Merge pull request #77 from semgrep/martin/upgrade-tree-sitter

Upgrade to tree-sitter 0.22.6

patch/tree-sitter-0.22.6/Makefile.patch

@@ -0,0 +1,53 @@
+--- ../../downloads/tree-sitter-0.22.6/Makefile.orig	2024-05-05 13:47:18.000000000 -0700
++++ ../../downloads/tree-sitter-0.22.6/Makefile	2024-06-18 19:52:35.772310737 -0700
+@@ -8,12 +8,12 @@
+
+ # collect sources
+ ifneq ($(AMALGAMATED),1)
+-	SRC := $(wildcard lib/src/*.c)
+-	# do not double-include amalgamation
+-	SRC := $(filter-out lib/src/lib.c,$(SRC))
++  SRC := $(wildcard lib/src/*.c)
++  # do not double-include amalgamation
++  SRC := $(filter-out lib/src/lib.c,$(SRC))
+ else
+-	# use amalgamated build
+-	SRC := lib/src/lib.c
++  # use amalgamated build
++  SRC := lib/src/lib.c
+ endif
+ OBJ := $(SRC:.c=.o)
+
+@@ -28,21 +28,20 @@
+ SONAME_MINOR := $(word 2,$(subst ., ,$(VERSION)))
+
+ # OS-specific bits
+-ifeq ($(OS),Windows_NT)
+-	$(error "Windows is not supported")
+-else ifeq ($(shell uname),Darwin)
+-	SOEXT = dylib
+-	SOEXTVER_MAJOR = $(SONAME_MAJOR).dylib
+-	SOEXTVER = $(SONAME_MAJOR).$(SONAME_MINOR).dylib
+-	LINKSHARED += -dynamiclib -Wl,-install_name,$(LIBDIR)/libtree-sitter.$(SONAME_MAJOR).dylib
++
++ifeq ($(shell uname),Darwin)
++  SOEXT = dylib
++  SOEXTVER_MAJOR = $(SONAME_MAJOR).dylib
++  SOEXTVER = $(SONAME_MAJOR).$(SONAME_MINOR).dylib
++  LINKSHARED += -dynamiclib -Wl,-install_name,$(LIBDIR)/libtree-sitter.$(SONAME_MAJOR).dylib
+ else
+-	SOEXT = so
+-	SOEXTVER_MAJOR = so.$(SONAME_MAJOR)
+-	SOEXTVER = so.$(SONAME_MAJOR).$(SONAME_MINOR)
+-	LINKSHARED += -shared -Wl,-soname,libtree-sitter.so.$(SONAME_MAJOR)
++  SOEXT = so
++  SOEXTVER_MAJOR = so.$(SONAME_MAJOR)
++  SOEXTVER = so.$(SONAME_MAJOR).$(SONAME_MINOR)
++  LINKSHARED += -shared -Wl,-soname,libtree-sitter.so.$(SONAME_MAJOR)
+ endif
+ ifneq ($(filter $(shell uname),FreeBSD NetBSD DragonFly),)
+-	PCLIBDIR := $(PREFIX)/libdata/pkgconfig
++  PCLIBDIR := $(PREFIX)/libdata/pkgconfig
+ endif
+
+ all: libtree-sitter.a libtree-sitter.$(SOEXT) tree-sitter.pc

scripts/download-tree-sitter

@@ -9,7 +9,7 @@ prog_name=$(basename "$0")
 # The official version of tree-sitter we use for the code generator and
 # for the runtime library. Please try to keep this as the single source
 # of truth.
-default_version="0.20.6"
+default_version="0.22.6"
 
 error() {
   echo "Current directory: $(pwd)" >&2
@@ -87,5 +87,23 @@ EOF
     if [[ ! -d "$src_dir" ]]; then
       error "Archive didn't unpack into the expected folder '$src_dir'."
     fi
+
+    case "$version" in
+      0.20.6)
+      ;;
+      0.22.6)
+        # See https://github.com/tree-sitter/tree-sitter/pull/3417
+        # regarding Windows/Cygwin support
+        patch --backup \
+          tree-sitter-0.22.6/Makefile \
+          ../patch/tree-sitter-0.22.6/Makefile.patch
+        ;;
+      *)
+        cat >&2 <<EOF
+*** tree-sitter version: $version
+*** Warning: tree-sitter 0.22.6 required patching the Makefile!
+*** Remove this code once we're fully migrated to a version > 0.22.6
+EOF
+    esac
   fi
 )

scripts/ocaml-tree-sitter-gen-c

@@ -42,6 +42,16 @@ else
   # This mechanism is used with languages whose base grammar is defined
   # in a tree-sitter-* submodule.
 
+  # tree-sitter 0.22.6 creates a dummy 'grammar.js' file if
+  # there isn't one already. This creates confusing situations
+  # because our input grammar as created by ocaml-tree-sitter is
+  # 'src/grammar.json', not the usual 'grammar.js'. Creating our
+  # own unparseable 'grammar.js' is a protection against using
+  # the wrong grammar.
+  # Tracked at https://github.com/tree-sitter/tree-sitter/issues/3415
+  echo "..... not the input grammar! Use 'tree-sitter generate src/grammar.json'" \
+    > grammar.js
+
   orig_grammar_json="$import_from"/src/grammar.json
   if [[ ! -e "$orig_grammar_json" ]]; then
     cat >&2 <<EOF