[printergui] Sanitize job names (ASCII only to be extra safe)

1 files changed, 279 insertions, 2 deletions

diff --git a/src/maingui/printergui.cpp b/src/maingui/printergui.cpp
index 2467dcc..e235007 100644
--- a/src/maingui/printergui.cpp
+++ b/src/maingui/printergui.cpp
@@ -12,6 +12,10 @@ static QStringList knownColorOptions = QStringList() << "ColorModel" << "XRXColo
 static QStringList knownDuplexOptions = QStringList() << "Duplex";
 static QStringList knownPageSizeOptions = QStringList() << "PageSize";
 
+static int isUtf8(const char *s);
+static void toLatin1(char *s);
+static char* cleanCupsString(const char* in);
+
 // ____________________________________________________________________________
 PrinterGui::PrinterGui(char *argv[], QWidget *parent) :
 	QDialog(parent),
@@ -293,7 +297,7 @@ void PrinterGui::on_buttonPrint_clicked()
 	}
 
 	cupsSetUser(this->user);
-	char jobtitle[200];
+	char jobtitle[250];
 	const char *docName;
 	docName = getenv("J");
 	if (docName == NULL) {
@@ -302,7 +306,11 @@ void PrinterGui::on_buttonPrint_clicked()
 	if (docName == NULL) {
 		docName = "Untitled";
 	}
-	snprintf(jobtitle, sizeof(jobtitle), "gui-%d-%s (%s)", (int)getpid(), this->user, docName);
+	char *cleanUser = cleanCupsString(this->user);
+	char *cleanTitle = cleanCupsString(docName);
+	snprintf(jobtitle, sizeof(jobtitle), "gui-%d-%s (%s)", (int)getpid(), cleanUser, cleanTitle);
+	free(cleanUser);
+	free(cleanTitle);
 
 	// Drucken
 	jobId = cupsPrintFile(dest->name,
@@ -337,3 +345,272 @@ void PrinterGui::bgTimer_timeout()
 	}
 }
 
+/**
+ * Checks if the given c string is UTF-8.
+ * Using as bool-function:
+ * Return value != 0 means (potentially) UTF-8,
+ * that is, it can be processed by UTF-8 functions without error.
+ * Using for in-detail detection:
+ * Return value 0 = not valid UTF-8
+ * Return value 1 = valid ASCII (which is a subset of UTF-8)
+ * Return value 2 = valid UTF-8, but not valid ASCII.
+ */
+static int isUtf8(const char *s)
+{
+	bool isAscii = true;
+	while (*s != '\0') {
+		if ((s[0] & 248) == 240 &&
+				(s[1] & 192) == 128 &&
+				(s[2] & 192) == 128 &&
+				(s[3] & 192) == 128) {
+			s += 4;
+			isAscii = false;
+		} else if ((s[0] & 240) == 224 &&
+				(s[1] & 192) == 128 &&
+				(s[2] & 192) == 128) {
+			s += 3;
+			isAscii = false;
+		} else if ((s[0] & 224) == 192 &&
+				(s[1] & 192) == 128) {
+			s += 2;
+			isAscii = false;
+		} else if ((s[0] & 128) != 0) {
+			// Invalid UTF-8 sequence, so no ASCII either
+			return 0;
+		} else {
+			s+= 1;
+		}
+	}
+	return isAscii ? 1 : 2;
+}
+
+static void toLatin1(char *s)
+{
+	char *d = s;
+	while (*s != '\0') {
+		if ((s[0] & 248) == 240 &&
+				(s[1] & 192) == 128 &&
+				(s[2] & 192) == 128 &&
+				(s[3] & 192) == 128) {
+			d[0] = '_';
+			s += 4;
+		} else if ((s[0] & 240) == 224 &&
+				(s[1] & 192) == 128 &&
+				(s[2] & 192) == 128) {
+			d[0] = '_';
+			s += 3;
+		} else if ((s[0] & 224) == 192 &&
+				(s[1] & 192) == 128) {
+			int c = ((s[0] & 31) << 6) | (s[1] & 63);
+			if (c > 255) {
+				d[0] = '_';
+			} else {
+				d[0] = (char)c;
+			}
+			s += 2;
+		} else if ((s[0] & 128) != 0) {
+			// Invalid UTF-8 sequence!?
+			d[0] = '_';
+			s+= 1;
+		} else {
+			d[0] = s[0];
+			s+= 1;
+		}
+		d += 1;
+	}
+	*d = '\0';
+}
+
+static char* cleanCupsString(const char* in)
+{
+	int enc = isUtf8(in);
+	char *out = strdup(in);
+	if (enc == 2) {
+		// UTF-8, turn to Latin-1
+		toLatin1(out);
+	}
+	// Try to convert umlauts etc., remove non-printing chars
+	unsigned char *tmp = (unsigned char*)out;
+	while (*tmp != '\0') {
+		if (*tmp < 0x20 || *tmp > 0x7E) {
+			switch (*tmp) {
+			case 0x84: // „
+			case 0x93: // “
+			case 0x94: // ”
+			case 0xAB: // «
+			case 0xBB: // »
+				*tmp = '"';
+				break;
+			case 0x91: // ‘
+			case 0x92: // ’
+			case 0xB4: // ´
+				*tmp = '\'';
+				break;
+			case 0xB7: // ·
+				*tmp = '*';
+				break;
+			case 0xF7: // ÷
+				*tmp = '/';
+				break;
+			case 0x82: // ‚
+			case 0xB8: // ¸
+				*tmp = ',';
+				break;
+			case 0x96: // –
+			case 0x97: // —
+			case 0xAC: // ¬
+				*tmp = '-';
+				break;
+			case 0xAF: // ¯
+			case 0xB0: // °
+				*tmp = '^';
+				break;
+			case 0x98: // ˜
+				*tmp = '~';
+				break;
+			case 0xB9: // ¹
+				*tmp = '1';
+				break;
+			case 0xB2: // ²
+				*tmp = '2';
+				break;
+			case 0xB3: // ³
+				*tmp = '3';
+				break;
+			case 0xC0: // À
+			case 0xC1: // Á
+			case 0xC2: // Â
+			case 0xC3: // Ã
+			case 0xC4: // Ä
+			case 0xC5: // Å
+			case 0xC6: // Æ
+				*tmp = 'A';
+				break;
+			case 0xAA: // ª
+			case 0xE0: // à
+			case 0xE1: // á
+			case 0xE2: // â
+			case 0xE3: // ã
+			case 0xE4: // ä
+			case 0xE5: // å
+			case 0xE6: // æ
+				*tmp = 'a';
+				break;
+			case 0xA9: // ©
+			case 0xC7: // Ç
+				*tmp = 'C';
+				break;
+			case 0xA2: // ¢
+			case 0xE7: // ç
+				*tmp = 'c';
+				break;
+			case 0xD0: // Ð
+				*tmp = 'D';
+				break;
+			case 0xF0: // ð
+				*tmp = 'd';
+				break;
+			case 0xC8: // È
+			case 0xC9: // É
+			case 0xCA: // Ê
+			case 0xCB: // Ë
+				*tmp = 'E';
+				break;
+			case 0xE8: // è
+			case 0xE9: // é
+			case 0xEA: // ê
+			case 0xEB: // ë
+				*tmp = 'e';
+				break;
+			case 0x83: // ƒ
+				*tmp = 'f';
+				break;
+			case 0xCC: // Ì
+			case 0xCD: // Í
+			case 0xCE: // Î
+			case 0xCF: // Ï
+				*tmp = 'I';
+				break;
+			case 0xA1: // ¡
+			case 0xEC: // ì
+			case 0xED: // í
+			case 0xEE: // î
+			case 0xEF: // ï
+				*tmp = 'i';
+				break;
+			case 0xD1: // Ñ
+				*tmp = 'N';
+				break;
+			case 0xF1: // ñ
+				*tmp = 'n';
+				break;
+			case 0xD2: // Ò
+			case 0xD3: // Ó
+			case 0xD4: // Ô
+			case 0xD5: // Õ
+			case 0xD6: // Ö
+			case 0xD8: // Ø
+				*tmp = 'O';
+				break;
+			case 0xF2: // ò
+			case 0xF3: // ó
+			case 0xF4: // ô
+			case 0xF5: // õ
+			case 0xF6: // ö
+			case 0xF8: // ø
+				*tmp ='o';
+				break;
+			case 0xA3: // £
+			case 0xA7: // §
+				*tmp = 'P';
+				break;
+			case 0xAE: // ®
+				*tmp = 'R';
+				break;
+			case 0x8A: // Š
+				*tmp = 'S';
+				break;
+			case 0x9A: // š
+			case 0xDF: // ß
+				*tmp = 's';
+				break;
+			case 0xDE: // Þ
+				*tmp = 'T';
+				break;
+			case 0xFE: // þ
+				*tmp = 't';
+				break;
+			case 0xD9: // Ù
+			case 0xDA: // Ú
+			case 0xDB: // Û
+			case 0xDC: // Ü
+				*tmp = 'U';
+				break;
+			case 0xB5: // µ
+			case 0xF9: // ù
+			case 0xFA: // ú
+			case 0xFB: // û
+			case 0xFC: // ü
+				*tmp = 'u';
+				break;
+			case 0xD7: // ×
+				*tmp = 'x';
+				break;
+			case 0xA5: // ¥
+			case 0xDD: // Ý
+				*tmp = 'Y';
+				break;
+			case 0xFD: // ý
+			case 0xFF: // ÿ
+				*tmp = 'y';
+				break;
+			default:
+				*tmp = '.';
+				break;
+			}
+		}
+		tmp++;
+	}
+	return out;
+}
+