Salesforce grammars for tree-sitter; includes Apex, SOQL, and SOSL languages. Also includes parser for the Salesforce Log files (sflog).
Try it out using our playground
If you are a Neo-vim user, the parsers and syntax highlights are part of nvim-treesitter plugin already, guidance.
npm install tree-sitter
npm install tree-sitter-sfapex// import libraries
import Parser from "tree-sitter";
import TsSfApex from "tree-sitter-sfapex";
// create a parser instance
const parser = new Parser();
// could be .apex, .soql, or .sosl
parser.setLanguage(TsSfApex.apex);
// we set the language to Apex so lets feed it some apex
const apexTree = parser.parse(`
/**
* block comment
*/
global class TestClass implements TestInterface {
public static String Prop1 = 'TestVal';
global Account setName(Account acct, String nameVal){
acct.Name = nameVal;
return acct;
}
}`);
// just a super simple example of printing the discovered nodes
// to see the anonymous nodes (syntax without formal names) set this to `true`
const includeAnonymousNodes = false;
console.log("APEX TREE");
printTree(apexTree.rootNode);
// do it with some SOQL this time
parser.setLanguage(TsSfApex.soql);
const soqlTree = parser.parse(`
SELECT Id, Name, Parent.Name,
TYPEOF Owner
WHEN User THEN Id, Username, FederationId
WHEN Group THEN Name
END,
(SELECT Id, Name FROM Contacts)
FROM Account
WHERE Name = 'Robots' AND Are_Coming__c = FALSE
`);
console.log("SOQL TREE");
printTree(soqlTree.rootNode);
function printTree(node, indent = 0) {
console.log(
" ".repeat(indent),
(node.isNamed ? "(" : "") + node.type + (node.isNamed ? ")" : "")
);
for (let c of includeAnonymousNodes ? node.children : node.namedChildren) {
printTree(c, indent + 2);
}
}
Most of the parsers are built and tested on large corpus of Apex, I still intend to write automated tests that parse large Apex libraries as part of evaluating the grammar.
Contributed by @finxxi, the sflog parser is very new and still being figured out. Feel free to contribute logs that fail to parse via the issues or by adding a failing test. I expect it to have a lot of development in terms of comprehention of individual log patterns and pulling out the types present there. For now it generally handles the main structure of log entries but doesn't parse the logged details very finely.
Please open an issue on this repo and we'll work through it.
Still figuring this out. By far the most useful contributions would be tests, if you have a scenario that doesn't work you can just provide the example or open a PR with a new failing test and I can figure out what to do about it.