, encoding = 'utf8' } = {}) {
super();
// initialize internal values
this._delimiter = delimiter instanceof RegExp ? delimiter : new RegExp(delimiter, 'g');
this._encoding = encoding;
this._buffer = '';
this._first = true;
}
_transform(chunk, encoding, callback) {
// convert input encoding into output encoding
// and append to internal buffer
if (encoding === 'buffer') {
this._buffer += chunk.toString(this._encoding);
} else if (encoding === this._encoding) {
this._buffer += chunk;
} else {
this._buffer += Buffer.from(chunk, encoding).toString(this._encoding);
}
let partialJSON = '';
// check if delimiter is found
if (this._delimiter.test(this._buffer)) {
// split internal buffer by delimiter
let sections = this._buffer.split(this._delimiter);
// put possibly incomplete section from array back into internal buffer
this._buffer = sections.pop();
// add each section to partial json array
sections.forEach(section => {
partialJSON += `${this._first ? '[' : ','}${JSON.stringify(section)}`;
this._first = false;
});
}
// push partial json array to readable stream
callback(null, partialJSON);
}
_flush(callback) {
// add remaining buffer as last section to json array
callback(null, `${this._first ? '[' : ','}${JSON.stringify(this._buffer)}]`);
}
}
用法示例:
const fs = require('fs');
let stream = fs.createReadStream('file.txt', 'utf8');
let transform = new Delimited({ delimiter: /\n\n(?=Part #\d)/g });
let json = '';
transform.on('data', (chunk) => json += chunk);
transform.on('end', () => console.log(JSON.parse(json)));
stream.pipe(transform);
const fs = require('fs');
let stream = fs.createReadStream('file.txt', 'utf8');
let transform = new Delimited(/\n\n(?=Part #\d)/g);
let array = [];
transform.on('data', (chunk) => array.push(chunk));
transform.on('end', () => console.log(array));
stream.pipe(transform);
串流
或者,如果您不希望将JSON传输到另一个文件、进程或作为客户端响应,则可以通过将输出流设置为objectMode:true
,将每个部分作为块发送:
const { Transform } = require('stream');
class Delimited extends Transform {
constructor(delimiter = /\r?\n/g) {
super({ objectMode: true });
// initialize internal values
this._delimiter = delimiter instanceof RegExp ? delimiter : new RegExp(delimiter, 'g');
this._encoding = 'utf8';
this._buffer = '';
this._first = true;
}
_transform(chunk, encoding, callback) {
// convert input encoding into output encoding
// and append to internal buffer
if (encoding === 'buffer') {
this._buffer += chunk.toString(this._encoding);
} else if (encoding === this._encoding) {
this._buffer += chunk;
} else {
this._buffer += Buffer.from(chunk, encoding).toString(this._encoding);
}
if (this._delimiter.test(this._buffer)) {
// split internal buffer by delimiter
let sections = this._buffer.split(this._delimiter);
// put possibly incomplete section from array back into internal buffer
this._buffer = sections.pop();
// push each section to readable stream in object mode
sections.forEach(this.push, this);
}
callback();
}
_flush(callback) {
// push remaining buffer to readable stream
callback(null, this._buffer);
}
}
用法示例:
const fs = require('fs');
let stream = fs.createReadStream('file.txt', 'utf8');
let transform = new Delimited({ delimiter: /\n\n(?=Part #\d)/g });
let json = '';
transform.on('data', (chunk) => json += chunk);
transform.on('end', () => console.log(JSON.parse(json)));
stream.pipe(transform);
const fs = require('fs');
let stream = fs.createReadStream('file.txt', 'utf8');
let transform = new Delimited(/\n\n(?=Part #\d)/g);
let array = [];
transform.on('data', (chunk) => array.push(chunk));
transform.on('end', () => console.log(array));
stream.pipe(transform);
像
fs.readFileSync('file.txt','utf-8').split('Part')这样的东西就足够了吗?
对于Node.js中的转换流来说,这是一个非常好的用例。流将缓冲数据直到下一部分,然后一次性发出整个部分。(将流的可读部分置于对象模式。)最好处理此流,这样在做任何事情之前,您就不需要将整个文件缓冲到内存中。@Italoyers它可以工作,但它会切断“手”字符串。@Brad我知道时间晚了,但请查看我的答案。谢谢你的主意!概念上很好,但是最后一行用括号调用回调有什么用呢?还有,你为什么要从这里输出JSON呢?不要做自己的JSON序列化,这里不需要JSON。。。只需在对象模式下输出即可。@Brad其想法是在可写流结束时调用\u flush()
,因此内部缓冲区中剩余的部分是最后一部分,因为在其中从未找到分隔符,因此在验证流结束时需要将其添加到\u flush()
。如果它不是序列化的,那么使用这种转换就没有什么意义了。如果您希望整个数据最终都缓冲在内存中,那么在这种情况下,您也可以在正常流的末尾使用.split(delimiter)
。这对于将JSON传输到另一个文件或客户端响应等更有用。@Brad I添加了objectMode:true
流。我想这就是你的意思。