我有一个格式为 .txt 的文件:
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);
所以它返回文件行的数组。如何获取以“Part”字符串开头的文件 block ?
var parts = _.filter(lines,function( s ) { return s.indexOf( 'Part' ) !== -1; });
类似这样,但不是获取以“Part”开头的字符串,我想要从“Part”字符串到下一个“Part”字符串的所有行。
最佳答案
JSON 流
根据 @Brad's suggestion , 这是一个从 stream.Transform
扩展而来的类将文件分隔为 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)}]`);
}
}
示例用法:
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);
字符串流
或者,如果您不想将 JSON 传输到另一个文件、进程或作为客户端响应,您可以通过将输出流设置为 objectMode: true
将每个部分作为一个 block 发出:
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);
}
}
示例用法:
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);
关于javascript - 如何根据单词作为键分隔符读取文本文件的 block ?,我们在Stack Overflow上找到一个类似的问题: https://stackoverflow.com/questions/43927439/