2776813df3
The Sep 2026 decomposition (PR #102117) makes internal import paths a non-API: names now live in the focused modules that define them. This commit is the ONLY thing keeping the old paths alive, so external plugins have time to update. It is deliberately a single, unsquashed commit: git revert <this sha> removes every shim, stub and manifest at once on the announced date. Nothing in-tree may depend on these pointers: scripts/check_compat_pointers.py (wired into lint.yml) fails CI if it does. What it adds (see COMPAT_MANIFEST.md, compat_manifest.json): - 332 facade modules get one delimited `PLUGIN-COMPAT` block appended at the end of the file - 1,172 moved names resolved lazily via a module `__getattr__` (PEP 562) — never a top-level import, so no import cycles; facades that already had `__getattr__` get a chained one - 592 third-party/stdlib names the old modules used to expose, with their original import statements - 266 public definitions that had been deleted as unused, restored byte-for-byte from the pre-decomposition tree (+40 private helpers and 16 imports pulled in only because a restored definition needs them) - 3 deleted modules recreated as re-export stubs (gateway/startup_watchdog, hermes_cli/observability/ relay_runtime, tools/environments/modal_utils) - private names (`_x`) get no pointer: they were never API (3,792 skipped) Verified: all 335 touched modules import under a fresh HERMES_HOME and every manifest name resolves; the lint reports zero in-tree uses; ruff clean; targeted suites unchanged.
109 lines
4.5 KiB
Python
109 lines
4.5 KiB
Python
"""Parallel.ai web search (sync ``Parallel`` SDK) + async extract (``AsyncParallel``).
|
|
|
|
Env: ``PARALLEL_API_KEY`` (https://parallel.ai), optional
|
|
``PARALLEL_SEARCH_MODE`` = agentic (default) | fast | one-shot.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import asyncio
|
|
import logging
|
|
import os
|
|
from typing import Any, Dict, List
|
|
|
|
from plugins.web._common import (
|
|
SEARCH_LIMIT_CAP, BaseWebSearchProvider, cached_sdk_client, document, keyless_extract, keyless_search,
|
|
keyless_variant_schema, page_error, provider_env, run_extract_async, run_search, search_ok, use_keyless, web_hit,
|
|
)
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
_MISSING_KEY = "PARALLEL_API_KEY environment variable not set. Get your API key at https://parallel.ai"
|
|
|
|
|
|
def _client(slot: str, cls_name: str) -> Any:
|
|
def _factory(api_key: str) -> Any:
|
|
import parallel # deliberately lazy
|
|
return getattr(parallel, cls_name)(api_key=api_key)
|
|
|
|
return cached_sdk_client(slot, "PARALLEL_API_KEY", _MISSING_KEY, "search.parallel", _factory)
|
|
|
|
|
|
def _get_sync_client() -> Any:
|
|
return _client("_parallel_client", "Parallel")
|
|
|
|
|
|
def _get_async_client() -> Any:
|
|
return _client("_async_parallel_client", "AsyncParallel")
|
|
|
|
|
|
def _resolve_search_mode() -> str:
|
|
mode = os.getenv("PARALLEL_SEARCH_MODE", "agentic").lower().strip()
|
|
return mode if mode in {"fast", "one-shot", "agentic"} else "agentic"
|
|
|
|
|
|
class ParallelWebSearchProvider(BaseWebSearchProvider):
|
|
"""Parallel.ai search + async extract provider."""
|
|
|
|
NAME = "parallel"
|
|
DISPLAY_NAME = "Parallel"
|
|
KEY_ENV = "PARALLEL_API_KEY"
|
|
EXTRACT = True
|
|
KEYLESS = True
|
|
|
|
def search(self, query: str, limit: int = 5) -> Dict[str, Any]:
|
|
def _body() -> Dict[str, Any]:
|
|
if use_keyless("parallel", provider_env("PARALLEL_API_KEY")):
|
|
return keyless_search("Parallel", "parallel", query, limit, logger)
|
|
mode = _resolve_search_mode()
|
|
logger.info("Parallel search: '%s' (mode=%s, limit=%d)", query, mode, limit)
|
|
response = _get_sync_client().beta.search(search_queries=[query], objective=query, mode=mode, max_results=min(limit, SEARCH_LIMIT_CAP))
|
|
return search_ok([
|
|
web_hit(r.url or "", r.title or "", " ".join(r.excerpts or []), i + 1)
|
|
for i, r in enumerate(response.results or [])
|
|
])
|
|
|
|
return run_search("Parallel", logger, _body, sdk=True)
|
|
|
|
async def extract(self, urls: List[str], **kwargs: Any) -> List[Dict[str, Any]]:
|
|
async def _body() -> List[Dict[str, Any]]:
|
|
if use_keyless("parallel", provider_env("PARALLEL_API_KEY")):
|
|
# Keyless ring is blocking HTTP — hop off the event loop.
|
|
return await asyncio.to_thread(keyless_extract, "Parallel", "parallel", urls, logger)
|
|
logger.info("Parallel extract: %d URL(s)", len(urls))
|
|
response = await _get_async_client().beta.extract(urls=urls, full_content=True)
|
|
results = [document(r.url or "", r.title or "", r.full_content or "\n\n".join(r.excerpts or [])) for r in response.results or []]
|
|
return results + [
|
|
{**page_error(e.url or "", e.content or e.error_type or "extraction failed"), "metadata": {"sourceURL": e.url or ""}}
|
|
for e in response.errors or []
|
|
]
|
|
|
|
return await run_extract_async("Parallel", logger, urls, _body, sdk=True)
|
|
|
|
def get_setup_schema(self) -> Dict[str, Any]:
|
|
return keyless_variant_schema(
|
|
"Parallel", "PARALLEL_API_KEY", "https://parallel.ai",
|
|
free_tag="Objective-tuned search + page extraction on Parallel's anonymous free tier. Rate-limited under burst load.",
|
|
paid_tag="Objective-tuned search + parallel page extraction via the Parallel SDK. Unthrottled, guaranteed service.",
|
|
)
|
|
|
|
|
|
# ---- BEGIN PLUGIN-COMPAT (revert-scheduled; see COMPAT_MANIFEST.md) ----
|
|
# Names external plugins imported from this module before the Sep 2026 decomposition.
|
|
# Internal code MUST NOT use these (scripts/check_compat_pointers.py fails CI if it does).
|
|
# The whole block is removed by reverting the commit that added it.
|
|
|
|
|
|
_PLUGIN_COMPAT_LAZY = {
|
|
'WebSearchProvider': ('agent.web_search_provider', 'WebSearchProvider'),
|
|
}
|
|
|
|
|
|
def __getattr__(name): # PEP 562 — lazy so no import cycles
|
|
target = _PLUGIN_COMPAT_LAZY.get(name)
|
|
if target is None:
|
|
raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
|
|
import importlib
|
|
return getattr(importlib.import_module(target[0]), target[1])
|
|
# ---- END PLUGIN-COMPAT ----
|