commit 8d8a40a3a31ed502712af8f2002ba63c067838be Author: kjuulh Date: Tue Jul 4 19:37:50 2023 +0200 feat: with base Signed-off-by: kjuulh diff --git a/.gitignore b/.gitignore new file mode 100644 index 0000000..297fdef --- /dev/null +++ b/.gitignore @@ -0,0 +1,2 @@ +target/ +node_modules/ diff --git a/Cargo.toml b/Cargo.toml new file mode 100644 index 0000000..c49f387 --- /dev/null +++ b/Cargo.toml @@ -0,0 +1,26 @@ +[package] +name = "tree-sitter-hurl" +description = "hurl grammar for the tree-sitter parsing library" +version = "0.0.1" +keywords = ["incremental", "parsing", "hurl"] +categories = ["parsing", "text-editors"] +repository = "https://github.com/tree-sitter/tree-sitter-hurl" +edition = "2018" +license = "MIT" + +build = "bindings/rust/build.rs" +include = [ + "bindings/rust/*", + "grammar.js", + "queries/*", + "src/*", +] + +[lib] +path = "bindings/rust/lib.rs" + +[dependencies] +tree-sitter = "~0.20.10" + +[build-dependencies] +cc = "1.0" diff --git a/binding.gyp b/binding.gyp new file mode 100644 index 0000000..42c7003 --- /dev/null +++ b/binding.gyp @@ -0,0 +1,19 @@ +{ + "targets": [ + { + "target_name": "tree_sitter_hurl_binding", + "include_dirs": [ + " +#include "nan.h" + +using namespace v8; + +extern "C" TSLanguage * tree_sitter_hurl(); + +namespace { + +NAN_METHOD(New) {} + +void Init(Local exports, Local module) { + Local tpl = Nan::New(New); + tpl->SetClassName(Nan::New("Language").ToLocalChecked()); + tpl->InstanceTemplate()->SetInternalFieldCount(1); + + Local constructor = Nan::GetFunction(tpl).ToLocalChecked(); + Local instance = constructor->NewInstance(Nan::GetCurrentContext()).ToLocalChecked(); + Nan::SetInternalFieldPointer(instance, 0, tree_sitter_hurl()); + + Nan::Set(instance, Nan::New("name").ToLocalChecked(), Nan::New("hurl").ToLocalChecked()); + Nan::Set(module, Nan::New("exports").ToLocalChecked(), instance); +} + +NODE_MODULE(tree_sitter_hurl_binding, Init) + +} // namespace diff --git a/bindings/node/index.js b/bindings/node/index.js new file mode 100644 index 0000000..df40586 --- /dev/null +++ b/bindings/node/index.js @@ -0,0 +1,19 @@ +try { + module.exports = require("../../build/Release/tree_sitter_hurl_binding"); +} catch (error1) { + if (error1.code !== 'MODULE_NOT_FOUND') { + throw error1; + } + try { + module.exports = require("../../build/Debug/tree_sitter_hurl_binding"); + } catch (error2) { + if (error2.code !== 'MODULE_NOT_FOUND') { + throw error2; + } + throw error1 + } +} + +try { + module.exports.nodeTypeInfo = require("../../src/node-types.json"); +} catch (_) {} diff --git a/bindings/rust/build.rs b/bindings/rust/build.rs new file mode 100644 index 0000000..c6061f0 --- /dev/null +++ b/bindings/rust/build.rs @@ -0,0 +1,40 @@ +fn main() { + let src_dir = std::path::Path::new("src"); + + let mut c_config = cc::Build::new(); + c_config.include(&src_dir); + c_config + .flag_if_supported("-Wno-unused-parameter") + .flag_if_supported("-Wno-unused-but-set-variable") + .flag_if_supported("-Wno-trigraphs"); + let parser_path = src_dir.join("parser.c"); + c_config.file(&parser_path); + + // If your language uses an external scanner written in C, + // then include this block of code: + + /* + let scanner_path = src_dir.join("scanner.c"); + c_config.file(&scanner_path); + println!("cargo:rerun-if-changed={}", scanner_path.to_str().unwrap()); + */ + + c_config.compile("parser"); + println!("cargo:rerun-if-changed={}", parser_path.to_str().unwrap()); + + // If your language uses an external scanner written in C++, + // then include this block of code: + + /* + let mut cpp_config = cc::Build::new(); + cpp_config.cpp(true); + cpp_config.include(&src_dir); + cpp_config + .flag_if_supported("-Wno-unused-parameter") + .flag_if_supported("-Wno-unused-but-set-variable"); + let scanner_path = src_dir.join("scanner.cc"); + cpp_config.file(&scanner_path); + cpp_config.compile("scanner"); + println!("cargo:rerun-if-changed={}", scanner_path.to_str().unwrap()); + */ +} diff --git a/bindings/rust/lib.rs b/bindings/rust/lib.rs new file mode 100644 index 0000000..8a67f8c --- /dev/null +++ b/bindings/rust/lib.rs @@ -0,0 +1,52 @@ +//! This crate provides hurl language support for the [tree-sitter][] parsing library. +//! +//! Typically, you will use the [language][language func] function to add this language to a +//! tree-sitter [Parser][], and then use the parser to parse some code: +//! +//! ``` +//! let code = ""; +//! let mut parser = tree_sitter::Parser::new(); +//! parser.set_language(tree_sitter_hurl::language()).expect("Error loading hurl grammar"); +//! let tree = parser.parse(code, None).unwrap(); +//! ``` +//! +//! [Language]: https://docs.rs/tree-sitter/*/tree_sitter/struct.Language.html +//! [language func]: fn.language.html +//! [Parser]: https://docs.rs/tree-sitter/*/tree_sitter/struct.Parser.html +//! [tree-sitter]: https://tree-sitter.github.io/ + +use tree_sitter::Language; + +extern "C" { + fn tree_sitter_hurl() -> Language; +} + +/// Get the tree-sitter [Language][] for this grammar. +/// +/// [Language]: https://docs.rs/tree-sitter/*/tree_sitter/struct.Language.html +pub fn language() -> Language { + unsafe { tree_sitter_hurl() } +} + +/// The content of the [`node-types.json`][] file for this grammar. +/// +/// [`node-types.json`]: https://tree-sitter.github.io/tree-sitter/using-parsers#static-node-types +pub const NODE_TYPES: &'static str = include_str!("../../src/node-types.json"); + +// Uncomment these to include any queries that this grammar contains + +// pub const HIGHLIGHTS_QUERY: &'static str = include_str!("../../queries/highlights.scm"); +// pub const INJECTIONS_QUERY: &'static str = include_str!("../../queries/injections.scm"); +// pub const LOCALS_QUERY: &'static str = include_str!("../../queries/locals.scm"); +// pub const TAGS_QUERY: &'static str = include_str!("../../queries/tags.scm"); + +#[cfg(test)] +mod tests { + #[test] + fn test_can_load_grammar() { + let mut parser = tree_sitter::Parser::new(); + parser + .set_language(super::language()) + .expect("Error loading hurl language"); + } +} diff --git a/corpus/requests.txt b/corpus/requests.txt new file mode 100644 index 0000000..54ef5ec --- /dev/null +++ b/corpus/requests.txt @@ -0,0 +1,23 @@ +=== +GET request +=== + +GET https://somewhere.com + +--- + +(source_file + (request_declaration + (request_literal))) + +=== +GET request +=== + +POST https://somewhere.com?query=something + +--- + +(source_file + (request_declaration + (request_literal))) diff --git a/grammar.js b/grammar.js new file mode 100644 index 0000000..1d281a4 --- /dev/null +++ b/grammar.js @@ -0,0 +1,41 @@ +module.exports = grammar({ + name: "hurl", + + rules: { + source_file: ($) => repeat($._statement), + + // statements + + _statement: ($) => + choice($._declaration_statement), + + _declaration_statement: ($) => + choice($.request_declaration), + + request_declaration: ($) => + seq($._literal, field("url", $._url)), + + _literal: ($) => choice($.request_literal), + + request_literal: (_$) => + choice( + "GET", + "POST", + "PUT", + "DELETE", + "CONNECT", + "OPTIONS", + "TRACE", + "PATCH", + "LINK", + "UNLINK", + "PURGE", + "LOCK", + "UNLOCK", + "PROPFIND", + "VIEW" + ), + + _url: ($) => /\S+/, + }, +}); diff --git a/package-lock.json b/package-lock.json new file mode 100644 index 0000000..91c2ddd --- /dev/null +++ b/package-lock.json @@ -0,0 +1,50 @@ +{ + "name": "tree-sitter-hurl", + "version": "1.0.0", + "lockfileVersion": 3, + "requires": true, + "packages": { + "": { + "name": "tree-sitter-hurl", + "version": "1.0.0", + "license": "ISC", + "dependencies": { + "nan": "^2.17.0" + }, + "devDependencies": { + "prettier": "^2.8.8", + "tree-sitter-cli": "^0.20.8" + } + }, + "node_modules/nan": { + "version": "2.17.0", + "resolved": "https://registry.npmjs.org/nan/-/nan-2.17.0.tgz", + "integrity": "sha512-2ZTgtl0nJsO0KQCjEpxcIr5D+Yv90plTitZt9JBfQvVJDS5seMl3FOvsh3+9CoYWXf/1l5OaZzzF6nDm4cagaQ==" + }, + "node_modules/prettier": { + "version": "2.8.8", + "resolved": "https://registry.npmjs.org/prettier/-/prettier-2.8.8.tgz", + "integrity": "sha512-tdN8qQGvNjw4CHbY+XXk0JgCXn9QiF21a55rBe5LJAU+kDyC4WQn4+awm2Xfk2lQMk5fKup9XgzTZtGkjBdP9Q==", + "dev": true, + "bin": { + "prettier": "bin-prettier.js" + }, + "engines": { + "node": ">=10.13.0" + }, + "funding": { + "url": "https://github.com/prettier/prettier?sponsor=1" + } + }, + "node_modules/tree-sitter-cli": { + "version": "0.20.8", + "resolved": "https://registry.npmjs.org/tree-sitter-cli/-/tree-sitter-cli-0.20.8.tgz", + "integrity": "sha512-XjTcS3wdTy/2cc/ptMLc/WRyOLECRYcMTrSWyhZnj1oGSOWbHLTklgsgRICU3cPfb0vy+oZCC33M43u6R1HSCA==", + "dev": true, + "hasInstallScript": true, + "bin": { + "tree-sitter": "cli.js" + } + } + } +} diff --git a/package.json b/package.json new file mode 100644 index 0000000..520d0d8 --- /dev/null +++ b/package.json @@ -0,0 +1,30 @@ +{ + "name": "tree-sitter-hurl", + "version": "1.0.0", + "description": "", + "main": "bindings/node", + "scripts": { + "test": "tree-sitter test", + "generate": "tree-sitter generate", + "format": "prettier -w --print-width=50 grammar.js", + "ci": "npm run generate && npm run test" + }, + "author": "", + "license": "ISC", + "dependencies": { + "nan": "^2.17.0" + }, + "devDependencies": { + "prettier": "^2.8.8", + "tree-sitter-cli": "^0.20.8" + }, + "tree-sitter": [ + { + "scope": "source.hurl", + "file-types": [ + "hurl" + ], + "injection-regex": "hurl" + } + ] +} diff --git a/queries/highlights.scm b/queries/highlights.scm new file mode 100644 index 0000000..837da55 --- /dev/null +++ b/queries/highlights.scm @@ -0,0 +1,17 @@ +[ + "GET" + "POST" + "PUT" + "DELETE" + "CONNECT" + "OPTIONS" + "TRACE" + "PATCH" + "LINK" + "UNLINK" + "PURGE" + "LOCK" + "UNLOCK" + "PROPFIND" + "VIEW" + ] @keyword diff --git a/src/grammar.json b/src/grammar.json new file mode 100644 index 0000000..8aa7b39 --- /dev/null +++ b/src/grammar.json @@ -0,0 +1,137 @@ +{ + "name": "hurl", + "rules": { + "source_file": { + "type": "REPEAT", + "content": { + "type": "SYMBOL", + "name": "_statement" + } + }, + "_statement": { + "type": "CHOICE", + "members": [ + { + "type": "SYMBOL", + "name": "_declaration_statement" + } + ] + }, + "_declaration_statement": { + "type": "CHOICE", + "members": [ + { + "type": "SYMBOL", + "name": "request_declaration" + } + ] + }, + "request_declaration": { + "type": "SEQ", + "members": [ + { + "type": "SYMBOL", + "name": "_literal" + }, + { + "type": "FIELD", + "name": "url", + "content": { + "type": "SYMBOL", + "name": "_url" + } + } + ] + }, + "_literal": { + "type": "CHOICE", + "members": [ + { + "type": "SYMBOL", + "name": "request_literal" + } + ] + }, + "request_literal": { + "type": "CHOICE", + "members": [ + { + "type": "STRING", + "value": "GET" + }, + { + "type": "STRING", + "value": "POST" + }, + { + "type": "STRING", + "value": "PUT" + }, + { + "type": "STRING", + "value": "DELETE" + }, + { + "type": "STRING", + "value": "CONNECT" + }, + { + "type": "STRING", + "value": "OPTIONS" + }, + { + "type": "STRING", + "value": "TRACE" + }, + { + "type": "STRING", + "value": "PATCH" + }, + { + "type": "STRING", + "value": "LINK" + }, + { + "type": "STRING", + "value": "UNLINK" + }, + { + "type": "STRING", + "value": "PURGE" + }, + { + "type": "STRING", + "value": "LOCK" + }, + { + "type": "STRING", + "value": "UNLOCK" + }, + { + "type": "STRING", + "value": "PROPFIND" + }, + { + "type": "STRING", + "value": "VIEW" + } + ] + }, + "_url": { + "type": "PATTERN", + "value": "\\S+" + } + }, + "extras": [ + { + "type": "PATTERN", + "value": "\\s" + } + ], + "conflicts": [], + "precedences": [], + "externals": [], + "inline": [], + "supertypes": [] +} + diff --git a/src/node-types.json b/src/node-types.json new file mode 100644 index 0000000..365e66d --- /dev/null +++ b/src/node-types.json @@ -0,0 +1,97 @@ +[ + { + "type": "request_declaration", + "named": true, + "fields": {}, + "children": { + "multiple": false, + "required": true, + "types": [ + { + "type": "request_literal", + "named": true + } + ] + } + }, + { + "type": "request_literal", + "named": true, + "fields": {} + }, + { + "type": "source_file", + "named": true, + "fields": {}, + "children": { + "multiple": true, + "required": false, + "types": [ + { + "type": "request_declaration", + "named": true + } + ] + } + }, + { + "type": "CONNECT", + "named": false + }, + { + "type": "DELETE", + "named": false + }, + { + "type": "GET", + "named": false + }, + { + "type": "LINK", + "named": false + }, + { + "type": "LOCK", + "named": false + }, + { + "type": "OPTIONS", + "named": false + }, + { + "type": "PATCH", + "named": false + }, + { + "type": "POST", + "named": false + }, + { + "type": "PROPFIND", + "named": false + }, + { + "type": "PURGE", + "named": false + }, + { + "type": "PUT", + "named": false + }, + { + "type": "TRACE", + "named": false + }, + { + "type": "UNLINK", + "named": false + }, + { + "type": "UNLOCK", + "named": false + }, + { + "type": "VIEW", + "named": false + } +] \ No newline at end of file diff --git a/src/parser.c b/src/parser.c new file mode 100644 index 0000000..7fa8403 --- /dev/null +++ b/src/parser.c @@ -0,0 +1,682 @@ +#include + +#if defined(__GNUC__) || defined(__clang__) +#pragma GCC diagnostic push +#pragma GCC diagnostic ignored "-Wmissing-field-initializers" +#endif + +#define LANGUAGE_VERSION 14 +#define STATE_COUNT 8 +#define LARGE_STATE_COUNT 5 +#define SYMBOL_COUNT 24 +#define ALIAS_COUNT 0 +#define TOKEN_COUNT 17 +#define EXTERNAL_TOKEN_COUNT 0 +#define FIELD_COUNT 1 +#define MAX_ALIAS_SEQUENCE_LENGTH 2 +#define PRODUCTION_ID_COUNT 2 + +enum { + anon_sym_GET = 1, + anon_sym_POST = 2, + anon_sym_PUT = 3, + anon_sym_DELETE = 4, + anon_sym_CONNECT = 5, + anon_sym_OPTIONS = 6, + anon_sym_TRACE = 7, + anon_sym_PATCH = 8, + anon_sym_LINK = 9, + anon_sym_UNLINK = 10, + anon_sym_PURGE = 11, + anon_sym_LOCK = 12, + anon_sym_UNLOCK = 13, + anon_sym_PROPFIND = 14, + anon_sym_VIEW = 15, + sym__url = 16, + sym_source_file = 17, + sym__statement = 18, + sym__declaration_statement = 19, + sym_request_declaration = 20, + sym__literal = 21, + sym_request_literal = 22, + aux_sym_source_file_repeat1 = 23, +}; + +static const char * const ts_symbol_names[] = { + [ts_builtin_sym_end] = "end", + [anon_sym_GET] = "GET", + [anon_sym_POST] = "POST", + [anon_sym_PUT] = "PUT", + [anon_sym_DELETE] = "DELETE", + [anon_sym_CONNECT] = "CONNECT", + [anon_sym_OPTIONS] = "OPTIONS", + [anon_sym_TRACE] = "TRACE", + [anon_sym_PATCH] = "PATCH", + [anon_sym_LINK] = "LINK", + [anon_sym_UNLINK] = "UNLINK", + [anon_sym_PURGE] = "PURGE", + [anon_sym_LOCK] = "LOCK", + [anon_sym_UNLOCK] = "UNLOCK", + [anon_sym_PROPFIND] = "PROPFIND", + [anon_sym_VIEW] = "VIEW", + [sym__url] = "_url", + [sym_source_file] = "source_file", + [sym__statement] = "_statement", + [sym__declaration_statement] = "_declaration_statement", + [sym_request_declaration] = "request_declaration", + [sym__literal] = "_literal", + [sym_request_literal] = "request_literal", + [aux_sym_source_file_repeat1] = "source_file_repeat1", +}; + +static const TSSymbol ts_symbol_map[] = { + [ts_builtin_sym_end] = ts_builtin_sym_end, + [anon_sym_GET] = anon_sym_GET, + [anon_sym_POST] = anon_sym_POST, + [anon_sym_PUT] = anon_sym_PUT, + [anon_sym_DELETE] = anon_sym_DELETE, + [anon_sym_CONNECT] = anon_sym_CONNECT, + [anon_sym_OPTIONS] = anon_sym_OPTIONS, + [anon_sym_TRACE] = anon_sym_TRACE, + [anon_sym_PATCH] = anon_sym_PATCH, + [anon_sym_LINK] = anon_sym_LINK, + [anon_sym_UNLINK] = anon_sym_UNLINK, + [anon_sym_PURGE] = anon_sym_PURGE, + [anon_sym_LOCK] = anon_sym_LOCK, + [anon_sym_UNLOCK] = anon_sym_UNLOCK, + [anon_sym_PROPFIND] = anon_sym_PROPFIND, + [anon_sym_VIEW] = anon_sym_VIEW, + [sym__url] = sym__url, + [sym_source_file] = sym_source_file, + [sym__statement] = sym__statement, + [sym__declaration_statement] = sym__declaration_statement, + [sym_request_declaration] = sym_request_declaration, + [sym__literal] = sym__literal, + [sym_request_literal] = sym_request_literal, + [aux_sym_source_file_repeat1] = aux_sym_source_file_repeat1, +}; + +static const TSSymbolMetadata ts_symbol_metadata[] = { + [ts_builtin_sym_end] = { + .visible = false, + .named = true, + }, + [anon_sym_GET] = { + .visible = true, + .named = false, + }, + [anon_sym_POST] = { + .visible = true, + .named = false, + }, + [anon_sym_PUT] = { + .visible = true, + .named = false, + }, + [anon_sym_DELETE] = { + .visible = true, + .named = false, + }, + [anon_sym_CONNECT] = { + .visible = true, + .named = false, + }, + [anon_sym_OPTIONS] = { + .visible = true, + .named = false, + }, + [anon_sym_TRACE] = { + .visible = true, + .named = false, + }, + [anon_sym_PATCH] = { + .visible = true, + .named = false, + }, + [anon_sym_LINK] = { + .visible = true, + .named = false, + }, + [anon_sym_UNLINK] = { + .visible = true, + .named = false, + }, + [anon_sym_PURGE] = { + .visible = true, + .named = false, + }, + [anon_sym_LOCK] = { + .visible = true, + .named = false, + }, + [anon_sym_UNLOCK] = { + .visible = true, + .named = false, + }, + [anon_sym_PROPFIND] = { + .visible = true, + .named = false, + }, + [anon_sym_VIEW] = { + .visible = true, + .named = false, + }, + [sym__url] = { + .visible = false, + .named = true, + }, + [sym_source_file] = { + .visible = true, + .named = true, + }, + [sym__statement] = { + .visible = false, + .named = true, + }, + [sym__declaration_statement] = { + .visible = false, + .named = true, + }, + [sym_request_declaration] = { + .visible = true, + .named = true, + }, + [sym__literal] = { + .visible = false, + .named = true, + }, + [sym_request_literal] = { + .visible = true, + .named = true, + }, + [aux_sym_source_file_repeat1] = { + .visible = false, + .named = false, + }, +}; + +enum { + field_url = 1, +}; + +static const char * const ts_field_names[] = { + [0] = NULL, + [field_url] = "url", +}; + +static const TSFieldMapSlice ts_field_map_slices[PRODUCTION_ID_COUNT] = { + [1] = {.index = 0, .length = 1}, +}; + +static const TSFieldMapEntry ts_field_map_entries[] = { + [0] = + {field_url, 1}, +}; + +static const TSSymbol ts_alias_sequences[PRODUCTION_ID_COUNT][MAX_ALIAS_SEQUENCE_LENGTH] = { + [0] = {0}, +}; + +static const uint16_t ts_non_terminal_alias_map[] = { + 0, +}; + +static const TSStateId ts_primary_state_ids[STATE_COUNT] = { + [0] = 0, + [1] = 1, + [2] = 2, + [3] = 3, + [4] = 4, + [5] = 5, + [6] = 6, + [7] = 7, +}; + +static bool ts_lex(TSLexer *lexer, TSStateId state) { + START_LEXER(); + eof = lexer->eof(lexer); + switch (state) { + case 0: + if (eof) ADVANCE(55); + if (lookahead == 'C') ADVANCE(39); + if (lookahead == 'D') ADVANCE(9); + if (lookahead == 'G') ADVANCE(10); + if (lookahead == 'L') ADVANCE(21); + if (lookahead == 'O') ADVANCE(42); + if (lookahead == 'P') ADVANCE(1); + if (lookahead == 'T') ADVANCE(43); + if (lookahead == 'U') ADVANCE(33); + if (lookahead == 'V') ADVANCE(20); + if (lookahead == '\t' || + lookahead == '\n' || + lookahead == '\r' || + lookahead == ' ') SKIP(0) + END_STATE(); + case 1: + if (lookahead == 'A') ADVANCE(51); + if (lookahead == 'O') ADVANCE(46); + if (lookahead == 'R') ADVANCE(38); + if (lookahead == 'U') ADVANCE(44); + END_STATE(); + case 2: + if (lookahead == 'A') ADVANCE(7); + END_STATE(); + case 3: + if (lookahead == 'C') ADVANCE(19); + END_STATE(); + case 4: + if (lookahead == 'C') ADVANCE(26); + END_STATE(); + case 5: + if (lookahead == 'C') ADVANCE(28); + END_STATE(); + case 6: + if (lookahead == 'C') ADVANCE(49); + END_STATE(); + case 7: + if (lookahead == 'C') ADVANCE(13); + END_STATE(); + case 8: + if (lookahead == 'D') ADVANCE(69); + END_STATE(); + case 9: + if (lookahead == 'E') ADVANCE(29); + END_STATE(); + case 10: + if (lookahead == 'E') ADVANCE(47); + END_STATE(); + case 11: + if (lookahead == 'E') ADVANCE(53); + END_STATE(); + case 12: + if (lookahead == 'E') ADVANCE(66); + END_STATE(); + case 13: + if (lookahead == 'E') ADVANCE(62); + END_STATE(); + case 14: + if (lookahead == 'E') ADVANCE(59); + END_STATE(); + case 15: + if (lookahead == 'E') ADVANCE(6); + END_STATE(); + case 16: + if (lookahead == 'E') ADVANCE(52); + END_STATE(); + case 17: + if (lookahead == 'F') ADVANCE(23); + END_STATE(); + case 18: + if (lookahead == 'G') ADVANCE(12); + END_STATE(); + case 19: + if (lookahead == 'H') ADVANCE(63); + END_STATE(); + case 20: + if (lookahead == 'I') ADVANCE(11); + END_STATE(); + case 21: + if (lookahead == 'I') ADVANCE(31); + if (lookahead == 'O') ADVANCE(4); + END_STATE(); + case 22: + if (lookahead == 'I') ADVANCE(35); + if (lookahead == 'O') ADVANCE(5); + END_STATE(); + case 23: + if (lookahead == 'I') ADVANCE(32); + END_STATE(); + case 24: + if (lookahead == 'I') ADVANCE(40); + END_STATE(); + case 25: + if (lookahead == 'K') ADVANCE(64); + END_STATE(); + case 26: + if (lookahead == 'K') ADVANCE(67); + END_STATE(); + case 27: + if (lookahead == 'K') ADVANCE(65); + END_STATE(); + case 28: + if (lookahead == 'K') ADVANCE(68); + END_STATE(); + case 29: + if (lookahead == 'L') ADVANCE(16); + END_STATE(); + case 30: + if (lookahead == 'L') ADVANCE(22); + END_STATE(); + case 31: + if (lookahead == 'N') ADVANCE(25); + END_STATE(); + case 32: + if (lookahead == 'N') ADVANCE(8); + END_STATE(); + case 33: + if (lookahead == 'N') ADVANCE(30); + END_STATE(); + case 34: + if (lookahead == 'N') ADVANCE(45); + END_STATE(); + case 35: + if (lookahead == 'N') ADVANCE(27); + END_STATE(); + case 36: + if (lookahead == 'N') ADVANCE(37); + END_STATE(); + case 37: + if (lookahead == 'N') ADVANCE(15); + END_STATE(); + case 38: + if (lookahead == 'O') ADVANCE(41); + END_STATE(); + case 39: + if (lookahead == 'O') ADVANCE(36); + END_STATE(); + case 40: + if (lookahead == 'O') ADVANCE(34); + END_STATE(); + case 41: + if (lookahead == 'P') ADVANCE(17); + END_STATE(); + case 42: + if (lookahead == 'P') ADVANCE(50); + END_STATE(); + case 43: + if (lookahead == 'R') ADVANCE(2); + END_STATE(); + case 44: + if (lookahead == 'R') ADVANCE(18); + if (lookahead == 'T') ADVANCE(58); + END_STATE(); + case 45: + if (lookahead == 'S') ADVANCE(61); + END_STATE(); + case 46: + if (lookahead == 'S') ADVANCE(48); + END_STATE(); + case 47: + if (lookahead == 'T') ADVANCE(56); + END_STATE(); + case 48: + if (lookahead == 'T') ADVANCE(57); + END_STATE(); + case 49: + if (lookahead == 'T') ADVANCE(60); + END_STATE(); + case 50: + if (lookahead == 'T') ADVANCE(24); + END_STATE(); + case 51: + if (lookahead == 'T') ADVANCE(3); + END_STATE(); + case 52: + if (lookahead == 'T') ADVANCE(14); + END_STATE(); + case 53: + if (lookahead == 'W') ADVANCE(70); + END_STATE(); + case 54: + if (lookahead == '\t' || + lookahead == '\n' || + lookahead == '\r' || + lookahead == ' ') SKIP(54) + if (lookahead != 0) ADVANCE(71); + END_STATE(); + case 55: + ACCEPT_TOKEN(ts_builtin_sym_end); + END_STATE(); + case 56: + ACCEPT_TOKEN(anon_sym_GET); + END_STATE(); + case 57: + ACCEPT_TOKEN(anon_sym_POST); + END_STATE(); + case 58: + ACCEPT_TOKEN(anon_sym_PUT); + END_STATE(); + case 59: + ACCEPT_TOKEN(anon_sym_DELETE); + END_STATE(); + case 60: + ACCEPT_TOKEN(anon_sym_CONNECT); + END_STATE(); + case 61: + ACCEPT_TOKEN(anon_sym_OPTIONS); + END_STATE(); + case 62: + ACCEPT_TOKEN(anon_sym_TRACE); + END_STATE(); + case 63: + ACCEPT_TOKEN(anon_sym_PATCH); + END_STATE(); + case 64: + ACCEPT_TOKEN(anon_sym_LINK); + END_STATE(); + case 65: + ACCEPT_TOKEN(anon_sym_UNLINK); + END_STATE(); + case 66: + ACCEPT_TOKEN(anon_sym_PURGE); + END_STATE(); + case 67: + ACCEPT_TOKEN(anon_sym_LOCK); + END_STATE(); + case 68: + ACCEPT_TOKEN(anon_sym_UNLOCK); + END_STATE(); + case 69: + ACCEPT_TOKEN(anon_sym_PROPFIND); + END_STATE(); + case 70: + ACCEPT_TOKEN(anon_sym_VIEW); + END_STATE(); + case 71: + ACCEPT_TOKEN(sym__url); + if (lookahead != 0 && + lookahead != '\t' && + lookahead != '\n' && + lookahead != '\r' && + lookahead != ' ') ADVANCE(71); + END_STATE(); + default: + return false; + } +} + +static const TSLexMode ts_lex_modes[STATE_COUNT] = { + [0] = {.lex_state = 0}, + [1] = {.lex_state = 0}, + [2] = {.lex_state = 0}, + [3] = {.lex_state = 0}, + [4] = {.lex_state = 0}, + [5] = {.lex_state = 54}, + [6] = {.lex_state = 0}, + [7] = {.lex_state = 54}, +}; + +static const uint16_t ts_parse_table[LARGE_STATE_COUNT][SYMBOL_COUNT] = { + [0] = { + [ts_builtin_sym_end] = ACTIONS(1), + [anon_sym_GET] = ACTIONS(1), + [anon_sym_POST] = ACTIONS(1), + [anon_sym_PUT] = ACTIONS(1), + [anon_sym_DELETE] = ACTIONS(1), + [anon_sym_CONNECT] = ACTIONS(1), + [anon_sym_OPTIONS] = ACTIONS(1), + [anon_sym_TRACE] = ACTIONS(1), + [anon_sym_PATCH] = ACTIONS(1), + [anon_sym_LINK] = ACTIONS(1), + [anon_sym_UNLINK] = ACTIONS(1), + [anon_sym_PURGE] = ACTIONS(1), + [anon_sym_LOCK] = ACTIONS(1), + [anon_sym_UNLOCK] = ACTIONS(1), + [anon_sym_PROPFIND] = ACTIONS(1), + [anon_sym_VIEW] = ACTIONS(1), + }, + [1] = { + [sym_source_file] = STATE(6), + [sym__statement] = STATE(2), + [sym__declaration_statement] = STATE(2), + [sym_request_declaration] = STATE(2), + [sym__literal] = STATE(7), + [sym_request_literal] = STATE(7), + [aux_sym_source_file_repeat1] = STATE(2), + [ts_builtin_sym_end] = ACTIONS(3), + [anon_sym_GET] = ACTIONS(5), + [anon_sym_POST] = ACTIONS(5), + [anon_sym_PUT] = ACTIONS(5), + [anon_sym_DELETE] = ACTIONS(5), + [anon_sym_CONNECT] = ACTIONS(5), + [anon_sym_OPTIONS] = ACTIONS(5), + [anon_sym_TRACE] = ACTIONS(5), + [anon_sym_PATCH] = ACTIONS(5), + [anon_sym_LINK] = ACTIONS(5), + [anon_sym_UNLINK] = ACTIONS(5), + [anon_sym_PURGE] = ACTIONS(5), + [anon_sym_LOCK] = ACTIONS(5), + [anon_sym_UNLOCK] = ACTIONS(5), + [anon_sym_PROPFIND] = ACTIONS(5), + [anon_sym_VIEW] = ACTIONS(5), + }, + [2] = { + [sym__statement] = STATE(3), + [sym__declaration_statement] = STATE(3), + [sym_request_declaration] = STATE(3), + [sym__literal] = STATE(7), + [sym_request_literal] = STATE(7), + [aux_sym_source_file_repeat1] = STATE(3), + [ts_builtin_sym_end] = ACTIONS(7), + [anon_sym_GET] = ACTIONS(5), + [anon_sym_POST] = ACTIONS(5), + [anon_sym_PUT] = ACTIONS(5), + [anon_sym_DELETE] = ACTIONS(5), + [anon_sym_CONNECT] = ACTIONS(5), + [anon_sym_OPTIONS] = ACTIONS(5), + [anon_sym_TRACE] = ACTIONS(5), + [anon_sym_PATCH] = ACTIONS(5), + [anon_sym_LINK] = ACTIONS(5), + [anon_sym_UNLINK] = ACTIONS(5), + [anon_sym_PURGE] = ACTIONS(5), + [anon_sym_LOCK] = ACTIONS(5), + [anon_sym_UNLOCK] = ACTIONS(5), + [anon_sym_PROPFIND] = ACTIONS(5), + [anon_sym_VIEW] = ACTIONS(5), + }, + [3] = { + [sym__statement] = STATE(3), + [sym__declaration_statement] = STATE(3), + [sym_request_declaration] = STATE(3), + [sym__literal] = STATE(7), + [sym_request_literal] = STATE(7), + [aux_sym_source_file_repeat1] = STATE(3), + [ts_builtin_sym_end] = ACTIONS(9), + [anon_sym_GET] = ACTIONS(11), + [anon_sym_POST] = ACTIONS(11), + [anon_sym_PUT] = ACTIONS(11), + [anon_sym_DELETE] = ACTIONS(11), + [anon_sym_CONNECT] = ACTIONS(11), + [anon_sym_OPTIONS] = ACTIONS(11), + [anon_sym_TRACE] = ACTIONS(11), + [anon_sym_PATCH] = ACTIONS(11), + [anon_sym_LINK] = ACTIONS(11), + [anon_sym_UNLINK] = ACTIONS(11), + [anon_sym_PURGE] = ACTIONS(11), + [anon_sym_LOCK] = ACTIONS(11), + [anon_sym_UNLOCK] = ACTIONS(11), + [anon_sym_PROPFIND] = ACTIONS(11), + [anon_sym_VIEW] = ACTIONS(11), + }, + [4] = { + [ts_builtin_sym_end] = ACTIONS(14), + [anon_sym_GET] = ACTIONS(14), + [anon_sym_POST] = ACTIONS(14), + [anon_sym_PUT] = ACTIONS(14), + [anon_sym_DELETE] = ACTIONS(14), + [anon_sym_CONNECT] = ACTIONS(14), + [anon_sym_OPTIONS] = ACTIONS(14), + [anon_sym_TRACE] = ACTIONS(14), + [anon_sym_PATCH] = ACTIONS(14), + [anon_sym_LINK] = ACTIONS(14), + [anon_sym_UNLINK] = ACTIONS(14), + [anon_sym_PURGE] = ACTIONS(14), + [anon_sym_LOCK] = ACTIONS(14), + [anon_sym_UNLOCK] = ACTIONS(14), + [anon_sym_PROPFIND] = ACTIONS(14), + [anon_sym_VIEW] = ACTIONS(14), + }, +}; + +static const uint16_t ts_small_parse_table[] = { + [0] = 1, + ACTIONS(16), 1, + sym__url, + [4] = 1, + ACTIONS(18), 1, + ts_builtin_sym_end, + [8] = 1, + ACTIONS(20), 1, + sym__url, +}; + +static const uint32_t ts_small_parse_table_map[] = { + [SMALL_STATE(5)] = 0, + [SMALL_STATE(6)] = 4, + [SMALL_STATE(7)] = 8, +}; + +static const TSParseActionEntry ts_parse_actions[] = { + [0] = {.entry = {.count = 0, .reusable = false}}, + [1] = {.entry = {.count = 1, .reusable = false}}, RECOVER(), + [3] = {.entry = {.count = 1, .reusable = true}}, REDUCE(sym_source_file, 0), + [5] = {.entry = {.count = 1, .reusable = true}}, SHIFT(5), + [7] = {.entry = {.count = 1, .reusable = true}}, REDUCE(sym_source_file, 1), + [9] = {.entry = {.count = 1, .reusable = true}}, REDUCE(aux_sym_source_file_repeat1, 2), + [11] = {.entry = {.count = 2, .reusable = true}}, REDUCE(aux_sym_source_file_repeat1, 2), SHIFT_REPEAT(5), + [14] = {.entry = {.count = 1, .reusable = true}}, REDUCE(sym_request_declaration, 2, .production_id = 1), + [16] = {.entry = {.count = 1, .reusable = true}}, REDUCE(sym_request_literal, 1), + [18] = {.entry = {.count = 1, .reusable = true}}, ACCEPT_INPUT(), + [20] = {.entry = {.count = 1, .reusable = true}}, SHIFT(4), +}; + +#ifdef __cplusplus +extern "C" { +#endif +#ifdef _WIN32 +#define extern __declspec(dllexport) +#endif + +extern const TSLanguage *tree_sitter_hurl(void) { + static const TSLanguage language = { + .version = LANGUAGE_VERSION, + .symbol_count = SYMBOL_COUNT, + .alias_count = ALIAS_COUNT, + .token_count = TOKEN_COUNT, + .external_token_count = EXTERNAL_TOKEN_COUNT, + .state_count = STATE_COUNT, + .large_state_count = LARGE_STATE_COUNT, + .production_id_count = PRODUCTION_ID_COUNT, + .field_count = FIELD_COUNT, + .max_alias_sequence_length = MAX_ALIAS_SEQUENCE_LENGTH, + .parse_table = &ts_parse_table[0][0], + .small_parse_table = ts_small_parse_table, + .small_parse_table_map = ts_small_parse_table_map, + .parse_actions = ts_parse_actions, + .symbol_names = ts_symbol_names, + .field_names = ts_field_names, + .field_map_slices = ts_field_map_slices, + .field_map_entries = ts_field_map_entries, + .symbol_metadata = ts_symbol_metadata, + .public_symbol_map = ts_symbol_map, + .alias_map = ts_non_terminal_alias_map, + .alias_sequences = &ts_alias_sequences[0][0], + .lex_modes = ts_lex_modes, + .lex_fn = ts_lex, + .primary_state_ids = ts_primary_state_ids, + }; + return &language; +} +#ifdef __cplusplus +} +#endif diff --git a/src/tree_sitter/parser.h b/src/tree_sitter/parser.h new file mode 100644 index 0000000..2b14ac1 --- /dev/null +++ b/src/tree_sitter/parser.h @@ -0,0 +1,224 @@ +#ifndef TREE_SITTER_PARSER_H_ +#define TREE_SITTER_PARSER_H_ + +#ifdef __cplusplus +extern "C" { +#endif + +#include +#include +#include + +#define ts_builtin_sym_error ((TSSymbol)-1) +#define ts_builtin_sym_end 0 +#define TREE_SITTER_SERIALIZATION_BUFFER_SIZE 1024 + +typedef uint16_t TSStateId; + +#ifndef TREE_SITTER_API_H_ +typedef uint16_t TSSymbol; +typedef uint16_t TSFieldId; +typedef struct TSLanguage TSLanguage; +#endif + +typedef struct { + TSFieldId field_id; + uint8_t child_index; + bool inherited; +} TSFieldMapEntry; + +typedef struct { + uint16_t index; + uint16_t length; +} TSFieldMapSlice; + +typedef struct { + bool visible; + bool named; + bool supertype; +} TSSymbolMetadata; + +typedef struct TSLexer TSLexer; + +struct TSLexer { + int32_t lookahead; + TSSymbol result_symbol; + void (*advance)(TSLexer *, bool); + void (*mark_end)(TSLexer *); + uint32_t (*get_column)(TSLexer *); + bool (*is_at_included_range_start)(const TSLexer *); + bool (*eof)(const TSLexer *); +}; + +typedef enum { + TSParseActionTypeShift, + TSParseActionTypeReduce, + TSParseActionTypeAccept, + TSParseActionTypeRecover, +} TSParseActionType; + +typedef union { + struct { + uint8_t type; + TSStateId state; + bool extra; + bool repetition; + } shift; + struct { + uint8_t type; + uint8_t child_count; + TSSymbol symbol; + int16_t dynamic_precedence; + uint16_t production_id; + } reduce; + uint8_t type; +} TSParseAction; + +typedef struct { + uint16_t lex_state; + uint16_t external_lex_state; +} TSLexMode; + +typedef union { + TSParseAction action; + struct { + uint8_t count; + bool reusable; + } entry; +} TSParseActionEntry; + +struct TSLanguage { + uint32_t version; + uint32_t symbol_count; + uint32_t alias_count; + uint32_t token_count; + uint32_t external_token_count; + uint32_t state_count; + uint32_t large_state_count; + uint32_t production_id_count; + uint32_t field_count; + uint16_t max_alias_sequence_length; + const uint16_t *parse_table; + const uint16_t *small_parse_table; + const uint32_t *small_parse_table_map; + const TSParseActionEntry *parse_actions; + const char * const *symbol_names; + const char * const *field_names; + const TSFieldMapSlice *field_map_slices; + const TSFieldMapEntry *field_map_entries; + const TSSymbolMetadata *symbol_metadata; + const TSSymbol *public_symbol_map; + const uint16_t *alias_map; + const TSSymbol *alias_sequences; + const TSLexMode *lex_modes; + bool (*lex_fn)(TSLexer *, TSStateId); + bool (*keyword_lex_fn)(TSLexer *, TSStateId); + TSSymbol keyword_capture_token; + struct { + const bool *states; + const TSSymbol *symbol_map; + void *(*create)(void); + void (*destroy)(void *); + bool (*scan)(void *, TSLexer *, const bool *symbol_whitelist); + unsigned (*serialize)(void *, char *); + void (*deserialize)(void *, const char *, unsigned); + } external_scanner; + const TSStateId *primary_state_ids; +}; + +/* + * Lexer Macros + */ + +#define START_LEXER() \ + bool result = false; \ + bool skip = false; \ + bool eof = false; \ + int32_t lookahead; \ + goto start; \ + next_state: \ + lexer->advance(lexer, skip); \ + start: \ + skip = false; \ + lookahead = lexer->lookahead; + +#define ADVANCE(state_value) \ + { \ + state = state_value; \ + goto next_state; \ + } + +#define SKIP(state_value) \ + { \ + skip = true; \ + state = state_value; \ + goto next_state; \ + } + +#define ACCEPT_TOKEN(symbol_value) \ + result = true; \ + lexer->result_symbol = symbol_value; \ + lexer->mark_end(lexer); + +#define END_STATE() return result; + +/* + * Parse Table Macros + */ + +#define SMALL_STATE(id) id - LARGE_STATE_COUNT + +#define STATE(id) id + +#define ACTIONS(id) id + +#define SHIFT(state_value) \ + {{ \ + .shift = { \ + .type = TSParseActionTypeShift, \ + .state = state_value \ + } \ + }} + +#define SHIFT_REPEAT(state_value) \ + {{ \ + .shift = { \ + .type = TSParseActionTypeShift, \ + .state = state_value, \ + .repetition = true \ + } \ + }} + +#define SHIFT_EXTRA() \ + {{ \ + .shift = { \ + .type = TSParseActionTypeShift, \ + .extra = true \ + } \ + }} + +#define REDUCE(symbol_val, child_count_val, ...) \ + {{ \ + .reduce = { \ + .type = TSParseActionTypeReduce, \ + .symbol = symbol_val, \ + .child_count = child_count_val, \ + __VA_ARGS__ \ + }, \ + }} + +#define RECOVER() \ + {{ \ + .type = TSParseActionTypeRecover \ + }} + +#define ACCEPT_INPUT() \ + {{ \ + .type = TSParseActionTypeAccept \ + }} + +#ifdef __cplusplus +} +#endif + +#endif // TREE_SITTER_PARSER_H_