web-scrapingtext-extractionpdftotextpdf-scraping
PDF scraping using textract module
I have a Node.js app that have to do some web scraping of online pdf. This is a piece of code:
var textract = require('textract');
const util = require('util');
var methods = {};
var urls = [
{year: '2016', link: 'http://www.url2016.pdf'},
{year: '2015', link: 'http://www.url2015.pdf'}
];
var result = [];
const textractFromUrl = util.promisify(textract.fromUrl);
methods.download = function(req, res) {
return extractText();
}
async function extractText() {
try {
var config = {
preserveLineBreaks: true
};
for(let url of urls) {
let text = await textractFromUrl(url.link, config);
switch(url.year) {
case '2015':
await extractTextType1(url, text);
break;
case '2016':
await extractTextType2(url, text);
break;
default:
console.log('Error: no switch case');
}
}
}
catch(err) {
console.log('catch block');
console.log(err);
}
}
As you can see, I use textrack package to scrape pdf.
When I run this app, I get:
catch block
{ Error: Error for type: [[ application/pdf ]], file: [[ C:\Users\myUserName\AppData\Local\Temp\96710848773.pdf ]], extractor for type exists, but failed to initialize. Message: INFO: 'pdftotext' does not appear to be installed, so textract will be unable to extract PDFs.
at extract (C:\Users\myUserName\projectPath\node_modules\textract\lib\extract.js:147:15)
at Timeout._onTimeout (C:\Users\myUserName\projectPath\node_modules\textract\lib\extract.js:155:7)
at ontimeout (timers.js:466:11)
at tryOnTimeout (timers.js:304:5)
at Timer.listOnTimeout (timers.js:267:5) typeNotFound: true }
In the npm textract module page, is written that PDF extraction requires pdftotext be installed, link.
So I go to http://www.foolabs.com/xpdf/download.html and I downloaded and istalled Download XpdfReader: Windows 64-bit.
I try again to run the app using node app.js (app.js is the main file of my app) but I get the same error so I download Download the Xpdf tools: Windows 64-bit.
It's a zip file, I extract that file and then I try to install pdftotext.exe but when I double click on pdftotext.exe, anything happens.
I also try to install that using Administrator privileges. Nothing.
I'm using Windows 10, 64 bit.
What I have to do?
EDIT 1
As recommended, I copied the pdftotext.exe file to C:\Windows\System32. Then I ran my program again and I have this problem:
catch block
TypeError: Cannot read property 'split' of undefined
at extractTextType4 (C:\myUserName\projectPath\file.js:301:28)
at extractText (C:\myUserName\projectPath\file.js:78:12)
at <anonymous>
at process._tickCallback (internal/process/next_tick.js:182:7)
Do I have to change the environment variables?
EDIT 2
My C:\myUserName\projectPath\file.js is this:
var textract = require('textract');
const util = require('util');
var utilFunc = require('../helpers/utilFunc.js'); // my lib
var postgreSQLlib = require('../middlewares/postgreSQLlib.js'); // my lib
// object of methods
var methods = {};
// object of pdf links
var urls = [
{year: '2014', link: 'http://www.salute.gov.it/imgs/C_17_tavole_20_allegati_iitemAllegati_0_fileAllegati_itemFile_2_file.pdf'},
{year: '2013', link: 'http://www.salute.gov.it/imgs/C_17_tavole_20_allegati_iitemAllegati_0_fileAllegati_itemFile_1_file.pdf'},
{year: '2012', link: 'http://www.salute.gov.it/imgs/C_17_tavole_20_allegati_iitemAllegati_5_fileAllegati_itemFile_0_file.pdf'}
];
var result = [];
const textractFromUrl = util.promisify(textract.fromUrl);
/**
* Do web scraping and save info on DB.
*/
methods.download = function(req, res) {
return extractText();
}
/**
* Switch between different type of pdf file.
*/
async function extractText() {
try {
var config = {
preserveLineBreaks: true
};
for(let url of urls) {
let text = await textractFromUrl(url.link, config);
switch(url.year) {
case '2012':
await extractTextType1(url, text);
break;
case '2013':
await extractTextType2(url, text);
break;
case '2014':
await extractTextType3(url, text);
break;
default:
console.log('Error: no switch case');
}
}
}
catch(err) {
console.log('catch block');
console.log(err);
}
}
/**
* Save data on DB.
*/
function saveOnObject(vaccines, regions, map, url) {
vaccines = map.shift(); // remove and return the first element
regions = map.shift(); // remove and return the first element
let promises = [];
for(var i = 0; i <= 21; i++) {
var line = map[i];
line.forEach(async function(value, index) {
// make the values uniform
var vac = utilFunc.makeUniform(vaccines[index], 'vaccine');
var reg = utilFunc.makeUniform(regions[i], 'region');
var perc = utilFunc.makeUniform(value, 'value');
// create json object
var obj = utilFunc.createJsonObjectCov(parseFloat(url.year), 'Italy', vac, reg, perc);
covDataItAll.push(obj);
});
}
return Promise.all(promises);
}
/**
* Extract text to pdf 2000-2012.
*/
function extractTextType1(url, text) {
var matrix = [];
var map = [];
var vaccines = [];
var regionsTemp = [];
var regions = [];
var regionLength = [1, 2, 1, 2, 3, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1];
// text to matrix
var textArray = text.split('\n');
for(var i = 0; i < 23; i++) {
matrix[i] = textArray[i].split(' ');
}
// create vaccines array
matrix[0].shift();
vaccines = matrix[0];
map[0] = vaccines;
// create regionsTemp and values arrays
for(var i = 0; i < regionLength.length; i++) { // i index for regionLength
var j = i + 1; // index for matrix
var indexToRemove = 0;
var numberToRemove = regionLength[i];
var region = matrix[j].splice(indexToRemove, numberToRemove);
regionsTemp.push(region);
map[j+1] = matrix[j];
}
// create regions array (merge some elements)
for(var i = 0; i < regionsTemp.length; i++) {
var region = '';
if(regionLength[i] > 1) {
region = regionsTemp[i].join(' ');
}
else {
region = regionsTemp[i].join('');
}
regions.push(region);
}
map[1] = regions;
// remove \r char from map
for(var i = 0; i < map.length; i++) {
for(var j = 0; j < map[i].length; j++) {
map[i][j] = map[i][j].replace(/\r/g, '');
}
}
return saveOnObject(vaccines, regions, map, url);
}
/**
* Extract text to pdf 2013.
*/
function extractTextType2(url, text) {
var matrix = [];
var map = [];
var vaccines = [];
var regions = [];
// text to matrix
var textArray = text.split('\n');
for(var i = 0; i < 36; i++) {
matrix[i] = textArray[i].split(' ');
}
// create vaccines array
vaccines.push(matrix[0][1].replace(/\(a\)/g, '').replace(/\(b\)/g, '').replace(/\(c\)/g, '').replace(/\r/g, ''));
for(var i = 1; i < 10; i++) {
vaccines.push(matrix[i][0].replace(/\(a\)/g, '').replace(/\(b\)/g, '').replace(/\(c\)/g, '').replace(/\r/g, ''));
}
var meningo = ''.concat(matrix[10][0], matrix[11]).replace(/\(a\)/g, '').replace(/\(b\)/g, '').replace(/\(c\)/g, '').replace(/\r/g, '');
vaccines.push(meningo);
var pneumo = ''.concat(matrix[12][0], ' ', matrix[13]).replace(/\(a\)/g, '').replace(/\(b\)/g, '').replace(/\(c\)/g, '').replace(/\r/g, '');
vaccines.push(pneumo);
map[0] = vaccines;
// create regions array
for(var i = 14; i < matrix.length; i++) {
regions.push(matrix[i][0]);
}
map[1] = regions;
// create values array
for(var i = 14; i < matrix.length; i++) {
matrix[i].shift();
map.push(matrix[i]);
}
// remove \r char from map
for(var i = 0; i < map.length; i++) {
for(var j = 0; j < map[i].length; j++) {
map[i][j] = map[i][j].replace(/\r/g, '');
}
}
return saveOnObject(vaccines, regions, map, url);
}
/**
* Extract text to pdf 2014.
*/
function extractTextType3(url, text) {
var matrix = [];
var map = [];
var vaccines = [];
var regionsTemp = [];
var regions = [];
var regionLength = [1, 2, 1, 3, 3, 1, 3, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2]; // array that contains the length of the regions (I need to correctly split the arrays)
// text to matrix
var textArray = text.split('\n');
for(var i = 0; i < 36; i++) {
matrix[i] = textArray[i].split(' ');
}
// create vaccines array
vaccines.push(matrix[0][2].replace(/\(a\)/g, '').replace(/\(b\)/g, '').replace(/\(c\)/g, '').replace(/\r/g, ''));
for(var i = 1; i < 10; i++) {
vaccines.push(matrix[i][0].replace(/\(a\)/g, '').replace(/\(b\)/g, '').replace(/\(c\)/g, '').replace(/\r/g, ''));
}
var meningo = ''.concat(matrix[10][0], ' ', matrix[10][1], ' ', matrix[11]).replace(/\(a\)/g, '').replace(/\(b\)/g, '').replace(/\(c\)/g, '').replace(/\r/g, '');
vaccines.push(meningo);
var pneumo = ''.concat(matrix[12][0], ' ', matrix[13]).replace(/\(a\)/g, '').replace(/\(b\)/g, '').replace(/\(c\)/g, '').replace(/\r/g, '');
vaccines.push(pneumo);
map[0] = vaccines;
// create regionsTemp and values arrays
for(var i = 0; i < regionLength.length; i++) { // i index for regionLength
var j = i + 14; // index for matrix
var indexToRemove = 0;
var numberToRemove = regionLength[i];
var region = matrix[j].splice(indexToRemove, numberToRemove);
regionsTemp.push(region);
map[i+2] = matrix[j];
}
// create regions array (merge some elements)
for(var i = 0; i < regionsTemp.length; i++) {
var region = '';
if(regionLength[i] > 1) {
region = regionsTemp[i].join(' ');
}
else {
region = regionsTemp[i].join('');
}
regions.push(region);
}
map[1] = regions;
// remove \r char from map
for(var i = 0; i < map.length; i++) {
for(var j = 0; j < map[i].length; j++) {
map[i][j] = map[i][j].replace(/\r/g, '');
}
}
return saveOnObject(vaccines, regions, map, url);
}
// exports methods
module.exports = methods;
Solution
Copy the the file to pdftotext.exe to either c:\windows or c:\windows\system32.
Basically any folder which is added to your PATH. Then it should work
You can see in the source code
The code assumes pdftotext to be in path
