Merge pull request #85 from oraios/pr/84

Better fallbacks for missing keys (selectionRange etc). Download clang executable if needed

src/multilspy/language_server.py

@@ -821,8 +821,10 @@ class LanguageServer:
 
         def turn_item_into_symbol_with_children(item: GenericDocumentSymbol):
             item = cast(multilspy_types.UnifiedSymbolInformation, item)
+            absolute_path = os.path.join(self.repository_root_path, relative_file_path)
+            
+            # handle missing entries in location
             if "location" not in item:
-                absolute_path = os.path.join(self.repository_root_path, relative_file_path)
                 uri = pathlib.Path(absolute_path).as_uri()
                 assert "range" in item
                 tree_location = multilspy_types.Location(
@@ -832,8 +834,19 @@ class LanguageServer:
                     relativePath=relative_file_path,
                 )
                 item['location'] = tree_location
+            location = item["location"]
+            if "absolutePath" not in location:
+                location["absolutePath"] = absolute_path
+            if "relativePath" not in location:
+                location["relativePath"] = relative_file_path
             if include_body:
                 item['body'] = self.retrieve_symbol_body(item)
+            # handle missing selectionRange
+            if "selectionRange" not in item:
+                if "range" in item:
+                    item["selectionRange"] = item["range"]
+                else:
+                    item["selectionRange"] = item["location"]["range"]
             item[LSPConstants.CHILDREN] = item.get(LSPConstants.CHILDREN, [])
 
         flat_all_symbol_list: List[multilspy_types.UnifiedSymbolInformation] = []
@@ -1044,15 +1057,20 @@ class LanguageServer:
         Returns the list of tuples (name, kind, line, column) of all top-level symbols in the file.
         """
         _, document_roots = await self.request_document_symbols(relative_file_path)
-        return [
-            (
-                root["name"],
-                root["kind"],
-                root["selectionRange"]["start"]["line"], # type: ignore
-                root["selectionRange"]["start"]["character"] # type: ignore
-            )
-            for root in document_roots
-        ]
+        result = []
+        for root in document_roots:
+            try:
+                result.append(
+                   ( root["name"],
+                    root["kind"],
+                    root["selectionRange"]["start"]["line"],
+                    root["selectionRange"]["start"]["character"],)
+                )
+            except KeyError as e:
+                raise KeyError(
+                    f"Could not process symbol of name {root.get('name', 'unknown')} in {relative_file_path=}"
+                ) from e
+        return result
 
     async def request_overview(self, within_relative_path: str) -> dict[str, list[tuple[str, multilspy_types.SymbolKind, int, int]]]:
         """

src/multilspy/language_servers/clangd_language_server/clangd_language_server.py

@@ -64,13 +64,19 @@ class ClangdLanguageServer(LanguageServer):
 
         clangd_ls_dir = os.path.join(os.path.dirname(__file__), "static/clangd")
         clangd_executable_path = os.path.join(clangd_ls_dir, "clangd_19.1.2", "bin", dependency["binaryName"])
-        if not os.path.exists(clangd_ls_dir):
-            os.makedirs(clangd_ls_dir)
+        if not os.path.exists(clangd_executable_path):
+            clangd_url = dependency["url"]
+            logger.log(f"Clangd executable not found at {clangd_executable_path}. Downloading from {clangd_url}", logging.INFO)
+            os.makedirs(clangd_ls_dir, exist_ok=True)
             if dependency["archiveType"] == "zip":
                 FileUtils.download_and_extract_archive(
-                    logger, dependency["url"], clangd_ls_dir, dependency["archiveType"]
+                    logger, clangd_url, clangd_ls_dir, dependency["archiveType"]
                 )
-        assert os.path.exists(clangd_executable_path)
+        if not os.path.exists(clangd_executable_path):
+            raise FileNotFoundError(
+                f"Clangd executable not found at {clangd_executable_path}.\n"
+                "Make sure you have installed clangd. See https://clangd.llvm.org/installation"
+            )
         os.chmod(clangd_executable_path, stat.S_IEXEC)
 
         return clangd_executable_path

src/multilspy/multilspy_utils.py

@@ -117,7 +117,10 @@ class FileUtils:
         """
         Reads the file at the given path and returns the contents as a string.
         """
-        encodings = ["utf-8-sig", "utf-16"]
+        if not os.path.exists(file_path):
+            logger.log(f"File read '{file_path}' failed: File does not exist.", logging.ERROR)
+            raise MultilspyException(f"File read '{file_path}' failed: File does not exist.")
+        encodings = ["utf-8-sig", "utf-16", "utf-8", "latin-1"]
         try:
             for encoding in encodings:
                 try:
@@ -125,6 +128,9 @@ class FileUtils:
                         return inp_file.read()
                 except UnicodeError:
                     continue
+            # Try system default encoding as a last resort
+            with open(file_path, "r") as inp_file:
+                return inp_file.read()
         except Exception as exc:
             logger.log(f"File read '{file_path}' failed: {exc}", logging.ERROR)
             raise MultilspyException("File read failed.") from None

src/serena/agent.py

@@ -73,7 +73,10 @@ class ProjectConfig(ToStringMixin):
 
     def __init__(self, config_dict: dict[str, Any], project_name: str, project_root: Path | None = None):
         self.project_name: str = project_name
-        self.language: Language = Language(config_dict["language"])
+        try:
+            self.language: Language = Language(config_dict["language"].lower())
+        except ValueError as e:
+            raise ValueError(f"Invalid language: {config_dict['language']}.\nValid languages are: {[l.value for l in Language]}") from e
         if project_root is None:
             project_root = Path(config_dict["project_root"])
         self.project_root: str = str(project_root.resolve())