Skip to content

Instantly share code, notes, and snippets.

@jamlfy

jamlfy/getTexts.js

Last active January 3, 2017 21:20
Download ZIP

Revisions

  1. AlejoNext revised this gist Jan 3, 2017. 1 changed file with 41 additions and 0 deletions.
    41 changes: 41 additions & 0 deletions phantomjs.log
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,41 @@
    2017-01-03T16:17:16 [DEBUG] CookieJar - Created but will not store cookies (use option '--cookies-file=<filename>' to enable persistent cookie storage)
    2017-01-03T16:17:22 [DEBUG] Set "http" proxy to: "" : 1080
    2017-01-03T16:17:22 [DEBUG] Phantom - execute: Configuration
    2017-01-03T16:17:22 [DEBUG] 0 objectName : ""
    2017-01-03T16:17:22 [DEBUG] 1 cookiesFile : ""
    2017-01-03T16:17:22 [DEBUG] 2 diskCacheEnabled : "false"
    2017-01-03T16:17:22 [DEBUG] 3 maxDiskCacheSize : "-1"
    2017-01-03T16:17:22 [DEBUG] 4 diskCachePath : ""
    2017-01-03T16:17:22 [DEBUG] 5 ignoreSslErrors : "false"
    2017-01-03T16:17:22 [DEBUG] 6 localUrlAccessEnabled : "true"
    2017-01-03T16:17:22 [DEBUG] 7 localToRemoteUrlAccessEnabled : "false"
    2017-01-03T16:17:22 [DEBUG] 8 outputEncoding : "UTF-8"
    2017-01-03T16:17:22 [DEBUG] 9 proxyType : "http"
    2017-01-03T16:17:22 [DEBUG] 10 proxy : ":1080"
    2017-01-03T16:17:22 [DEBUG] 11 proxyAuth : ":"
    2017-01-03T16:17:22 [DEBUG] 12 scriptEncoding : "UTF-8"
    2017-01-03T16:17:22 [DEBUG] 13 webSecurityEnabled : "true"
    2017-01-03T16:17:22 [DEBUG] 14 offlineStoragePath : ""
    2017-01-03T16:17:22 [DEBUG] 15 localStoragePath : ""
    2017-01-03T16:17:22 [DEBUG] 16 localStorageDefaultQuota : "-1"
    2017-01-03T16:17:22 [DEBUG] 17 offlineStorageDefaultQuota : "-1"
    2017-01-03T16:17:22 [DEBUG] 18 printDebugMessages : "true"
    2017-01-03T16:17:22 [DEBUG] 19 javascriptCanOpenWindows : "true"
    2017-01-03T16:17:22 [DEBUG] 20 javascriptCanCloseWindows : "true"
    2017-01-03T16:17:22 [DEBUG] 21 sslProtocol : "default"
    2017-01-03T16:17:22 [DEBUG] 22 sslCiphers : "ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES128-GCM-SHA256:DHE-RSA-AES128-GCM-SHA256:ECDHE-ECDSA-AES256-SHA:ECDHE-ECDSA-AES128-SHA:ECDHE-RSA-AES128-SHA:ECDHE-RSA-AES256-SHA:ECDHE-ECDSA-RC4-SHA:ECDHE-RSA-RC4-SHA:DHE-RSA-AES128-SHA:DHE-DSS-AES128-SHA:DHE-RSA-AES256-SHA:AES128-GCM-SHA256:AES128-SHA:AES256-SHA:DES-CBC3-SHA:RC4-SHA:RC4-MD5"
    2017-01-03T16:17:22 [DEBUG] 23 sslCertificatesPath : ""
    2017-01-03T16:17:22 [DEBUG] 24 sslClientCertificateFile : ""
    2017-01-03T16:17:22 [DEBUG] 25 sslClientKeyFile : ""
    2017-01-03T16:17:22 [DEBUG] 26 sslClientKeyPassphrase : ""
    2017-01-03T16:17:22 [DEBUG] 27 webdriver : ":"
    2017-01-03T16:17:22 [DEBUG] 28 webdriverLogFile : ""
    2017-01-03T16:17:22 [DEBUG] 29 webdriverLogLevel : "INFO"
    2017-01-03T16:17:22 [DEBUG] 30 webdriverSeleniumGridHub : ""
    2017-01-03T16:17:22 [DEBUG] Phantom - execute: Script & Arguments
    2017-01-03T16:17:22 [DEBUG] script: "import/getText.js"
    2017-01-03T16:17:22 [DEBUG] Phantom - execute: Starting normal mode
    2017-01-03T16:17:22 [DEBUG] WebPage - setupFrame ""
    2017-01-03T16:17:22 [DEBUG] FileSystem - _open: ":/modules/fs.js" QMap(("mode", QVariant(QString, "r")))
    2017-01-03T16:17:23 [DEBUG] FileSystem - _open: ":/modules/system.js" QMap(("mode", QVariant(QString, "r")))
    2017-01-03T16:17:23 [DEBUG] FileSystem - _open: ":/modules/webpage.js" QMap(("mode", QVariant(QString, "r")))
  2. AlejoNext revised this gist Jan 3, 2017. 1 changed file with 51 additions and 43 deletions.
    94 changes: 51 additions & 43 deletions getTexts.js
    View file
    Original file line number Diff line number Diff line change
    @@ -1,23 +1,29 @@
    var webPage = require('webpage');
    var HOME = 'http://my.domain.com/';
    var stepIndex = 0;
    /*** HOME ***/
    var HOME = 'http://www.plotandesign.net/';

    var HOME_CONTENT = 'Where ara the info in home';
    var POST_MENU = 'Go to menu';
    var TOP_MENU = 'Top Menu';
    var CONTENT_POST = 'Where are the info in pages';
    /*** WHERE ***/
    var HOME_CONTENT = '#content .row2 > .main > .wrapper > .col1.cols',
    POST_MENU = 'ul.left a',
    TOP_MENU = '#header .menu li > a',
    CONTENT_POST = '#content .row2 .main > .wrapper';

    var PAGES = {};
    var POST = [];
    /*** DATA ***/
    var PAGES = {},
    POST = [],
    stepIndex = 0;

    function posibleError (msg, line, source) {
    function posibleJS (msg, line, source) {
    console.log('>', msg);
    }

    function posibleAlert (msg) {
    console.log('<', msg);
    }

    function posibleError (resourceError) {
    console.error(resourceError.errorString)
    }

    /**
    * [getPostOrPages description]
    * @param {String} name Selector
    @@ -29,15 +35,11 @@ function getPostOrPages (name) {
    for (var i = 0; i < node.length; i++) {
    if(node[i] && !node[i].body && node[i].tagName ){
    if(node[i].tagName == 'DIV'){
    var z = clean($(node[i]
    .innerHTML
    .replace(/\t|\n|\s{2,}/gim, '')
    .replace(/(\<br\>)/gim, '</p><p>')));

    var z = clean($(node[i].innerHTML.replace(/\t|\n|\s{2,}/gim, '').replace(/<br>/gim, '</p><p>')));
    for (var w = 0; w < z.length; w++) {
    newNode.push(z[w]);
    }
    } else if(/P|H[1-6]|UL|I|TABLE|LI|IMG/i.test(node[i].tagName)){
    } else if( /P|H[1-6]|UL|I|TABLE|LI|IMG/i.test(node[i].tagName) ){
    newNode.push(node[i]);
    }
    }
    @@ -46,16 +48,16 @@ function getPostOrPages (name) {
    return newNode;
    }

    var post = { text : [] };
    var child = clean($(name).children());
    var post = { text : [] };

    for (var i = 0; i < child.length; i++) {
    if(child[i] && child[i].tagName && !child[i].body && child[i].innerHTML.length){
    if( child[i] && child[i].tagName && !child[i].body && child[i].innerHTML.length ){
    if( child[i].tagName != 'BR' && child[i].tagName != 'H1' ){
    post.text.push(child[i].outerHTML);
    } else if ( child[i].tagName == 'H1' ) {
    post.text.push(child[i].outerHTML);
    } else if ( child[i].tagName == 'H1' ) {
    post.title = child[i].innerText;
    } else if ( child[i].tagName == 'I' && !post.autor ) {
    } else if ( child[i].tagName == 'I' && !post.autor ) {
    post.autor = child[i].innerText;
    }
    }
    @@ -70,27 +72,27 @@ function getPostOrPages (name) {
    * @return {Array} Element
    */
    function getHome(name) {
    var data = [];
    var post = {};
    var cild = $(name).children();
    for (var i = 0; i < cild.length; i++) {
    if(cild[i].tagName === 'H2'){
    var data = [],
    post = {},
    child = $(name).children();

    for (var i = 0; i < child.length; i++) {
    if(child[i].tagName === 'H2'){
    if(post.title){
    data.push(post);
    }
    post = { title : cild[i].innerText, text : [] };
    post = { title : child[i].innerText, text : [] };
    } else {
    if(post.text && cild[i].tagName != 'BR' ){
    if(cild[i].tagName != 'I'){
    post.text.push(cild[i]
    .outerHTML
    .replace(/\t|\n|(\s){2,}|(\<br\>){2,}/gim, ''));
    if(post.text && child[i].tagName != 'BR' ){
    if(child[i].tagName != 'I'){
    post.text.push(child[i].outerHTML.replace(/\t|\n|\s{2,}/gim, '').replace(/<br>/gim, '</p><p>')));
    } else {
    post.autor = cild[i].innerText;
    post.autor = child[i].innerText;
    }
    }
    }
    }

    if(post.title){
    data.push(post);
    }
    @@ -104,11 +106,13 @@ function getHome(name) {
    * @return {Array} Array URL
    */
    function getPostMenu (name) {
    var urs = [];
    var items = $(name);
    var urs = [],
    items = $(name);

    for (var i = 0; i < items.length; i++) {
    urs.push(items[i].href);
    }

    return urs;
    }

    @@ -128,11 +132,14 @@ function getMenu (i, top) {
    * @param {Boolean} isPost [description]
    */
    function startPage(url, isPost) {
    var page = webPage.create();
    var webPage = require('webpage');
    var page = webPage.create(),
    newUrl;

    page.onConsoleMessage = posibleError;
    page.onConsoleMessage = posibleJS;
    page.onAlert = posibleAlert;

    page.onResourceError = posibleError;

    stepIndex++;

    if(url){
    @@ -142,8 +149,6 @@ function startPage(url, isPost) {

    page.injectJs('jquery.min.js');

    var newUrl;

    if(!phantom.state){
    if(isPost || url.indexOf('index.html') < 0){
    PAGES[ url ] = page.evaluate(getPostOrPages, CONTENT_POST);
    @@ -161,10 +166,10 @@ function startPage(url, isPost) {
    PAGES[ url ].page = true;
    }

    newUrl = page.evaluate(getMenu, stepIndex, TOP_MENU));
    newUrl = page.evaluate(getMenu, stepIndex, TOP_MENU);
    }

    if(!newUrl){
    if(!newUrl && !isPost){
    stepIndex = 0;
    newUrl = POST[stepIndex];
    isPost = true;
    @@ -174,12 +179,15 @@ function startPage(url, isPost) {
    } else {
    phantom.state();
    }
    } else {
    console.log('mierda');
    }
    });
    } else {
    console.log(JSON.stringify(PAGES, null, '\t'));
    phantom.exit();
    }
    }

    console.log('Start :', HOME);
    startPage(HOME + 'index.html');

  3. AlejoNext revised this gist Jan 3, 2017. 1 changed file with 45 additions and 12 deletions.
    57 changes: 45 additions & 12 deletions getTexts.js
    View file
    Original file line number Diff line number Diff line change
    @@ -10,7 +10,6 @@ var CONTENT_POST = 'Where are the info in pages';
    var PAGES = {};
    var POST = [];


    function posibleError (msg, line, source) {
    console.log('>', msg);
    }
    @@ -19,13 +18,22 @@ function posibleAlert (msg) {
    console.log('<', msg);
    }

    /**
    * [getPostOrPages description]
    * @param {String} name Selector
    * @return {Object} Post/Pages
    */
    function getPostOrPages (name) {
    function clean (node) {
    var newNode = [];
    for (var i = 0; i < node.length; i++) {
    if(node[i] && !node[i].body && node[i].tagName ){
    if(node[i].tagName == 'DIV'){
    var z = clean($(node[i].innerHTML.replace(/\t|\n|\s{2,}/gim, '').replace(/(\<br\>)/gim, '</p><p>')));
    var z = clean($(node[i]
    .innerHTML
    .replace(/\t|\n|\s{2,}/gim, '')
    .replace(/(\<br\>)/gim, '</p><p>')));

    for (var w = 0; w < z.length; w++) {
    newNode.push(z[w]);
    }
    @@ -42,18 +50,25 @@ function getPostOrPages (name) {
    var child = clean($(name).children());

    for (var i = 0; i < child.length; i++) {
    if(child[i] && child[i].tagName != 'BR' && child[i].tagName != 'H1' && !child[i].body && child[i].innerHTML.length ){
    post.text.push(child[i].outerHTML);
    } else if ( child[i] && child[i].tagName == 'H1' ) {
    post.title = child[i].innerText;
    } else if ( child[i] && child[i].tagName == 'I' && !post.autor ) {
    post.autor = child[i].innerText;
    if(child[i] && child[i].tagName && !child[i].body && child[i].innerHTML.length){
    if( child[i].tagName != 'BR' && child[i].tagName != 'H1' ){
    post.text.push(child[i].outerHTML);
    } else if ( child[i].tagName == 'H1' ) {
    post.title = child[i].innerText;
    } else if ( child[i].tagName == 'I' && !post.autor ) {
    post.autor = child[i].innerText;
    }
    }
    }

    return post;
    }

    /**
    * [getHome description]
    * @param {String} name Selector
    * @return {Array} Element
    */
    function getHome(name) {
    var data = [];
    var post = {};
    @@ -67,7 +82,9 @@ function getHome(name) {
    } else {
    if(post.text && cild[i].tagName != 'BR' ){
    if(cild[i].tagName != 'I'){
    post.text.push(cild[i].outerHTML.replace(/\t|\n|(\s){2,}|(\<br\>){2,}/gim, ''));
    post.text.push(cild[i]
    .outerHTML
    .replace(/\t|\n|(\s){2,}|(\<br\>){2,}/gim, ''));
    } else {
    post.autor = cild[i].innerText;
    }
    @@ -81,7 +98,12 @@ function getHome(name) {
    return data;
    }

    function getPost (name) {
    /**
    * [getPostMenu description]
    * @param {String} name Selector
    * @return {Array} Array URL
    */
    function getPostMenu (name) {
    var urs = [];
    var items = $(name);
    for (var i = 0; i < items.length; i++) {
    @@ -90,10 +112,21 @@ function getPost (name) {
    return urs;
    }

    /**
    * [getMenu description]
    * @param {Number} i Number in array
    * @param {String} top Selector
    * @return {String} Go to
    */
    function getMenu (i, top) {
    return $(top)[i].href;
    }

    /**
    * [startPage description]
    * @param {String} url Go to
    * @param {Boolean} isPost [description]
    */
    function startPage(url, isPost) {
    var page = webPage.create();

    @@ -121,7 +154,7 @@ function startPage(url, isPost) {
    newUrl = POST[stepIndex];
    } else {
    if(url.indexOf('index.html') >= 0){
    POST = page.evaluate(getPost, POST_MENU);
    POST = page.evaluate(getPostMenu, POST_MENU);
    PAGES[ url ] = page.evaluate(getHome, HOME_CONTENT);
    PAGES[ url ].list = true;
    } else {
    @@ -145,8 +178,8 @@ function startPage(url, isPost) {
    });
    } else {
    console.log(JSON.stringify(PAGES, null, '\t'));
    phantom.kill();
    }
    }

    startPage(HOME + 'index.html');

  4. AlejoNext revised this gist Jan 3, 2017. 1 changed file with 1 addition and 68 deletions.
    69 changes: 1 addition & 68 deletions getTexts.js
    View file
    Original file line number Diff line number Diff line change
    @@ -10,73 +10,6 @@ var CONTENT_POST = 'Where are the info in pages';
    var PAGES = {};
    var POST = [];

    function posibleError (msg, line, source) {
    console.log('>', msg);
    }

    function posibleAlert (msg) {
    console.log('<', msg);
    }

    function getPostOrPages (content) {
    function clean (node) {
    var newNode = [];
    for (var i = 0; i < node.length; i++) {
    if(node[i] && !node[i].body && node[i].tagName ){
    if(node[i].tagName == 'DIV'){
    var z = clean($(node[i].innerHTML.replace(/\t|\n|\s{2,}/gim, '').replace(/(\<br\>)/gim, '</p><p>')));
    for (var w = 0; w < z.length; w++) {
    newNode.push(z[w]);
    }
    } else if(/P|H[1-6]|UL|I|TABLE|LI|IMG/i.test(node[i].tagName)){
    newNode.push(node[i]);
    }
    }
    }

    return newNode;
    }

    var post = { text : [] };
    var child = clean($(content).children());

    for (var i = 0; i < child.length; i++) {
    if(child[i] && child[i].tagName != 'BR' && child[i].tagName != 'H1' && !child[i].body && child[i].innerHTML.length ){
    post.text.push(child[i].outerHTML);
    } else if ( child[i] && child[i].tagName == 'H1' ) {
    post.title = child[i].innerText;
    } else if ( child[i] && child[i].tagName == 'I' && !post.autor ) {
    post.autor = child[i].innerText;
    }
    }

    return post;
    }

    function getHome(content) {
    var data = [];
    var post = {};
    var cild = $(content).children();
    for (var i = 0; i < cild.length; i++) {
    if(cild[i].tagName === 'H2'){
    if(post.title){
    data.push(post);
    }
    post = { title : cild[i].innerText, text : [] };
    } else {
    if(post.text && cild[i].tagName != 'BR' ){
    if(cild[i].tagName != 'I'){
    post.text.push(cild[i].outerHTML.replace(/\t|\n|(\s){2,}|(\<br\>){2,}/gim, ''));
    } else {
    post.autor = cild[i].innerText;
    }
    }
    }
    }
    if(post.title){
    data.push(post);
    }


    function posibleError (msg, line, source) {
    console.log('>', msg);
    @@ -216,4 +149,4 @@ function startPage(url, isPost) {
    }
    }

    startPage(HOME + 'index.html');
    startPage(HOME + 'index.html');
  5. AlejoNext revised this gist Jan 3, 2017. 1 changed file with 86 additions and 15 deletions.
    101 changes: 86 additions & 15 deletions getTexts.js
    View file
    Original file line number Diff line number Diff line change
    @@ -77,30 +77,103 @@ function getHome(content) {
    data.push(post);
    }


    function posibleError (msg, line, source) {
    console.log('>', msg);
    }

    function posibleAlert (msg) {
    console.log('<', msg);
    }

    function getPostOrPages (name) {
    function clean (node) {
    var newNode = [];
    for (var i = 0; i < node.length; i++) {
    if(node[i] && !node[i].body && node[i].tagName ){
    if(node[i].tagName == 'DIV'){
    var z = clean($(node[i].innerHTML.replace(/\t|\n|\s{2,}/gim, '').replace(/(\<br\>)/gim, '</p><p>')));
    for (var w = 0; w < z.length; w++) {
    newNode.push(z[w]);
    }
    } else if(/P|H[1-6]|UL|I|TABLE|LI|IMG/i.test(node[i].tagName)){
    newNode.push(node[i]);
    }
    }
    }

    return newNode;
    }

    var post = { text : [] };
    var child = clean($(name).children());

    for (var i = 0; i < child.length; i++) {
    if(child[i] && child[i].tagName != 'BR' && child[i].tagName != 'H1' && !child[i].body && child[i].innerHTML.length ){
    post.text.push(child[i].outerHTML);
    } else if ( child[i] && child[i].tagName == 'H1' ) {
    post.title = child[i].innerText;
    } else if ( child[i] && child[i].tagName == 'I' && !post.autor ) {
    post.autor = child[i].innerText;
    }
    }

    return post;
    }

    function getHome(name) {
    var data = [];
    var post = {};
    var cild = $(name).children();
    for (var i = 0; i < cild.length; i++) {
    if(cild[i].tagName === 'H2'){
    if(post.title){
    data.push(post);
    }
    post = { title : cild[i].innerText, text : [] };
    } else {
    if(post.text && cild[i].tagName != 'BR' ){
    if(cild[i].tagName != 'I'){
    post.text.push(cild[i].outerHTML.replace(/\t|\n|(\s){2,}|(\<br\>){2,}/gim, ''));
    } else {
    post.autor = cild[i].innerText;
    }
    }
    }
    }
    if(post.title){
    data.push(post);
    }

    return data;
    }

    function getPost (post) {
    function getPost (name) {
    var urs = [];
    var items = $(post);
    var items = $(name);
    for (var i = 0; i < items.length; i++) {
    urs.push(items[i].href);
    }
    return urs;
    }

    function getMenu (i, top) {
    return $(top)[i].href;
    }

    function startPage(url, isPost) {
    var page = webPage.create();

    page.onConsoleMessage = posibleError;
    page.onAlert = posibleAlert;

    stepIndex++;

    if(url){
    page.open(url, function(status){
    console.log(status);
    if (status === 'success') {
    console.log('Start :', url);

    page.injectJs('jquery.min.js');

    var newUrl;
    @@ -115,16 +188,20 @@ function startPage(url, isPost) {
    newUrl = POST[stepIndex];
    } else {
    if(url.indexOf('index.html') >= 0){
    POST = page.evaluate(getPost, POST_MENU)
    POST = page.evaluate(getPost, POST_MENU);
    PAGES[ url ] = page.evaluate(getHome, HOME_CONTENT);
    PAGES[ url ].list = true;
    } else {
    PAGES[ url ].page = true;
    }

    newUrl = page.evaluate(function(i, top) {
    return $(top)[i].href;
    }, stepIndex, TOP_MENU));
    newUrl = page.evaluate(getMenu, stepIndex, TOP_MENU));
    }

    if(!newUrl){
    stepIndex = 0;
    newUrl = POST[stepIndex];
    isPost = true;
    }

    startPage(newUrl, isPost);
    @@ -134,14 +211,8 @@ function startPage(url, isPost) {
    }
    });
    } else {
    if(isPost){
    console.log(JSON.stringify(PAGES, null, '\t'));
    phantom.exit();
    phantom.kill();
    } else {
    stepIndex = 0;
    startPage(POST[stepIndex], true);
    }
    console.log(JSON.stringify(PAGES, null, '\t'));
    phantom.kill();
    }
    }

  6. AlejoNext renamed this gist Jan 3, 2017. 1 changed file with 0 additions and 0 deletions.
    0 gistfile1.txt → getTexts.js
    View file
    File renamed without changes.
  7. AlejoNext created this gist Jan 3, 2017.
    148 changes: 148 additions & 0 deletions gistfile1.txt
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,148 @@
    var webPage = require('webpage');
    var HOME = 'http://my.domain.com/';
    var stepIndex = 0;

    var HOME_CONTENT = 'Where ara the info in home';
    var POST_MENU = 'Go to menu';
    var TOP_MENU = 'Top Menu';
    var CONTENT_POST = 'Where are the info in pages';

    var PAGES = {};
    var POST = [];

    function posibleError (msg, line, source) {
    console.log('>', msg);
    }

    function posibleAlert (msg) {
    console.log('<', msg);
    }

    function getPostOrPages (content) {
    function clean (node) {
    var newNode = [];
    for (var i = 0; i < node.length; i++) {
    if(node[i] && !node[i].body && node[i].tagName ){
    if(node[i].tagName == 'DIV'){
    var z = clean($(node[i].innerHTML.replace(/\t|\n|\s{2,}/gim, '').replace(/(\<br\>)/gim, '</p><p>')));
    for (var w = 0; w < z.length; w++) {
    newNode.push(z[w]);
    }
    } else if(/P|H[1-6]|UL|I|TABLE|LI|IMG/i.test(node[i].tagName)){
    newNode.push(node[i]);
    }
    }
    }

    return newNode;
    }

    var post = { text : [] };
    var child = clean($(content).children());

    for (var i = 0; i < child.length; i++) {
    if(child[i] && child[i].tagName != 'BR' && child[i].tagName != 'H1' && !child[i].body && child[i].innerHTML.length ){
    post.text.push(child[i].outerHTML);
    } else if ( child[i] && child[i].tagName == 'H1' ) {
    post.title = child[i].innerText;
    } else if ( child[i] && child[i].tagName == 'I' && !post.autor ) {
    post.autor = child[i].innerText;
    }
    }

    return post;
    }

    function getHome(content) {
    var data = [];
    var post = {};
    var cild = $(content).children();
    for (var i = 0; i < cild.length; i++) {
    if(cild[i].tagName === 'H2'){
    if(post.title){
    data.push(post);
    }
    post = { title : cild[i].innerText, text : [] };
    } else {
    if(post.text && cild[i].tagName != 'BR' ){
    if(cild[i].tagName != 'I'){
    post.text.push(cild[i].outerHTML.replace(/\t|\n|(\s){2,}|(\<br\>){2,}/gim, ''));
    } else {
    post.autor = cild[i].innerText;
    }
    }
    }
    }
    if(post.title){
    data.push(post);
    }

    return data;
    }

    function getPost (post) {
    var urs = [];
    var items = $(post);
    for (var i = 0; i < items.length; i++) {
    urs.push(items[i].href);
    }
    return urs;
    }

    function startPage(url, isPost) {
    var page = webPage.create();
    page.onConsoleMessage = posibleError;
    page.onAlert = posibleAlert;

    stepIndex++;

    if(url){
    page.open(url, function(status){
    console.log(status);
    if (status === 'success') {
    console.log('Start :', url);
    page.injectJs('jquery.min.js');

    var newUrl;

    if(!phantom.state){
    if(isPost || url.indexOf('index.html') < 0){
    PAGES[ url ] = page.evaluate(getPostOrPages, CONTENT_POST);
    }

    if(isPost){
    PAGES[ url ].post = true;
    newUrl = POST[stepIndex];
    } else {
    if(url.indexOf('index.html') >= 0){
    POST = page.evaluate(getPost, POST_MENU)
    PAGES[ url ] = page.evaluate(getHome, HOME_CONTENT);
    PAGES[ url ].list = true;
    } else {
    PAGES[ url ].page = true;
    }

    newUrl = page.evaluate(function(i, top) {
    return $(top)[i].href;
    }, stepIndex, TOP_MENU));
    }

    startPage(newUrl, isPost);
    } else {
    phantom.state();
    }
    }
    });
    } else {
    if(isPost){
    console.log(JSON.stringify(PAGES, null, '\t'));
    phantom.exit();
    phantom.kill();
    } else {
    stepIndex = 0;
    startPage(POST[stepIndex], true);
    }
    }
    }

    startPage(HOME + 'index.html');