Node add-on for memory reading and writing! (finally!)
NOTE: version 3 of this library introduces breaking changes that are incompatible with previous versions. The notable change is that when reading memory, writing memory and pattern scanning you are required to pass the handle through for the process (that is given when doing memoryjs.openProcess). This allows for multi-process support.
- open a process
- close the process (handle)
- list all open processes
- list all modules associated with a process
- find a certain module associated with a process
- read memory
- write to memory
- read/write buffers
- change memory protection
- pattern scanning
TODO:
- nothing (suggestions welcome)
This is a Node add-on (last tested to be working on v6.9.1
) and therefore requires node-gyp to use.
You may also need to follow these steps.
npm install memoryjs
When using memoryjs, the target process should match the platform architecture of the Node version running. For example if you want to target a 64 bit process, you should try and use a 64 bit version of Node.
You also need to recompile the library and target the platform you want. Head to the memoryjs node module directory, open up a terminal and to run the compile scripts, type:
npm run build32
if you want to target 32 bit processes
npm run build64
if you want to target 64 bit processes
If you are planning to use this module with Node Webkit or Electron, take a look at Liam Mitchell's build notes here.
Initialise:
const memoryjs = require('memoryjs');
const processName = "csgo.exe";
Open a process (sync):
const processObject = memoryjs.openProcess(processIdentifier);
Open a process (async):
memoryjs.openProcess(processIdentifier, (error, processObject) => {
});
Get all processes (sync):
const processes = memoryjs.getProcesses();
Get all processes (async):
memoryjs.getProcesses((error, processes) => {
});
See the Documentation section of this README to see what a process object looks like.
Find a module (sync):
const module = memoryjs.findModule(moduleName, processId);
Find a module (async):
memoryjs.findModule(moduleName, processId, (error, module) => {
});
Get all modules (sync):
const modules = memoryjs.getModules(processId);
Get all modules (async):
memoryjs.getModules(processId, (error, modules) => {
});
See the Documentation section of this README to see what a module object looks like.
Read from memory (sync):
const value = memoryjs.readMemory(handle, address, dataType);
Read from memory (async):
memoryjs.readMemory(handle, address, dataType, (error, value) => {
});
Read buffer from memory (sync):
const buffer = memoryjs.readBuffer(handle, address, size);
Read buffer from memory (async):
memoryjs.readBuffer(handle, address, size, (error, buffer) => {
});
Write to memory:
memoryjs.writeMemory(handle, address, value, dataType);
Write buffer to memory:
memoryjs.writeBuffer(handle, address, buffer);
See the Documentation section of this README to see what values dataType
can be.
Set protection of memory:
var oldProtection = memoryjs.setProtection(handle, address, size, protection);
See the Documentation section of this README to see what values protection
can be.
Pattern scanning (sync):
const offset = memoryjs.findPattern(handle, moduleName, signature, signatureType, patternOffset, addressOffset);
Pattern scanning (async):
memoryjs.findPattern(handle, moduleName, signature, signatureType, patternOffset, addressOffset, (error, offset) => {
})
{ dwSize: 304,
th32ProcessID: 10316,
cntThreads: 47,
th32ParentProcessID: 7804,
pcPriClassBase: 8,
szExeFile: "csgo.exe",
modBaseAddr: 1673789440,
handle: 808 }
The handle
and modBaseAddr
properties are only available when opening a process and not when listing processes.
{ modBaseAddr: 468123648,
modBaseSize: 80302080,
szExePath: 'c:\\program files (x86)\\steam\\steamapps\\common\\counter-strike global offensive\\csgo\\bin\\client.dll',
szModule: 'client.dll',
th32ProcessID: 10316 }
When using the write or read functions, the data type (dataType) parameter can either be a string and be one of the following:
"int", "dword", "short", "long", "float", "double", "bool", "boolean", "ptr", "pointer", "str", "string", "vec3", "vector3", "vec4", "vector4"
or can reference constants from within the library:
memoryjs.INT, memoryjs.DWORD, memoryjs.SHORT, memoryjs.LONG, memoryjs.FLOAT, memoryjs.DOUBLE, memoryjs.BOOL, memoryjs.BOOLEAN, memoryjs.PTR, memoryjs.POINTER, memoryjs.STR, memoryjs.STRING, memoryjs.VEC3, memoryjs.VECTOR3, memoryjs.VEC4, memoryjs.VECTOR4
This is simply used to denote the type of data being read or written.
Vector3 is a data structure of three floats:
const vector3 = { x: 0.0, y: 0.0, z: 0.0 };
memoryjs.writeMemory(address, vector3);
Vector4 is a data structure of four floats:
const vector4 = { w: 0.0, x: 0.0, y: 0.0, z: 0.0 };
memoryjs.writeMemory(address, vector4);
If you have a structure you want to write to memory, you can use buffers. For an example on how to do this, view the buffers example.
To write a structure to memory, you can use the concentrate library to describe the structure as a buffer
and then write the buffer to memory using the writeBuffer
function.
To read a structure from memory, you will need to read a buffer from memory using the readBuffer
function, and then you can use the dissolve library to parse the buffer into a structure.
In either case you don't need to use the two libraries mentioned above, they just make it easy to turn your structure into a buffer, and your buffer into a structure.
Protection type is a bit flag DWORD value.
This parameter should reference a constant from the library:
memoryjs.PAGE_NOACCESS, memoryjs.PAGE_READONLY, memoryjs.PAGE_READWRITE, memoryjs.PAGE_WRITECOPY, memoryjs.PAGE_EXECUTE, memoryjs.PAGE_EXECUTE_READ, memoryjs.PAGE_EXECUTE_READWRITE, memoryjs.PAGE_EXECUTE_WRITECOPY, memoryjs.PAGE_GUARD, memoryjs.PAGE_NOCACHE, memoryjs.PAGE_WRITECOMBINE, memoryjs.PAGE_ENCLAVE_THREAD_CONTROL, memoryjs.PAGE_TARGETS_NO_UPDATE, memoryjs.PAGE_TARGETS_INVALID, memoryjs.PAGE_ENCLAVE_UNVALIDATED
Refer to MSDN's Memory Protection Constants for more information.
You can use this library to read either a "string", or "char*" and to write a string.
In both cases you want to get the address of the char array:
std::string str1 = "hello";
std::cout << "Address: 0x" << hex << (DWORD) str1.c_str() << dec << std::endl;
char* str2 = "hello";
std::cout << "Address: 0x" << hex << (DWORD) str2 << dec << std::endl;
From here you can simply use this address to write and read memory.
There is one caveat when reading a string in memory however, due to the fact that the library does not know how long the string is, it will continue reading until it finds the first null-terminator. To prevent an infinite loop, it will stop reading if it has not found a null-terminator after 1 million characters.
One way to bypass this limitation in the future would be to allow a parameter to let users set the maximum character count.
When pattern scanning, flags need to be raised for the signature types. The signature type parameter needs to be one of the following:
0x0
or memoryjs.NORMAL
which denotes a normal signature.
0x1
or memoryjs.READ
which will read the memory at the address.
0x2
or memoryjs.SUBSTRACT
which will subtract the image base from the address.
To raise multiple flags, use the bitwise OR operator: memoryjs.READ | memoryjs.SUBTRACT
.
opens a process to be able to read from and write to it
- processIdentifier (string/int) - the identifier of the process to open, can either be a name ('csgo.exe') or an id (3270)
- callback (function) - has two parameters:
- err (string) - error message (empty if there were no errors)
- processObject (object) - information about the process
returns process object (object) either directly or via the callback
closes the handle on the opened process
- handle (int) - the handle of the process to close
collects information about all the running processes
- callback (function) - has two parameters:
- err (string) - error message (empty if there were no errors)
- processes (array) - array of process object (JSON)
returns an array of process object (JSON) for all the running processes
finds a module associated with a given process
- moduleName (string) - the name of the module to find
- processId (int) - the id of the process in which to find the module
- callback (function) - has two parameters:
- err (string) - error message (empty if there were no errors)
- module (object) - information about the module
returns module object (object) either directly or via the callback
gets all modules associated with a given process
- processId (int) - the id of the process in which to find the module
- callback (function) - has two parameters:
- err (string) - error message (empty if there were no errors)
- modules (array) - array of module object (JSON)
returns an array of module object (object) for all the modules found
reads the memory at a given address
- handle (int) - the handle of the process, given to you by the process object retrieved when opening the process
- address (int) - the address in memory to read from
- dataType (string) - the data type to read into (definitions can be found at the top of this section)
- callback (function) - has two parameters:
- err (string) - error message (empty if there were no errors)
- value (any data type) - the value stored at the given address in memory
returns the value that has been read from memory
reads size
bytes of memory at the given address
- handle (int) - the handle of the process, given to you by the process object retrieved when opening the process
- address (int) - the address in memory to read from
- size (int) - the number of bytes to read into the buffer
- callback (function) - has two parameters:
- err (string) - error message (empty if there were no errors)
- value (buffer) - the bytes read from memory at the given address in a buffer
returns the bytes read from memory at the given address in a buffer
writes to an address in memory
- handle (int) - the handle of the process, given to you by the process object retrieved when opening the process
- address (int) - the address in memory to write to
- value (any data type) - the data type of value must be either
number
,string
orboolean
and is the value that will be written to the address in memory - dataType (string) the data type of the value (definitions can be found at the top of this section)
- callback (function) - has one parameter:
- err (string) - error message (empty if there were no errors)
writes size
bytes of memory to the given address
- handle (int) - the handle of the process, given to you by the process object retrieved when opening the process
- address (int) - the address in memory to write to
- buffer (buffer) - the buffer to write to memory
pattern scans memory to find an offset
- handle (int) - the handle of the process, given to you by the process object retrieved when opening the process
- moduleName (string) - the name of the module to pattern scan (module.szModule)
- signature (string) - the actual signature mask (in the form
A9 ? ? ? A3 ?
) - signatureType (int) - flags for signature types (definitions can be found at the top of this section)
- patternOffset (int) - offset will be added to the address (before reading, if
memoryjs.READ
is raised) - addressOffset (int) - offset will be added to the address returned
- callback (function) - has two parameters:
- err (string) - error message (empty if there were no errors)
- offset (int) - value of the offset found (will return -1 if the module was not found, -2 if the pattern found no address)
returns the value of the offset found
sets the protection of the memory address
- handle (int) - the handle of the process, given to you by the process object retrieved when opening the process
- address (int) - the address in memory to write to
- size (int) -
number
of bytes at the address to change the protection of. - protection (int) the protection type to set this if a bit flag. See Documentation
returns old protection value.