Update rizin-shell-parser to tree-sitter 0.19.4 (#928)

.prettierignore

@@ -0,0 +1 @@
+shlr/rizin-shell-parser/bindings/node/index.js

librz/core/cmd.c

@@ -6091,9 +6091,15 @@ static RzCmdStatus core_cmd_tsr2cmd(RzCore *core, const char *cstr, bool split_l
 	ts_symbols_init(core->rcmd);
 
 	TSParser *parser = ts_parser_new();
-	ts_parser_set_language(parser, (TSLanguage *)core->rcmd->language);
+	bool language_ok = ts_parser_set_language(parser, (TSLanguage *)core->rcmd->language);
+	rz_return_val_if_fail(language_ok, RZ_CMD_STATUS_INVALID);
 
 	TSTree *tree = ts_parser_parse_string(parser, NULL, input, strlen(input));
+	if (!tree) {
+		rz_warn_if_reached();
+		return RZ_CMD_STATUS_INVALID;
+	}
+
 	TSNode root = ts_tree_root_node(tree);
 
 	RzCmdStatus res = RZ_CMD_STATUS_INVALID;

meson.build

@@ -7,6 +7,7 @@ pkgconfig_mod = import('pkgconfig')
 # Python scripts used during the build process
 create_tags_rz_py = files('sys/create_tags_rz.py')
 syscall_preprocessing_py = files('sys/syscall_preprocessing.py')
+tree_sitter_wrap_py = files('sys/meson_tree_sitter_generate.py')
 
 # Get rizin version
 rizin_version = 'unknown-error'

shlr/meson.build

@@ -212,25 +212,7 @@ endif
 
 
 # new rizin shell parser
-shell_parser_path = 'rizin-shell-parser'
-shell_parser_files = [
-  join_paths(shell_parser_path, 'src/parser.c'),
-  join_paths(shell_parser_path, 'src/scanner.c'),
-]
-
-shell_parser_inc = [platform_inc, include_directories('rizin-shell-parser/src/tree_sitter')]
-
-libshell_parser = static_library('shell_parser', shell_parser_files,
-  include_directories: shell_parser_inc,
-  dependencies: tree_sitter_dep.partial_dependency(includes: true),
-  implicit_include_directories: true
-)
-
-shell_parser_dep = declare_dependency(
-  link_with: libshell_parser,
-  include_directories: shell_parser_inc,
-  dependencies: tree_sitter_dep
-)
+subdir('rizin-shell-parser')
 
 
 # handle bochs dependency

shlr/packagefiles/tree-sitter-0.19.4/meson.build

@@ -0,0 +1,29 @@
+project('tree-sitter', 'c')
+
+cc = meson.get_compiler('c')
+
+tree_sitter_cflags = ''
+if cc.has_argument('-std=gnu99')
+  tree_sitter_cflags = '-std=gnu99'
+elif cc.has_argument('-std=c99')
+  tree_sitter_cflags = '-std=c99'
+endif
+
+tree_sitter_path = 'tree-sitter'
+
+tree_sitter_files = ['lib/src/lib.c']
+
+tree_sitter_inc = [include_directories('lib/src'), include_directories('lib/include')]
+
+libtree_sitter = static_library('tree_sitter', tree_sitter_files,
+  include_directories: tree_sitter_inc,
+  implicit_include_directories: false,
+  c_args: tree_sitter_cflags,
+  install: not meson.is_subproject()
+)
+
+tree_sitter_dep = declare_dependency(
+  compile_args: tree_sitter_cflags,
+  link_with: libtree_sitter,
+  include_directories: tree_sitter_inc
+)

shlr/rizin-shell-parser/Cargo.toml

@@ -0,0 +1,26 @@
+[package]
+name = "tree-sitter-rzcmd"
+description = "rzcmd grammar for the tree-sitter parsing library"
+version = "0.0.1"
+keywords = ["incremental", "parsing", "rzcmd"]
+categories = ["parsing", "text-editors"]
+repository = "https://github.com/tree-sitter/tree-sitter-javascript"
+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 = "0.17"
+
+[build-dependencies]
+cc = "1.0"

shlr/rizin-shell-parser/binding.gyp

@@ -9,7 +9,7 @@
       "sources": [
         "src/parser.c",
         "src/scanner.c",
-        "src/binding.cc"
+        "bindings/node/binding.cc"
       ],
       "cflags_c": [
         "-std=c99 -ggdb -O0",

shlr/rizin-shell-parser/bindings/node/binding.cc

@@ -19,7 +19,7 @@ void Init(Local<Object> exports, Local<Object> module) {
   Local<Object> instance = constructor->NewInstance(Nan::GetCurrentContext()).ToLocalChecked();
   Nan::SetInternalFieldPointer(instance, 0, tree_sitter_rzcmd());
 
-  Nan::Set(instance, Nan::New("name").ToLocalChecked(), Nan::New("r2cmd").ToLocalChecked());
+  Nan::Set(instance, Nan::New("name").ToLocalChecked(), Nan::New("rzcmd").ToLocalChecked());
   Nan::Set(module, Nan::New("exports").ToLocalChecked(), instance);
 }
 

shlr/rizin-shell-parser/bindings/node/index.js

@@ -0,0 +1,19 @@
+try {
+  module.exports = require("../../build/Release/tree_sitter_rzcmd_binding");
+} catch (error1) {
+  if (error1.code !== 'MODULE_NOT_FOUND') {
+    throw error1;
+  }
+  try {
+    module.exports = require("../../build/Debug/tree_sitter_rzcmd_binding");
+  } catch (error2) {
+    if (error2.code !== 'MODULE_NOT_FOUND') {
+      throw error2;
+    }
+    throw error1
+  }
+}
+
+try {
+  module.exports.nodeTypeInfo = require("../../src/node-types.json");
+} catch (_) {}

shlr/rizin-shell-parser/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());
+    */
+}

shlr/rizin-shell-parser/bindings/rust/lib.rs

@@ -0,0 +1,52 @@
+//! This crate provides rzcmd 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_rzcmd::language()).expect("Error loading rzcmd 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_rzcmd() -> 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_rzcmd() }
+}
+
+/// 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 rzcmd language");
+    }
+}

shlr/rizin-shell-parser/index.js

@@ -1,13 +0,0 @@
-try {
-  module.exports = require("./build/Release/tree_sitter_rzcmd_binding");
-} catch (error) {
-  try {
-    module.exports = require("./build/Debug/tree_sitter_rzcmd_binding");
-  } catch (_) {
-    throw error
-  }
-}
-
-try {
-  module.exports.nodeTypeInfo = require("./src/node-types.json");
-} catch (_) {}

shlr/rizin-shell-parser/meson.build

@@ -0,0 +1,15 @@
+grammar_js = files('grammar.js')
+
+subdir('src')
+
+libshell_parser = static_library('shell_parser', shell_parser_files,
+  include_directories: shell_parser_inc,
+  dependencies: tree_sitter_dep.partial_dependency(includes: true),
+  implicit_include_directories: true
+)
+
+shell_parser_dep = declare_dependency(
+  link_with: libshell_parser,
+  include_directories: shell_parser_inc,
+  dependencies: tree_sitter_dep
+)

shlr/rizin-shell-parser/package-lock.json

@@ -1,5 +1,5 @@
 {
-  "name": "tree-sitter-r2cmd",
+  "name": "tree-sitter-rzcmd",
   "version": "1.0.0",
   "lockfileVersion": 1,
   "requires": true,
@@ -10,9 +10,9 @@
       "integrity": "sha512-INOFj37C7k3AfaNTtX8RhsTw7qRy7eLET14cROi9+5HAVbbHuIWUHEauBv5qT4Av2tWasiTY1Jw6puUNqRJXQg=="
     },
     "tree-sitter-cli": {
-      "version": "0.16.8",
-      "resolved": "https://registry.npmjs.org/tree-sitter-cli/-/tree-sitter-cli-0.16.8.tgz",
-      "integrity": "sha512-gMVZ0tpSCk1DQoyrSXlRaFZ4HztyVTKqYBTXitYKHmM+mrEIznDGV70ycqsCQYDjCnE461naMViGOItDGJDIHw==",
+      "version": "0.19.4",
+      "resolved": "https://registry.npmjs.org/tree-sitter-cli/-/tree-sitter-cli-0.19.4.tgz",
+      "integrity": "sha512-p2kxjuoQeauXBu5eE+j7c5BMCRXmc17EiAswnnWn3ieUlHXBkA0Z7vRnaCSElDR34MhZnSgqgzuuzQk0cDqCjw==",
       "dev": true
     }
   }

shlr/rizin-shell-parser/package.json

@@ -1,8 +1,8 @@
 {
-  "name": "tree-sitter-r2cmd",
+  "name": "tree-sitter-rzcmd",
   "version": "1.0.0",
   "description": "Tree-Sitter grammar for parsing rizin commands",
-  "main": "index.js",
+  "main": "bindings/node",
   "scripts": {
     "test": "echo \"Error: no test specified\" && exit 1"
   },
@@ -12,6 +12,6 @@
     "nan": "^2.14.0"
   },
   "devDependencies": {
-    "tree-sitter-cli": "^0.16.8"
+    "tree-sitter-cli": "^0.19.4"
   }
 }

shlr/rizin-shell-parser/src/grammar.json

@@ -3766,6 +3766,7 @@
     }
   ],
   "conflicts": [],
+  "precedences": [],
   "externals": [
     {
       "type": "SYMBOL",

shlr/rizin-shell-parser/src/meson.build

@@ -0,0 +1,14 @@
+tree_sitter_bin = find_program('tree-sitter', required: false)
+node_bin = find_program('node', required: false)
+if tree_sitter_bin.found() and node_bin.found() and tree_sitter_dep.type_name() != 'internal'
+  parser_c = custom_target('parser_src_c',
+    command: [tree_sitter_wrap_py, tree_sitter_bin, '@OUTDIR@/..', '@INPUT@'],
+    input: [grammar_js],
+    output: 'parser.c',
+  )
+else
+  parser_c = files('parser.c')
+endif
+
+shell_parser_files = [files('scanner.c'), parser_c]
+shell_parser_inc = [platform_inc, include_directories('tree_sitter')]

shlr/rizin-shell-parser/src/node-types.json

@@ -323,7 +323,7 @@
     "fields": {},
     "children": {
       "multiple": true,
-      "required": true,
+      "required": false,
       "types": [
         {
           "type": "arged_command",
@@ -6492,13 +6492,7 @@
   {
     "type": "legacy_quoted_command",
     "named": true,
-    "fields": {
-      "string": {
-        "multiple": false,
-        "required": false,
-        "types": []
-      }
-    }
+    "fields": {}
   },
   {
     "type": "macro_args",

shlr/rizin-shell-parser/src/tree_sitter/parser.h

@@ -13,6 +13,8 @@ extern "C" {
 #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;
@@ -30,11 +32,10 @@ typedef struct {
   uint16_t length;
 } TSFieldMapSlice;
 
-typedef uint16_t TSStateId;
-
 typedef struct {
-  bool visible : 1;
-  bool named : 1;
+  bool visible;
+  bool named;
+  bool supertype;
 } TSSymbolMetadata;
 
 typedef struct TSLexer TSLexer;
@@ -56,21 +57,21 @@ typedef enum {
   TSParseActionTypeRecover,
 } TSParseActionType;
 
-typedef struct {
-  union {
-    struct {
-      TSStateId state;
-      bool extra : 1;
-      bool repetition : 1;
-    } shift;
-    struct {
-      TSSymbol symbol;
-      int16_t dynamic_precedence;
-      uint8_t child_count;
-      uint8_t production_id;
-    } reduce;
-  } params;
-  TSParseActionType type : 4;
+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 {
@@ -82,7 +83,7 @@ typedef union {
   TSParseAction action;
   struct {
     uint8_t count;
-    bool reusable : 1;
+    bool reusable;
   } entry;
 } TSParseActionEntry;
 
@@ -92,13 +93,24 @@ struct TSLanguage {
   uint32_t alias_count;
   uint32_t token_count;
   uint32_t external_token_count;
-  const char **symbol_names;
-  const TSSymbolMetadata *symbol_metadata;
-  const uint16_t *parse_table;
-  const TSParseActionEntry *parse_actions;
-  const TSLexMode *lex_modes;
-  const TSSymbol *alias_sequences;
+  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 **symbol_names;
+  const char **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;
@@ -111,14 +123,6 @@ struct TSLanguage {
     unsigned (*serialize)(void *, char *);
     void (*deserialize)(void *, const char *, unsigned);
   } external_scanner;
-  uint32_t field_count;
-  const TSFieldMapSlice *field_map_slices;
-  const TSFieldMapEntry *field_map_entries;
-  const char **field_names;
-  uint32_t large_state_count;
-  const uint16_t *small_parse_table;
-  const uint32_t *small_parse_table_map;
-  const TSSymbol *public_symbol_map;
 };
 
 /*
@@ -167,66 +171,50 @@ struct TSLanguage {
 
 #define ACTIONS(id) id
 
-#define SHIFT(state_value)                \
-  {                                       \
-    {                                     \
-      .params = {                         \
-        .shift = {                        \
-          .state = state_value            \
-        }                                 \
-      },                                  \
-      .type = TSParseActionTypeShift      \
-    }                                     \
-  }
+#define SHIFT(state_value)            \
+  {{                                  \
+    .shift = {                        \
+      .type = TSParseActionTypeShift, \
+      .state = state_value            \
+    }                                 \
+  }}
 
 #define SHIFT_REPEAT(state_value)     \
-  {                                   \
-    {                                 \
-      .params = {                     \
-        .shift = {                    \
-          .state = state_value,       \
-          .repetition = true          \
-        }                             \
-      },                              \
-      .type = TSParseActionTypeShift  \
+  {{                                  \
+    .shift = {                        \
+      .type = TSParseActionTypeShift, \
+      .state = state_value,           \
+      .repetition = true              \
     }                                 \
-  }
-
-#define RECOVER()                        \
-  {                                      \
-    { .type = TSParseActionTypeRecover } \
-  }
+  }}
 
 #define SHIFT_EXTRA()                 \
-  {                                   \
-    {                                 \
-      .params = {                     \
-        .shift = {                    \
-          .extra = true               \
-        }                             \
-      },                              \
-      .type = TSParseActionTypeShift  \
+  {{                                  \
+    .shift = {                        \
+      .type = TSParseActionTypeShift, \
+      .extra = true                   \
     }                                 \
-  }
+  }}
 
 #define REDUCE(symbol_val, child_count_val, ...) \
-  {                                              \
-    {                                            \
-      .params = {                                \
-        .reduce = {                              \
-          .symbol = symbol_val,                  \
-          .child_count = child_count_val,        \
-          __VA_ARGS__                            \
-        },                                       \
-      },                                         \
-      .type = TSParseActionTypeReduce            \
-    }                                            \
-  }
+  {{                                             \
+    .reduce = {                                  \
+      .type = TSParseActionTypeReduce,           \
+      .symbol = symbol_val,                      \
+      .child_count = child_count_val,            \
+      __VA_ARGS__                                \
+    },                                           \
+  }}
 
-#define ACCEPT_INPUT()                  \
-  {                                     \
-    { .type = TSParseActionTypeAccept } \
-  }
+#define RECOVER()                    \
+  {{                                 \
+    .type = TSParseActionTypeRecover \
+  }}
+
+#define ACCEPT_INPUT()              \
+  {{                                \
+    .type = TSParseActionTypeAccept \
+  }}
 
 #ifdef __cplusplus
 }

shlr/tree-sitter

@@ -1 +1 @@
-Subproject commit 1dc127e5dae7550a58a09886a19fdea19196bcf1
+Subproject commit 56c7c6b39d908c2df059e2c7f75860f819010671

sys/meson_tree_sitter_generate.py

@@ -0,0 +1,41 @@
+#!/usr/bin/env python
+#
+# SPDX-FileCopyrightText: 2021 ret2libc <sirmy15@gmail.com>
+# SPDX-License-Identifier: LGPL-3.0-only
+
+""" Python script to execute tree-sitter generate in the right output folder,
+necessary because tree-sitter generate does not allow to specify it """
+
+import os
+import subprocess
+import sys
+
+
+def parse():
+    if len(sys.argv) <= 3:
+        print(
+            "Usage: {} <tree-sitter-path> <output-dir> <grammar.js>".format(sys.argv[0])
+        )
+        sys.exit(1)
+
+    tree_sitter_exe = sys.argv[1]
+    output_dir = sys.argv[2]
+    grammar_js = sys.argv[3]
+
+    return tree_sitter_exe, output_dir, grammar_js
+
+
+def main():
+    tree_sitter_exe, output_dir, grammar_js = parse()
+
+    grammar_js = os.path.abspath(grammar_js)
+    output_dir = os.path.abspath(output_dir)
+    tree_sitter_exe = os.path.abspath(tree_sitter_exe)
+
+    subprocess.run(
+        [tree_sitter_exe, "generate", grammar_js], check=True, cwd=output_dir
+    )
+
+
+if __name__ == "__main__":
+    main()