Lua Debugger that can Attach to Process - lua

My company has a program that uses Lua embedded in its runtime, loading up .lua files from disk and executing functions defined in them repeatedly.
Is there a way to attach to the running process and set breakpoints in my .lua files? (I'd accept either gdb-style command-line debugging as part of the Lua distribution, or perhaps a third-party IDE that provides Visual-Studio-like GUI breakpoints.)
Or is what I'm asking for entirely nonsensical and impossible given the nature of the runtime loading up random files from disk?
Edit: Looks like it's not nonsensical, given that Lua's own debug.getinfo() function can determine the source file for a given function, and debug.sethook() allows a callback for each new line of code entered. So, it's reasonable to load source code from disk and be able to tell when the interpreter is executing a particular line of code from that file. The question remains: how do I latch onto an existing process that has a Lua interpreter and inject my own trace function (which can then watch for file/line number pairs and pause execution)?

If you can modify the .lua files, you can insert the following call just before anything you need to debug:
require 'remdebug.engine'.start()
It starts the RemDebug Lua debugger engine and tries to connect to a controller. If it cannot connect, it will just continue running as normal. I did some fixes to the debugger engine, such as dealing with temporary variables, and my student is working on a debugger GUI (due next year).
In the meantime, you can try if Lua Development Tools works for you. It features a debugger similar to RemDebug, which should be possible to set up as follows:
require("debugger")(host, port, idekey)
Alternatively, you can use SciTE-debug, which is an extension to the SciTE editor, and can serve as a controller to RemDebug. Just make sure you insert the call to remdebug.engine.start somewhere in your Lua code and insert this into the SciTE output window:
:debug.target=remote.lua
When you start your program, SciTE should show the source and current line.

I've been using Decoda editor for that. It allows you to attach to a running C++ application, after that it detects that you're running a Lua Interpreter within your C++ code and show your Lua source code, where you can add beakpoints and inspect variables as usual.

This is an alternative I use after much searching. If you have an external executable that loads lua, I got this working in a few minutes. The op is very responsive, it has an interactive debugger which loads your code you can place debug points interactively. It doesn't have an editor, but I use scite or crimson editor and start the executable, one line in your main lua module enables the debugger.
http://www.cushy-code.com/grld/ - this link seems dead now
I've moved to eclipse.org/ldt it has an ide and integrated debugger, recommended
hth

The Lua plugin for IntelliJ has a working debugger with no special setup required other than pointing to your Lua interpreter.
Here's a screencast of it:
http://www.screencast.com/t/CBWIkoZPg

Similar to what Michal Kottman described, I have implemented a debugger based on RemDebug, but with additional fixes and features (on github: https://github.com/pkulchenko/MobDebug).
You can update your .lua file with require("mobdebug").start("localhost", 8171) at the point where you want the debugging to start. You can then use the command line debugger to execute commands, set breakpoints, evaluate/execute expressions and so on.
As an alternative, you can use ZeroBrane Studio IDE, which integrates with the debugger and gives you a front-end to load your code and execute same debugger commands in a nice GUI. If you want to see the IDE in action, I have a simple demo here: http://notebook.kulchenko.com/zerobrane/live-coding-in-lua-bret-victor-style.

You should probably use Decoda.
Go to Debug -> Processes -> Attach to attach your process. This should work fine.

Well the easiest way is this, thanks to the genius author
https://github.com/slembcke/debugger.lua
you don't need to setup a remote debug server ,just require one file,and simplely call dbg() and it will pause,just like gdb
an tutorial is also shipped with it, check it out.

Related

Can't debug using Informix 4GL Interactive Debugger

I am pretty new to Informix and I have a program that I am adding some functionality to.
It seems that the program has some existing issues with it though.
When I run make -f makefile.mk I get success and the .4ge gets generated and I am able to run it.
However I am trying to get the program to run within informix 4gl interactive debugger but I get the error: Invalid module name [main] specified.
Any assistance would be greatly appreciated. Unfortunately I am unable to share code as the program contains confidential information
The Informix-4GL Interactive Debugger (ID) is for debugging programs compiled with the Informix-4GL Rapid Development System (RDS). The object files created by RDS (fglpc) have the extension .4go (I4GL p-code object file) and the executables are conventionally given the extension .4gi (I4GL p-code interpretable file — run using fglgo or ID's fgldb).
By contrast, the plain Informix-4GL (c-code) system uses an I4GL compiler to generate first ESQL/C code and then C code, and a C compiler to create regular object files (.o) and to create its executables, which are conventionally given the extension .4ge (I4GL c-code executable).
The ID cannot debug c-code executables. It can only debug p-code interpretable files.
On the face of it, therefore, your problem is that you are using the wrong tool for the job. Either you need to compile with RDS and create an interpretable, or you need to use a C code debugger such as GDB. However, be warned that debugging I4GL code with GDB is mainly an exercise in frustration as the bulk of the code is a series of function calls to library functions — or is an incredibly tortuous sequence of goto statements if you're debugging inside an I4GL report function. It is machine-generated C code; it is not intended to be comprehensible to humans.

How SciTE IDE for lua automatically or manually format my code

I am new in lua,need basic type of help.After install BabeLua extension on Visual Studio,they indent and everything for me. Want to work with SciTE IDE, It's a pain in having to indent my code all the time so looking a way which can format my lua code manually and automatically.
Note: try to use source-code-formatter and beautifier.I failed to use them in my lua module.How to use them in lua?They are workable or not?
I think the idea with those beautifiers you reference is that you can run them as an external program: save your current buffer to a file, run through the filter to beautify, then read the result back into the current buffer in SciTE.
I can vouch for the second program as I used it in the past to re-indent some of my code; I ended up re-implementing it in Perl as it didn't handle all the cases I was interested in.
If you want to integrate it into SciTE, you can probably strip some of the io functions and just use functions to read buffer content in SciTE (like GetLine) and then modify the indentation (probably using GetLineIndentation and SetLineIndentation). I've implemented a very similar logic in my Lua IDE, which is using the same editor component that is used in SciTE.

Dart: Possible to exchange source code on the fly in a running system?

In this article it says: "The Dart VM reads and executes source code, which means there is no compile step between edit and run.". Does that mean that you can exchange source-code on the fly in a running Dart system like in Erlang? Maybe the compiler is removed from the runtime system and then this is no longer possible. So that's why I'm asking.
Dart is run "natively" only in Dartium, which is a flavour of Chrome with DartVM. When you develop an application you still need to compile it it to JavaScript. This way you get fast development lifecycle and in the end you can compile code to JS. Because it's compiled code there is lots more room for compiler to run optimisations on the code. So from my perspective, the compiler is still there and I don't think you would be able to replace code at runtime.
You can send around source code and run it, but it would need to be in a separate isolate. Isolates do have some relationship to Erlang concepts.
The Dart VM doesn't support hot swapping (Called live edit in V8). However, based on mailing list discussions, it sounds like this is something that the authors do want to support in the future.
However, as the others have mentioned, it is possible to dynamically load code into another isolate.

not enough space for environment appears when executing ".exe" file

I am trying to use an application called CLUT.exe which is an old application for MS-DOS that can be used to reindex NTX files for DBF databases.
(This is not the main topic, but I am just writing this if someone wants to test the app and don't trust at all about the content).
The problem starts when trying to run the command line version through console (cmd.exe) and this error appears:
C:\>CLUT.exe [arg1] [arg2] [arg3]
run-time error R6009
- not enough space for environment
So, according to what I've searched, this could be a possible solution:
http://support.microsoft.com/default.aspx?scid=kb;en-us;230205
but it doesn't work and every alternative that I found to solve this over the internet is the same.
Another alternative could be to make right-click in the .exe file, go to Properties then Memory tab and increase the Initial environment memory from Auto to the max value but it doesn't work too.
Well, I am stuck and no "possible" solution is working for me. If someone is interested, knows more about this issue and want to test, you can download the application from here (click "Free Download" green button):
http://www.filebasket.com/free/Development-Clipper-programming-language/clut-exe/13996.html
or directly from my DropBox:
https://dl.dropbox.com/u/15208254/stackoverflow/clut_214.rar
Just to know, I am using Windows 7 and the CLUT.exe application is a Clipper based app (old programming language) that may run under windows console (cmd.exe).
Wikipedia does mention other dos emulators but, oddly, doesn't mention BOCHS.
Reindexing NTX files is not a difficult thing to do, and can be done with tools other than CLUT. For example, many of the utilities listed on this part of Download32 could be used. Otherwise, you could write your own using Harbour Project or xHarbour. Or contact me off list and I'll cook up something in Clipper 5.3.
LATER
If I read the README correctly for CLUT, it's a replacement for the DBU utility that comes with Clipper 5.x. I can supply you with a build of that if you're unsuccessful with other approaches.

How to generate the Symbols file of a PowerBuilder program for dump analysis?

How to generate the Symbols file of a PowerBuilder program for dump analysis?
Are you running into a problem with a compiled PowerBuilder application?
I've never used the Microsoft debugger, however, I have used Dependency Walker and ran my app from within that environment. Usually that helps us if we have missing deployables, etc.
If you need a stack dump, have you thought about running the app from the Run window and adding the /debug option after the name of the app? That creates a call stack log that shows all your commands being called, etc. Probably would get you what you need.
Please advise.
You can install Dr. Watson (drwtsn32) to automatically create a crash dump when/if your application dies.
That said, the dump will usually only be useful if you are calling into native DLL's. Otherwise the stack will just include various PBVM calls that will be difficult to correlate back to actual source code lines.

Resources