Input to program without command-line arguments

You can use tools to display all the hardcoded strings inside the application, and hopefully the filename will show up.

Check OllyDBG on Windows and strings on Linux (it's a cmd-line application).

A more complex solution: fire up your favorite asm debugger (OllyDBG, GDB) and get your hands dirty.

In the .NET world, this is achieved with the use of an app.config file included with your source code. This is copied to the output folder with a name following the convention "MyApp.exe.config" where your application is "MyApp.exe".

I don't know if there are similar conventions for other languages.

Popular convention seems to be for programs to use standard streams for stuff like that:

In computer programming, standard streams are preconnected input and output channels between a computer program and its environment (typically a text terminal) when it begins execution...

You can see how it works using whatever command interpreter is available in your OS. In Windows, you may launch DOS shell and type help. In Unix you launch shell and type man.

You could look at COM, as implemented by Microsoft. In that system, the OS loads your DLL into memory when someone needs an object of your type. If anyone needs to pass information to that object, it can use the standard IUnknown::QueryInterface(GUID) method to obtain an well-known interface, and then pass information via that interface.

Of course, if your object doesn't have such an interface, it'd return E_NO_INTERFACE. But that's no different (just eassier to debug) than a CLI program ignoring its arguments.

Many Operating Systems already have the concept of "environment variables". This is a set of named values that are passed to every process, in addition to the command line arguments. Using C, you'd access them with getenv().

Similar to the /? convention for command lines, the OS writer can define a convention that the presence of a HELP variable in the environment means that a program should show usage information. In C code: if (getenv("HELP")) { ShowHelp(); }

I don't quite understand what the OP is really asking, but it is fairly common for command line programs to use environmental variables as inputs in lieu of command line parameters.

Of course that begs the question of where did the environmental variables come from.

If you have to instruct an application about its job without parameters I think you can call that...configuration in the more old meaning of the term. Depending on the environment (Windows or Linux?) and the language/framework (Java? Any .NET language?) your choice will be really different. You speak about

an environment with no command-line arguments

so I guess you do not have a console application. Moreover I guess (again) that your application when launched will do its job and then terminate. So your question is "how users may change application options?".

You can use a configuration file, what - in this case a more advanced - user will expect depends on the environment. I don't think it's a good solution for end-users or non technical personnel. They have to know or to guess that in the application directory (or in the user home directory or something like that) there is a configuration file of some sort.

You can provide some kind of help (man page, plain text file, HTML help or HTML file) to teach your user what to search and where. I think you should always do this but sometimes technical details aren't clear for every user and you give them a chance to break your application in many many ways because a wrong configuration (they may save your configuration file with the wrong encoding, break its syntax, write something stupid).

Finally the solution I prefer: you may provide two applications, one to do the job and another one for configuration purposes. Configuration is still saved somewhere (so advanced users may enjoy more advanced esoteric options) and normal users will have a user-friendly interface (like xconfig, for example).

The format used to store these data is up to you. Usually I think to be coherent with the environment is the most important thing but how your application will be used is also important: you may use the Registry (in Windows), XML configuration files, JSON configuration files, INI-like configuration files. It depends on the kind of data you need to store. Sometimes a simple text file for command line arguments used as response file is enough.

UPDATED

If your application may have multiple configurations and multiple instances you need to make the configuration file runnable. How is environment dependant. In Windows, for example, you may register a ".MyApplication.config" file extension. Users won't run (with double click...) the application itself but the configuration file, application will be run by OS with the configuration as parameter.
Configuration will be a document and user will see that document, not the application (moreover you may provide extra verbs for editing).

You can create .bat file with your arguments in it.

I'm going for a straight-man answer.

You read the user's manual. Then you use some sort of input device like a keyboard or mouse to control the program. If the program requires a large amount of data, you can direct it to a file. Directing a program to a file has varied widely through the ages, but has standardized to OS typical popup windows.

If it's a command line program, with no command line arguments, then you are dependent on the program's interface. At some point you probably have to type in the path to the file name.