feat(workflows): add from_json expression filter (#2961)

* feat(workflows): add from_json expression filter

Step outputs captured as strings could never become typed values in
templates - the filter set was default/join/map/contains only, so e.g.
a fan-out items: could never consume a step's JSON stdout. Add an
arg-less from_json pipe filter with parse-or-raise semantics: invalid
JSON or non-string input raises a clear ValueError rather than passing
through silently.

Fixes #2960

* fix(expressions): make from_json strict — reject any arguments

Address review (#2961): from_json('x') and from_json() previously fell through to a silent passthrough of the unparsed value. Reject any parenthesized form with a clear error so mis-wired templates fail loudly. Rename test to ...parses_object (JSON under test is an object) and add coverage for the strict no-arguments behavior.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs(workflows): document the from_json expression filter

Address Copilot review: the user-facing filter references omitted the
newly added `from_json` filter. Add it to the ARCHITECTURE.md filter table
(with the `{{ steps.emit.output.stdout | from_json }}` example) and to the
filter enumerations in workflows/README.md and docs/reference/workflows.md
so the docs match the evaluator's capabilities.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(workflows): make from_json strictness reject trailing tokens; fix docstring

Address Copilot review:
- Strictness only rejected parenthesized forms, so typos like
  `| from_json)` or `| from_json extra` still fell through to the
  unknown-filter path and silently returned the unparsed value. Match on
  the leading filter token and require the whole filter to be exactly
  `from_json`, so every mis-wired form raises. Extend the rejection test to
  cover the trailing-token cases.
- The module docstring claimed "no imports", which is misleading now that
  the module imports `json`. Reword to state the actual sandbox guarantee:
  templates cannot do file I/O, import modules, or run arbitrary code.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
Huy Do
2026-06-18 01:43:26 +07:00
committed by GitHub
parent 00bff788c9
commit affbf5ead5
5 changed files with 93 additions and 4 deletions

View File

@@ -1,11 +1,13 @@
"""Sandboxed expression evaluator for workflow templates.
Provides a safe Jinja2 subset for evaluating expressions in workflow YAML.
No file I/O, no imports, no arbitrary code execution.
Templates cannot perform file I/O, import modules, or run arbitrary code
the evaluator only walks the namespace and applies a fixed set of filters.
"""
from __future__ import annotations
import json
import re
from typing import Any
@@ -57,6 +59,23 @@ def _filter_contains(value: Any, substring: str) -> bool:
return False
def _filter_from_json(value: Any) -> Any:
"""Parse a JSON string into a typed value (list/dict/scalar).
Raises ``ValueError`` on non-string input or invalid JSON — a parse
failure here means the pipeline wiring is wrong, and silently
passing the unparsed value through would hide it.
"""
if not isinstance(value, str):
raise ValueError(
f"from_json: expected a JSON string, got {type(value).__name__}"
)
try:
return json.loads(value)
except json.JSONDecodeError as exc:
raise ValueError(f"from_json: invalid JSON: {exc}") from exc
# -- Expression resolution ------------------------------------------------
_EXPR_PATTERN = re.compile(r"\{\{(.+?)\}\}")
@@ -122,7 +141,7 @@ def _evaluate_simple_expression(expr: str, namespace: dict[str, Any]) -> Any:
- Comparisons: ``==``, ``!=``, ``>``, ``<``, ``>=``, ``<=``
- Boolean operators: ``and``, ``or``, ``not``
- ``in``, ``not in``
- Pipe filters: ``| default('...')``, ``| join(', ')``, ``| contains('...')``, ``| map('...')``
- Pipe filters: ``| default('...')``, ``| join(', ')``, ``| contains('...')``, ``| from_json``, ``| map('...')``
- String and numeric literals
"""
expr = expr.strip()
@@ -140,6 +159,22 @@ def _evaluate_simple_expression(expr: str, namespace: dict[str, Any]) -> Any:
value = _evaluate_simple_expression(parts[0].strip(), namespace)
filter_expr = parts[1].strip()
# `from_json` is strict: it takes no arguments and tolerates no
# trailing tokens. Match on the leading filter name and require the
# whole filter to be exactly `from_json`, so every mis-wired form
# (`from_json()`, `from_json('x')`, `from_json)`, `from_json extra`)
# fails loudly instead of silently falling through to the
# unknown-filter path and returning the unparsed value. (filter_expr
# is already stripped above.)
leading = re.match(r"\w+", filter_expr)
if leading and leading.group(0) == "from_json":
if filter_expr != "from_json":
raise ValueError(
"from_json: expected '| from_json' with no arguments or "
f"trailing tokens, got '| {filter_expr}'"
)
return _filter_from_json(value)
# Parse filter name and argument
filter_match = re.match(r"(\w+)\((.+)\)", filter_expr)
if filter_match: