Support unmodified Xpdf binaries

Use stdout redirection scripts for pdfinfo and, on Windows, a script to
run pdftotext hidden, which together allow for all unmodified binaries
(including, probably, symlinked system ones, though I didn't test that).

On Windows, using a .vbs does cause a brief wait cursor. The stock
pdfinfo needs the redirection script anyway, so that's unavoidable, but
on the async branch I think we'll be able to switch to pdf.js for the
page count, at which point maybe I'll try to remember how I modified the
Windows binaries to be hidden and use a modified version of pdftotext to
avoid VBScript. (We use the stock pdftotext elsewhere already.)

chrome/content/zotero/preferences/preferences_search.js

@@ -162,46 +162,40 @@ Zotero_Preferences.Search = {
 	 * if a newer version is available
 	 */
 	checkPDFToolsDownloadVersion: function () {
-		var url = Zotero.Fulltext.pdfToolsDownloadBaseURL
-					+ Zotero.platform.replace(' ', '-') + '.latest';
+		var url = Zotero.Fulltext.pdfToolsDownloadBaseURL + 'latest.json';
 		
 		// Find latest version for this platform
 		var self = this;
 		var sent = Zotero.HTTP.doGet(url, function (xmlhttp) {
 			try {
-				if (xmlhttp.status == 200) {
+				if (xmlhttp.status != 200) {
+					throw new Error("Unexpected response code " + xmlhttp.status);
+				}
+				
+				var platform = Zotero.platform.replace(/\s/g, '-');
+				var json = JSON.parse(xmlhttp.responseText);
+				var latestVersion = json[platform] || json['default'];
+				
+				Zotero.debug("Latest PDF tools version for " + platform + " is " + latestVersion);
+				
 				var converterIsRegistered = Zotero.Fulltext.pdfConverterIsRegistered();
 				var infoIsRegistered = Zotero.Fulltext.pdfInfoIsRegistered();
 				var bothRegistered = converterIsRegistered && infoIsRegistered;
 				
-					var converterVersion = xmlhttp.responseText.split(/\s/)[0];
-					var infoVersion = xmlhttp.responseText.split(/\s/)[1];
-					
-					// Install if not installed, version unknown, Xpdf 3.02 (to upgrade to Poppler),
-					// or outdated
-					var converterVersionAvailable = converterVersion &&
-						(!converterIsRegistered ||
+				// Install if not installed, version unknown, outdated, or
+				// Xpdf 3.02/3.04 (to upgrade to Poppler),
+				var converterVersionAvailable = (!converterIsRegistered ||
 						Zotero.Fulltext.pdfConverterVersion == 'UNKNOWN'
-								|| (converterVersion != '3.02' && Zotero.Fulltext.pdfConverterVersion == '3.02')
-								||  converterVersion > Zotero.Fulltext.pdfConverterVersion);
-					var infoVersionAvailable = infoVersion &&
-						(!infoIsRegistered ||
+							|| latestVersion > Zotero.Fulltext.pdfConverterVersion
+							|| (latestVersion != '3.02' && Zotero.Fulltext.pdfConverterVersion == '3.02')
+							|| (latestVersion != '3.02' && latestVersion != '3.04' && Zotero.Fulltext.pdfConverterVersion == '3.04'));
+				var infoVersionAvailable = (!infoIsRegistered ||
 						Zotero.Fulltext.pdfInfoVersion == 'UNKNOWN'
-								|| (infoVersion != '3.02' && Zotero.Fulltext.pdfInfoVersion == '3.02')
-								|| infoVersion > Zotero.Fulltext.pdfInfoVersion);
+							|| latestVersion > Zotero.Fulltext.pdfInfoVersion
+							|| (latestVersion != '3.02' && Zotero.Fulltext.pdfInfoVersion == '3.02')
+							|| (latestVersion != '3.02' && latestVersion != '3.04' && Zotero.Fulltext.pdfInfoVersion == '3.04'));
 				var bothAvailable = converterVersionAvailable && infoVersionAvailable;
 				
-					/*
-					Zotero.debug(converterIsRegistered);
-					Zotero.debug(infoIsRegistered);
-					Zotero.debug(converterVersion);
-					Zotero.debug(infoVersion);
-					Zotero.debug(Zotero.Fulltext.pdfConverterVersion);
-					Zotero.debug(Zotero.Fulltext.pdfInfoVersion);
-					Zotero.debug(converterVersionAvailable);
-					Zotero.debug(infoVersionAvailable);
-					*/
-					
 				// Up to date -- disable update button
 				if (!converterVersionAvailable && !infoVersionAvailable) {
 					var button = document.getElementById('pdftools-update-button');
@@ -221,12 +215,12 @@ Zotero_Preferences.Search = {
 					
 					if (converterVersionAvailable) {
 						let tvp = Zotero.getString('zotero.preferences.search.pdf.toolVersionPlatform',
-								[Zotero.Fulltext.pdfConverterName, converterVersion]);
+							[Zotero.Fulltext.pdfConverterName, latestVersion]);
 						msg += '- ' + tvp + '\n';
 					}
 					if (infoVersionAvailable) {
 						let tvp = Zotero.getString('zotero.preferences.search.pdf.toolVersionPlatform',
-								[Zotero.Fulltext.pdfInfoName, infoVersion]);
+							[Zotero.Fulltext.pdfInfoName, latestVersion]);
 						msg += '- ' + tvp + '\n';
 					}
 					msg += '\n';
@@ -244,24 +238,17 @@ Zotero_Preferences.Search = {
 						null, null, null, {});
 					
 					if (index == 0) {
-							var installVersions = {
-								converter: converterVersionAvailable ?
-									converterVersion : null,
-								info: infoVersionAvailable ?
-									infoVersion : null
-							};
-							
 						document.getElementById('pdftools-update-button').disabled = true;
 						var str = Zotero.getString('zotero.preferences.search.pdf.downloading');
 						document.getElementById('pdftools-update-button').setAttribute('label', str);
 						
 						if (converterVersionAvailable && infoVersionAvailable) {
-								Zotero.Fulltext.downloadPDFTool('converter', converterVersion, function (success) {
+							Zotero.Fulltext.downloadPDFTool('converter', latestVersion, function (success) {
 								if (!success) {
 									self.onPDFToolsDownloadError("Error downloading pdftotext");
 									return;
 								}
-									Zotero.Fulltext.downloadPDFTool('info', infoVersion, function (success) {
+								Zotero.Fulltext.downloadPDFTool('info', latestVersion, function (success) {
 									if (!success) {
 										self.onPDFToolsDownloadError("Error downloading pdfinfo");
 										return;
@@ -271,7 +258,7 @@ Zotero_Preferences.Search = {
 							});
 						}
 						else if (converterVersionAvailable) {
-								Zotero.Fulltext.downloadPDFTool('converter', converterVersion, function (success) {
+							Zotero.Fulltext.downloadPDFTool('converter', latestVersion, function (success) {
 								if (!success) {
 									self.onPDFToolsDownloadError("Error downloading pdftotext");
 									return;
@@ -280,7 +267,7 @@ Zotero_Preferences.Search = {
 							});
 						}
 						else {
-								Zotero.Fulltext.downloadPDFTool('info', infoVersion, function (success) {
+							Zotero.Fulltext.downloadPDFTool('info', latestVersion, function (success) {
 								if (!success) {
 									self.onPDFToolsDownloadError("Error downloading pdfinfo");
 									return;
@@ -291,11 +278,6 @@ Zotero_Preferences.Search = {
 					}
 				}
 			}
-				// Version not found for platform
-				else if (xmlhttp.status == 404) {
-					self.onPDFToolsDownloadError(404);
-				}
-			}
 			catch (e) {
 				self.onPDFToolsDownloadError(e);
 			}

chrome/content/zotero/xpcom/fulltext.js

@@ -55,7 +55,7 @@ Zotero.Fulltext = new function(){
 	//this.clearItemContent = clearItemContent;
 	this.purgeUnusedWords = purgeUnusedWords;
 	
-	this.__defineGetter__("pdfToolsDownloadBaseURL", function() { return 'http://www.zotero.org/download/xpdf/'; });
+	this.__defineGetter__("pdfToolsDownloadBaseURL", function() { return 'https://www.zotero.org/download/xpdf/'; });
 	this.__defineGetter__("pdfToolsName", function() { return 'Xpdf'; });
 	this.__defineGetter__("pdfToolsURL", function() { return 'http://www.foolabs.com/xpdf/'; });
 	this.__defineGetter__("pdfConverterName", function() { return 'pdftotext'; });
@@ -82,9 +82,11 @@ Zotero.Fulltext = new function(){
 
 	var _pdfConverterVersion = null;
 	var _pdfConverterFileName = null;
+	var _pdfConverterScript = null; // nsIFile of hidden window script on Windows
 	var _pdfConverter = null; // nsIFile to executable
 	var _pdfInfoVersion = null;
 	var _pdfInfoFileName = null;
+	var _pdfInfoScript = null; // nsIFile of redirection script
 	var _pdfInfo = null; // nsIFile to executable
 	
 	var _idleObserverIsRegistered = false;
@@ -179,7 +181,7 @@ Zotero.Fulltext = new function(){
 				var fileName = this.pdfInfoFileName;
 			}
 			
-			var spec = this.pdfToolsDownloadBaseURL + fileName + '-' + version;
+			var spec = this.pdfToolsDownloadBaseURL + version + "/" + fileName;
 			var uri = ioService.newURI(spec, null, null);
 			
 			var file = Zotero.getTempDirectory();
@@ -202,15 +204,43 @@ Zotero.Fulltext = new function(){
 				
 				Zotero.File.putContentsAsync(file, is)
 				.then(function () {
+					var scriptExt = _getScriptExtension();
+					// On Windows, write out script to hide pdftotext console window
+					if (tool == 'converter') {
+						if (Zotero.isWin) {
+							var content = Zotero.File.getContentsFromURL('resource://zotero/hide.' + scriptExt);
+							var scriptFile = Zotero.getTempDirectory();
+							scriptFile.append('pdftotext.' + scriptExt);
+							Zotero.File.putContents(scriptFile, content);
+						}
+					}
+					// Write out output redirection script for pdfinfo
+					else if (tool == 'info') {
+						var content = Zotero.File.getContentsFromURL('resource://zotero/redirect.' + scriptExt);
+						var scriptFile = Zotero.getTempDirectory();
+						scriptFile.append('pdfinfo.' + scriptExt);
+						Zotero.File.putContents(scriptFile, content);
+					}
+					
 					// Set permissions to 755
 					if (Zotero.isMac) {
 						file.permissions = 33261;
+						if (scriptFile) {
+							scriptFile.permissions = 33261;
+						}
 					}
 					else if (Zotero.isLinux) {
 						file.permissions = 493;
+						if (scriptFile) {
+							scriptFile.permissions = 493;
+						}
 					}
 					
 					var destDir = Zotero.getZoteroDirectory()
+					// Move redirect script and executable into data dir
+					if (scriptFile) {
+						scriptFile.moveTo(destDir, null);
+					}
 					file.moveTo(destDir, null);
 					
 					// Write the version number to a file
@@ -223,6 +253,11 @@ Zotero.Fulltext = new function(){
 					if (callback) {
 						callback(true);
 					}
+				})
+				.catch(function (e) {
+					Zotero.debug(e, 1);
+					Components.utils.reportError(e);
+					callback(false);
 				});
 			});
 		}
@@ -274,11 +309,43 @@ Zotero.Fulltext = new function(){
 			return false;
 		}
 		
+		// If scripts exist, use those instead
+		switch (tool) {
+		case 'converter':
+			if (Zotero.isWin) {
+				var script = Zotero.getZoteroDirectory();
+				script.append('pdftotext.' + _getScriptExtension())
+				if (script.exists()) {
+					Zotero.debug(script.leafName + " registered");
+					_pdfConverterScript = script;
+				}
+			}
+			break;
+		
+		case 'info':
+			var script = Zotero.getZoteroDirectory();
+			script.append('pdfinfo.' + _getScriptExtension())
+			// The redirection script is necessary to run pdfinfo
+			if (!script.exists()) {
+				Zotero.debug(script.leafName + " not found -- PDF statistics disabled");
+				return false;
+			}
+			Zotero.debug(toolName + " redirection script registered");
+			_pdfInfoScript = script;
+			break;
+		}
+		
 		var versionFile = exec.parent;
 		versionFile.append(fileName + '.version');
 		if (versionFile.exists()) {
+			try {
 				var version = Zotero.File.getSample(versionFile).split(/[\r\n\s]/)[0];
 			}
+			catch (e) {
+				Zotero.debug(e, 1);
+				Components.utils.reportError(e);
+			}
+		}
 		if (!version) {
 			var version = 'UNKNOWN';
 		}
@@ -295,7 +362,7 @@ Zotero.Fulltext = new function(){
 				break;
 		}
 		
-		Zotero.debug(toolName + ' version ' + version + ' registered at ' + exec.path);
+		Zotero.debug(toolName + ' version ' + version + ' registered');
 		
 		return true;
 	}
@@ -529,16 +596,19 @@ Zotero.Fulltext = new function(){
 		}
 		cacheFile.append(this.pdfConverterCacheFile);
 		
-		if (_pdfInfo) {
+		if (_pdfInfoScript) {
 			var infoFile = cacheFile.parent;
 			infoFile.append(this.pdfInfoCacheFile);
-			Zotero.debug('Running pdfinfo "' + file.path + '" "' + infoFile.path + '"');
 			
-			var proc = Components.classes["@mozilla.org/process/util;1"].
-					createInstance(Components.interfaces.nsIProcess);
-			proc.init(_pdfInfo);
+			var args = [_pdfInfo.path, file.path, infoFile.path];
+			
+			Zotero.debug("Running " + _pdfInfoScript.path + ' '
+				+ args.map(arg => "'" + arg + "'").join(' '));
+			
+			var proc = Components.classes["@mozilla.org/process/util;1"]
+				.createInstance(Components.interfaces.nsIProcess);
+			proc.init(_pdfInfoScript);
 			
-			var args = [file.path, infoFile.path];
 			try {
 				proc.runw(true, args, args.length);
 				var totalPages = this.getTotalPagesFromFile(itemID);
@@ -551,15 +621,12 @@ Zotero.Fulltext = new function(){
 			Zotero.debug(this.pdfInfoName + " is not available");
 		}
 		
-		Zotero.debug('Running pdftotext -enc UTF-8 -nopgbrk '
-			+ (allPages ? '' : '-l ' + maxPages) + ' "' + file.path + '" "'
-			+ cacheFile.path + '"');
+		var args = []
+		if (_pdfConverterScript) {
+			args.push(_pdfConverter.path);
+		}
+		args.push('-enc', 'UTF-8', '-nopgbrk');
 		
-		var proc = Components.classes["@mozilla.org/process/util;1"].
-				createInstance(Components.interfaces.nsIProcess);
-		proc.init(_pdfConverter);
-		
-		var args = ['-enc', 'UTF-8', '-nopgbrk'];
 		if (allPages) {
 			if (totalPages) {
 				var pagesIndexed = totalPages;
@@ -570,6 +637,20 @@ Zotero.Fulltext = new function(){
 			var pagesIndexed = Math.min(maxPages, totalPages);
 		}
 		args.push(file.path, cacheFile.path);
+		
+		var proc = Components.classes["@mozilla.org/process/util;1"]
+			.createInstance(Components.interfaces.nsIProcess);
+		if (_pdfConverterScript) {
+			Zotero.debug("Running " + _pdfConverterScript.path + ' '
+				+ args.map(arg => "'" + arg + "'").join(' '));
+			proc.init(_pdfConverterScript);
+		}
+		else {
+			Zotero.debug("Running " + _pdfConverter.path + ' '
+				+ args.map(arg => "'" + arg + "'").join(' '));
+			proc.init(_pdfConverter);
+		}
+		
 		try {
 			proc.runw(true, args, args.length);
 		}
@@ -1679,4 +1760,9 @@ Zotero.Fulltext = new function(){
 			return w;
 		});
 	}
+	
+	function _getScriptExtension() {
+		return Zotero.isWin ? 'vbs' : 'sh';
+	}
+
 }

chrome/content/zotero/xpcom/schema.js

@@ -1590,6 +1590,9 @@ Zotero.Schema = new function(){
 		var styleUpdates = xmlhttp.responseXML.getElementsByTagName('style');
 		
 		var updatePDFTools = function () {
+			// No updates for PPC
+			if (Zotero.platform == 'MacPPC') return;
+			
 			let pdfToolsUpdates = xmlhttp.responseXML.getElementsByTagName('pdftools');
 			if (pdfToolsUpdates.length) {
 				let availableVersion = pdfToolsUpdates[0].getAttribute('version');

resource/hide.vbs

@@ -0,0 +1,24 @@
+Option Explicit
+
+Dim WshShell, fso, exe, args, I
+
+Set WshShell = Wscript.CreateObject("Wscript.Shell")
+Set fso = WScript.CreateObject("Scripting.FileSystemObject")
+
+If WScript.Arguments.Count = 0 Then
+  WScript.Echo "Usage: hidden.vbs program.exe [args...]"
+  WScript.Quit 1
+End If
+
+exe = WScript.Arguments(0)
+If Not(fso.FileExists(exe)) Then
+  WScript.Echo "Executable not found: " & exe
+  WScript.Quit 1
+End If
+
+args = ""
+For I = 1 to WScript.Arguments.Count - 1
+  args = args & " " & chr(34) & WScript.Arguments(I) & chr(34)
+Next
+
+WshShell.Run exe & args, 0, true

resource/redirect.sh

@@ -0,0 +1,6 @@
+#!/bin/sh
+if [ -z "$1" ] || [ -z "$2" ] || [ -z "$3" ]; then
+    echo "Usage: $0 cmd source output.txt"
+    exit 1
+fi
+"$1" "$2" > "$3"

resource/redirect.vbs

@@ -0,0 +1,33 @@
+Option Explicit
+
+Dim WshShell
+Dim fso
+Dim exe, src, dest
+
+Set WshShell = Wscript.CreateObject("Wscript.Shell")
+Set fso = WScript.CreateObject("Scripting.FileSystemObject")
+
+If Not(Wscript.Arguments.Count = 3) Then
+  Wscript.Echo "Usage: redirect.vbs <.exe file> <source file> <text file>"
+  WScript.Quit 1
+End If
+
+exe = WScript.Arguments(0)
+src = WScript.Arguments(1)
+dest = WScript.Arguments(2)
+If Not(fso.FileExists(exe)) Then
+  WScript.Echo "Executable not found: " & exe
+  WScript.Quit 1
+End If
+
+If Not(fso.FileExists(src)) Then
+  WScript.Echo "Source file not found: " & src
+  WScript.Quit 1
+End If
+
+If Not(fso.FolderExists(Left(dest, InstrRev(dest, "\")))) Then
+  WScript.Echo "Destination folder not found: " & Left(dest, InstrRev(dest, "\"))
+  WScript.Quit 1
+End If
+
+WshShell.Run "%comspec% /c " & exe & " " & chr(34) & src & chr(34) & " > " & chr(34) & dest & chr(34), 0, true