Files
nx_post_support/server/src/tools/navigation.py
T
Christoph 5b027a2717 feat(def_flow): recognize .def names propagated via variables and wrapper procs
Add a new def_flow analysis module that follows .def block template
and address names through local variables and proc parameters, and a
wrapper-table builder to resolve proc arguments that forward .def names.
Derived names are resolved only for hover/definition (not for rename).

Integrate this into the LSP:
- lsp_server: add _word_at and _tcl_def_symbol helpers; fallback to
  derived_def_symbol when direct NX-argument navigation fails for hover,
  goto-definition and references; return proper ranges for hover.
- lsp_tclserver: cache and expose a def_wrapper_table built from index
  def_flows (with cache invalidation on index generation).

Also add unit tests for def_flow and update CHANGELOG to note hover/
definition and completion preview improvements for derived names.
2026-09-25 08:14:59 +02:00

873 lines
29 KiB
Python

from __future__ import annotations
from dataclasses import dataclass
from pathlib import Path
import lsprotocol.types as lsp
from tclint.syntax_tree import Command, List, Node, QuotedWord, Script, VarSub
from tools.def_flow import ( # noqa: F401 (re-exported)
DEF_ADDRESS,
DEF_BLOCK_TEMPLATE,
DEF_SYMBOL_KINDS,
ProcDefFlows,
command_flow_facts,
def_argument_kinds,
proc_def_flows,
)
from tools.def_flow import def_name as _def_name
from tools.variable_names import array_key_parts, variable_name
ROOT_NAMESPACE = "::"
@dataclass(frozen=True)
class SymbolIdentity:
kind: str
name: str
scope: str | None = None
@dataclass(frozen=True)
class SymbolOccurrence:
identity: SymbolIdentity
range: lsp.Range
placeholder: str
is_definition: bool = False
symbol_kind: lsp.SymbolKind = lsp.SymbolKind.Variable
container_name: str | None = None
fallback_identity: SymbolIdentity | None = None
caller: SymbolIdentity | None = None
declaration_range: lsp.Range | None = None
array_element: str | None = None
array_parts: tuple[str | None, ...] = ()
array_template_parts: tuple[str | None, ...] = ()
@dataclass(frozen=True)
class FileSymbolIndex:
path: str
uri: str
occurrences: tuple[SymbolOccurrence, ...]
document_range: lsp.Range | None = None
# Procs whose parameters reach .def arguments: (qualified proc name, flows).
def_flows: tuple[tuple[str, ProcDefFlows], ...] = ()
@dataclass(frozen=True)
class _Scope:
filepath: str
namespace: str = ROOT_NAMESPACE
proc_name: str | None = None
global_variables: tuple[tuple[str, str], ...] = ()
namespace_variables: tuple[tuple[str, str], ...] = ()
def _without_array_index(name: str) -> str:
return name.split("(", 1)[0]
def _basename(name: str) -> str:
return _without_array_index(name).rsplit("::", 1)[-1]
def _qualify(name: str, namespace: str) -> str:
name = _without_array_index(name)
if name.startswith("::"):
return name
if namespace == ROOT_NAMESPACE:
return f"::{name}"
return f"{namespace}::{name}"
def _namespace_of(qualified_name: str) -> str:
parent = qualified_name.rsplit("::", 1)[0]
return parent or ROOT_NAMESPACE
def _display_name(identity: SymbolIdentity) -> str:
if identity.kind == "variable" and identity.scope is not None:
return identity.name
return identity.name.removeprefix("::")
def _container_name(identity: SymbolIdentity) -> str | None:
if identity.scope is not None:
_, _, proc_name = identity.scope.partition("::proc::")
return proc_name.removeprefix("::") or None
qualified = identity.name.removeprefix("::")
if "::" not in qualified:
return None
return qualified.rsplit("::", 1)[0]
def _static_contents(node: Node | None) -> str | None:
value = getattr(node, "contents", None)
return value if isinstance(value, str) else None
def _name_range(node: Node, raw_name: str, *, variable_sub: bool = False) -> lsp.Range:
if variable_sub:
line, column = node.pos
column += 2 if getattr(node, "braced", False) else 1
else:
position = getattr(node, "contents_pos", None) or node.pos
if isinstance(node, QuotedWord) and node.contents is None and node.children:
position = node.children[0].pos
line, column = position
normalized = _without_array_index(raw_name)
prefix_length = normalized.rfind("::") + 2 if "::" in normalized else 0
start_character = column - 1 + prefix_length
name = normalized[prefix_length:]
return lsp.Range(
start=lsp.Position(line=line - 1, character=start_character),
end=lsp.Position(line=line - 1, character=start_character + len(name)),
)
def _node_range(node: Node) -> lsp.Range | None:
if node.pos is None or node.end_pos is None:
return None
return lsp.Range(
start=lsp.Position(line=node.pos[0] - 1, character=node.pos[1] - 1),
end=lsp.Position(
line=node.end_pos[0] - 1,
character=node.end_pos[1] - 1,
),
)
def _proc_identity(raw_name: str, namespace: str) -> SymbolIdentity:
return SymbolIdentity(kind="proc", name=_qualify(raw_name, namespace))
def _proc_fallback(raw_name: str, namespace: str) -> SymbolIdentity | None:
if raw_name.startswith("::") or "::" in raw_name or namespace == ROOT_NAMESPACE:
return None
return SymbolIdentity(kind="proc", name=_qualify(raw_name, ROOT_NAMESPACE))
def _variable_identity(raw_name: str, scope: _Scope) -> SymbolIdentity:
normalized = _without_array_index(raw_name)
if normalized.startswith("::"):
return SymbolIdentity(kind="variable", name=normalized)
if "::" in normalized:
return SymbolIdentity(
kind="variable", name=_qualify(normalized, scope.namespace)
)
if scope.proc_name is None:
return SymbolIdentity(
kind="variable", name=_qualify(normalized, scope.namespace)
)
for alias, target in scope.global_variables + scope.namespace_variables:
if normalized == alias:
return SymbolIdentity(kind="variable", name=target)
return SymbolIdentity(
kind="variable",
name=normalized,
scope=f"{scope.filepath}::proc::{scope.proc_name}",
)
def _variable_command_nodes(command: Command) -> list[tuple[Node, bool]]:
routine = _static_contents(command.routine)
if routine == "set" and command.args:
return [(command.args[0], len(command.args) >= 2)]
if routine in {"incr", "append", "lappend"} and command.args:
return [(command.args[0], True)]
if routine == "lset" and command.args:
return [(command.args[0], False)]
if routine == "unset":
return [
(argument, False)
for argument in command.args
if not (_static_contents(argument) or "").startswith("-")
]
if routine == "array" and len(command.args) >= 2:
return [
(command.args[1], _static_contents(command.args[0]) == "set")
]
if routine == "dict" and len(command.args) >= 2:
subcommand = _static_contents(command.args[0])
if subcommand in {"set", "unset", "append", "incr", "lappend", "update", "with"}:
return [
(
command.args[1],
subcommand in {"set", "append", "incr", "lappend"},
)
]
return []
def _binding_nodes(node: Node) -> list[Node]:
if isinstance(node, List):
return list(node.children)
return [node]
def _variable_binding_nodes(command: Command) -> list[Node]:
routine = _static_contents(command.routine)
if routine in {"foreach", "lmap"} and len(command.args) >= 3:
return [
variable
for variable_list in command.args[:-1:2]
for variable in _binding_nodes(variable_list)
]
if routine == "lassign" and len(command.args) >= 2:
return list(command.args[1:])
if routine == "catch" and len(command.args) >= 2:
return list(command.args[1:3])
if (
routine == "dict"
and command.args
and _static_contents(command.args[0]) == "update"
):
return list(command.args[3:-1:2])
return []
def _variable_declaration_nodes(command: Command) -> list[Node]:
routine = _static_contents(command.routine)
if routine == "global":
return list(command.args)
if routine == "variable":
return list(command.args[::2])
return []
def _scan_proc_imports(
node: Node, namespace: str
) -> tuple[dict[str, str], dict[str, str]]:
global_variables: dict[str, str] = {}
namespace_variables: dict[str, str] = {}
def walk(current: Node) -> None:
if isinstance(current, Command):
routine = _static_contents(current.routine)
if routine == "proc":
return
if routine == "global":
for argument in current.args:
name = _static_contents(argument)
if name:
global_variables[_basename(name)] = _qualify(
name, ROOT_NAMESPACE
)
elif routine == "variable":
for argument in current.args[::2]:
name = _static_contents(argument)
if name:
namespace_variables[_basename(name)] = _qualify(
name, namespace
)
for child in getattr(current, "children", []):
walk(child)
walk(node)
return global_variables, namespace_variables
def build_file_symbol_index(
filepath: str, uri: str, tree: Node
) -> FileSymbolIndex:
occurrences: list[SymbolOccurrence] = []
# Most occurrences repeat a few identities; sharing one object per identity
# keeps the index (and its persistent cache) small.
identities: dict[SymbolIdentity, SymbolIdentity] = {}
def_flows: list[tuple[str, ProcDefFlows]] = []
# Flow facts of the procs being walked, innermost last.
flow_facts: list[list] = []
def shared(identity: SymbolIdentity | None) -> SymbolIdentity | None:
return None if identity is None else identities.setdefault(identity, identity)
def add_proc(
node: Node,
raw_name: str,
scope: _Scope,
*,
is_definition: bool,
declaration_range: lsp.Range | None = None,
) -> None:
identity = shared(_proc_identity(raw_name, scope.namespace))
caller = None
if not is_definition:
caller = (
SymbolIdentity(kind="proc", name=scope.proc_name)
if scope.proc_name is not None
else SymbolIdentity(kind="file", name=filepath)
)
occurrences.append(
SymbolOccurrence(
identity=identity,
fallback_identity=(
None
if is_definition
else shared(_proc_fallback(raw_name, scope.namespace))
),
range=_name_range(node, raw_name),
placeholder=_basename(raw_name),
is_definition=is_definition,
symbol_kind=lsp.SymbolKind.Function,
container_name=_container_name(identity),
caller=shared(caller),
declaration_range=declaration_range,
)
)
def add_variable(
node: Node,
raw_name: str,
scope: _Scope,
*,
is_definition: bool,
variable_sub: bool = False,
identity: SymbolIdentity | None = None,
) -> None:
symbol_identity = shared(identity or _variable_identity(raw_name, scope))
occurrences.append(
SymbolOccurrence(
identity=symbol_identity,
range=_name_range(node, raw_name, variable_sub=variable_sub),
placeholder=_basename(raw_name),
is_definition=is_definition,
symbol_kind=lsp.SymbolKind.Variable,
container_name=_container_name(symbol_identity),
array_element=(
raw_name.split("(", 1)[1][:-1]
if "(" in raw_name and raw_name.endswith(")")
else None
),
array_parts=array_key_parts(node),
array_template_parts=array_key_parts(node, preserve_variables=True),
)
)
def walk_embedded(node: Node, scope: _Scope) -> None:
if isinstance(node, Script):
walk_script(node, scope)
return
if isinstance(node, Command):
walk_command(node, scope)
return
if isinstance(node, VarSub):
raw_name = getattr(node, "value", None)
if isinstance(raw_name, str):
add_variable(
node,
raw_name,
scope,
is_definition=False,
variable_sub=True,
)
for child in getattr(node, "children", []):
walk_embedded(child, scope)
def walk_proc(command: Command, scope: _Scope) -> None:
if len(command.args) < 3:
return
raw_name = _static_contents(command.args[0])
body = command.args[2]
if raw_name is None or not isinstance(body, Script):
return
add_proc(
command.args[0],
raw_name,
scope,
is_definition=True,
declaration_range=_node_range(command),
)
proc_identity = _proc_identity(raw_name, scope.namespace)
proc_namespace = _namespace_of(proc_identity.name)
global_variables, namespace_variables = _scan_proc_imports(
body, proc_namespace
)
proc_scope = _Scope(
filepath=filepath,
namespace=proc_namespace,
proc_name=proc_identity.name,
global_variables=tuple(sorted(global_variables.items())),
namespace_variables=tuple(sorted(namespace_variables.items())),
)
parameters = command.args[1]
parameter_names = []
for parameter in getattr(parameters, "children", []):
parameter_node = parameter
if isinstance(parameter, List) and parameter.children:
parameter_node = parameter.children[0]
parameter_name = _static_contents(parameter_node)
parameter_names.append(parameter_name or "")
if parameter_name:
add_variable(
parameter_node,
parameter_name,
proc_scope,
is_definition=True,
)
flow_facts.append([])
walk_script(body, proc_scope)
facts = flow_facts.pop()
flows = proc_def_flows(facts, parameter_names, proc_namespace) if facts else ()
if flows:
def_flows.append((proc_identity.name, flows))
def walk_namespace(command: Command, scope: _Scope) -> bool:
if len(command.args) < 3 or _static_contents(command.args[0]) != "eval":
return False
raw_name = _static_contents(command.args[1])
body = command.args[2]
if raw_name is None or not isinstance(body, Script):
return False
namespace = _qualify(raw_name, scope.namespace)
identity = SymbolIdentity(kind="namespace", name=namespace)
occurrences.append(
SymbolOccurrence(
identity=identity,
range=_name_range(command.args[1], raw_name),
placeholder=_basename(raw_name),
is_definition=True,
symbol_kind=lsp.SymbolKind.Namespace,
container_name=_container_name(identity),
)
)
walk_script(
body,
_Scope(filepath=filepath, namespace=namespace),
)
return True
def walk_command(command: Command, scope: _Scope) -> None:
routine = _static_contents(command.routine)
if routine == "proc":
walk_proc(command, scope)
return
if routine == "namespace" and walk_namespace(command, scope):
return
if routine:
add_proc(command.routine, routine, scope, is_definition=False)
if flow_facts:
flow_facts[-1].extend(command_flow_facts(command))
for node, kind in def_argument_kinds(command):
name = _def_name(node)
if name is None:
continue
line, column = node.contents_pos
occurrences.append(
SymbolOccurrence(
identity=shared(SymbolIdentity(kind=kind, name=name)),
range=lsp.Range(
start=lsp.Position(line=line - 1, character=column - 1),
end=lsp.Position(line=line - 1, character=column - 1 + len(name)),
),
placeholder=name,
symbol_kind=lsp.SymbolKind.Struct if kind == DEF_BLOCK_TEMPLATE else lsp.SymbolKind.Field,
)
)
declaration_nodes = _variable_declaration_nodes(command)
declaration_ids = {id(node) for node in declaration_nodes}
for node in declaration_nodes:
raw_name = _static_contents(node)
if raw_name:
is_definition = routine == "variable" and scope.proc_name is None
if routine == "global":
identity = SymbolIdentity(
kind="variable",
name=_qualify(raw_name, ROOT_NAMESPACE),
)
else:
identity = SymbolIdentity(
kind="variable",
name=_qualify(raw_name, scope.namespace),
)
add_variable(
node,
raw_name,
scope,
is_definition=is_definition,
identity=identity,
)
for node, is_definition in _variable_command_nodes(command):
if id(node) in declaration_ids:
continue
raw_name = variable_name(node)
if raw_name:
add_variable(
node,
raw_name,
scope,
is_definition=is_definition,
)
for node in _variable_binding_nodes(command):
raw_name = _static_contents(node)
if raw_name:
add_variable(node, raw_name, scope, is_definition=True)
for argument in command.args:
walk_embedded(argument, scope)
def walk_script(script: Node, scope: _Scope) -> None:
for child in getattr(script, "children", []):
walk_embedded(child, scope)
walk_script(tree, _Scope(filepath=filepath))
return FileSymbolIndex(
path=filepath,
uri=uri,
occurrences=tuple(occurrences),
document_range=_node_range(tree),
def_flows=tuple(def_flows),
)
def definition_identities(indexes: dict[str, FileSymbolIndex]) -> set[SymbolIdentity]:
return {
occurrence.identity
for index in indexes.values()
for occurrence in index.occurrences
if occurrence.is_definition
}
def resolve_identity(
occurrence: SymbolOccurrence, definitions: set[SymbolIdentity]
) -> SymbolIdentity:
if occurrence.identity in definitions or occurrence.fallback_identity is None:
return occurrence.identity
if occurrence.fallback_identity in definitions:
return occurrence.fallback_identity
return occurrence.identity
def _may_resolve_to(occurrence: SymbolOccurrence, identity: SymbolIdentity) -> bool:
"""Cheap name pre-filter; resolve_identity only returns one of these two."""
name = identity.name
fallback = occurrence.fallback_identity
return occurrence.identity.name == name or (
fallback is not None and fallback.name == name
)
def symbol_at_position(
index: FileSymbolIndex,
position: lsp.Position,
definitions: set[SymbolIdentity],
) -> tuple[SymbolOccurrence, SymbolIdentity] | None:
for occurrence in index.occurrences:
start = occurrence.range.start
end = occurrence.range.end
if (
position.line == start.line == end.line
and start.character <= position.character < end.character
):
return occurrence, resolve_identity(occurrence, definitions)
return None
def matching_occurrences(
identity: SymbolIdentity,
indexes: dict[str, FileSymbolIndex],
definitions: set[SymbolIdentity],
) -> list[tuple[FileSymbolIndex, SymbolOccurrence]]:
matches = []
for index in indexes.values():
for occurrence in index.occurrences:
if not _may_resolve_to(occurrence, identity):
continue
if resolve_identity(occurrence, definitions) == identity:
matches.append((index, occurrence))
return matches
def document_highlights(
index: FileSymbolIndex,
identity: SymbolIdentity,
definitions: set[SymbolIdentity],
) -> list[lsp.DocumentHighlight]:
"""Return all occurrences of one symbol in the active document."""
highlights = []
for occurrence in index.occurrences:
if not _may_resolve_to(occurrence, identity):
continue
if resolve_identity(occurrence, definitions) != identity:
continue
kind = lsp.DocumentHighlightKind.Text
if identity.kind == "variable":
kind = (
lsp.DocumentHighlightKind.Write
if occurrence.is_definition
else lsp.DocumentHighlightKind.Read
)
highlights.append(lsp.DocumentHighlight(range=occurrence.range, kind=kind))
return sorted(
highlights,
key=lambda highlight: (
highlight.range.start.line,
highlight.range.start.character,
highlight.range.end.line,
highlight.range.end.character,
),
)
def workspace_symbols(
indexes: dict[str, FileSymbolIndex], query: str
) -> list[lsp.SymbolInformation]:
query = query.casefold()
results = []
seen = set()
for index in indexes.values():
for occurrence in index.occurrences:
identity = occurrence.identity
if not occurrence.is_definition:
continue
if identity.kind == "variable" and identity.scope is not None:
continue
name = _display_name(identity)
if query and query not in name.casefold():
continue
key = identity
if key in seen:
continue
seen.add(key)
results.append(
lsp.SymbolInformation(
name=name,
kind=occurrence.symbol_kind,
location=lsp.Location(uri=index.uri, range=occurrence.range),
container_name=occurrence.container_name,
)
)
return sorted(results, key=lambda symbol: symbol.name.casefold())
_CALL_HIERARCHY_DATA_KIND = "nx-post-support.call-hierarchy"
def _call_hierarchy_data(identity: SymbolIdentity) -> dict[str, str]:
return {
"source": _CALL_HIERARCHY_DATA_KIND,
"kind": identity.kind,
"name": identity.name,
}
def call_hierarchy_identity(item: lsp.CallHierarchyItem) -> SymbolIdentity | None:
"""Restore the symbol identity carried by a call hierarchy item."""
data = item.data
if (
not isinstance(data, dict)
or data.get("source") != _CALL_HIERARCHY_DATA_KIND
):
return None
kind = data.get("kind")
name = data.get("name")
if kind not in {"proc", "file"} or not isinstance(name, str):
return None
return SymbolIdentity(kind=kind, name=name)
def _proc_definitions(
indexes: dict[str, FileSymbolIndex],
) -> dict[SymbolIdentity, list[tuple[FileSymbolIndex, SymbolOccurrence]]]:
definitions: dict[
SymbolIdentity, list[tuple[FileSymbolIndex, SymbolOccurrence]]
] = {}
for index in indexes.values():
for occurrence in index.occurrences:
if occurrence.is_definition and occurrence.identity.kind == "proc":
definitions.setdefault(occurrence.identity, []).append(
(index, occurrence)
)
return definitions
def _unique_proc_definition(
identity: SymbolIdentity,
proc_definitions: dict[
SymbolIdentity, list[tuple[FileSymbolIndex, SymbolOccurrence]]
],
) -> tuple[FileSymbolIndex, SymbolOccurrence] | None:
matches = proc_definitions.get(identity, [])
if len(matches) != 1:
return None
return matches[0]
def _file_item(
identity: SymbolIdentity,
indexes: dict[str, FileSymbolIndex],
) -> lsp.CallHierarchyItem | None:
if identity.kind != "file":
return None
index = indexes.get(identity.name)
if index is None:
return None
range_ = index.document_range or lsp.Range(
start=lsp.Position(line=0, character=0),
end=lsp.Position(line=0, character=0),
)
selection_range = lsp.Range(start=range_.start, end=range_.start)
return lsp.CallHierarchyItem(
name=Path(index.path).name,
kind=lsp.SymbolKind.File,
uri=index.uri,
range=range_,
selection_range=selection_range,
detail=str(Path(index.path).parent),
data=_call_hierarchy_data(identity),
)
def _call_hierarchy_item(
identity: SymbolIdentity,
indexes: dict[str, FileSymbolIndex],
proc_definitions: dict[
SymbolIdentity, list[tuple[FileSymbolIndex, SymbolOccurrence]]
],
) -> lsp.CallHierarchyItem | None:
if identity.kind == "file":
return _file_item(identity, indexes)
definition = _unique_proc_definition(identity, proc_definitions)
if definition is None:
return None
index, occurrence = definition
basename = _basename(identity.name)
symbol_kind = (
lsp.SymbolKind.Event
if basename.startswith("MOM_")
else lsp.SymbolKind.Function
)
return lsp.CallHierarchyItem(
name=_display_name(identity),
kind=symbol_kind,
uri=index.uri,
range=occurrence.declaration_range or occurrence.range,
selection_range=occurrence.range,
detail=Path(index.path).name,
data=_call_hierarchy_data(identity),
)
def call_hierarchy_items(
identity: SymbolIdentity,
indexes: dict[str, FileSymbolIndex],
) -> list[lsp.CallHierarchyItem]:
"""Build the hierarchy item for one unambiguous workspace procedure."""
item = _call_hierarchy_item(identity, indexes, _proc_definitions(indexes))
return [item] if item is not None else []
def _range_key(range_: lsp.Range) -> tuple[int, int, int, int]:
return (
range_.start.line,
range_.start.character,
range_.end.line,
range_.end.character,
)
def _item_key(item: lsp.CallHierarchyItem) -> tuple[str, str, int, int]:
return (
item.name.casefold(),
item.uri,
item.selection_range.start.line,
item.selection_range.start.character,
)
def incoming_call_hierarchy(
identity: SymbolIdentity,
indexes: dict[str, FileSymbolIndex],
definitions: set[SymbolIdentity],
) -> list[lsp.CallHierarchyIncomingCall]:
"""Return statically resolved workspace procedures that call ``identity``."""
proc_definitions = _proc_definitions(indexes)
if _unique_proc_definition(identity, proc_definitions) is None:
return []
grouped: dict[SymbolIdentity, list[lsp.Range]] = {}
for _, occurrence in matching_occurrences(identity, indexes, definitions):
if occurrence.is_definition or occurrence.caller is None:
continue
caller = occurrence.caller
if caller.kind == "proc" and (
_unique_proc_definition(caller, proc_definitions) is None
):
continue
grouped.setdefault(caller, []).append(occurrence.range)
results = []
for caller, ranges in grouped.items():
item = _call_hierarchy_item(caller, indexes, proc_definitions)
if item is not None:
results.append(
lsp.CallHierarchyIncomingCall(
from_=item,
from_ranges=sorted(ranges, key=_range_key),
)
)
return sorted(results, key=lambda call: _item_key(call.from_))
def outgoing_call_hierarchy(
identity: SymbolIdentity,
indexes: dict[str, FileSymbolIndex],
definitions: set[SymbolIdentity],
) -> list[lsp.CallHierarchyOutgoingCall]:
"""Return statically resolved workspace procedures called by ``identity``."""
proc_definitions = _proc_definitions(indexes)
if identity.kind == "proc":
if _unique_proc_definition(identity, proc_definitions) is None:
return []
elif identity.kind == "file":
if identity.name not in indexes:
return []
else:
return []
grouped: dict[SymbolIdentity, list[lsp.Range]] = {}
for index in indexes.values():
for occurrence in index.occurrences:
if occurrence.is_definition or occurrence.caller != identity:
continue
callee = resolve_identity(occurrence, definitions)
if _unique_proc_definition(callee, proc_definitions) is None:
continue
grouped.setdefault(callee, []).append(occurrence.range)
results = []
for callee, ranges in grouped.items():
item = _call_hierarchy_item(callee, indexes, proc_definitions)
if item is not None:
results.append(
lsp.CallHierarchyOutgoingCall(
to=item,
from_ranges=sorted(ranges, key=_range_key),
)
)
return sorted(results, key=lambda call: _item_key(call.to))