From d3193afc2559e7d84ed2d76664a650dc03b4c6ef Mon Sep 17 00:00:00 2001
From: Riley Bruins <ribru17@hotmail.com>
Date: Thu, 19 Sep 2024 13:08:22 -0700
Subject: fix(treesitter): remove duplicate symbol names in language.inspect()

**Problems:**

- `vim.treesitter.language.inspect()` returns duplicate
  symbol names, sometimes up to 6 of one kind in the case of `markdown`
- The list-like `symbols` table can have holes and is thus not even a
  valid msgpack table anyway, mentioned in a test

**Solution:** Return symbols as a map, rather than a list, where field
names are the names of the symbol. The boolean value associated with the
field encodes whether or not the symbol is named.

Note that anonymous nodes are surrounded with double quotes (`"`) to
prevent potential collisions with named counterparts that have the same
identifier.
---
 test/functional/treesitter/language_spec.lua | 24 +++++++++++-------------
 test/functional/treesitter/parser_spec.lua   |  4 ++--
 2 files changed, 13 insertions(+), 15 deletions(-)

(limited to 'test')

diff --git a/test/functional/treesitter/language_spec.lua b/test/functional/treesitter/language_spec.lua
index e1e34fcecc..633a2dc725 100644
--- a/test/functional/treesitter/language_spec.lua
+++ b/test/functional/treesitter/language_spec.lua
@@ -51,7 +51,7 @@ describe('treesitter language API', function()
   it('inspects language', function()
     local keys, fields, symbols = unpack(exec_lua(function()
       local lang = vim.treesitter.language.inspect('c')
-      local keys, symbols = {}, {}
+      local keys = {}
       for k, v in pairs(lang) do
         if type(v) == 'boolean' then
           keys[k] = v
@@ -60,12 +60,7 @@ describe('treesitter language API', function()
         end
       end
 
-      -- symbols array can have "holes" and is thus not a valid msgpack array
-      -- but we don't care about the numbers here (checked in the parser test)
-      for _, v in pairs(lang.symbols) do
-        table.insert(symbols, v)
-      end
-      return { keys, lang.fields, symbols }
+      return { keys, lang.fields, lang.symbols }
     end))
 
     eq({ fields = true, symbols = true, _abi_version = true, _wasm = false }, keys)
@@ -79,16 +74,19 @@ describe('treesitter language API', function()
     eq(true, fset['initializer'])
 
     local has_named, has_anonymous
-    for _, s in pairs(symbols) do
-      eq('string', type(s[1]))
-      eq('boolean', type(s[2]))
-      if s[1] == 'for_statement' and s[2] == true then
+    for symbol, named in pairs(symbols) do
+      eq('string', type(symbol))
+      eq('boolean', type(named))
+      if symbol == 'for_statement' and named == true then
         has_named = true
-      elseif s[1] == '|=' and s[2] == false then
+      elseif symbol == '"|="' and named == false then
         has_anonymous = true
       end
     end
-    eq({ true, true }, { has_named, has_anonymous })
+    eq(
+      { has_named = true, has_anonymous = true },
+      { has_named = has_named, has_anonymous = has_anonymous }
+    )
   end)
 
   it(
diff --git a/test/functional/treesitter/parser_spec.lua b/test/functional/treesitter/parser_spec.lua
index c8829f4785..2f8d204d36 100644
--- a/test/functional/treesitter/parser_spec.lua
+++ b/test/functional/treesitter/parser_spec.lua
@@ -42,13 +42,13 @@ describe('treesitter parser API', function()
     eq('function_definition', exec_lua('return child:type()'))
     eq(true, exec_lua('return child:named()'))
     eq('number', type(exec_lua('return child:symbol()')))
-    eq({ 'function_definition', true }, exec_lua('return lang.symbols[child:symbol()]'))
+    eq(true, exec_lua('return lang.symbols[child:type()]'))
 
     exec_lua('anon = root:descendant_for_range(0,8,0,9)')
     eq('(', exec_lua('return anon:type()'))
     eq(false, exec_lua('return anon:named()'))
     eq('number', type(exec_lua('return anon:symbol()')))
-    eq({ '(', false }, exec_lua('return lang.symbols[anon:symbol()]'))
+    eq(false, exec_lua([=[return lang.symbols[string.format('"%s"', anon:type())]]=]))
 
     exec_lua('descendant = root:descendant_for_range(1,2,1,12)')
     eq('<node declaration>', exec_lua('return tostring(descendant)'))
-- 
cgit