feat: closes GH-50 - @rewrite blocks (#106)

* feat: GH-50 @rewrite blocks * remove failing test and put in GH ticket
kylegoetz · Nov 7, 2024 · 920c8e1 · 920c8e1
1 parent e18194c
commit 920c8e1
Show file tree

Hide file tree

Showing 15 changed files with 210,171 additions and 156,619 deletions.
diff --git a/.gitignore b/.gitignore
@@ -1,25 +1,38 @@
 !*/
 node_modules
+.jj
 *.log
 log
 log.html
 .unisonHistory
+.gitattributes
 
 # OS-specific
 .DS_Store
 
 # IDE-specific
 .vscode
+.editorconfig
 
 # app specific
+Makefile
+*.swift
+pyproject.toml
 src/*
 !src/tree_sitter
 !src/tree_sitter/parser.h
 scratch.u
 *.wasm
+*.py
+bindings/c
+bindings/go
+bindings/python
+bindings/swift
 target/
 build/
 Cargo.lock
+setup.py
+test.html
 
 !grammar.json
 !parser.c

diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -2,19 +2,29 @@
 
 All notable changes to this project will be documented in this file.
 
-## [2.0.0] - 2024-10-29
+## [2.0.0] - 2024-11-06
 
 ### New
 
-- #86 - add `force`, in which `x()` is parsed as `(force (regular_identifier) (unit))`
+- GH-50 - term rewrite blocks
+- GH-85 - a built-in hash is a single unit, no longer `*PATH_SEGMENT . REGULAR_ID`
+- GH-86 - add `force`, in which `x()` is parsed as `(force (regular_identifier) (unit))`
+- GH-90 - liberal recognition of `_layout_semicolon` at ends of lines with equal indentation levels, which helps with parsing in many places
+- GH-91 - infix symops can have namespace/path prefix
 
 ### Fixed
 
-- #82 - allow doc block wherever
-- #84 - Symbols in paths now work
-- #85 - built-ins are now just `built_in_hash` without any child nodes
+- GH-80 - parsing folds of more than three hyphens
+- GH-84 - Symbols in paths now work
+- GH-85 - built-ins are now just `built_in_hash` without any child nodes
+- GH-92 - test watch expressions can only have `test` and `test.io` as prefix for `>`
+- GH-95 - doc blocks can only appear two ways (syntax sugar before a top-level `binding` and in similar places as a `literal`)
+- GH-96 - enable built-in hashes
+- GH-97 - separate a term `binding` from a top-level `binding` (latter has the doc block syntax sugar)
+- GH-98 - +4 is correctly tokenized as `int` rather than `symboly_id . nat`
 - #?? - pattern matching guards now are part of a layouted block
-
+- GH-103 - `match` scrutinee now correctly categorized as `_term` rather than `_expression`
+- GH-105 - two-char symbols ending in `&` or `|` now parse as two-char symbols
 
 ### Changed
 

diff --git a/grammar/blocks.js b/grammar/blocks.js
@@ -28,4 +28,36 @@ module.exports = {
 
   // handle [block] with [block]
   handler: ($) => seq($.handle, $._block, $.with, $._block),
+
+  rewrite_block: $ => seq(
+    $.rewrite,
+    $._layout_start,
+    seq($._rewrite_type, repeat(seq($._layout_semicolon, $._rewrite_type))),
+    $._layout_end),
+
+  _rewrite_type: $ => choice(
+    $.rewrite_term,
+    $.rewrite_case,
+    $.rewrite_type,
+  ),
+  _rewrite_term_like: $ => seq(
+    $._term,
+    $.rewrite_arrow,
+    $._layout_start,
+    $.__layout_block
+  ),
+
+  rewrite_term: $ => seq($.term, $._rewrite_term_like),
+  rewrite_case: $ => seq($.case, $._rewrite_term_like),
+  rewrite_type: $ => seq(
+    $.signature,
+    optional(seq(
+      repeat1($._prefix_definition_name),
+      alias('.', $.dot))),
+    $._computation_type,
+    $.rewrite_arrow,
+    $._layout_start,
+    $._computation_type,
+    $._layout_end,
+  ),
 };
diff --git a/grammar/conflicts.js b/grammar/conflicts.js
@@ -15,4 +15,5 @@ module.exports = ($) => [
   [$._literal, $._link],
   [$._infix_app_or_boolean_op],
   [$._pattern_constructor, $.effect_bind],
+  [$._identifier, $._wordy_definition_name],
 ];
diff --git a/grammar/delayed-computation.js b/grammar/delayed-computation.js
@@ -39,6 +39,7 @@ module.exports = {
     choice(
       $.exp_let,
       $.handler,
+      $.rewrite_block,
       $.exp_if,
       $._pattern_matching, // this is match and lam case
     ),

diff --git a/grammar/reserved.js b/grammar/reserved.js
@@ -22,4 +22,9 @@ module.exports = {
   with: ($) => "with",
   open_parens: $ => prec.right('('),
   close_parens: $ => ')',
+  rewrite: $ => '@rewrite',
+  term: $ => 'term',
+  case: $ => 'case',
+  signature: $ => 'signature',
+  rewrite_arrow: $ => '==>',
 };
diff --git a/src/grammar.json b/src/grammar.json
@@ -1,5 +1,4 @@
 {
-  "$schema": "https://tree-sitter.github.io/tree-sitter/assets/schemas/grammar.schema.json",
   "name": "unison",
   "rules": {
     "unison": {
@@ -2240,6 +2239,10 @@
           "type": "SYMBOL",
           "name": "handler"
         },
+        {
+          "type": "SYMBOL",
+          "name": "rewrite_block"
+        },
         {
           "type": "SYMBOL",
           "name": "exp_if"
@@ -2686,6 +2689,170 @@
         }
       ]
     },
+    "rewrite_block": {
+      "type": "SEQ",
+      "members": [
+        {
+          "type": "SYMBOL",
+          "name": "rewrite"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "_layout_start"
+        },
+        {
+          "type": "SEQ",
+          "members": [
+            {
+              "type": "SYMBOL",
+              "name": "_rewrite_type"
+            },
+            {
+              "type": "REPEAT",
+              "content": {
+                "type": "SEQ",
+                "members": [
+                  {
+                    "type": "SYMBOL",
+                    "name": "_layout_semicolon"
+                  },
+                  {
+                    "type": "SYMBOL",
+                    "name": "_rewrite_type"
+                  }
+                ]
+              }
+            }
+          ]
+        },
+        {
+          "type": "SYMBOL",
+          "name": "_layout_end"
+        }
+      ]
+    },
+    "_rewrite_type": {
+      "type": "CHOICE",
+      "members": [
+        {
+          "type": "SYMBOL",
+          "name": "rewrite_term"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "rewrite_case"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "rewrite_type"
+        }
+      ]
+    },
+    "_rewrite_term_like": {
+      "type": "SEQ",
+      "members": [
+        {
+          "type": "SYMBOL",
+          "name": "_term"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "rewrite_arrow"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "_layout_start"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "__layout_block"
+        }
+      ]
+    },
+    "rewrite_term": {
+      "type": "SEQ",
+      "members": [
+        {
+          "type": "SYMBOL",
+          "name": "term"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "_rewrite_term_like"
+        }
+      ]
+    },
+    "rewrite_case": {
+      "type": "SEQ",
+      "members": [
+        {
+          "type": "SYMBOL",
+          "name": "case"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "_rewrite_term_like"
+        }
+      ]
+    },
+    "rewrite_type": {
+      "type": "SEQ",
+      "members": [
+        {
+          "type": "SYMBOL",
+          "name": "signature"
+        },
+        {
+          "type": "CHOICE",
+          "members": [
+            {
+              "type": "SEQ",
+              "members": [
+                {
+                  "type": "REPEAT1",
+                  "content": {
+                    "type": "SYMBOL",
+                    "name": "_prefix_definition_name"
+                  }
+                },
+                {
+                  "type": "ALIAS",
+                  "content": {
+                    "type": "STRING",
+                    "value": "."
+                  },
+                  "named": true,
+                  "value": "dot"
+                }
+              ]
+            },
+            {
+              "type": "BLANK"
+            }
+          ]
+        },
+        {
+          "type": "SYMBOL",
+          "name": "_computation_type"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "rewrite_arrow"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "_layout_start"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "_computation_type"
+        },
+        {
+          "type": "SYMBOL",
+          "name": "_layout_end"
+        }
+      ]
+    },
     "type_signature": {
       "type": "SEQ",
       "members": [
@@ -3882,6 +4049,26 @@
       "type": "STRING",
       "value": ")"
     },
+    "rewrite": {
+      "type": "STRING",
+      "value": "@rewrite"
+    },
+    "term": {
+      "type": "STRING",
+      "value": "term"
+    },
+    "case": {
+      "type": "STRING",
+      "value": "case"
+    },
+    "signature": {
+      "type": "STRING",
+      "value": "signature"
+    },
+    "rewrite_arrow": {
+      "type": "STRING",
+      "value": "==>"
+    },
     "wordy_id": {
       "type": "PREC_RIGHT",
       "value": 0,
@@ -4491,6 +4678,10 @@
     [
       "_pattern_constructor",
       "effect_bind"
+    ],
+    [
+      "_identifier",
+      "_wordy_definition_name"
     ]
   ],
   "precedences": [