Use tree-sitter 0.20

* Handle words containing bare '#'

Only a word beginning with a '#' starts a comment. A word can contain
'#' character without escaping as long as it is not the first character.

See: Bash Reference Manual section '3.1.3 Comments'
https://www.gnu.org/software/bash/manual/bash.html#Comments

* Regenerate

Cargo.toml

@@ -1,7 +1,7 @@
 [package]
 name = "tree-sitter-bash"
 description = "bash grammar for the tree-sitter parsing library"
-version = "0.19.0"
+version = "0.20.0"
 keywords = ["incremental", "parsing", "bash"]
 categories = ["parsing", "text-editors"]
 repository = "https://github.com/tree-sitter/tree-sitter-bash"
@@ -23,7 +23,7 @@ include = [
 path = "bindings/rust/lib.rs"
 
 [dependencies]
-tree-sitter = "0.19"
+tree-sitter = "0.20"
 
 [build-dependencies]
 cc = "1.0"

bindings/rust/lib.rs

@@ -35,7 +35,7 @@ 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 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");

corpus/literals.txt

@@ -355,3 +355,18 @@ echo -ne "\033k$1\033\\" > /dev/stderr
   (redirected_statement
     (command (command_name (word)) (word) (string (simple_expansion (variable_name))))
     (file_redirect (word))))
+
+================================================================================
+Words containing bare '#'
+================================================================================
+
+curl -# localhost #comment without space
+nix build nixpkgs#hello -v # comment with space
+
+--------------------------------------------------------------------------------
+
+(program
+  (command (command_name (word)) (word) (word))
+  (comment)
+  (command (command_name (word)) (word) (word) (word))
+  (comment))

corpus/zsh.txt

@@ -0,0 +1,35 @@
+# See https://zsh.sourceforge.io/Doc/Release/Expansion.html#Parameter-Expansion-Flags
+=============================
+Parameter Expansion Flags
+=============================
+echo ${(v)V}
+echo ${(s.:.)V}
+echo ${(@)V}
+echo ${()V}
+
+---
+
+(program
+  (command (command_name (word)) (expansion (expansion_flags) (variable_name)))
+  (command (command_name (word)) (expansion (expansion_flags) (variable_name)))
+  (command (command_name (word)) (expansion (expansion_flags) (variable_name)))
+  (command (command_name (word)) (expansion (expansion_flags) (variable_name))))
+
+=============================
+Parameter Expansion Flags Quotes
+=============================
+echo "${(v_sldkfj_sdklfj)V}"
+
+---
+
+(program
+  (command (command_name (word)) (string (expansion (expansion_flags) (variable_name)))))
+=============================
+Parameter Expansion Invalid Flags
+=============================
+echo "${(())V}"
+
+---
+
+(program
+(command (command_name (word)) (string (expansion (ERROR) (expansion_flags) (ERROR (word))))))

grammar.js

@@ -8,7 +8,6 @@ const SPECIAL_CHARACTERS = [
   '|', '&', ';',
   '\\',
   '\\s',
-  '#',
 ];
 
 module.exports = grammar({
@@ -491,8 +490,12 @@ module.exports = grammar({
 
     string_expansion: $ => seq('$', choice($.string, $.raw_string)),
 
+    // See https://zsh.sourceforge.io/Doc/Release/Expansion.html#Parameter-Expansion-Flags
+    expansion_flags: ($) => seq("(", repeat(/[^()]/), ")"),
+
     expansion: $ => seq(
       '${',
+      optional($.expansion_flags),
       optional(choice('#', '!')),
       optional(choice(
         seq(
@@ -537,10 +540,16 @@ module.exports = grammar({
 
     _special_variable_name: $ => alias(choice('*', '@', '?', '-', '$', '0', '_'), $.special_variable_name),
 
-    word: $ => token(repeat1(choice(
+    word: $ => token(seq(
+      choice(
+        noneOf('#', ...SPECIAL_CHARACTERS),
+        seq('\\', noneOf('\\s'))
+      ),
+      repeat(choice(
         noneOf(...SPECIAL_CHARACTERS),
         seq('\\', noneOf('\\s'))
-    ))),
+      ))
+    )),
 
     test_operator: $ => token(prec(1, seq('-', /[a-zA-Z]+/))),
 

package.json

@@ -30,7 +30,8 @@
       "scope": "source.bash",
       "file-types": [
         "sh",
-        "bash"
+        "bash",
+        "zsh"
       ]
     }
   ]

queries/highlights.scm

@@ -46,6 +46,7 @@
   ">>"
   "<"
   "|"
+  (expansion_flags)
 ] @operator
 
 (

src/grammar.json

@@ -431,11 +431,20 @@
     },
     "while_statement": {
       "type": "SEQ",
+      "members": [
+        {
+          "type": "CHOICE",
           "members": [
             {
               "type": "STRING",
               "value": "while"
             },
+            {
+              "type": "STRING",
+              "value": "until"
+            }
+          ]
+        },
         {
           "type": "FIELD",
           "name": "condition",
@@ -2144,6 +2153,26 @@
         }
       ]
     },
+    "expansion_flags": {
+      "type": "SEQ",
+      "members": [
+        {
+          "type": "STRING",
+          "value": "("
+        },
+        {
+          "type": "REPEAT",
+          "content": {
+            "type": "PATTERN",
+            "value": "[^()]"
+          }
+        },
+        {
+          "type": "STRING",
+          "value": ")"
+        }
+      ]
+    },
     "expansion": {
       "type": "SEQ",
       "members": [
@@ -2151,6 +2180,18 @@
           "type": "STRING",
           "value": "${"
         },
+        {
+          "type": "CHOICE",
+          "members": [
+            {
+              "type": "SYMBOL",
+              "name": "expansion_flags"
+            },
+            {
+              "type": "BLANK"
+            }
+          ]
+        },
         {
           "type": "CHOICE",
           "members": [
@@ -2461,13 +2502,38 @@
     "word": {
       "type": "TOKEN",
       "content": {
-        "type": "REPEAT1",
+        "type": "SEQ",
+        "members": [
+          {
+            "type": "CHOICE",
+            "members": [
+              {
+                "type": "PATTERN",
+                "value": "[^#'\"<>{}\\[\\]()`$|&;\\\\\\s]"
+              },
+              {
+                "type": "SEQ",
+                "members": [
+                  {
+                    "type": "STRING",
+                    "value": "\\"
+                  },
+                  {
+                    "type": "PATTERN",
+                    "value": "[^\\s]"
+                  }
+                ]
+              }
+            ]
+          },
+          {
+            "type": "REPEAT",
             "content": {
               "type": "CHOICE",
               "members": [
                 {
                   "type": "PATTERN",
-              "value": "[^'\"<>{}\\[\\]()`$|&;\\\\\\s#]"
+                  "value": "[^'\"<>{}\\[\\]()`$|&;\\\\\\s]"
                 },
                 {
                   "type": "SEQ",
@@ -2485,6 +2551,8 @@
               ]
             }
           }
+        ]
+      }
     },
     "test_operator": {
       "type": "TOKEN",

src/node-types.json

@@ -638,6 +638,10 @@
           "type": "concatenation",
           "named": true
         },
+        {
+          "type": "expansion_flags",
+          "named": true
+        },
         {
           "type": "regex",
           "named": true
@@ -657,6 +661,11 @@
       ]
     }
   },
+  {
+    "type": "expansion_flags",
+    "named": true,
+    "fields": {}
+  },
   {
     "type": "file_redirect",
     "named": true,
@@ -1639,6 +1648,10 @@
     "type": "unsetenv",
     "named": false
   },
+  {
+    "type": "until",
+    "named": false
+  },
   {
     "type": "variable_name",
     "named": true