From 60afe4fd53d4b2d0aab2e452022334f47affb5c0 Mon Sep 17 00:00:00 2001
From: Alex Carney <alcarneyme@gmail.com>
Date: Sat, 5 Oct 2024 21:50:45 +0100
Subject: [PATCH] fix: return ``None`` for non ``file:`` URIs

---
 docs/source/howto/migrate-to-v2.rst |  5 +++++
 docs/source/reference/uris.rst      | 16 ++++++++++++++++
 pygls/uris.py                       | 10 +++++++---
 pygls/workspace/text_document.py    | 18 ++++++++++--------
 pygls/workspace/workspace.py        | 16 +++++++++-------
 5 files changed, 47 insertions(+), 18 deletions(-)
 create mode 100644 docs/source/reference/uris.rst

diff --git a/docs/source/howto/migrate-to-v2.rst b/docs/source/howto/migrate-to-v2.rst
index 182d9fb9..7a4b7de8 100644
--- a/docs/source/howto/migrate-to-v2.rst
+++ b/docs/source/howto/migrate-to-v2.rst
@@ -25,6 +25,11 @@ Python Support
 
 *pygls v2* removes support for Python 3.8 and adds support for Python 3.13 (with the GIL, you are welcome to try the free-threaded version just note that it has not been tested yet!)
 
+URI Handling
+------------
+
+The :func:`pygls.uris.to_fs_path` will now return ``None`` for URIs that do not have a ``file:`` scheme.
+
 
 Removed Deprecated Functions
 ----------------------------
diff --git a/docs/source/reference/uris.rst b/docs/source/reference/uris.rst
new file mode 100644
index 00000000..98f94966
--- /dev/null
+++ b/docs/source/reference/uris.rst
@@ -0,0 +1,16 @@
+URIs
+====
+
+.. currentmodule:: pygls.uris
+
+.. autofunction:: from_fs_path
+
+.. autofunction:: to_fs_path
+
+.. autofunction:: uri_scheme
+
+.. autofunction:: uri_with
+
+.. autofunction:: urlparse
+
+.. autofunction:: urlunparse
diff --git a/pygls/uris.py b/pygls/uris.py
index 8c40f70b..b0c34f1d 100644
--- a/pygls/uris.py
+++ b/pygls/uris.py
@@ -21,6 +21,8 @@
 
 https://github.com/Microsoft/vscode-uri/blob/e59cab84f5df6265aed18ae5f43552d3eef13bb9/lib/index.ts
 """
+from __future__ import annotations
+
 from typing import Optional, Tuple
 
 import re
@@ -75,20 +77,22 @@ def from_fs_path(path: str):
         return None
 
 
-def to_fs_path(uri: str):
+def to_fs_path(uri: str) -> str | None:
     """
     Returns the filesystem path of the given URI.
 
     Will handle UNC paths and normalize windows drive letters to lower-case.
     Also uses the platform specific path separator. Will *not* validate the
     path for invalid characters and semantics.
-    Will *not* look at the scheme of this URI.
     """
     try:
         # scheme://netloc/path;parameters?query#fragment
         scheme, netloc, path, _, _, _ = urlparse(uri)
 
-        if netloc and path and scheme == "file":
+        if scheme != "file":
+            return None
+
+        if netloc and path:
             # unc path: file://shares/c$/far/boo
             value = f"//{netloc}{path}"
 
diff --git a/pygls/workspace/text_document.py b/pygls/workspace/text_document.py
index 7beb6ca4..023569a3 100644
--- a/pygls/workspace/text_document.py
+++ b/pygls/workspace/text_document.py
@@ -19,12 +19,13 @@
 import io
 import logging
 import os
+import pathlib
 import re
 from typing import List, Optional, Pattern
 
 from lsprotocol import types
 
-from pygls.uris import to_fs_path
+from pygls.uris import urlparse, to_fs_path
 from .position_codec import PositionCodec
 
 # TODO: this is not the best e.g. we capture numbers
@@ -47,9 +48,10 @@ def __init__(
     ):
         self.uri = uri
         self.version = version
-        path = to_fs_path(uri)
-        if path is None:
-            raise Exception("`path` cannot be None")
+
+        if (path := to_fs_path(uri)) is None:
+            _, _, path, *_ = urlparse(uri)
+
         self.path = path
         self.language_id = language_id
         self.filename: Optional[str] = os.path.basename(self.path)
@@ -177,10 +179,10 @@ def offset_at_position(self, client_position: types.Position) -> int:
 
     @property
     def source(self) -> str:
-        if self._source is None:
-            with io.open(self.path, "r", encoding="utf-8") as f:
-                return f.read()
-        return self._source
+        if self._source is None and self.path is not None:
+            return pathlib.Path(self.path).read_text(encoding="utf-8")
+
+        return self._source or ""
 
     def word_at_position(
         self,
diff --git a/pygls/workspace/workspace.py b/pygls/workspace/workspace.py
index 0bd68e5b..246af1d2 100644
--- a/pygls/workspace/workspace.py
+++ b/pygls/workspace/workspace.py
@@ -47,10 +47,7 @@ def __init__(
         self._root_uri = root_uri
         if self._root_uri is not None:
             self._root_uri_scheme = uri_scheme(self._root_uri)
-            root_path = to_fs_path(self._root_uri)
-            if root_path is None:
-                raise Exception("Couldn't get `root_path` from `root_uri`")
-            self._root_path = root_path
+            self._root_path = to_fs_path(self._root_uri)
         else:
             self._root_path = None
         self._sync_kind = sync_kind
@@ -151,9 +148,14 @@ def get_text_document(self, doc_uri: str) -> TextDocument:
         return self._text_documents.get(doc_uri) or self._create_text_document(doc_uri)
 
     def is_local(self):
-        return (
-            self._root_uri_scheme == "" or self._root_uri_scheme == "file"
-        ) and os.path.exists(self._root_path)
+
+        if self._root_uri_scheme not in {"", "file"}:
+            return False
+
+        if (path := self._root_path) is None:
+            return False
+
+        return os.path.exists(path)
 
     def put_notebook_document(self, params: types.DidOpenNotebookDocumentParams):
         notebook = params.notebook_document