I am extending an existing open-source project, which already has a very advanced Lexer/Parser/TreeParser in ANTLR. I would like to adhere to the current directory structure, so I created my own directory where I would like to put my new (already written) grammar files. They are importing these three original grammar files and to compile the code, I can use java org.antlr.Tool, where I can specify one directory to search for imported grammars with -lib option argument.
My problem is that these three original imported grammar files have some imports themselves and again in different directory. To illustrate (inheritance/importing going to top):
Abstract syntax directory: lexer parser tree parser
Original syntax directory: lexer parser tree parser
My new syntax directory: lexer parser tree parser
This is the hierarchy. The trouble I am encountering is how to specify "Abstract syntax directory" and also "Original syntax directory" at once with the -lib option on the command line for the ANTLR tool (or any other solution which would allow me to compile my grammar importing the original one with the given directories structure).
I tried -lib directory1 directory2, that just appears to ignore the directory2 (and then tries to compile it, which it cannot, because it is a directory, not an ANTLR grammar). I tried specifying "-lib directory1 -lib directory2", the "-lib directory2" just overwrites the "-lib directory1" then.
It is what I believe a fairly basic need to specify multiple directories to search in, in any more complicated system of grammars, therefore I am sure I missed something. I just cannot google anything useful out.
IF you need any more details, I am happy to provide them, it is my master thesis extending an open source project, so I do not need to keep anything secret. :) If it is not possible, I can live without it, but would really like to keep the consistency of the original project.
The default location for the grammar files is in the director src/main/antlr4/ . The plugin will produce . java files for the generated parser in the output director target/generated-sources/antlr4/ .
ANTLR 2 accepts three types of grammar specifications -- parsers, lexers, and tree-parsers (also called tree-walkers). Because ANTLR 2 uses LL(k) analysis for all three grammar variants, the grammar specifications are similar, and the generated lexers and parsers behave similarly.
ANTLR (ANother Tool for Language Recognition) is a tool for processing structured text. It does this by giving us access to language processing primitives like lexers, grammars, and parsers as well as the runtime to process text against them. It's often used to build tools and frameworks.
In fact, there are context-free grammars that you can "specify" with ANTLR that it cannot process correctly, which is true of most parser generators. (For ANTLR, this includes grammars with indirect left recursion, ambiguity, arbitrary lookahead, etc.)
Looking at the source of the org.antlr.Tool
class:
else if (args[i].equals("-lib")) {
if (i + 1 >= args.length) {
System.err.println("missing library directory with -lib option; ignoring");
}
else {
i++;
libDirectory = args[i];
if (libDirectory.endsWith("/") || libDirectory.endsWith("\\")) {
libDirectory = libDirectory.substring(0,libDirectory.length()-1);
}
File outDir = new File(libDirectory);
if (!outDir.exists()) {
ErrorManager.error(ErrorManager.MSG_DIR_NOT_FOUND,libDirectory);
libDirectory = ".";
}
}
}
it appears just one -lib
directory is being read.
I'm assuming your (implied) question is hereby answered (that it is not possible to point to more than 1 -lib
directory).
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With