Inserting into DB with parameters safe from SQL injection? - c#

I been reading a bit about SQL injection and I want to be sure my code is lets say "safe" from it, I was planning on using RegExp validators to check the user input but another post in here suggested only using parametrized querys, well I'm using them but I want to be sure my code is safe, is it?
using ( SqlConnection dataConnection = new SqlConnection(myConnectionString) )
{
using ( SqlCommand dataCommand = dataConnection.CreateCommand() )
{
dataCommand.CommandText = "INSERT INTO Lines (Name, CreationTime) " +
"VALUES (#LineName, #CurrentDateTime)";
dataCommand.Parameters.AddWithValue("#LineName", TextBox2.Text);
dataCommand.Parameters.AddWithValue("#CurrentDateTime", DateTime.Now.ToString());
dataConnection.Open();
//do other DB stuff
I chop the last part to make the post shorter, the rest is just trying and catching exceptions and closing db connection as well as providing user feedback on inserting successful.

Your code is fine, it is protected from injection because the values are passed as parameters not string literals. However, if you are writing this type of data access yourself, have you considered creating SqlParameter objects and explicitly setting the type, size etc, and adding the parameters to the command? AddWithValue will work just fine, but SQL Server will have to determine the type, a little, but unnecessary overhead.

Well, you could always try to inject a SQL statement into the textbox, that will probably give you a quicker, definite answer.

Yes, that's reasonably safe. So long as you don't use "sanitized" variables from a prepared statement to generate dynamic sql later, you're usually ok. The fact that you're using a prepared statement will take care of dealing with escape characters and other simple methods of injection.
I wouldn't forgo any other validation though...

Related

Cant insert character values from textbox to database

Trying to Insert into database from textboxes, but it will only accept integers and not characters - what could the problem be?
string sCMD = string.Format("INSERT INTO [Table] ([Item], [Des1],[Des2], [Prodline], [ANR], [STime]) VALUES({0},{1},'0',{2},{3},{4})"
,txtText1.Text, txtText2.Text, txText3.Text, txtText4.Text, txtText5.Text);
The name "" is not permitted in this context. Valid expressions are constants, constant expressions, and (in some contexts) variables. Column names are not permitted.
Here's something to consider:
using (SqlConnection c = new SqlConnection(connString))
{
c.Open();
using (SqlCommand cmd = new SqlCommand("INSERT INTO [Table] ([Item], [Des1],[Des2], [Prodline], [ANR], [STime]) VALUES(#Item,#Des1,'0',#ProdLine,#ANR,#STime)", c))
{
cmd.Parameters.AddWithValue("#Item", txtText1.Text);
cmd.Parameters.AddWithValue("#Des1", txtText2.Text);
cmd.Parameters.AddWithValue("#ProdLine", txText3.Text);
cmd.Parameters.AddWithValue("#ANR", txtText4.Text);
cmd.Parameters.AddWithValue("#STime", txtText5.Text);
cmd.ExecuteNonQuery();
}
}
If you were to send the SQL statement you're building to the server it might look something like this:
INSERT INTO [Table] ([Item], [Des1],[Des2], [Prodline], [ANR], [STime])
VALUES(Item Name,Here is my description.,'0',Prod line,ANR value,some time)
Clearly that's going to fail. It's not even surrounding the values with single quotes. Now, all of them may not be character, I get that, but you get my point.
But to add insult to injury it's wide open to SQL Injection the way you have it written. With the parameterized approach it's not.
This is a very dangerous line of code.
For starters, you are creating a string with no delimiters whatsoever for the values, which means that the only values that could pass withour issue are integers. Anything else needs to be delimited, which you don't do in this String.Format statement.
Worse, you are trying to create a sql statement from direct user input. Any weird value can cause your code to fail, or worse, cause execution of unwanted code. Imagine what would happen in a user entered eg 'sdf','sdfs',sdf' as the FIRST value. The resulting string would have 3 correct values for the first three columns that came from the first textboxt.
Now image what would happen if the user entered something like 'blablabla';DROP TABLE sometable;--. This would cause the Delete command to execute. This is a standard SQL injection attack scenario.
There are many more problems, eg when you try to pass numbers with floats, dates or any type whose conversion to string depends on your locale.
Instead of trying to construct a SQL query by string concatenation, you should use parameterized queries, as described in Give me parameterized SQL or give me death. The resulting code is easier to write, performs much faster, has no conversion errors and is not subject to SQL injection attacks,

How should I handle the single quote in a form to help protect against SQL injection, using WebMatrix razor (C#)

Currently I simply don't allow apostrophe's at all (along with other character's as you can see) with this, reiterated for each field:
foreach(char c in Comments)
{
if(c=='\'' || c=='$' || c=='\"' || c=='&' || c=='%' || c=='#' || c=='-' || c=='<' || c=='>')
{
errorMessage = "You have entered at least one invalid character in the \"Comments\" field. Invalid characters are: [\'], [\"], [&], [$], [#], [-], [<], [>], and [%]";
}
}
I've been coding this for a while, and I am getting better, but the real problem is that, while I am sure there is a way to effectively "strip-out" or otherwise validate the user input, I am not sure which approach is best and probably wouldn't until a security crisis was imminent.
If I have to, I will settle on simply never allowing single quotes into the form at all (as it is now), however this may aggravate anyone named say... Bill O'Reilly for the name field, etc., etc.
There are probably other (well I don't know what to call them, 'plug-ins?' 'outside programs?') that would do the job for me, but that is virtually useless since not only do I have no idea how to integrate that, I would have no idea how to tailor it specifically to my web interface/database.
Is there any code that could help me detect a sql injection apostrophe (by the characters surrounding it maybe?) from a normal apostrophe? Keep in mind some of my fields can hold up to 500 characters (textareas) and one up to 1,000.
Thanks for any ideas or help!
No amount of input encoding/cleanup will be as safe as parametrized queries.
See SqlCommand.Parameters for details on parametrized queries.
string commandText = "SELECT * FROM Sales WHERE CustomerID = #ID;";
using (SqlConnection connection = new SqlConnection(connectionString))
{
SqlCommand command = new SqlCommand(commandText, connection);
command.Parameters.Add("#ID", SqlDbType.Int);
command.Parameters["#ID"].Value = customerID;
var reader = command.ExecuteReader();
//.....
}
SQL Injections is not a problem with the input containing specific characters, it's a problem with how you handle the input.
By disallowing certain characters you can stop the obvious ways to cause SQL injections, but it's virtually impossible to use that to stop all possible ways.
If encoded correctly, there are no character that causes problems. The best way of doing that for database calls is to use parameterised queries, so that the database driver takes care of encoding the correct characters according to the data type and the specific database.
Also, you need to encode the values correctly when you use them later on, like HTML encoding strings that are put in HTML code, URL encoding strings that are used in an URL (and both for strings that are put in an URL in the HTML code.)
You should use parameterised queries to prevent SQL Injection as other people have already said.
Alexei Levenkov provides a good example of using ADO.NET parameters, but more commonly, you will use the Database Helper when working with WebMatrix Razor pages (ASP.NET Web Pages Framework) where parameter handling is slightly different. The Database.Query method (Query(string commandText, params object[] parameters) takes a string representing the SQL to be executed, and an array of objects, representing the parameter values to be passed to the SQL. The Database helper expects parameter markers to start at #0, and increment by 1 each time e.g.
var sql = "SELECT * From MyTable WHERE TheDate > #0 AND ID > #1";
Then you pass actual values in the following manner:
var data = Database.Open("MyDb").Query(sql, Request["date"], Request["id"]);
Internally, the Database class takes care of matching values to placeholders and creating ADO.NET parameters for you.

Why do we need SqlCeCommand.Parameters.AddWithValue() to Insert a value?

I have a C# WPF desktop application which uses SQL Compact 3.5 as its embedded database.
In the insertion function it has
using (SqlCeCommand com = new SqlCeCommand(
"INSERT INTO FooTable VALUES(#num)", con))
{
com.Parameters.AddWithValue("#num", num);
com.ExecuteNonQuery();
}
I don't get what the com.Parameters.AddWithValue() is about. I commented out this line of code and the insertion function run exactly the same. I thought ExecuteNonQuery carries out the insertion, so what is this Parameters.AddWithValue thing?
#num is a TSQL parameter. Without AddWithValue(#num, num) this is neither defined nor assigned a value. It simply will not work with the parameter omitted, and even if it did: where would it get your chosen value (num) from? The absolute best it could do would be to use null which was not your intent; more typically it would simply fail to execute (are you sure you aren't swallowing an exception somewhere?).
Note that concatenating the value into the string itself is not recommended; it would cause a SQL injection risk, and can reduce performance (plan re-use; not sure this applies to CE though - CE might very well not bother with cached plans).

Replace ' with \' in all textboxes in my program

So throughout my program I have probably 50 or so text boxes in various places. I know how to replace a single quote with something (in this case ' -> \'), but I am thinking there must be a better way to do this than go through and add specific code for every single text box. I have to do this because when this stuff is getting sent to the database, if there is a single quote, it gives an error. Is there a way to change the default TextBox control behavior so that all textboxes in the program automatically replace all single quotes with \'?
EDIT:
string statement = "select count(*) from users where username='#username'";
MySqlCommand command = new MySqlCommand(statement, conn);
command.Parameters.AddWithValue("#username", username);
if (Convert.ToInt32(command.ExecuteScalar()) == 1)
I have been playing with the paramerterized code and this is what I have right now. From how I understand it, the statement string is basically the same as before, but where I used to have the variable "username" I know use a parameter (which I called #username), then the command is created. Then in the parameters.addwithvalue, it replaces the parameter username, with whatever is in the variable username. Unfortunately, this is not working for me, and I don't really see how it helps because username is still just getting stuck in the command?
EDIT: Found the problem, in the statement you don't need to put single quotes around '#username'
so it should be:
string statement = "select count(*) from users where username=#username";
Don't use concatenation to build SQL queries. Use proper parametrized queries. This will make repeated queries a bit faster and will also eliminate input sanitizing code (replacing ' with \' for example) and SQL injection attacks.
You should be using parameterized queries, not only to resolve the problem you have, but also to reduce your exposure to SQL injection. When you use string concatenation to build SQL queries you are suseptable to SQL injection attackes.
U can use onKeyUp javascript function or asp.net OnTextChanged event to create function that will change quotes.

Avoiding SQL injection without parameters

We are having another discussion here at work about using parametrized sql queries in our code. We have two sides in the discussion: Me and some others that say we should always use parameters to safeguard against sql injections and the other guys that don't think it is necessary. Instead they want to replace single apostrophes with two apostrophes in all strings to avoid sql injections. Our databases are all running Sql Server 2005 or 2008 and our code base is running on .NET framework 2.0.
Let me give you a simple example in C#:
I want us to use this:
string sql = "SELECT * FROM Users WHERE Name=#name";
SqlCommand getUser = new SqlCommand(sql, connection);
getUser.Parameters.AddWithValue("#name", userName);
//... blabla - do something here, this is safe
While the other guys want to do this:
string sql = "SELECT * FROM Users WHERE Name=" + SafeDBString(name);
SqlCommand getUser = new SqlCommand(sql, connection);
//... blabla - are we safe now?
Where the SafeDBString function is defined as follows:
string SafeDBString(string inputValue)
{
return "'" + inputValue.Replace("'", "''") + "'";
}
Now, as long as we use SafeDBString on all string values in our queries we should be safe. Right?
There are two reasons to use the SafeDBString function. First, it is the way it has been done since the stone ages, and second, it is easier to debug the sql statements since you see the excact query that is run on the database.
So then. My question is whether it really is enough to use the SafeDBString function to avoid sql injection attacks. I have been trying to find examples of code that breaks this safety measure, but I can't find any examples of it.
Is there anybody out there that can break this? How would you do it?
EDIT:
To summarize the replies so far:
Nobody has found a way to get around the SafeDBString on Sql Server 2005 or 2008 yet. That is good, I think?
Several replies pointed out that you get a performance gain when using parametrized queries. The reason is that the query plans can be reused.
We also agree that using parametrized queries give more readable code that is easier to maintain
Further it is easier to always use parameters than to use various versions of SafeDBString, string to number conversions and string to date conversions.
Using parameters you get automatic type conversion, something that is especially useful when we are working with dates or decimal numbers.
And finally: Don't try to do security yourself as JulianR wrote. The database vendors spend lots of time and money on security. There is no way we can do better and no reason we should try to do their job.
So while nobody was able to break the simple security of the SafeDBString function I got lots of other good arguments. Thanks!
I think the correct answer is:
Don't try to do security yourself. Use whatever trusted, industry standard library there is available for what you're trying to do, rather than trying to do it yourself. Whatever assumptions you make about security, might be incorrect. As secure as your own approach may look (and it looks shaky at best), there's a risk you're overlooking something and do you really want to take that chance when it comes to security?
Use parameters.
And then somebody goes and uses " instead of '. Parameters are, IMO, the only safe way to go.
It also avoids a lot of i18n issues with dates/numbers; what date is 01/02/03? How much is 123,456? Do your servers (app-server and db-server) agree with each-other?
If the risk factor isn't convincing to them, how about performance? The RDBMS can re-use the query plan if you use parameters, helping performance. It can't do this with just the string.
The argument is a no-win. If you do manage to find a vulnerability, your co-workers will just change the SafeDBString function to account for it and then ask you to prove that it's unsafe all over again.
Given that parametrized queries are an undisputed programming best practice, the burden of proof should be on them to state why they aren't using a method that is both safer and better performing.
If the issue is rewriting all the legacy code, the easy compromise would be to use parametrized queries in all new code, and refactor old code to use them when working on that code.
My guess is the actual issue is pride and stubbornness, and there's not much more you can do about that.
First of all, your sample for the "Replace" version is wrong. You need to put apostrophes around the text:
string sql = "SELECT * FROM Users WHERE Name='" + SafeDBString(name) & "'";
SqlCommand getUser = new SqlCommand(sql, connection);
So that's one other thing parameters do for you: you don't need to worry about whether or not a value needs to be enclosed in quotes. Of course, you could build that into the function, but then you need to add a lot of complexity to the function: how to know the difference between 'NULL' as null and 'NULL' as just a string, or between a number and a string that just happens to contain a lot of digits. It's just another source for bugs.
Another thing is performance: parameterized query plans are often cached better than concatenated plans, thus perhaps saving the server a step when running the query.
Additionally, escaping single quotes isn't good enough. Many DB products allow alternate methods for escaping characters that an attacker could take advantage of. In MySQL, for example, you can also escape a single quote with a backslash. And so the following "name" value would blow up MySQL with just the SafeDBString() function, because when you double the single quote the first one is still escaped by the backslash, leaving the 2nd one "active":
x\' OR 1=1;--
Also, JulianR brings up a good point below: NEVER try to do security work yourself. It's so easy to get security programming wrong in subtle ways that appear to work, even with thorough testing. Then time passes and a year later your find out your system was cracked six months ago and you never even knew it until just then.
Always rely as much as possible on the security libraries provided for your platform. They will be written by people who do security code for a living, much better tested than what you can manage, and serviced by the vendor if a vulnerability is found.
So I'd say:
1) Why are you trying to re-implement something that's built in? it's there, readily available, easy to use and already debugged on a global scale. If future bugs are found in it, they'll be fixed and available to everyone very quickly without you having to do anything.
2) What processes are in place to guarantee that you never miss a call to SafeDBString? Missing it in just 1 place could open up a whole host of issues. How much are you going to eyeball these things, and consider how much wasted that effort is when the accepted correct answer is so easy to reach.
3) How certain are you that you've covered off every attack vector that Microsoft(the author of the DB and the access library) knows about in your SafeDBString implementation ...
4) How easy is it to read the structure of the sql? The example uses + concatenation, parameters are very like string.Format, which is more readable.
Also, there are 2 ways of working out what was actually run - roll your own LogCommand function, a simple function with no security concerns, or even look at an sql trace to work out what the database thinks is really going on.
Our LogCommand function is simply:
string LogCommand(SqlCommand cmd)
{
StringBuilder sb = new StringBuilder();
sb.AppendLine(cmd.CommandText);
foreach (SqlParameter param in cmd.Parameters)
{
sb.Append(param.ToString());
sb.Append(" = \"");
sb.Append(param.Value.ToString());
sb.AppendLine("\"");
}
return sb.ToString();
}
Right or wrong, it gives us the information we need without security issues.
With parameterised queries you get more than protection against sql injection. You also get better execution plan caching potential. If you use the sql server query profiler you can still see the 'exact sql that is run on the database' so you're not really losing anything in terms of debugging your sql statements either.
I have used both approaches to avoid SQL injection attacks and definitely prefer parametrized queries. When I have used concatenated queries I have used a library function to escape the variables (like mysql_real_escape_string) and wouldn't be confident I have covered everything in a proprietary implementation (as it seems you are too).
You aren't able to easily do any type checking of the user input without using parameters.
If you use the SQLCommand and SQLParameter classes to make you're DB calls, you can still see the SQL query that's being executed. Look at the SQLCommand's CommandText property.
I'm always a litle suspect of the roll-your-own approach to preventing SQL injection when parameterized queries are so easy to use. Second, just because "it's always been done that way" doesn't mean it's the right way to do it.
This is only safe if you're guaranteed that you're going to pass in a string.
What if you're not passing in a string at some point? What if you pass just a number?
http://www.mywebsite.com/profile/?id=7;DROP DATABASE DB
Would ultimately become:
SELECT * FROM DB WHERE Id = 7;DROP DATABASE DB
I'd use stored procedures or functions for everything, so the question wouldn't arise.
Where I have to put SQL into code, I use parameters, which is the only thing that makes sense. Remind the dissenters that there are hackers smarter than they are, and with better incentive to break the code that's trying to outsmart them. Using parameters, it's simply not possible, and it's not like it's difficult.
Agree hugely on the security issues.
Another reason to use parameters is for efficiency.
Databases will always compile your query and cache it, then re-use the cached query (which is obviously faster for subsequent requests).
If you use parameters then even if you use different parameters the database will re-use your cached query as it matches based on the SQL string before binding the parameters.
If however you don't bind parameters then the SQL string changes on every request (that has different parameters) and it will never match what's in your cache.
For the reasons already given, parameters are a very good idea. But we hate using them because creating the param and assigning its name to a variable for later use in a query is a triple indirection head wreck.
The following class wraps the stringbuilder that you will commonly use for building SQL requests. It lets you write paramaterized queries without ever having to create a parameter, so you can concentrate on the SQL. Your code will look like this...
var bldr = new SqlBuilder( myCommand );
bldr.Append("SELECT * FROM CUSTOMERS WHERE ID = ").Value(myId, SqlDbType.Int);
//or
bldr.Append("SELECT * FROM CUSTOMERS WHERE NAME LIKE ").FuzzyValue(myName, SqlDbType.NVarChar);
myCommand.CommandText = bldr.ToString();
Code readability, I hope you agree, is greatly improved, and the output is a proper parameterized query.
The class looks like this...
using System;
using System.Collections.Generic;
using System.Text;
using System.Data;
using System.Data.SqlClient;
namespace myNamespace
{
/// <summary>
/// Pour le confort et le bonheur, cette classe remplace StringBuilder pour la construction
/// des requêtes SQL, avec l'avantage qu'elle gère la création des paramètres via la méthode
/// Value().
/// </summary>
public class SqlBuilder
{
private StringBuilder _rq;
private SqlCommand _cmd;
private int _seq;
public SqlBuilder(SqlCommand cmd)
{
_rq = new StringBuilder();
_cmd = cmd;
_seq = 0;
}
//Les autres surcharges de StringBuilder peuvent être implémenté ici de la même façon, au besoin.
public SqlBuilder Append(String str)
{
_rq.Append(str);
return this;
}
/// <summary>
/// Ajoute une valeur runtime à la requête, via un paramètre.
/// </summary>
/// <param name="value">La valeur à renseigner dans la requête</param>
/// <param name="type">Le DBType à utiliser pour la création du paramètre. Se référer au type de la colonne cible.</param>
public SqlBuilder Value(Object value, SqlDbType type)
{
//get param name
string paramName = "#SqlBuilderParam" + _seq++;
//append condition to query
_rq.Append(paramName);
_cmd.Parameters.Add(paramName, type).Value = value;
return this;
}
public SqlBuilder FuzzyValue(Object value, SqlDbType type)
{
//get param name
string paramName = "#SqlBuilderParam" + _seq++;
//append condition to query
_rq.Append("'%' + " + paramName + " + '%'");
_cmd.Parameters.Add(paramName, type).Value = value;
return this;
}
public override string ToString()
{
return _rq.ToString();
}
}
}
From the very short time I've had to investigate SQL injection problems, I can see that making a value 'safe' also means that you're shutting the door to situations where you might actually want apostrophes in your data - what about someone's name, eg O'Reilly.
That leaves parameters and stored procedures.
And yes, you should always try to implement code in the best way you know now - not just how its always been done.
Here are a couple of articles that you might find helpful in convincing your co-workers.
http://www.sommarskog.se/dynamic_sql.html
http://unixwiz.net/techtips/sql-injection.html
Personally I prefer to never allow any dynamic code to touch my database, requiring all contact to be through sps (and not one which use dynamic SQl). This means nothing excpt what I have given users permission to do can be done and that internal users (except the very few with production access for admin purposes) cannot directly access my tables and create havoc, steal data or commit fraud. If you run a financial application, this is the safest way to go.
It can be broken, however the means depends on exact versions/patches etc.
One that has already been brought up is the overflow/truncation bug that can be exploited.
Another future means would be finding bugs similar to other databases - for example the MySQL/PHP stack suffered an escaping problem because certain UTF8 sequences could be used to manipulate the replace function - the replace function would be tricked into introducing the injection characters.
At the end of the day, the replacement security mechanism relies on expected but not intended functionality. Since the functionality was not the intended purpose of the code, there is a high probablity that some discovered quirk will break your expected functionality.
If you have a lot of legacy code, the replace method could be used as a stopgap to avoid lengthy rewriting and testing. If you are writing new code, there is no excuse.
Always use parameterized queries where possible. Sometimes even a simple input without the use of any weird characters can already create an SQL-injection if its not identified as a input for a field in the database.
So just let the database do its work of identifying the input itself, not to mention it also saves allot of trouble when you need to actually insert weird characters that otherwise would be escaped or changed. It can even save some valuable runtime in the end for not having to calculate the input.
I did not see any other answsers address this side of the 'why doing it yourself is bad', but consider a SQL Truncation attack.
There is also the QUOTENAME T-SQL function that can be helpful if you can't convince them to use params. It catches a lot (all?) of the escaped qoute concerns.
2 years later, I recidivated... Anyone who finds parameters a pain is welcome to try my VS Extension, QueryFirst. You edit your request in a real .sql file (Validation, Intellisense). To add a parameter, you just type it directly into your SQL, starting with the '#'. When you save the file, QueryFirst will generate wrapper classes to let you run the query and access the results. It will look up the DB type of your parameter and map it to a .net type, which you will find as an input to the generated Execute() methods. Could not be simpler. Doing it the right way is radically quicker and easier than doing it any other way, and creating a sql injection vulnerability becomes impossible, or at least perversely difficult. There are other killer advantages, like being able to delete columns in your DB and immediately see compile errors in your application.
legal disclaimer : I wrote QueryFirst
Here are a few reasons to use parameterized queries:
Security - The database access layer knows how to remove or escape items that are not allowed in data.
Separation of concerns - My code is not responsible for transforming the data into a format that the database likes.
No redundancy - I don't need to include an assembly or class in every project that does this database formatting/escaping; it's built in to the class library.
There were few vulnerability(I can't remember which database it was) that is related to buffer overflow of the SQL statement.
What I want to say is, SQL-Injection is more then just "escape the quote", and you have no idea what will come next.
Another important consideration is keeping track of escaped and unescaped data. There are tons and tons of applications, Web and otherwise, that don't seem to properly keep track of when data is raw-Unicode, &-encoded, formatted HTML, et cetera. It's obvious that it will become difficult to keep track of which strings are ''–encoded and which aren't.
It's also a problem when you end up changing the type of some variable — perhaps it used to be an integer, but now it's a string. Now you have a problem.

Categories

Resources