From 039b671973536a7f6f146f0842e58a956c601c59 Mon Sep 17 00:00:00 2001
From: Joseph Ferano <joseph@ferano.io>
Date: Sat, 11 Nov 2023 15:15:47 +0700
Subject: [PATCH] Playing around with treesitter

---
 .gitignore    |  1 +
 fide.py       | 26 ++++++++++++++++++++++--
 fide.todo     |  7 ++++++-
 treesitter.py | 56 +++++++++++++++++++++++++++++++++++++++++++++++++++
 4 files changed, 87 insertions(+), 3 deletions(-)
 create mode 100644 treesitter.py

diff --git a/.gitignore b/.gitignore
index c9e0e9c..3ded613 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,3 +1,4 @@
 /cosmic-text/
 /__pycache__/
 /nim-version/fide
+/build/
diff --git a/fide.py b/fide.py
index 5529287..59f90b5 100755
--- a/fide.py
+++ b/fide.py
@@ -155,8 +155,30 @@ def main():
                 #     console.runsource(src[3:])
 
         for i,curr_line in enumerate(current_buffer.lines):
-            render = text_renderer.render(curr_line.piece.get_text(), True, 'black')
-            screen.blit(render, (0, i * line_height))
+            line_syntax = syntax_highlights.get_line(i)
+            for grammar in line_syntax:
+                if grammar.name == 'string':
+                    color = 'green'
+                else:
+                    color = 'black'
+                # Does this even make sense? Do we even need a piece
+                # table anymore? Why can't we just use a parsed
+                # grammar and iterate that and print out the text?
+
+                # I think it makes sense to just grab the string from
+                # the treesitter grammar What we need to do is iterate
+                # over each child, keeping track of our position As we
+                # iterate, we start at row 0, col 0, add any
+                # whitespace needed so get the first token. If we were
+                # to do it that way, then we have to generate the
+                # whitespace and draw it as well.
+
+                # The thing I'm not clear on is what is the best way
+                # to identify a token, because we need to tell when to
+                # actually write the thing out
+                string = curr_line.piece.string_at(grammar.start, grammar.end - grammar.start)
+                render = text_renderer.render(string, True, color)
+                screen.blit(render, (0, i * line_height))
 
         if cursor_on and cursor_flash < 0.5:
             cursor_draw(cursor)
diff --git a/fide.todo b/fide.todo
index 15e4b42..09b5135 100644
--- a/fide.todo
+++ b/fide.todo
@@ -10,7 +10,12 @@
 * DONE Keymap fallthrough/precedence stacking
 * DONE Pass the file you want to edit as a command line argument
 * DONE Save buffer function with a keybinding
-* TODO Add reverse parsing of keymaps to show users what they're typing
+* DONE Install, Import and play around with Treesitter
+* TODO Change the color of string literals
+* TODO Incorporate color regions into the buffer/line objects
+* TODO Create a simple color map based on treesitter keywords
+* TODO Walk the AST to generate the color regions
+* TODO Update treesitter AST as the user edits the file
 * TODO Add reverse parsing of keymaps to show users what they're typing
 
 
diff --git a/treesitter.py b/treesitter.py
new file mode 100644
index 0000000..73e2aa6
--- /dev/null
+++ b/treesitter.py
@@ -0,0 +1,56 @@
+from tree_sitter import Language, Parser, Node
+from typing import Tuple, List
+from dataclasses import dataclass
+
+# Language.build_library(
+#   # Store the library in the `build` directory
+#   'build/compiled-languages.so',
+
+#   # Include one or more languages
+#   [
+#     '/home/joe/Repositories/tree-sitter/languages/tree-sitter-cpp',
+#     '/home/joe/Repositories/tree-sitter/languages/tree-sitter-c',
+#     '/home/joe/Repositories/tree-sitter/languages/tree-sitter-rust',
+#     '/home/joe/Repositories/tree-sitter/languages/tree-sitter-python',
+#     # 'vendor/tree-sitter-javascript',
+#     # 'vendor/tree-sitter-python'
+#   ]
+# )
+C_LANGUAGE = Language('build/compiled-languages.so', 'c')
+CPP_LANGUAGE = Language('build/compiled-languages.so', 'cpp')
+PYTHON_LANGUAGE = Language('build/compiled-languages.so', 'python')
+# PY_LANGUAGE = Language('build/my-languages.so', 'python')
+
+cpp_parser = Parser()
+cpp_parser.set_language(CPP_LANGUAGE)
+python_parser = Parser()
+python_parser.set_language(PYTHON_LANGUAGE)
+
+with open('fide.py', 'r') as f:
+    text = f.read()
+    # tree = cpp_parser.parse(bytes(text, 'utf-8'))
+    tree = python_parser.parse(bytes(text, 'utf-8'))
+    root: Node = tree.root_node
+
+    def dfs(node: Node, indent_level):
+        indent = "".join([' ' * (indent_level * 4)])
+        # if node.grammar_name != node.type:
+        #     print(node)
+        print(indent, '-', f"({node.type})", # node.range,
+              node.start_point, node.end_point, node.text)
+        for n in node.children:
+            dfs(n, indent_level + 1)
+    dfs(root, 0)
+
+@dataclass
+class ParsedRange:
+    name: str
+    line_num: int
+    crange: Tuple[int, int]
+    
+def parse_file(text: str) -> List[ParsedRange]:
+    lines = [] * len()
+
+
+
+