Usuario:Ignacio Rodríguez/TemplateScriptTest.js

/*

This page defines a TemplateScript library. It's not meant to be referenced
directly. See [[Wikisource:TemplateScript]] for usage.



*/
/* global $, pathoschild */

/**
 * TemplateScript adds configurable templates and scripts to the sidebar, and adds an example regex editor.
 * @see https://meta.wikimedia.org/wiki/TemplateScript
 * @update-token [[File:Pathoschild/templatescript.js]]
 */
// <nowiki>
$.ajax('//tools-static.wmflabs.org/meta/scripts/pathoschild.templatescript.js', { dataType:'script', cache:true }).then(function() {
	/*********
	** Define library
	*********/
	pathoschild.TemplateScript.library.define({
		key: 'wikisource.proofreading',
		name: 'Herramientas de transcripción',
		url: '//es.wikisource.org/wiki/Wikisource:TemplateScript',
		description: 'Una serie de herramientas para facilitar la <a href="/wiki/Ayuda:Páginas de transcripción">transcripción en el espacio de nombres <tt>Página:</tt></a>  Incluye herramientas para mejorar el OCR, agregar encabezados, y otras herramientas de formato.',
		categories: [
			{
				name: 'Herramientas de transcripción',
				scripts: [
					{ key: 'add-header', name: 'Agregar encabezado', scriptUrl:'Special:MyPage/titulos.js', script: function(editor) { addPageHeader(editor); }, forNamespaces: 'page', accessKey:'3' },
					{ key: 'add-footer', name: 'Agregar pie de página', scriptUrl:'Special:MyPage/titulos.js', script: function(editor) { addPageFooter(editor); }, forNamespaces: 'page', accessKey:'4' },
					{ key: 'ocr', name: 'OCR (reconocimiento automático de texto)', script: function(editor) { do_hocr(); }, forNamespaces: 'page', accessKey:'1' },
					{ key: 'cleanup-ocr', name: 'Limpiar OCR', script: function(editor) { pageCleanup(editor); }, forNamespaces: 'page', accessKey:'2' },
					{ key: 'make-refs', name: 'Hacer referencias', script: function(editor) { makeReference(editor); }, forNamespaces: 'page', accessKey:'h' },
					{ key: 'smallcaps', name: 'A versalitas', script: function(editor) { smallcaps(editor); }, forNamespaces: 'page' },
					{ key: 'uppercase', name: 'A mayúsculas', script: function(editor) { upper(editor); }, forNamespaces: 'page' },
					{ key: 'lowercase', name: 'A minúsculas', script: function(editor) { lower(editor); }, forNamespaces: 'page' }
				]
			}
		]
	});

	/*********
	** Page context
	*********/
	var state = {
		initialised: false,  // whether the page context has been initialised
		page: {
			number: null,   // the djvu page number extracted from the URL
			proofed: null
		},
		specialFormats: [] // work-specific header template formats
	};

	/*********
	** Private methods
	*********/
	/**
	 * Initialise the data needed by the page tools.
	 */
	var _initialise = function() {
		// only initialise once
		if(state.initialised)
			return;
		state.initialised = true;

		// get page metadata
		//var pn = /\.(?:djvu|pdf)\/([0-9]+)&action=edit/g.exec(location.href);
		var pn = obtenerPn();
		var pq = document.getElementById('pagequality');
		state.page = {
			number: pn !== null ? parseInt(pn[1], 10) : null,
			proofed: pq && pq.getAttribute('class') && pq.getAttribute('class').match(/quality0|quality[2-4]/)
		};

		// get user-defined work formats
		// expected format:
		//   {
		//      title: /History of England /,
		//      evenHeader: '{{rh|...}}',
		//      oddHeader: '{{rh|...}}',
		//      footer: '',
		//      footerWithReferences: '{{smallrefs}}'
	 	//   }
		
		if(window.specialFormats)
			//state.specialFormats = state.specialFormats.concat(window.specialFormats);
			state.specialFormats = window.specialFormats.concat(state.specialFormats);
	};
	
	/**
	 * Convert the text to title case based on English rules.
	 * @param {string} text The text to convert.
	 */
	var obtenerPn = function(){
		var pnfc= /\.(?:djvu|pdf)\/([0-9]+)/g.exec($("#firstHeading").html());
		window.pn = pnfc !== null ? parseInt(pnfc[1], 10) : null;
		return window.pn;
	};
	
	var _titlecase = function(text) {
		// split text into individual words and examine them one by one
		var words = text.toLowerCase().split(" ");
		$.each(function(i, word) {
			switch(word) {
				case "a":
				case "á":
				case "e":
				case "é":
				case "i":
				case "y":
				case "o":
				case "ó":
				case "u":
				case "el":
				case "la":
				case "los":
				case "las":
				case "un":
				case "una":
				case "unos":
				case "de":
				case "del":
					break; // don't capitalise articles, "to" as part of an infinitive, prepositions or short conjunctions
				default: // capitalise everything else
					words[i] = word.substring(0, 1).toUpperCase() + word.substring(1, words[i].length);
					break;
			}
		});

		// capitalise first word regardless
		words[0] = words[0].substring(0, 1).toUpperCase() + words[0].substring(1, words[0].length);

		// capitalise last word regardless
		var last = words.length-1;
		words[last] = words[last].substring(0, 1).toUpperCase() + words[last].substring(1, words[last].length);

		// reconstruct title
		return words.join(' ');
	};

	/*********
	** Script methods
	*********/
	/**
	 * Add a {{running header}} template to the page.
	 * @param {object} editor The script helpers for the page.
	 */
	var addPageHeader = function(editor) {
		_initialise();
		obtenerPn();
		
		if(window.pn === null)
			return;
		
		var isEven = (window.pn % 2 === 0);
		
		var generic = true;
		var headertext = '';
		
		for (var f in state.specialFormats) {
			var format = state.specialFormats[f];
			if (mw.config.get('wgTitle').match(format.title)) {
				headertext = isEven ? format.evenHeader : format.oddHeader;
				generic = false;
				break;
			}
		}

		// no special header matched, use a generic running header
		if (generic) {
			if (isEven)
				headertext = '{{RH|left='+state.page.number+'|center=}}'; // assume verso, with page number at left
			else
				headertext = '{{RH|center=|right='+state.page.number+'}}';
		}
		
		$('#wpHeaderTextbox').val(function(i, val) {
			return $.trim(val + '\n' + headertext);
		});
	};

	/**
	 * Clean up OCR errors in the text, and push <noinclude> content at the top
	 * & bottom of the page into the header & footer boxes respectively.
	 * @param {object} editor The script helpers for the page.
	 */
	var pageCleanup = function(editor) {
		_initialise();
		
		// push <noinclude> content at the top & bottom into the header & footer
		if (editor.get().match(/^<noinclude\>/)) {
			var text = editor.get();
			var e = text.indexOf("</noinclude>");
			$('#wpHeaderTextbox').val(function(i, val) {
				return $.trim(val + "\n" + text.substr(11, e-11).replace(/^\s+|\s+$/g, ''));
			});
			editor.set(text.substr(e+12));
		}
		if (editor.get().match(/<\/noinclude\>$/)) {
			var text = editor.get();
			var s = text.lastIndexOf("<noinclude>");
			$('#wpFooterTextbox').val(function(i, val) {
				return $.trim(text.substr(s+11, text.length-s-11-12).replace(/^\s+|\s+$/g, '') + "\n" + val);
			});
			editor.set(text.substr(0, s));
		}
		
		// clean up text
		editor
			// remove trailing spaces at the end of each line
			.replace(/ +\n/g, '\n')

			// remove trailing whitespace preceding a hard line break
			.replace(/ +<br *\/?>/g, '<br />')

			// remove trailing whitespace and numerals at the end of page text
			// (numerals are nearly always page numbers in the footer)
			.replace(/[\s\d]+$/g, '')

			// remove trailing spaces at the end of refs
			.replace(/ +<\/ref>/g, '</ref>')
	
			// remove trailing spaces at the end of template calls
			.replace(/ +}}/g, '}}')
	
			// convert double-hyphen to mdash (avoiding breaking HTML comment syntax)
			.replace(/([^\!])--([^>])/g, '$1—$2')
	
			// remove spacing around mdash, but only if it has spaces on both sides
			// (we don't want to remove the trailing space from "...as follows:— ",
			// bearing in mind that the space will already be gone if at end of line).
			.replace(/ +— +/g, '—')
	
			// join words that are hyphenated across a line break
			// (but leave "|-" table syntax alone)
			.replace(/([^\|])-\n/g, '$1');

		// clean up pages if they don't have <poem>
		if (!editor.contains('<poem>')) {
			editor
				// lines that start with " should probably be new lines,
				// if the previous line ends in punctuation,
				// other than a comma or semicolon
				// and let's get rid of trailing space while we're at it
				.replace(/([^\n\w,;])\n\" */g, '$1\n\n"')
	
				// lines that end with " should probably precede a new line,
				// unless preceded by a comma,
				// or unless the new line starts with a lower-case letter;
				// and let's get rid of preceding space while we're at it
				.replace(/([^,])\ *\"\n([^a-z\n])/g, '$1"\n\n$2')
	
				// remove single line breaks; preserve multiple.
				// but not if there's a tag, template or table syntax either side of the line break
				.replace(/([^>}\|\n])\n([^:#\*<{\|\n])/g, '$1 $2')
	
				// collapse sequences of spaces into a single space
				.replace(/  +/g, ' ');
		}
		
		// more page cleanup
		editor
			// dump spurious hard breaks at the end of paragraphs
			.replace(/<br *\/?>\n\n/g, '\n\n')

			// remove unwanted spaces around punctuation marks
			.replace(/ ([;:\?!,\.])/g, '$1')
	
			// unicodify
			.replace(/&mdash;/g, '—')
			.replace(/&ndash;/g, '–')
			.replace(/&quot;/g, '"')
	
			// straighten quotes and apostrophes.
			.replace(/[“”]/g, '"')
			.replace(/[‘’`]/g, '\'')
	
			//OCR fixes
			// números
			.replace(/1([^0-9\.\s,;:])/g, 'l$1')
			.replace(/0([^0-9\.\s,;:])/g, 'o$1')
			.replace(/[58~]([^0-9\.\s,;:])/g, 's$1')
			.replace(/([a-zA-Z])6/g, '$1ó')
			.replace(/([a-zA-Z])0/g, '$1o')
			.replace(/6([^0-9\.\s,;:])/g, 'é$1')
	
			.replace(/_/g, '—')
			.replace(/ \(l/g,' d')
			.replace(/I\)/g,'P')
			.replace(/([a-záéíóú])U/g, '$1u')
			
			//partículas y consonantongos que no existen, o casi no existen en español
			.replace(/c([ ,.\n])/,'e$1') // c final --> e
			.replace(/ hl([aeiouáéíóú])/g, 'bl$1')
			.replace(/hr/g, 'br')
			.replace(/cb/g, 'ch')
			.replace(/Cb/g, 'Ch')
			.replace(/([Aa])1/g, '$1l')
			.replace(/tn/g, 'tu')
			.replace(/([ae])iite/g,'$1nte')
			.replace(/([ae])ute(s?[ ,.\n])/g, '$1nte$2')
			.replace(/([scgp])nn/g,'$1un')
			.replace(/clel/g, 'del')
			.replace(/cou/g, 'con')
			.replace(/nii/g, 'mi')
			.replace(/ iia/g, ' na')
			.replace(/cne/g, 'cue')
			.replace(/cleb/g, 'deb')
			.replace(/snf/g, 'suf')
			.replace(/aiito([ ,.\n])/g, 'anto$1')
			.replace(/ln([dn])/g, 'lu$1')
			.replace(/rn([pb])/g, 'm$1')
			.replace(/ iuc/g, ' inc')
			.replace(/ rn/g, ' m')
			.replace(/ nne/g, ' nue')
			.replace(/clia/g, 'cha')
			.replace(/ lia/g, ' ha')
			.replace(/([^u])irn/g, '$1im')
			.replace(/([ae])iid/g, '$1nd')
			.replace(/cli([jrd])/g, 'di$1')
			.replace(/clist/g, 'dist')
			.replace(/cnto([ ,.\n])/,'ento$1')
			.replace(/nndo([ ,.\n])/,'ando$1')
			.replace(/ii([ ,.\n])/,'n$1')
			.replace(/cl([ ,.\n])/,'d$1')
			.replace(/fc/g, 'fe')
			.replace(/gnl/g,'gul')
			.replace(/nlt/g,'nst')
			.replace(/pne/g,'pue')
			.replace(/oue/g,'one')
			.replace(/xol/g,'xcl')
			.replace(/ gn/g,' gu')
			.replace(/dn([oa])/g,'du$1')
			.replace(/ oo/g,' co')
			.replace(/([n])eoe/g,'$1ece')
			.replace(/ncs/g,'nes')
			
			//fi --> ñ
			.replace(/ afio/g,' año')
			.replace(/ifia/g,'iña')
			.replace(/afier/g,'añer')
			.replace(/tafia/g,'taña')
			.replace(/([Ss])efial/g,'$1eñal')
			.replace(/([Ss])efior/g,'$1eñor')
			.replace(/efio([ .,\n])/g,'eño$1')
			.replace(/mafio([ .,\n])/g,'maño$1')
			.replace(/trafio/g,'traño')
			
			// OCR a español (gracias Aleator)
			
			
			.replace(/([Qq])(?:ll|n|li)/g, '$1u')
			.replace(/([Qq])(?:o|a|tl|u)([eié])/g, '$1u$2')
			.replace(/([Qq])(?:o|a|tl|u)á/g, '$1ué')
			.replace(/([Qq])[aou][ceors]/g, '$1ue')
			.replace(/([Qq])lll/g, '$1ue')
			.replace(/d([A-ZÁÉÍÓÚ])/g,'¿$1')
			.replace(/ d([qcdhp])/g,' ¿$1')
			.replace(/([a-zó])[Pp]"/g,'$1?"')
			.replace(/([a-zó])[Pp]([ .,\n])/g,'$1?$2')
			
			.replace(/ DO /g,' no ')
			.replace(/ UTI /g,' un ')
			.replace(/ u n /g,' un ')
			.replace(/ n /g,' a ')
			.replace(/ d /g,' el ')
			
			.replace(/]([a-záéíóú])/g,'l$1')
			
			.replace(/fi/g,'fi')
			.replace(/fl/g,'fl')
			
			.replace(/((?!para).{4})noi([ao])/g,'$1nci$2')
			
			.replace(/au([ ,.\n])/g, 'an$1')
			.replace(/aha([ ,.\n])/g, 'aba$1')
			.replace(/ahan([ ,.\n])/g, 'aban$1')
			.replace(/(a[iïíìl!1I]g[anou][nu])/g, 'algun')
			.replace(/(ag[un]a)/g, 'agua')
			.replace(/ afio /g, ' año ')
			.replace(/ ao/g, ' au')
			.replace(/ oen/g, ' cen')
			.replace(/c[iïíìl!1I][56d][un]([ ,.\n])/g, 'ción$1')
			.replace(/oio[un]/g, 'cion ')
			.replace(/oió[un]/g, 'ción ')
			.replace(/(c[iïíìl!1I][56d][ns][ce][8s] )/g, 'ciones ')
			.replace(/( corno )/g, ' como ')
			.replace(/( co[un][iïíìl!1I]o )/g, ' como ')
			.replace(/( eomo )/g, ' como ')
			.replace(/( [oce][op]n )/g, ' con ')
			.replace(/(cb)/g, 'ch')
			.replace(/c[ïíl!1I]o([ ,.\n])/g, 'cio$1')
			.replace(/( c[li]e )/g, ' de ')
			.replace(/( d[ocs] )/g, ' de ')
			.replace(/( ¿c )/g, ' de ')
			.replace(/( c[li][ce]l )/g, ' del ')
			.replace(/( d[ce][iïíìl!1I] )/g, ' del ')
			.replace(/(d[ce][s8])/g, 'des')
			.replace(/(d[iïíìl!1I])/g, 'di')
			.replace(/([BE][iïíìl!1I] )/g, 'El ')
			.replace(/(EUa )/g, 'Ella ')
			.replace(/ eua /g, ' ella')
			.replace(/( [ce][iïíìl!1I] )/g, ' el ')
			.replace(/( [ce][iïíìl!1I][iïíìl!1I]a)/g, ' ella')
			.replace(/( [ce][iïíìl!1I][iïíìl!1I]o)/g, ' ello')
			.replace(/(eU)/g, 'ell')
			.replace(/(cm)/g, 'em')
			.replace(/[BE][nqu] /g, 'En ')
			.replace(/ cn/g, ' en')
			.replace(/[ce][un]do([ ,.\n])/g, 'endo$1')
			.replace(/ [ce][un] /g, ' en ')
			.replace(/[ce]n[lt] /g, 'ent ')
			.replace(/[ce][un][lt]o[un][ce][ce]s/g, 'entonces')
			.replace(/E[un][lt]o[un][ce][ce]s/g, 'Entonces')
			.replace(/ [ce]n[it]r[ce]/g, ' entre')
			.replace(/ [ce][as] /g, ' es ')
			.replace(/[ce][8s][ce]ri/g, 'escri')
			.replace(/ [ce][s8][lt]([aáà])([ t])/g, ' est$1$2')
			.replace(/ [ce][8s][tl]([aoe])/g, ' est$1')
			.replace(/ [ce]t[ce]([.,;:])/g, ' etc$1')
			.replace(/ é[8s][tl]([aoe])/g, ' ést$1')
			.replace(/exlr/g, 'extr')
			.replace(/(g[iïíìl!1I][6é]s)/g, 'glés')
			.replace(/g[nu][óé]/g, 'gué')
			.replace(/gn[ce]/g, 'gne')
			.replace(/gu[ce]/g, 'gue')
			.replace(/ha[nu] /g, 'han ')
			.replace(/ ba /g, ' ha ')
			.replace(/ h[ns]b/g, ' hab')
			.replace(/[bh][oe]mbr[ce]/g, 'hombre')
			.replace(/(horn)/g, 'hom')
			.replace(/ [iïíìl!1]a /g, ' la ')
			.replace(/ [iïíìl!1]a[a8s] /g, ' las ')
			.replace(/ [iïíìl!1]o /g, ' lo ')
			.replace(/ [iïíìl!1]o[a8s] /g, ' los ')
			.replace(/(lloa([ ,.\n]))/g, 'llos$1')
			.replace(/(llaa([ ,.\n]))/g, 'llas$1')
			.replace(/(loient[ec]([ ,.\n]))/g, 'lmente$1')
			.replace(/m[ce][un]t[ce]([ ,.\n])/g, 'mente$1')
			.replace(/m[ce]ot[ce]([ ,.\n])/g, 'mente$1')
			.replace(/ mne/g, ' mue')
			.replace(/(Q[anou][ce])/g, 'Que')
			.replace(/([çq][anou][ceoa])/g, 'que')
			.replace(/([çq]ii[ceo])/g, 'que')
			.replace(/q[a-záéíóú]e/g, 'que')
			.replace(/q[a-záéíóú]é/g, 'qué')
			.replace(/([çq][anou][óé])/g, 'qué')
			.replace(/ [sB][ec] /g, ' se ')
			.replace(/ sn /g, ' su ')
			.replace(/ tau /g, ' tan ')
			.replace(/ [un][un] /g, ' un ')
			.replace(/ [un](?:[un]|ii)a /g, ' una ')
			.replace(/ o[un]a /g, ' una ')
			.replace(/ \'?"?[lv] /g, ' y ')
			.replace(/(•)/g, '.')
			.replace(/[lt!]'/g, 'r')
			.replace(/i([A-ZÁÉÍÓÚ])/g,'¡$1')
			
			//paréntesis en palabras
			.replace(/\(\)/g,'o')
			.replace(/\(\'\)/g,'ó')
			.replace(/l\)([a-záéíóú])/g,'b$1')
			.replace(/\(\(/g,'«')
			.replace(/\)\)/g,'»')
			
			//espacio después de punto o coma
			.replace(/(\.)([A-ZÁÉÍÓÚ])/g,'$1 $2')
			.replace(/([,;:!\?])([a-záéíóú])/g,'$1 $2')
			.replace(/([a-záéíóú][a-záéíóú][a-záéíóú])\. ([a-záéíóú][a-záéíóú])/g,'$1, $2')
			.replace(/([a-záéíóú])\'([a-záéíóú])/g,'$1 $2')
			.replace(/\.-/g, '.—')
			.replace(/\^ /g, ', ')
			.replace(/ P /g, '? ')
			.replace(/([¿¡]) /g, '$1');
	

	};
	
	/**
	 * As you work your way through the page, when you encounter a reference, just mark it with <ref></ref> tags and continue.
	 * Once you've got to the end of the page and proofed the references, simply highlight each reference in turn,
	 * and use this function to move it to its proper position.
	 * @param {object} editor The script helpers for the page.
	 */
	var makeReference = function(editor) {
		_initialise();
		
		var editbox = $('#wpTextbox1').get(0);
		editbox.focus();
		var refStart = editbox.selectionStart;
		var refEnd = editbox.selectionEnd;

		var firstref = editbox.value.indexOf('<ref></ref>');
		if (firstref != -1) {
			editbox.value = editbox.value.slice(0,firstref+5)
			              + editbox.value.slice(refStart, refEnd)
			              + editbox.value.slice(firstref+5, refStart)
			              + editbox.value.slice(refEnd);
		}
		addPageFooter(editor);
	};

	/**
	 * Insert formatted references into the footer box if needed.
	 * @param {object} editor The script helpers for the page.
	 */
	var addPageFooter = function(editor) {
		_initialise();
		
		var editbox = $('#wpTextbox1').get(0);
		var footerbox;
		var generic;
		var format;
		var f;
		if (editbox.value.indexOf("<ref") == -1 && editbox.value.indexOf("{{#tag:ref") == -1) {
			// page contains no refs
			generic = true;
			for (f in state.specialFormats) {
				format = state.specialFormats[f];
				if (mw.config.get('wgTitle').match(format.title)) {
					footerbox = format.footer;
					generic = false;
					break;
				}
			}

			// no special footer matched, use just strip out the references tag
			if (generic)
				footerbox = '';
		}
		
		else {
			generic = true;
			for (f in state.specialFormats) {
				format = state.specialFormats[f];
				if (mw.config.get('wgTitle').match(format.title)) {
					footerbox = format.footerWithReferences;
					generic = false;
					break;
				}
			}

			// no special footer matched, so use a generic ref tag
			if (generic)
				footerbox  = '{{listaref}}';
			}
		
		$('#wpFooterTextbox').val(footerbox);
		
	};

	/**
	 * Mark the selected text with {{sc}}. If the text is uppercase, it will be converted to titlecase.
	 * @param {object} editor The script helpers for the page.
	 */
	var smallcaps = function(editor) {
		_initialise();
		
		editor.replaceSelection(function(text) {
			// Applying small-caps to all-caps text is pointless...
			// ... unless the all-caps is OCR of text that is actually small-caps.
			// Check if text is all-caps, and if it is, convert it to title case before applying small-caps.
			if (text == text.toUpperCase())
				text = _titlecase(text);
			
			return '{{may|' + text + '}}';
		});
	};

	/**
	 * Convert the text to uppercase.
	 * @param {object} editor The script helpers for the page.
	 */
	var upper = function(editor) {
		_initialise();
		
		editor.replaceSelection(function(text) {
			return text.toUpperCase();
		});
	};
	var lower = function(editor) {
		_initialise();
		
		editor.replaceSelection(function(text) {
			return text.toLowerCase();
		});
	};
	

/*global $, mw*/

/*
 * Query an ocr for a given Page:, first try to get the hocr text layer as it's available
 * for most book, fast and of a better quality. If it fails, try the older and slower
 * ocr method. hocr fail around 1/5000 books. ocr should never fails as it use the image
 * visible on the Page:.
 */

var lang = mw.config.get( 'wgContentLanguage' );

function disable_input(set)
{
	if (set) {
		$(document).keyup(function(e) {
			if (e.which == 27) { disable_input(false); }
		});
	}

	set ? $('#wsOcr1').off('click') : $('#wsOcr1').on('click', do_hocr);

	$('#wpTextbox1').prop('disabled', set);
}

function ocr_callback(data) {
	if (data.error) {
		alert(data.text);
	} else {
		// Checking if tb is disabled is required with chrome as ESC doesn't kill
		// the query.
		var tb = document.getElementById("wpTextbox1");
		if (tb.disabled)
			tb.value = data.text;
	}

	disable_input(false);
}

function hocr_callback(data) {
	if (data.error) {
		// Fallback to the slow way.
		disable_input(false);
		do_ocr();
		return;
	} else {
		// Checking if tb is disabled is required with chrome as ESC doesn't kill
		// the query.
		var tb = document.getElementById("wpTextbox1");
		if (tb.disabled) {
                        localStorage.ws_hOCR = data.text;

			var text = $(data.text).text();
			// Ugly as hell.
			text = text.replace(/[ ]*\n[ ]*/g, '\n')
				.replace(/\n\n\n\n/g, '@_@_@_@_@_@')
				.replace(/\n\n/g, '\n')
				.replace(/@_@_@_@_@_@/g, '\n\n')
				.replace(/\n\n\n/g, '\n\n');
			tb.value = $.trim(text);
		}
	}

	disable_input(false);
}

function do_hocr() {
	disable_input(true);

	var request_url = '//tools.wmflabs.org/phetools/hocr_cgi.py?cmd=hocr&book='
		+ encodeURIComponent(mw.config.get('wgTitle')) + '&lang=' + lang + '&user=' + mw.config.get('wgUserName');

	$.getJSON(request_url).done(hocr_callback).fail(do_ocr);
}

function do_ocr() {
	if ($( '.prp-page-image img' ).length) {
		disable_input(true);

		// server side can't use protocol relative url, request it as https:
		var url_image = 'https:' + $( '.prp-page-image img' ).attr('src');

		var request_url = "//tools.wmflabs.org/phetools/ocr.php?cmd=ocr&url="+url_image+"&lang="+lang+"&user="+mw.config.get('wgUserName');

		$.getJSON( request_url ).done( ocr_callback );
	}
}
});
// </nowiki>