goyimnose/Medios-Macina
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
Files
6076ea307bc9bd68a5a9eeb22d4bfdd8f1277ba9
Medios-Macina/cmdlet/_shared.py
Nose 6076ea307b kl
2026-01-11 18:56:26 -08:00

3149 lines
105 KiB
Python
Raw Blame History

""" """
from __future__ import annotations
import base64
import hashlib
import json
import re
import shutil
import sys
import tempfile
from collections.abc import Iterable as IterableABC
from SYS.logger import log, debug
from pathlib import Path
from typing import Any, Callable, Dict, Iterable, List, Optional, Sequence, Set
from dataclasses import dataclass, field
from SYS import models
from SYS import pipeline as pipeline_context
from SYS.result_table import ResultTable
from SYS.rich_display import stderr_console as get_stderr_console
from rich.prompt import Confirm
from contextlib import AbstractContextManager, nullcontext
@dataclass
class CmdletArg:
"""Represents a single cmdlet argument with optional enum choices."""
name: str
"""Argument name, e.g., '-path' or 'location'"""
type: str = "string"
"""Argument type: 'string', 'int', 'flag', 'enum', etc."""
required: bool = False
"""Whether this argument is required"""
description: str = ""
"""Human-readable description of the argument"""
choices: List[str] = field(default_factory=list)
"""Optional list of valid choices for enum/autocomplete, e.g., ['hydrus', 'local', '0x0.st']"""
alias: str = ""
"""Optional alias for the argument name, e.g., 'loc' for 'location'"""
handler: Optional[Any] = None
"""Optional handler function/callable for processing this argument's value"""
variadic: bool = False
"""Whether this argument accepts multiple values (consumes remaining positional args)"""
usage: str = ""
"""dsf"""
requires_db: bool = False
"""Whether this argument requires the local DB/library root to be configured."""
# Query-mapping support:
# Some cmdlets use a unified `-query` string. When configured, individual args
# can be populated from fields inside `-query` (e.g., -query "hash:<sha256>").
query_key: Optional[str] = None
"""Field name inside -query that maps to this argument (e.g., 'hash')."""
query_aliases: List[str] = field(default_factory=list)
"""Additional field names inside -query that map to this argument."""
query_only: bool = False
"""When True, do not accept a dedicated CLI flag for this arg; only map from -query."""
def resolve(self, value: Any) -> Any:
"""Resolve/process the argument value using the handler if available.
Args:
value: The raw argument value to process
Returns:
Processed value from handler, or original value if no handler
Example:
# For STORAGE arg with a handler
storage_path = SharedArgs.STORAGE.resolve('local') # Returns Path(tempfile.gettempdir())
"""
if self.handler is not None and callable(self.handler):
return self.handler(value)
return value
def to_flags(self) -> tuple[str, ...]:
"""Generate all flag variants (short and long form) for this argument.
Returns a tuple of all valid flag forms for this argument, including:
- Long form with double dash: --name
- Single dash multi-char form: -name (for convenience)
- Short form with single dash: -alias (if alias exists)
For flags, also generates negation forms:
- --no-name, -name (negation of multi-char form)
- --no-name, -nalias (negation with alias)
Returns:
Tuple of flag strings, e.g., ('--archive', '-archive', '-arch')
or for flags: ('--archive', '-archive', '-arch', '--no-archive', '-narch')
Example:
archive_flags = SharedArgs.ARCHIVE.to_flags()
# Returns: ('--archive', '-archive', '-arch', '--no-archive', '-narch')
storage_flags = SharedArgs.STORAGE.to_flags()
# Returns: ('--storage', '-storage', '-s')
"""
flags = [
f"--{self.name}",
f"-{self.name}"
] # Both double-dash and single-dash variants
# Add short form if alias exists
if self.alias:
flags.append(f"-{self.alias}")
# Add negation forms for flag type
if self.type == "flag":
flags.append(f"--no-{self.name}")
flags.append(f"-no{self.name}") # Single-dash negation variant
if self.alias:
flags.append(f"-n{self.alias}")
return tuple(flags)
def QueryArg(
name: str,
*,
key: Optional[str] = None,
aliases: Optional[Sequence[str]] = None,
type: str = "string",
required: bool = False,
description: str = "",
choices: Optional[Sequence[str]] = None,
handler: Optional[Any] = None,
query_only: bool = True,
) -> CmdletArg:
"""Create an argument that can be populated from `-query` fields.
By default, this does NOT create a dedicated flag (query_only=True). This is
useful for deprecating bloat flags like `-hash` while still making `hash:` a
first-class, documented, reusable field.
"""
return CmdletArg(
name=str(name),
type=str(type or "string"),
required=bool(required),
description=str(description or ""),
choices=list(choices or []),
handler=handler,
query_key=str(key or name).strip().lower()
if str(key or name).strip() else None,
query_aliases=[
str(a).strip().lower() for a in (aliases or []) if str(a).strip()
],
query_only=bool(query_only),
)
# ============================================================================
# SHARED ARGUMENTS - Reusable argument definitions across cmdlet
# ============================================================================
class SharedArgs:
"""Registry of shared CmdletArg definitions used across multiple cmdlet.
This class provides a centralized location for common arguments so they're
defined once and used consistently everywhere. Reduces duplication and ensures
all cmdlet handle the same arguments identically.
Example:
CMDLET = Cmdlet(
name="my-cmdlet",
summary="Does something",
usage="my-cmdlet",
args=[
SharedArgs.QUERY, # Use predefined shared arg (e.g., -query "hash:<sha256>")
SharedArgs.LOCATION, # Use another shared arg
CmdletArg(...), # Mix with custom args
]
)
"""
# NOTE: This project no longer exposes a dedicated -hash flag.
# Use SharedArgs.QUERY with `hash:` syntax instead (e.g., -query "hash:<sha256>").
STORE = CmdletArg(
name="store",
type="enum",
choices=[], # Dynamically populated via get_store_choices()
description="Selects store",
query_key="store",
)
PATH = CmdletArg(
name="path",
type="string",
choices=[], # Dynamically populated via get_store_choices()
description="selects store",
)
URL = CmdletArg(
name="url",
type="string",
description="http parser",
)
PROVIDER = CmdletArg(
name="provider",
type="string",
description="selects provider",
)
@staticmethod
@staticmethod
def get_store_choices(config: Optional[Dict[str, Any]] = None, force: bool = False) -> List[str]:
"""Get list of available store backend names.
This method returns the cached list of available backends from the most
recent startup check. Stores that failed to initialize are filtered out.
Users must restart to refresh the list if stores are enabled/disabled.
Args:
config: Optional config dict. Used if force=True or no cache exists.
force: If True, force a fresh check of the backends.
Returns:
List of backend names (e.g., ['default', 'test', 'home', 'work'])
Only includes backends that successfully initialized at startup.
Example:
SharedArgs.STORE.choices = SharedArgs.get_store_choices(config)
"""
# Use the cached startup check result if available (unless force=True)
if not force and hasattr(SharedArgs, "_cached_available_stores"):
return SharedArgs._cached_available_stores or []
# Refresh the cache
SharedArgs._refresh_store_choices_cache(config)
return SharedArgs._cached_available_stores or []
@staticmethod
def _refresh_store_choices_cache(config: Optional[Dict[str, Any]] = None) -> None:
"""Refresh the cached store choices list. Should be called once at startup.
This performs the actual StoreRegistry initialization check and caches the result.
Subsequent calls to get_store_choices() will use this cache.
Args:
config: Config dict. If not provided, will try to load from config module.
"""
try:
if config is None:
try:
from SYS.config import load_config
config = load_config()
except Exception:
SharedArgs._cached_available_stores = []
return
# Initialize registry once to filter disabled stores
from Store.registry import Store as StoreRegistry
try:
registry = StoreRegistry(config=config, suppress_debug=True)
available = registry.list_backends()
SharedArgs._cached_available_stores = available or []
except Exception:
# If registry creation fails, fallback to configured names
from Store.registry import list_configured_backend_names
SharedArgs._cached_available_stores = list_configured_backend_names(config) or []
except Exception:
SharedArgs._cached_available_stores = []
LOCATION = CmdletArg(
"location",
type="enum",
choices=["hydrus",
"0x0",
"local"],
required=True,
description="Destination location",
)
DELETE = CmdletArg(
"delete",
type="flag",
description="Delete the file and its .tag after successful operation.",
)
# Metadata arguments
ARTIST = CmdletArg(
"artist",
type="string",
description="Filter by artist name (case-insensitive, partial match).",
)
ALBUM = CmdletArg(
"album",
type="string",
description="Filter by album name (case-insensitive, partial match).",
)
TRACK = CmdletArg(
"track",
type="string",
description="Filter by track title (case-insensitive, partial match).",
)
# Library/Search arguments
LIBRARY = CmdletArg(
"library",
type="string",
choices=["hydrus",
"local",
"soulseek",
"libgen",
"ftp"],
description="Search library or source location.",
)
TIMEOUT = CmdletArg(
"timeout",
type="integer",
description="Search or operation timeout in seconds."
)
LIMIT = CmdletArg(
"limit",
type="integer",
description="Maximum number of results to return."
)
# Path/File arguments
PATH = CmdletArg("path", type="string", description="File or directory path.")
OUTPUT = CmdletArg("output", type="string", description="Output file path.")
# Generic arguments
QUERY = CmdletArg(
"query",
type="string",
description="Unified query string (e.g., hash:<sha256>, hash:{<h1>,<h2>}).",
)
REASON = CmdletArg(
"reason",
type="string",
description="Reason or explanation for the operation."
)
ARCHIVE = CmdletArg(
"archive",
type="flag",
description=
"Archive the URL to Wayback Machine, Archive.today, and Archive.ph (requires URL argument in cmdlet).",
alias="arch",
)
@staticmethod
def resolve_storage(
storage_value: Optional[str],
default: Optional[Path] = None
) -> Path:
"""Resolve a storage location name to a filesystem Path.
Maps storage identifiers to their actual filesystem paths.
This project has been refactored to use system temporary directories
for all staging/downloads by default.
Args:
storage_value: One of 'hydrus', 'local', 'ftp', or None (currently unified to temp)
default: Path to return if storage_value is None (defaults to temp directory)
Returns:
Resolved Path object for the storage location (typically system temp)
Example:
# In a cmdlet:
storage_path = SharedArgs.resolve_storage(parsed.get('storage'))
# Returns Path(tempfile.gettempdir())
"""
# We no longer maintain a hardcoded map for 'hydrus' (~/.hydrus) or 'local' (~/Videos).
# Everything defaults to the system temp directory unless a specific default is provided.
# This ensures environment independence.
if default is not None:
return default
return Path(tempfile.gettempdir())
@classmethod
def get(cls, name: str) -> Optional[CmdletArg]:
"""Get a shared argument by name.
Args:
name: Uppercase name like 'HASH', 'LOCATION', etc.
Returns:
CmdletArg if found, None otherwise
Example:
arg = SharedArgs.get('QUERY') # Returns SharedArgs.QUERY
"""
try:
return getattr(cls, name.upper())
except AttributeError:
return None
@dataclass
class Cmdlet:
"""Represents a cmdlet with metadata and arguments.
Example:
cmd = Cmdlet(
name="add-file",
summary="Upload a media file",
usage="add-file <location>",
aliases=["add-file-alias"],
args=[
CmdletArg("location", required=True, description="Destination location"),
CmdletArg("-delete", type="flag", description="Delete after upload"),
],
details=[
"- This is a detail line",
"- Another detail",
]
)
# Access properties
log(cmd.name) # "add-file"
log(cmd.summary) # "Upload a media file"
log(cmd.args[0].name) # "location"
"""
name: str
""""""
summary: str
"""One-line summary of the cmdlet"""
usage: str
"""Usage string, e.g., 'add-file <location> [-delete]'"""
alias: List[str] = field(default_factory=list)
"""List of aliases for this cmdlet, e.g., ['add', 'add-f']"""
arg: List[CmdletArg] = field(default_factory=list)
"""List of arguments accepted by this cmdlet"""
detail: List[str] = field(default_factory=list)
"""Detailed explanation lines (for help text)"""
examples: List[str] = field(default_factory=list)
"""Example invocations shown in `.help`."""
# Execution function: func(result, args, config) -> int
exec: Optional[Callable[[Any,
Sequence[str],
Dict[str,
Any]],
int]] = field(default=None)
def _collect_names(self) -> List[str]:
"""Collect primary name plus aliases, de-duplicated and normalized."""
names: List[str] = []
if self.name:
names.append(self.name)
for alias in self.alias or []:
if alias:
names.append(alias)
for alias in getattr(self, "aliases", None) or []:
if alias:
names.append(alias)
seen: Set[str] = set()
deduped: List[str] = []
for name in names:
key = name.replace("_", "-").lower()
if key in seen:
continue
seen.add(key)
deduped.append(name)
return deduped
def register(self) -> "Cmdlet":
"""Register this cmdlet's exec under its name and aliases."""
if not callable(self.exec):
return self
try:
from . import (
register_callable as _register_callable,
) # Local import to avoid circular import cost
except Exception:
return self
names = self._collect_names()
if not names:
return self
_register_callable(names, self.exec)
return self
def get_flags(self, arg_name: str) -> set[str]:
"""Generate -name and --name flag variants for an argument.
Args:
arg_name: The argument name (e.g., 'library', 'tag', 'size')
Returns:
Set containing both single-dash and double-dash variants
(e.g., {'-library', '--library'})
Example:
if low in cmdlet.get_flags('library'):
# handle library flag
"""
return {f"-{arg_name}",
f"--{arg_name}"}
def build_flag_registry(self) -> Dict[str, set[str]]:
"""Build a registry of all flag variants for this cmdlet's arguments.
Automatically generates all -name and --name variants for each argument.
Useful for parsing command-line arguments without hardcoding flags.
Returns:
Dict mapping argument names to their flag sets
(e.g., {'library': {'-library', '--library'}, 'tag': {'-tag', '--tag'}})
Example:
flags = cmdlet.build_flag_registry()
if low in flags.get('library', set()):
# handle library
elif low in flags.get('tag', set()):
# handle tag
"""
return {
arg.name: self.get_flags(arg.name)
for arg in self.arg
}
# Tag groups cache (loaded from JSON config file)
_TAG_GROUPS_CACHE: Optional[Dict[str, List[str]]] = None
_TAG_GROUPS_MTIME: Optional[float] = None
# Path to tag groups configuration (set by caller or lazily discovered)
TAG_GROUPS_PATH: Optional[Path] = None
def set_tag_groups_path(path: Path) -> None:
"""Set the path to the tag groups JSON file."""
global TAG_GROUPS_PATH
TAG_GROUPS_PATH = path
def parse_cmdlet_args(args: Sequence[str],
cmdlet_spec: Dict[str,
Any] | Cmdlet) -> Dict[str,
Any]:
"""Parse command-line arguments based on cmdlet specification.
Extracts argument values from command-line tokens using the argument names
and types defined in the cmdlet metadata. Automatically supports single-dash
and double-dash variants of flag names. Arguments without dashes in definition
are treated as positional arguments.
Args:
args: Command-line arguments (e.g., ["-path", "/home/file.txt", "-foo", "bar"])
cmdlet_spec: Cmdlet metadata dict with "args" key containing list of arg specs,
or a Cmdlet object. Each arg spec should have at least "name" key.
Argument names can be defined with or without prefixes.
Returns:
Dict mapping canonical arg names to their parsed values. If an arg is not
provided, it will not be in the dict. Lookup will normalize prefixes.
Example:
cmdlet = {
"args": [
{"name": "path", "type": "string"}, # Positional - matches bare value or -path/--path
{"name": "count", "type": "int"} # Positional - matches bare value or -count/--count
]
}
result = parse_cmdlet_args(["value1", "-count", "5"], cmdlet)
# result = {"path": "value1", "count": "5"}
"""
result: Dict[str,
Any] = {}
# Only accept Cmdlet objects
if not isinstance(cmdlet_spec, Cmdlet):
raise TypeError(f"Expected Cmdlet, got {type(cmdlet_spec).__name__}")
# Build arg specs from cmdlet
arg_specs: List[CmdletArg] = cmdlet_spec.arg
positional_args: List[CmdletArg] = [] # args without prefix in definition
flagged_args: List[CmdletArg] = [] # args with prefix in definition
query_mapped_args: List[CmdletArg] = []
arg_spec_map: Dict[str,
str] = {} # prefix variant -> canonical name (without prefix)
for spec in arg_specs:
name = spec.name
if not name:
continue
# Track args that can be populated from -query.
try:
if getattr(spec, "query_key", None):
query_mapped_args.append(spec)
except Exception:
pass
name_str = str(name)
canonical_name = name_str.lstrip("-")
# Query-only args do not register dedicated flags/positionals.
try:
if bool(getattr(spec, "query_only", False)):
continue
except Exception:
pass
# Determine if this is positional (no dashes in original definition)
if "-" not in name_str:
positional_args.append(spec)
else:
flagged_args.append(spec)
# Register all prefix variants for flagged lookup
arg_spec_map[canonical_name.lower()] = canonical_name # bare name
arg_spec_map[f"-{canonical_name}".lower()] = canonical_name # single dash
arg_spec_map[f"--{canonical_name}".lower()] = canonical_name # double dash
# Parse arguments
i = 0
positional_index = 0 # Track which positional arg we're on
while i < len(args):
token = str(args[i])
token_lower = token.lower()
# Legacy guidance: -hash/--hash was removed in favor of -query "hash:...".
# However, some cmdlets may explicitly re-introduce a -hash flag.
if token_lower in {"-hash",
"--hash"} and token_lower not in arg_spec_map:
try:
log(
'Legacy flag -hash is no longer supported. Use: -query "hash:<sha256>"',
file=sys.stderr,
)
except Exception:
pass
i += 1
continue
# Check if this token is a known flagged argument
if token_lower in arg_spec_map:
canonical_name = arg_spec_map[token_lower]
spec = next(
(
s for s in arg_specs
if str(s.name).lstrip("-").lower() == canonical_name.lower()
),
None,
)
# Check if it's a flag type (which doesn't consume next value, just marks presence)
is_flag = spec and spec.type == "flag"
if is_flag:
# For flags, just mark presence without consuming next token
result[canonical_name] = True
i += 1
else:
# For non-flags, consume next token as the value
if i + 1 < len(args) and not str(args[i + 1]).startswith("-"):
value = args[i + 1]
# Check if variadic
is_variadic = spec and spec.variadic
if is_variadic:
if canonical_name not in result:
result[canonical_name] = []
elif not isinstance(result[canonical_name], list):
result[canonical_name] = [result[canonical_name]]
result[canonical_name].append(value)
else:
result[canonical_name] = value
i += 2
else:
i += 1
# Otherwise treat as positional if we have positional args remaining
elif positional_index < len(positional_args):
positional_spec = positional_args[positional_index]
canonical_name = str(positional_spec.name).lstrip("-")
is_variadic = positional_spec.variadic
if is_variadic:
# For variadic args, append to a list
if canonical_name not in result:
result[canonical_name] = []
elif not isinstance(result[canonical_name], list):
# Should not happen if logic is correct, but safety check
result[canonical_name] = [result[canonical_name]]
result[canonical_name].append(token)
# Do not increment positional_index so subsequent tokens also match this arg
# Note: Variadic args should typically be the last positional argument
i += 1
else:
result[canonical_name] = token
positional_index += 1
i += 1
else:
# Unknown token, skip it
i += 1
# Populate query-mapped args from the unified -query string.
try:
raw_query = result.get("query")
except Exception:
raw_query = None
if query_mapped_args and raw_query is not None:
try:
from SYS.cli_syntax import parse_query as _parse_query
parsed_query = _parse_query(str(raw_query))
fields = parsed_query.get("fields",
{}) if isinstance(parsed_query,
dict) else {}
norm_fields = (
{
str(k).strip().lower(): v
for k, v in fields.items()
} if isinstance(fields,
dict) else {}
)
except Exception:
norm_fields = {}
for spec in query_mapped_args:
canonical_name = str(getattr(spec, "name", "") or "").lstrip("-")
if not canonical_name:
continue
# Do not override explicit flags.
if canonical_name in result and result.get(canonical_name) not in (None,
""):
continue
try:
key = str(getattr(spec, "query_key", "") or "").strip().lower()
aliases = getattr(spec, "query_aliases", None)
alias_list = [
str(a).strip().lower() for a in (aliases or []) if str(a).strip()
]
except Exception:
key = ""
alias_list = []
candidates = [k for k in [key, canonical_name] + alias_list if k]
val = None
for k in candidates:
if k in norm_fields:
val = norm_fields.get(k)
break
if val is None:
continue
try:
result[canonical_name] = spec.resolve(val)
except Exception:
result[canonical_name] = val
return result
def normalize_hash(hash_hex: Optional[str]) -> Optional[str]:
"""Normalize a hash string to lowercase, or return None if invalid.
Args:
hash_hex: String that should be a hex hash
Returns:
Lowercase hash string, or None if input is not a string or is empty
"""
if not isinstance(hash_hex, str):
return None
text = hash_hex.strip().lower()
if not text:
return None
if len(text) != 64:
return None
if not all(ch in "0123456789abcdef" for ch in text):
return None
return text
def parse_hash_query(query: Optional[str]) -> List[str]:
"""Parse a unified query string for `hash:` into normalized SHA256 hashes.
Supported examples:
- hash:<h1>
- hash:<h1>,<h2>,<h3>
- Hash: <h1> <h2> <h3>
- hash:{<h1>, <h2>}
Returns:
List of unique normalized 64-hex SHA256 hashes.
"""
import re
q = str(query or "").strip()
if not q:
return []
m = re.match(r"^hash(?:es)?\s*:\s*(.+)$", q, flags=re.IGNORECASE)
if not m:
return []
rest = (m.group(1) or "").strip()
if rest.startswith("{") and rest.endswith("}"):
rest = rest[1:-1].strip()
if rest.startswith("[") and rest.endswith("]"):
rest = rest[1:-1].strip()
raw_parts = [p.strip() for p in re.split(r"[\s,]+", rest) if p.strip()]
out: List[str] = []
for part in raw_parts:
h = normalize_hash(part)
if not h:
continue
if h not in out:
out.append(h)
return out
def parse_single_hash_query(query: Optional[str]) -> Optional[str]:
"""Parse `hash:` query and require exactly one hash."""
hashes = parse_hash_query(query)
if len(hashes) != 1:
return None
return hashes[0]
def get_hash_for_operation(
override_hash: Optional[str],
result: Any,
field_name: str = "hash"
) -> Optional[str]:
"""Get normalized hash from override or result object, consolidating common pattern.
Eliminates repeated pattern: normalize_hash(override) if override else normalize_hash(get_field(result, ...))
Args:
override_hash: Hash passed as command argument (takes precedence)
result: Object containing hash field (fallback)
field_name: Name of hash field in result object (default: "hash")
Returns:
Normalized hash string, or None if neither override nor result provides valid hash
"""
if override_hash:
return normalize_hash(override_hash)
hash_value = (
get_field(result,
field_name) or getattr(result,
field_name,
None) or getattr(result,
"hash",
None)
)
return normalize_hash(hash_value)
def fetch_hydrus_metadata(
config: Any,
hash_hex: str,
*,
store_name: Optional[str] = None,
hydrus_client: Any = None,
**kwargs,
) -> tuple[Optional[Dict[str,
Any]],
Optional[int]]:
"""Fetch metadata from Hydrus for a given hash, consolidating common fetch pattern.
Eliminates repeated boilerplate: client initialization, error handling, metadata extraction.
Args:
config: Configuration object (passed to hydrus_wrapper.get_client)
hash_hex: File hash to fetch metadata for
store_name: Optional Hydrus store name. When provided, do not fall back to a global/default Hydrus client.
hydrus_client: Optional explicit Hydrus client. When provided, takes precedence.
**kwargs: Additional arguments to pass to client.fetch_file_metadata()
Common: include_service_keys_to_tags, include_notes, include_file_url, include_duration, etc.
Returns:
Tuple of (metadata_dict, error_code)
- metadata_dict: Dict from Hydrus (first item in metadata list) or None if unavailable
- error_code: 0 on success, 1 on any error (suitable for returning from cmdlet execute())
"""
from API import HydrusNetwork
hydrus_wrapper = HydrusNetwork
client = hydrus_client
if client is None:
if store_name:
# Store specified: do not fall back to a global/default Hydrus client.
try:
from Store import Store
store = Store(config)
backend = store[str(store_name)]
candidate = getattr(backend, "_client", None)
if candidate is not None and hasattr(candidate, "fetch_file_metadata"):
client = candidate
except Exception as exc:
log(f"Hydrus client unavailable for store '{store_name}': {exc}")
client = None
if client is None:
log(f"Hydrus client unavailable for store '{store_name}'")
return None, 1
else:
try:
client = hydrus_wrapper.get_client(config)
except Exception as exc:
log(f"Hydrus client unavailable: {exc}")
return None, 1
if client is None:
log("Hydrus client unavailable")
return None, 1
try:
payload = client.fetch_file_metadata(hashes=[hash_hex], **kwargs)
except Exception as exc:
log(f"Hydrus metadata fetch failed: {exc}")
return None, 1
items = payload.get("metadata") if isinstance(payload, dict) else None
meta = items[0] if (
isinstance(items,
list) and items and isinstance(items[0],
dict)
) else None
return meta, 0
def get_field(obj: Any, field: str, default: Optional[Any] = None) -> Any:
"""Extract a field from either a dict or object with fallback default.
Handles both dict.get(field) and getattr(obj, field) access patterns.
Also handles lists by accessing the first element.
For PipeObjects, checks the extra field as well.
Used throughout cmdlet to uniformly access fields from mixed types.
Args:
obj: Dict, object, or list to extract from
field: Field name to retrieve
default: Value to return if field not found (default: None)
Returns:
Field value if found, otherwise the default value
Examples:
get_field(result, "hash") # From dict or object
get_field(result, "table", "unknown") # With default
"""
# Handle lists by accessing the first element
if isinstance(obj, list):
if not obj:
return default
obj = obj[0]
if isinstance(obj, dict):
return obj.get(field, default)
else:
# Try direct attribute access first
value = getattr(obj, field, None)
if value is not None:
return value
# For PipeObjects, also check the extra field
extra_val = getattr(obj, "extra", None)
if isinstance(extra_val, dict):
return extra_val.get(field, default)
return default
def should_show_help(args: Sequence[str]) -> bool:
"""Check if help flag was passed in arguments.
Consolidates repeated pattern of checking for help flags across cmdlet.
Args:
args: Command arguments to check
Returns:
True if any help flag is present (-?, /?, --help, -h, help, --cmdlet)
Examples:
if should_show_help(args):
log(json.dumps(CMDLET, ensure_ascii=False, indent=2))
return 0
"""
try:
return any(
str(a).lower() in {"-?",
"/?",
"--help",
"-h",
"help",
"--cmdlet"} for a in args
)
except Exception:
return False
def looks_like_hash(candidate: Optional[str]) -> bool:
"""Check if a string looks like a SHA256 hash (64 hex chars).
Args:
candidate: String to test
Returns:
True if the string is 64 lowercase hex characters
"""
if not isinstance(candidate, str):
return False
text = candidate.strip().lower()
return len(text) == 64 and all(ch in "0123456789abcdef" for ch in text)
def pipeline_item_local_path(item: Any) -> Optional[str]:
"""Extract local file path from a pipeline item.
Supports both dataclass objects with .path attribute and dicts.
Returns None for HTTP/HTTPS url.
Args:
item: Pipeline item (PipelineItem dataclass, dict, or other)
Returns:
Local file path string, or None if item is not a local file
"""
path_value: Optional[str] = None
if hasattr(item, "path"):
path_value = getattr(item, "path", None)
elif isinstance(item, dict):
raw = item.get("path") or item.get("url")
path_value = str(raw) if raw is not None else None
if not isinstance(path_value, str):
return None
text = path_value.strip()
if not text:
return None
if text.lower().startswith(("http://", "https://")):
return None
return text
def collect_relationship_labels(
payload: Any,
label_stack: List[str] | None = None,
mapping: Dict[str,
str] | None = None
) -> Dict[str,
str]:
"""Recursively extract hash-to-label mappings from nested relationship data.
Walks through nested dicts/lists looking for sha256-like strings (64 hex chars)
and builds a mapping from hash to its path in the structure.
Example:
data = {
"duplicates": [
"abc123...", # Will be mapped to "duplicates"
{"type": "related", "items": ["def456..."]} # Will be mapped to "duplicates / type / items"
]
}
result = collect_relationship_labels(data)
# result = {"abc123...": "duplicates", "def456...": "duplicates / type / items"}
Args:
payload: Nested data structure (dict, list, string, etc.)
label_stack: Internal use - tracks path during recursion
mapping: Internal use - accumulates hash->label mappings
Returns:
Dict mapping hash strings to their path labels
"""
if label_stack is None:
label_stack = []
if mapping is None:
mapping = {}
if isinstance(payload, dict):
for key, value in payload.items():
next_stack = label_stack
if isinstance(key, str) and key:
formatted = key.replace("_", " ").strip()
next_stack = label_stack + [formatted]
collect_relationship_labels(value, next_stack, mapping)
elif isinstance(payload, (list, tuple, set)):
for value in payload:
collect_relationship_labels(value, label_stack, mapping)
elif isinstance(payload, str) and looks_like_hash(payload):
hash_value = payload.lower()
if label_stack:
label = " / ".join(item for item in label_stack if item)
else:
label = "related"
mapping.setdefault(hash_value, label)
return mapping
def parse_tag_arguments(arguments: Sequence[str]) -> List[str]:
"""Parse tag arguments from command line tokens.
- Supports comma-separated tags.
- Supports pipe namespace shorthand: "artist:A|B|C" -> artist:A, artist:B, artist:C.
Args:
arguments: Sequence of argument strings
Returns:
List of normalized tag strings (empty strings filtered out)
"""
def _expand_pipe_namespace(text: str) -> List[str]:
parts = text.split("|")
expanded: List[str] = []
last_ns: Optional[str] = None
for part in parts:
segment = part.strip()
if not segment:
continue
if ":" in segment:
ns, val = segment.split(":", 1)
ns = ns.strip()
val = val.strip()
last_ns = ns or last_ns
if last_ns and val:
expanded.append(f"{last_ns}:{val}")
elif ns or val:
expanded.append(f"{ns}:{val}".strip(":"))
else:
if last_ns:
expanded.append(f"{last_ns}:{segment}")
else:
expanded.append(segment)
return expanded
tags: List[str] = []
for argument in arguments:
for token in argument.split(","):
text = token.strip()
if not text:
continue
# Expand namespace shorthand with pipes
pipe_expanded = _expand_pipe_namespace(text)
for entry in pipe_expanded:
candidate = entry.strip()
if not candidate:
continue
if ":" in candidate:
ns, val = candidate.split(":", 1)
ns = ns.strip()
val = val.strip()
candidate = f"{ns}:{val}" if ns or val else ""
if candidate:
tags.append(candidate)
return tags
def fmt_bytes(n: Optional[int]) -> str:
"""Format bytes as human-readable with 1 decimal place (MB/GB).
Args:
n: Number of bytes, or None
Returns:
Formatted string like "1.5 MB" or "2.0 GB", or "unknown"
"""
if n is None or n < 0:
return "unknown"
gb = n / (1024.0 * 1024.0 * 1024.0)
if gb >= 1.0:
return f"{gb:.1f} GB"
mb = n / (1024.0 * 1024.0)
return f"{mb:.1f} MB"
def _normalise_tag_group_entry(value: Any) -> Optional[str]:
"""Internal: Normalize a single tag group entry."""
if not isinstance(value, str):
value = str(value)
text = value.strip()
return text or None
def _load_tag_groups() -> Dict[str, List[str]]:
"""Load tag group definitions from JSON file with caching."""
global _TAG_GROUPS_CACHE, _TAG_GROUPS_MTIME, TAG_GROUPS_PATH
# Auto-discover adjective.json if not set
if TAG_GROUPS_PATH is None:
# Try to find adjective.json in the script directory or helper subdirectory
try:
script_dir = Path(__file__).parent.parent
# Check root directory
candidate = script_dir / "adjective.json"
if candidate.exists():
TAG_GROUPS_PATH = candidate
else:
# Check helper directory
candidate = script_dir / "helper" / "adjective.json"
if candidate.exists():
TAG_GROUPS_PATH = candidate
except Exception:
pass
if TAG_GROUPS_PATH is None:
return {}
path = TAG_GROUPS_PATH
try:
stat_result = path.stat()
except FileNotFoundError:
_TAG_GROUPS_CACHE = {}
_TAG_GROUPS_MTIME = None
return {}
except OSError as exc:
log(f"Failed to read tag groups: {exc}", file=sys.stderr)
_TAG_GROUPS_CACHE = {}
_TAG_GROUPS_MTIME = None
return {}
mtime = stat_result.st_mtime
if _TAG_GROUPS_CACHE is not None and _TAG_GROUPS_MTIME == mtime:
return _TAG_GROUPS_CACHE
try:
payload = json.loads(path.read_text(encoding="utf-8"))
except (OSError, json.JSONDecodeError) as exc:
log(f"Invalid tag group JSON ({path}): {exc}", file=sys.stderr)
_TAG_GROUPS_CACHE = {}
_TAG_GROUPS_MTIME = mtime
return {}
groups: Dict[str,
List[str]] = {}
if isinstance(payload, dict):
for key, value in payload.items():
if not isinstance(key, str):
continue
name = key.strip().lower()
if not name:
continue
members: List[str] = []
if isinstance(value, list):
for entry in value:
normalised = _normalise_tag_group_entry(entry)
if normalised:
members.append(normalised)
elif isinstance(value, str):
normalised = _normalise_tag_group_entry(value)
if normalised:
members.extend(
token.strip() for token in normalised.split(",")
if token.strip()
)
if members:
groups[name] = members
_TAG_GROUPS_CACHE = groups
_TAG_GROUPS_MTIME = mtime
return groups
def expand_tag_groups(raw_tags: Iterable[str]) -> List[str]:
"""Expand tag group references (e.g., {my_group}) into member tags.
Tag groups are defined in JSON and can be nested. Groups are referenced
with curly braces: {group_name}.
Args:
raw_tags: Sequence of tag strings, some may reference groups like "{group_name}"
Returns:
List of expanded tags with group references replaced
"""
groups = _load_tag_groups()
if not groups:
return [tag for tag in raw_tags if isinstance(tag, str) and tag.strip()]
def _expand(tokens: Iterable[str], seen: Set[str]) -> List[str]:
result: List[str] = []
for token in tokens:
if not isinstance(token, str):
continue
candidate = token.strip()
if not candidate:
continue
if candidate.startswith("{") and candidate.endswith("}") and len(candidate
) > 2:
name = candidate[1:-1].strip().lower()
if not name:
continue
if name in seen:
log(
f"Tag group recursion detected for {{{name}}}; skipping",
file=sys.stderr
)
continue
members = groups.get(name)
if not members:
log(f"Unknown tag group {{{name}}}", file=sys.stderr)
result.append(candidate)
continue
result.extend(_expand(members,
seen | {name}))
else:
result.append(candidate)
return result
return _expand(raw_tags, set())
def first_title_tag(source: Optional[Iterable[str]]) -> Optional[str]:
"""Find the first tag starting with "title:" in a collection.
Args:
source: Iterable of tag strings
Returns:
First title: tag found, or None
"""
if not source:
return None
for item in source:
if not isinstance(item, str):
continue
candidate = item.strip()
if candidate and candidate.lower().startswith("title:"):
return candidate
return None
def apply_preferred_title(tags: List[str], preferred: Optional[str]) -> List[str]:
"""Replace any title: tags with a preferred title tag.
Args:
tags: List of tags (may contain multiple "title:" entries)
preferred: Preferred title tag to use (full "title: ..." format)
Returns:
List with old title tags removed and preferred title added (at most once)
"""
if not preferred:
return tags
preferred_clean = preferred.strip()
if not preferred_clean:
return tags
preferred_lower = preferred_clean.lower()
filtered: List[str] = []
has_preferred = False
for tag in tags:
candidate = tag.strip()
if not candidate:
continue
if candidate.lower().startswith("title:"):
if candidate.lower() == preferred_lower:
if not has_preferred:
filtered.append(candidate)
has_preferred = True
continue
filtered.append(candidate)
if not has_preferred:
filtered.append(preferred_clean)
return filtered
# ============================================================================
# PIPEOBJECT UTILITIES (for chainable cmdlet and multi-action pipelines)
# ============================================================================
def create_pipe_object_result(
source: str,
identifier: str,
file_path: str,
cmdlet_name: str,
title: Optional[str] = None,
hash_value: Optional[str] = None,
is_temp: bool = False,
parent_hash: Optional[str] = None,
tag: Optional[List[str]] = None,
**extra: Any,
) -> Dict[str,
Any]:
"""Create a PipeObject-compatible result dict for pipeline chaining.
This is a helper to emit results in the standard format that downstream
cmdlet can process (filter, tag, cleanup, etc.).
Args:
source: Source system (e.g., 'local', 'hydrus', 'download')
identifier: Unique ID from source
file_path: Path to the file
cmdlet_name: Name of the cmdlet that created this (e.g., 'download-data', 'screen-shot')
title: Human-readable title
hash_value: SHA-256 hash of file (for integrity)
is_temp: If True, this is a temporary/intermediate artifact
parent_hash: Hash of the parent file in the chain (for provenance)
tag: List of tag values to apply
**extra: Additional fields
Returns:
Dict with all PipeObject fields for emission
"""
result: Dict[str, Any] = {
"source": source,
"id": identifier,
"path": file_path,
"action": f"cmdlet:{cmdlet_name}", # Format: cmdlet:cmdlet_name
}
if title:
result["title"] = title
if hash_value:
result["hash"] = hash_value
if is_temp:
result["is_temp"] = True
if parent_hash:
result["parent_hash"] = parent_hash
if tag:
result["tag"] = tag
# Canonical store field: use source for compatibility
try:
if source:
result["store"] = source
except Exception:
pass
# Add any extra fields
result.update(extra)
return result
def mark_as_temp(pipe_object: Dict[str, Any]) -> Dict[str, Any]:
"""Mark a PipeObject dict as temporary (intermediate artifact).
Args:
pipe_object: Result dict from cmdlet emission
Returns:
Modified dict with is_temp=True
"""
pipe_object["is_temp"] = True
return pipe_object
def set_parent_hash(pipe_object: Dict[str, Any], parent_hash: str) -> Dict[str, Any]:
"""Set the parent_hash for provenance tracking.
Args:
pipe_object: Result dict
parent_hash: Parent file's hash
Returns:
Modified dict with parent_hash set to the hash
"""
pipe_object["parent_hash"] = parent_hash
return pipe_object
def get_pipe_object_path(pipe_object: Any) -> Optional[str]:
"""Extract file path from PipeObject, dict, or pipeline-friendly object."""
if pipe_object is None:
return None
for attr in ("path", "target"):
if hasattr(pipe_object, attr):
value = getattr(pipe_object, attr)
if value:
return value
if isinstance(pipe_object, dict):
for key in ("path", "target"):
value = pipe_object.get(key)
if value:
return value
return None
def _extract_flag_value(args: Sequence[str], *flags: str) -> Optional[str]:
"""Return the value for the first matching flag in args.
This is intentionally lightweight (no cmdlet spec required) so callers in CLI/pipeline
can share the same behavior.
"""
if not args:
return None
want = {str(f).strip().lower()
for f in flags if str(f).strip()}
if not want:
return None
try:
tokens = [str(a) for a in args]
except Exception:
tokens = list(args) # type: ignore[list-item]
for i, tok in enumerate(tokens):
low = str(tok).strip().lower()
if low in want:
if i + 1 >= len(tokens):
return None
nxt = str(tokens[i + 1])
# Allow paths like "-"? Treat missing value as None.
if not nxt.strip():
return None
# Don't consume another flag as value.
if nxt.startswith("-"):
return None
return nxt
return None
def _unique_destination_path(dest: Path) -> Path:
"""Generate a non-colliding destination path by appending " (N)"."""
try:
if not dest.exists():
return dest
except Exception:
return dest
parent = dest.parent
stem = dest.stem
suffix = dest.suffix
for i in range(1, 10_000):
candidate = parent / f"{stem} ({i}){suffix}"
try:
if not candidate.exists():
return candidate
except Exception:
return candidate
return dest
def _print_live_safe_stderr(message: str) -> None:
"""Print to stderr without breaking Rich Live progress output."""
try:
from SYS.rich_display import stderr_console # type: ignore
except Exception:
return
cm = None
try:
from SYS import pipeline as _pipeline_ctx # type: ignore
suspend = getattr(_pipeline_ctx, "suspend_live_progress", None)
cm = suspend() if callable(suspend) else None
except Exception:
cm = None
try:
from contextlib import nullcontext
except Exception:
nullcontext = None # type: ignore
if cm is None:
cm = nullcontext() if callable(nullcontext) else None
try:
if cm is not None:
with cm:
stderr_console.print(str(message))
else:
stderr_console.print(str(message))
except Exception:
return
def apply_output_path_from_pipeobjects(
*,
cmd_name: str,
args: Sequence[str],
emits: Sequence[Any],
) -> List[Any]:
"""If the user supplied `-path`, move emitted temp/PATH files there.
This enables a dynamic pattern:
- Any cmdlet can include `SharedArgs.PATH`.
- If it emits a file-backed PipeObject (`path` exists on disk) and the item is
a temp/PATH artifact, then `-path <dest>` will save it to that location.
Rules:
- Only affects items whose `action` matches the current cmdlet.
- Only affects items that look like local artifacts (`is_temp` True or `store` == PATH).
- Updates the emitted object's `path` (and `target` when it points at the same file).
"""
dest_raw = _extract_flag_value(args, "-path", "--path")
if not dest_raw:
return list(emits or [])
# Guard: users sometimes pass a URL into -path by mistake (e.g. `-path https://...`).
# Treat that as invalid for filesystem moves and avoid breaking Rich Live output.
try:
dest_str = str(dest_raw).strip()
if "://" in dest_str:
_print_live_safe_stderr(
f"Ignoring -path value that looks like a URL: {dest_str}"
)
return list(emits or [])
except Exception:
pass
cmd_norm = str(cmd_name or "").replace("_", "-").strip().lower()
if not cmd_norm:
return list(emits or [])
try:
dest_hint_dir = str(dest_raw).endswith(("/", "\\"))
except Exception:
dest_hint_dir = False
try:
dest_path = Path(str(dest_raw)).expanduser()
except Exception:
return list(emits or [])
items = list(emits or [])
# Identify which emitted items are actually file artifacts produced by this cmdlet.
artifact_indices: List[int] = []
artifact_paths: List[Path] = []
for idx, item in enumerate(items):
action = str(get_field(item, "action", "") or "").strip().lower()
if not action.startswith("cmdlet:"):
continue
action_name = action.split(":", 1)[-1].strip().lower()
if action_name != cmd_norm:
continue
store = str(get_field(item, "store", "") or "").strip().lower()
is_temp = bool(get_field(item, "is_temp", False))
if not (is_temp or store == "path"):
continue
src_str = get_pipe_object_path(item)
if not src_str:
continue
try:
src = Path(str(src_str)).expanduser()
except Exception:
continue
try:
if not src.exists() or not src.is_file():
continue
except Exception:
continue
artifact_indices.append(idx)
artifact_paths.append(src)
if not artifact_indices:
return items
# Decide whether the destination is a directory or a single file.
if len(artifact_indices) > 1:
# Multiple artifacts: always treat destination as a directory.
if dest_path.suffix:
dest_dir = dest_path.parent
else:
dest_dir = dest_path
try:
dest_dir.mkdir(parents=True, exist_ok=True)
except Exception as exc:
_print_live_safe_stderr(
f"Failed to create destination directory: {dest_dir} ({exc})"
)
return items
for idx, src in zip(artifact_indices, artifact_paths):
final = dest_dir / src.name
final = _unique_destination_path(final)
try:
if src.resolve() == final.resolve():
_apply_saved_path_update(
items[idx],
old_path=str(src),
new_path=str(final)
)
_print_saved_output_panel(items[idx], final)
continue
except Exception:
pass
try:
shutil.move(str(src), str(final))
except Exception as exc:
_print_live_safe_stderr(f"Failed to save output to {final}: {exc}")
continue
_apply_saved_path_update(items[idx], old_path=str(src), new_path=str(final))
_print_saved_output_panel(items[idx], final)
return items
# Single artifact: destination can be a directory or a concrete file path.
src = artifact_paths[0]
idx = artifact_indices[0]
final: Path
try:
if dest_hint_dir or (dest_path.exists() and dest_path.is_dir()):
final = dest_path / src.name
else:
final = dest_path
except Exception:
final = dest_path
try:
final.parent.mkdir(parents=True, exist_ok=True)
except Exception as exc:
_print_live_safe_stderr(
f"Failed to create destination directory: {final.parent} ({exc})"
)
return items
final = _unique_destination_path(final)
try:
if src.resolve() != final.resolve():
shutil.move(str(src), str(final))
except Exception as exc:
_print_live_safe_stderr(f"Failed to save output to {final}: {exc}")
return items
_apply_saved_path_update(items[idx], old_path=str(src), new_path=str(final))
_print_saved_output_panel(items[idx], final)
return items
def _print_saved_output_panel(item: Any, final_path: Path) -> None:
"""When -path is used, print a Rich panel summarizing the saved output.
Shows: Title, Location, Hash.
Best-effort: reads existing fields first to avoid recomputing hashes.
"""
try:
from rich.panel import Panel # type: ignore
from rich.table import Table # type: ignore
from SYS.rich_display import stderr_console # type: ignore
except Exception:
return
# If Rich Live progress is active, pause it while printing so the panel
# doesn't get overwritten/truncated by Live's cursor control.
try:
from SYS import pipeline as _pipeline_ctx # type: ignore
suspend = getattr(_pipeline_ctx, "suspend_live_progress", None)
cm = suspend() if callable(suspend) else None
except Exception:
cm = None
try:
from contextlib import nullcontext
except Exception:
nullcontext = None # type: ignore
if cm is None:
cm = nullcontext() if callable(nullcontext) else None
try:
location = str(final_path)
except Exception:
location = ""
title = ""
try:
title = str(get_field(item, "title") or get_field(item, "name") or "").strip()
except Exception:
title = ""
if not title:
try:
title = str(final_path.stem or final_path.name)
except Exception:
title = ""
file_hash = ""
try:
file_hash = str(get_field(item,
"hash") or get_field(item,
"sha256") or "").strip()
except Exception:
file_hash = ""
if not file_hash:
try:
from SYS.utils import sha256_file # type: ignore
file_hash = str(sha256_file(final_path) or "").strip()
except Exception:
file_hash = ""
grid = Table.grid(padding=(0, 1))
grid.add_column(justify="right", style="bold")
grid.add_column()
grid.add_row("Title", title or "(unknown)")
grid.add_row("Location", location or "(unknown)")
grid.add_row("Hash", file_hash or "(unknown)")
try:
if cm is not None:
with cm:
stderr_console.print(Panel(grid, title="Saved", expand=False))
else:
stderr_console.print(Panel(grid, title="Saved", expand=False))
except Exception:
return
def _apply_saved_path_update(item: Any, *, old_path: str, new_path: str) -> None:
"""Update a PipeObject-like item after its backing file has moved."""
old_str = str(old_path)
new_str = str(new_path)
if isinstance(item, dict):
try:
if str(item.get("path") or "") == old_str:
item["path"] = new_str
except Exception:
pass
try:
if str(item.get("target") or "") == old_str:
item["target"] = new_str
except Exception:
pass
try:
extra = item.get("extra")
if isinstance(extra, dict):
if str(extra.get("target") or "") == old_str:
extra["target"] = new_str
if str(extra.get("path") or "") == old_str:
extra["path"] = new_str
except Exception:
pass
return
# models.PipeObject or PipeObject-ish
try:
if getattr(item, "path", None) == old_str:
setattr(item, "path", new_str)
except Exception:
pass
try:
extra = getattr(item, "extra", None)
if isinstance(extra, dict):
if str(extra.get("target") or "") == old_str:
extra["target"] = new_str
if str(extra.get("path") or "") == old_str:
extra["path"] = new_str
except Exception:
pass
def get_pipe_object_hash(pipe_object: Any) -> Optional[str]:
"""Extract file hash from PipeObject, dict, or pipeline-friendly object."""
if pipe_object is None:
return None
for attr in ("hash",
):
if hasattr(pipe_object, attr):
value = getattr(pipe_object, attr)
if value:
return value
if isinstance(pipe_object, dict):
for key in ("hash",
):
value = pipe_object.get(key)
if value:
return value
return None
def normalize_result_input(result: Any) -> List[Dict[str, Any]]:
"""Normalize input result to a list of dicts.
Handles:
- None -> []
- Dict -> [dict]
- List of dicts -> list as-is
- PipeObject -> [dict]
- List of PipeObjects -> list of dicts
Args:
result: Result from piped input
Returns:
List of result dicts (may be empty)
"""
if result is None:
return []
# Single dict
if isinstance(result, dict):
return [result]
# List - convert each item to dict if needed
if isinstance(result, list):
output = []
for item in result:
if isinstance(item, dict):
output.append(item)
elif hasattr(item, "to_dict"):
output.append(item.to_dict())
else:
# Try as-is
output.append(item)
return output
# PipeObject or other object with to_dict
if hasattr(result, "to_dict"):
return [result.to_dict()]
# Fallback: wrap it
if isinstance(result, dict):
return [result]
return []
def filter_results_by_temp(results: List[Any], include_temp: bool = False) -> List[Any]:
"""Filter results by temporary status.
Args:
results: List of result dicts or PipeObjects
include_temp: If True, keep temp files; if False, exclude them
Returns:
Filtered list
"""
if include_temp:
return results
filtered = []
for result in results:
is_temp = False
# Check PipeObject
if hasattr(result, "is_temp"):
is_temp = result.is_temp
# Check dict
elif isinstance(result, dict):
is_temp = result.get("is_temp", False)
if not is_temp:
filtered.append(result)
return filtered
def merge_sequences(*sources: Optional[Iterable[Any]],
case_sensitive: bool = True) -> list[str]:
"""Merge iterable sources while preserving order and removing duplicates."""
seen: set[str] = set()
merged: list[str] = []
for source in sources:
if not source:
continue
if isinstance(source, str) or not isinstance(source, IterableABC):
iterable = [source]
else:
iterable = source
for value in iterable:
if value is None:
continue
text = str(value).strip()
if not text:
continue
key = text if case_sensitive else text.lower()
if key in seen:
continue
seen.add(key)
merged.append(text)
return merged
def collapse_namespace_tags(
tags: Optional[Iterable[Any]],
namespace: str,
prefer: str = "last"
) -> list[str]:
"""Reduce tags so only one entry for a given namespace remains.
Keeps either the first or last occurrence (default last) while preserving overall order
for non-matching tags. Useful for ensuring a single title: tag.
"""
if not tags:
return []
ns = str(namespace or "").strip().lower()
if not ns:
return list(tags) if isinstance(tags, list) else list(tags)
prefer_last = str(prefer or "last").lower() != "first"
ns_prefix = ns + ":"
items = list(tags)
if prefer_last:
kept: list[str] = []
seen_ns = False
for tag in reversed(items):
text = str(tag)
if text.lower().startswith(ns_prefix):
if seen_ns:
continue
seen_ns = True
kept.append(text)
kept.reverse()
return kept
else:
kept_ns = False
result: list[str] = []
for tag in items:
text = str(tag)
if text.lower().startswith(ns_prefix):
if kept_ns:
continue
kept_ns = True
result.append(text)
return result
def collapse_namespace_tag(
tags: Optional[Iterable[Any]],
namespace: str,
prefer: str = "last"
) -> list[str]:
"""Singular alias for collapse_namespace_tags.
Some cmdlet prefer the singular name; keep behavior centralized.
"""
return collapse_namespace_tags(tags, namespace, prefer=prefer)
def extract_tag_from_result(result: Any) -> list[str]:
tag: list[str] = []
if isinstance(result, models.PipeObject):
tag.extend(result.tag or [])
if isinstance(result.extra, dict):
extra_tag = result.extra.get("tag")
if isinstance(extra_tag, list):
tag.extend(extra_tag)
elif isinstance(extra_tag, str):
tag.append(extra_tag)
elif hasattr(result, "tag"):
# Handle objects with tag attribute (e.g. SearchResult)
val = getattr(result, "tag")
if isinstance(val, (list, set, tuple)):
tag.extend(val)
elif isinstance(val, str):
tag.append(val)
if isinstance(result, dict):
raw_tag = result.get("tag")
if isinstance(raw_tag, list):
tag.extend(raw_tag)
elif isinstance(raw_tag, str):
tag.append(raw_tag)
extra = result.get("extra")
if isinstance(extra, dict):
extra_tag = extra.get("tag")
if isinstance(extra_tag, list):
tag.extend(extra_tag)
elif isinstance(extra_tag, str):
tag.append(extra_tag)
return merge_sequences(tag, case_sensitive=True)
def extract_title_from_result(result: Any) -> Optional[str]:
"""Extract the title from a result dict or PipeObject."""
if isinstance(result, models.PipeObject):
return result.title
elif hasattr(result, "title"):
return getattr(result, "title")
elif isinstance(result, dict):
return result.get("title")
return None
def extract_url_from_result(result: Any) -> list[str]:
url: list[str] = []
def _extend(candidate: Any) -> None:
if not candidate:
return
if isinstance(candidate, list):
url.extend(candidate)
elif isinstance(candidate, str):
url.append(candidate)
if isinstance(result, models.PipeObject):
_extend(result.extra.get("url"))
_extend(result.extra.get("url")) # Also check singular url
if isinstance(result.metadata, dict):
_extend(result.metadata.get("url"))
_extend(result.metadata.get("url"))
_extend(result.metadata.get("url"))
if isinstance(getattr(result, "full_metadata", None), dict):
fm = getattr(result, "full_metadata", None)
if isinstance(fm, dict):
_extend(fm.get("url"))
_extend(fm.get("url"))
_extend(fm.get("url"))
elif hasattr(result, "url") or hasattr(result, "url"):
# Handle objects with url/url attribute
_extend(getattr(result, "url", None))
_extend(getattr(result, "url", None))
if isinstance(result, dict):
_extend(result.get("url"))
_extend(result.get("url"))
_extend(result.get("url"))
fm = result.get("full_metadata")
if isinstance(fm, dict):
_extend(fm.get("url"))
_extend(fm.get("url"))
_extend(fm.get("url"))
extra = result.get("extra")
if isinstance(extra, dict):
_extend(extra.get("url"))
_extend(extra.get("url"))
_extend(extra.get("url"))
return merge_sequences(url, case_sensitive=True)
def extract_relationships(result: Any) -> Optional[Dict[str, Any]]:
if isinstance(result, models.PipeObject):
relationships = result.get_relationships()
return relationships or None
if isinstance(result, dict):
relationships = result.get("relationships")
if isinstance(relationships, dict) and relationships:
return relationships
return None
def extract_duration(result: Any) -> Optional[float]:
duration = None
if isinstance(result, models.PipeObject):
duration = result.duration
elif isinstance(result, dict):
duration = result.get("duration")
if duration is None:
metadata = result.get("metadata")
if isinstance(metadata, dict):
duration = metadata.get("duration")
if duration is None:
return None
try:
return float(duration)
except (TypeError, ValueError):
return None
def coerce_to_pipe_object(
value: Any,
default_path: Optional[str] = None
) -> models.PipeObject:
"""Normalize any incoming result to a PipeObject for single-source-of-truth state.
Uses hash+store canonical pattern.
"""
# Debug: Print ResultItem details if coming from search_file.py
try:
from SYS.logger import is_debug_enabled, debug
if (is_debug_enabled() and hasattr(value,
"__class__")
and value.__class__.__name__ == "ResultItem"):
debug("[ResultItem -> PipeObject conversion]")
debug(f" title={getattr(value, 'title', None)}")
debug(f" target={getattr(value, 'target', None)}")
debug(f" hash={getattr(value, 'hash', None)}")
debug(f" media_kind={getattr(value, 'media_kind', None)}")
debug(f" tag={getattr(value, 'tag', None)}")
debug(f" tag_summary={getattr(value, 'tag_summary', None)}")
debug(f" size_bytes={getattr(value, 'size_bytes', None)}")
debug(f" duration_seconds={getattr(value, 'duration_seconds', None)}")
debug(f" relationships={getattr(value, 'relationships', None)}")
debug(f" url={getattr(value, 'url', None)}")
debug(
f" full_metadata keys={list(getattr(value, 'full_metadata', {}).keys()) if hasattr(value, 'full_metadata') and value.full_metadata else []}"
)
except Exception:
pass
if isinstance(value, models.PipeObject):
return value
known_keys = {
"hash",
"store",
"tag",
"title",
"url",
"source_url",
"duration",
"metadata",
"warnings",
"path",
"relationships",
"is_temp",
"action",
"parent_hash",
}
# Convert common object-like results into a dict so we can preserve fields like
# hash/store/url when they come from result tables (e.g., get-url emits UrlItem).
#
# Priority:
# 1) explicit to_dict()
# 2) best-effort attribute extraction for known PipeObject-ish fields
if hasattr(value, "to_dict"):
value = value.to_dict()
elif not isinstance(value, dict):
try:
obj_map: Dict[str,
Any] = {}
for k in (
"hash",
"store",
"provider",
"prov",
"tag",
"title",
"url",
"source_url",
"duration",
"duration_seconds",
"metadata",
"full_metadata",
"warnings",
"path",
"target",
"relationships",
"is_temp",
"action",
"parent_hash",
"extra",
"media_kind",
):
if hasattr(value, k):
obj_map[k] = getattr(value, k)
if obj_map:
value = obj_map
except Exception:
pass
if isinstance(value, dict):
# Extract hash and store (canonical identifiers)
hash_val = value.get("hash")
store_val = value.get("store") or "PATH"
if not store_val or store_val == "PATH":
try:
extra_store = value.get("extra",
{}).get("store")
except Exception:
extra_store = None
if extra_store:
store_val = extra_store
# If no hash, try to compute from path or use placeholder
if not hash_val:
path_val = value.get("path")
if path_val:
try:
from SYS.utils import sha256_file
from pathlib import Path
hash_val = sha256_file(Path(path_val))
except Exception:
hash_val = "unknown"
else:
hash_val = "unknown"
# Extract title from filename if not provided
title_val = value.get("title")
if not title_val:
path_val = value.get("path")
if path_val:
try:
from pathlib import Path
title_val = Path(path_val).stem
except Exception:
pass
extra = {
k: v
for k, v in value.items() if k not in known_keys
}
# Extract URL: prefer direct url field, then url list
from SYS.metadata import normalize_urls
url_list = normalize_urls(value.get("url"))
url_val = url_list[0] if url_list else None
if len(url_list) > 1:
extra["url"] = url_list
# Extract relationships
rels = value.get("relationships") or {}
# Canonical tag: accept list or single string
tag_val: list[str] = []
if "tag" in value:
raw_tag = value["tag"]
if isinstance(raw_tag, list):
tag_val = [str(t) for t in raw_tag if t is not None]
elif isinstance(raw_tag, str):
tag_val = [raw_tag]
# Consolidate path: prefer explicit path key, but NOT target if it's a URL
path_val = value.get("path")
# Only use target as path if it's not a URL (url should stay in url field)
if not path_val and "target" in value:
target = value["target"]
if target and not (isinstance(target,
str) and (target.startswith("http://")
or target.startswith("https://"))):
path_val = target
# If the path value is actually a URL, move it to url_val and clear path_val
try:
if isinstance(path_val,
str) and (path_val.startswith("http://")
or path_val.startswith("https://")):
# Prefer existing url_val if present, otherwise move path_val into url_val
if not url_val:
url_val = path_val
path_val = None
except Exception:
pass
# Extract media_kind if available
if "media_kind" in value:
extra["media_kind"] = value["media_kind"]
pipe_obj = models.PipeObject(
hash=hash_val,
store=store_val,
provider=str(
value.get("provider") or value.get("prov") or extra.get("provider")
or ""
).strip() or None,
tag=tag_val,
title=title_val,
url=url_val,
source_url=value.get("source_url"),
duration=value.get("duration") or value.get("duration_seconds"),
metadata=value.get("metadata") or value.get("full_metadata") or {},
warnings=list(value.get("warnings") or []),
path=path_val,
relationships=rels,
is_temp=bool(value.get("is_temp",
False)),
action=value.get("action"),
parent_hash=value.get("parent_hash"),
extra=extra,
)
return pipe_obj
# Fallback: build from path argument or bare value
hash_val = "unknown"
path_val = default_path or getattr(value, "path", None)
url_val: Optional[str] = None
title_val = None
# If the raw value is a string, treat it as either a URL or a file path.
# This is important for @-selection results that are plain URL strings.
if isinstance(value, str):
s = value.strip()
if s.lower().startswith(("http://", "https://")):
url_val = s
path_val = None
else:
path_val = s
if path_val and path_val != "unknown":
try:
from SYS.utils import sha256_file
from pathlib import Path
path_obj = Path(path_val)
hash_val = sha256_file(path_obj)
# Extract title from filename (without extension)
title_val = path_obj.stem
except Exception:
pass
# When coming from a raw URL string, mark it explicitly as URL.
# Otherwise treat it as a local path.
store_val = "URL" if url_val else "PATH"
pipe_obj = models.PipeObject(
hash=hash_val,
store=store_val,
provider=None,
path=str(path_val) if path_val and path_val != "unknown" else None,
title=title_val,
url=url_val,
source_url=url_val,
tag=[],
extra={},
)
return pipe_obj
def propagate_metadata(
previous_items: Sequence[Any],
new_items: Sequence[Any]
) -> List[Any]:
"""Merge metadata/tags from previous pipeline stage into new items.
Implements "sticky metadata": items generated by a transformation (download, convert)
should inherit rich info (lyrics, art, tags) from their source.
Strategies:
A. Hash Match: If inputs/outputs share a hash, they are the same item.
B. Index Match: If lists are same length, assume 1:1 mapping (heuristic).
C. Explicit Parent: If output has `parent_hash`, link to input with that hash.
"""
if not previous_items or not new_items:
return list(new_items)
try:
prev_normalized = [coerce_to_pipe_object(p) for p in previous_items]
except Exception:
return list(new_items)
prev_by_hash: Dict[str, models.PipeObject] = {}
for p_obj in prev_normalized:
if p_obj.hash and p_obj.hash != "unknown":
prev_by_hash[p_obj.hash] = p_obj
normalized: List[models.PipeObject] = []
# Pre-calculate length matching for heuristic
is_same_length = len(new_items) == len(prev_normalized)
for i, item in enumerate(new_items):
try:
obj = coerce_to_pipe_object(item)
except Exception:
normalized.append(item) # Should not happen given coerce guards
continue
parent: Optional[models.PipeObject] = None
# Strategy A: Precise Hash Match
if obj.hash in prev_by_hash:
parent = prev_by_hash[obj.hash]
# Strategy B: Index Match (Heuristic)
if not parent and is_same_length:
parent = prev_normalized[i]
# Strategy C: Explicit Parent Hash
if not parent and obj.parent_hash and obj.parent_hash in prev_by_hash:
parent = prev_by_hash[obj.parent_hash]
if parent:
# 1. Tags: Merge unique tags
if parent.tag:
if not obj.tag:
obj.tag = list(parent.tag)
else:
curr_tags = {str(t).lower() for t in obj.tag}
for pt in parent.tag:
if str(pt).lower() not in curr_tags:
obj.tag.append(pt)
# 2. Metadata: Merge missing keys
if parent.metadata:
if not obj.metadata:
obj.metadata = parent.metadata.copy()
else:
for mk, mv in parent.metadata.items():
if mk not in obj.metadata:
obj.metadata[mk] = mv
# 3. Source URL: Propagate if missing
if parent.source_url and not obj.source_url:
obj.source_url = parent.source_url
elif parent.url and not obj.source_url and not obj.url:
# If parent had a URL and child has none, it's likely the source
obj.source_url = parent.url
# 4. Relationships: Merge missing keys
if parent.relationships:
if not obj.relationships:
obj.relationships = parent.relationships.copy()
else:
for rk, rv in parent.relationships.items():
if rk not in obj.relationships:
obj.relationships[rk] = rv
# 5. Extra (Notes/etc): Merge missing keys
# Important for passing 'notes' payload (lyrics, captions)
if parent.extra:
if not obj.extra:
obj.extra = parent.extra.copy()
else:
# Recursive merge for 'notes' dict specifically?
# For now just shallow merge keys, but handle 'notes' specially if valid.
for ek, ev in parent.extra.items():
if ek not in obj.extra:
obj.extra[ek] = ev
elif ek == "notes" and isinstance(ev, dict) and isinstance(obj.extra[ek], dict):
# Merge notes dict
for nk, nv in ev.items():
if nk not in obj.extra[ek]:
obj.extra[ek][nk] = nv
normalized.append(obj)
return normalized
def register_url_with_local_library(
pipe_obj: models.PipeObject,
config: Dict[str,
Any]
) -> bool:
"""Register url with a file in the local library database.
This is called automatically by download cmdlet to ensure url are persisted
without requiring a separate add-url step in the pipeline.
Args:
pipe_obj: PipeObject with path and url
config: Config dict containing local library path
Returns:
True if url were registered, False otherwise
"""
try:
from SYS.config import get_local_storage_path
from API.folder import API_folder_store
file_path = get_field(pipe_obj, "path")
url_field = get_field(pipe_obj, "url", [])
urls: List[str] = []
if isinstance(url_field, str):
urls = [u.strip() for u in url_field.split(",") if u.strip()]
elif isinstance(url_field, (list, tuple)):
urls = [u for u in url_field if isinstance(u, str) and u.strip()]
if not file_path or not urls:
return False
path_obj = Path(file_path)
if not path_obj.exists():
return False
storage_path = get_local_storage_path(config)
if not storage_path:
return False
with API_folder_store(storage_path) as db:
file_hash = db.get_file_hash(path_obj)
if not file_hash:
return False
metadata = db.get_metadata(file_hash) or {}
existing_url = metadata.get("url") or []
# Add any new url
changed = False
for u in urls:
if u not in existing_url:
existing_url.append(u)
changed = True
if changed:
metadata["url"] = existing_url
db.save_metadata(path_obj, metadata)
return True
return True # url already existed
except Exception:
return False
def resolve_tidal_manifest_path(item: Any) -> Optional[str]:
"""Persist the Tidal manifest from search results and return a local path."""
metadata = None
if isinstance(item, dict):
metadata = item.get("full_metadata") or item.get("metadata")
else:
metadata = getattr(item, "full_metadata", None) or getattr(item, "metadata", None)
if not isinstance(metadata, dict):
return None
existing_path = metadata.get("_tidal_manifest_path")
if existing_path:
try:
resolved = Path(str(existing_path))
if resolved.is_file():
return str(resolved)
except Exception:
pass
existing_url = metadata.get("_tidal_manifest_url")
if existing_url and isinstance(existing_url, str):
candidate = existing_url.strip()
if candidate:
return candidate
raw_manifest = metadata.get("manifest")
if not raw_manifest:
# When piping directly from the Tidal search table, we may only have a track id.
# Fetch track details from the proxy so downstream stages can decode the manifest.
try:
already = bool(metadata.get("_tidal_track_details_fetched"))
except Exception:
already = False
track_id = metadata.get("trackId") or metadata.get("id")
if track_id is None:
try:
if isinstance(item, dict):
candidate_path = item.get("path") or item.get("url")
else:
candidate_path = getattr(item, "path", None) or getattr(item, "url", None)
except Exception:
candidate_path = None
if candidate_path:
m = re.search(
r"(tidal|hifi):(?://)?track[\\/](\d+)",
str(candidate_path),
flags=re.IGNORECASE,
)
if m:
track_id = m.group(2)
if (not already) and track_id is not None:
try:
track_int = int(track_id)
except Exception:
track_int = None
if track_int and track_int > 0:
try:
import httpx
resp = httpx.get(
"https://tidal-api.binimum.org/track/",
params={"id": str(track_int)},
timeout=10.0,
)
resp.raise_for_status()
payload = resp.json()
data = payload.get("data") if isinstance(payload, dict) else None
if isinstance(data, dict) and data:
try:
metadata.update(data)
except Exception:
pass
try:
metadata["_tidal_track_details_fetched"] = True
except Exception:
pass
if not metadata.get("url"):
try:
resp_info = httpx.get(
"https://tidal-api.binimum.org/info/",
params={"id": str(track_int)},
timeout=10.0,
)
resp_info.raise_for_status()
info_payload = resp_info.json()
info_data = info_payload.get("data") if isinstance(info_payload, dict) else None
if isinstance(info_data, dict) and info_data:
try:
for k, v in info_data.items():
if k not in metadata:
metadata[k] = v
except Exception:
pass
try:
if info_data.get("url"):
metadata["url"] = info_data.get("url")
except Exception:
pass
except Exception:
pass
except Exception:
pass
raw_manifest = metadata.get("manifest")
if not raw_manifest:
return None
manifest_str = "".join(str(raw_manifest or "").split())
if not manifest_str:
return None
manifest_bytes: bytes
try:
manifest_bytes = base64.b64decode(manifest_str, validate=True)
except Exception:
try:
manifest_bytes = base64.b64decode(manifest_str, validate=False)
except Exception:
try:
manifest_bytes = manifest_str.encode("utf-8")
except Exception:
return None
if not manifest_bytes:
return None
head = (manifest_bytes[:1024] or b"").lstrip()
if head.startswith((b"{", b"[")):
try:
text = manifest_bytes.decode("utf-8", errors="ignore")
payload = json.loads(text)
urls = payload.get("urls") or []
selected_url = None
for candidate in urls:
if isinstance(candidate, str):
candidate = candidate.strip()
if candidate:
selected_url = candidate
break
if selected_url:
try:
metadata["_tidal_manifest_url"] = selected_url
except Exception:
pass
return selected_url
try:
metadata["_tidal_manifest_error"] = "JSON manifest contained no urls"
except Exception:
pass
log(
f"[tidal] JSON manifest for track {metadata.get('trackId') or metadata.get('id')} had no playable urls",
file=sys.stderr,
)
except Exception as exc:
try:
metadata["_tidal_manifest_error"] = (
f"Failed to parse JSON manifest: {exc}"
)
except Exception:
pass
log(
f"[tidal] Failed to parse JSON manifest for track {metadata.get('trackId') or metadata.get('id')}: {exc}",
file=sys.stderr,
)
return None
looks_like_mpd = (
head.startswith(b"<?xml")
or head.startswith(b"<MPD")
or b"<MPD" in head
)
if not looks_like_mpd:
manifest_mime = str(metadata.get("manifestMimeType") or "").strip().lower()
try:
metadata["_tidal_manifest_error"] = (
f"Decoded manifest is not an MPD XML (mime: {manifest_mime or 'unknown'})"
)
except Exception:
pass
try:
log(
f"[tidal] Decoded manifest is not an MPD XML for track {metadata.get('trackId') or metadata.get('id')} (mime {manifest_mime or 'unknown'})",
file=sys.stderr,
)
except Exception:
pass
return None
manifest_hash = str(metadata.get("manifestHash") or "").strip()
track_id = metadata.get("trackId") or metadata.get("id")
identifier = manifest_hash or hashlib.sha256(manifest_bytes).hexdigest()
identifier_safe = re.sub(r"[^A-Za-z0-9_-]+", "_", identifier)[:64]
if not identifier_safe:
identifier_safe = hashlib.sha256(manifest_bytes).hexdigest()[:12]
track_safe = "tidal"
if track_id is not None:
track_safe = re.sub(r"[^A-Za-z0-9_-]+", "_", str(track_id))[:32]
if not track_safe:
track_safe = "tidal"
# Persist as .mpd for DASH manifests.
ext = "mpd"
manifest_dir = Path(tempfile.gettempdir()) / "medeia" / "tidal"
try:
manifest_dir.mkdir(parents=True, exist_ok=True)
except Exception:
pass
filename = f"tidal-{track_safe}-{identifier_safe[:24]}.{ext}"
target_path = manifest_dir / filename
try:
with open(target_path, "wb") as fh:
fh.write(manifest_bytes)
metadata["_tidal_manifest_path"] = str(target_path)
if isinstance(item, dict):
if item.get("full_metadata") is metadata:
item["full_metadata"] = metadata
elif item.get("metadata") is metadata:
item["metadata"] = metadata
else:
extra = getattr(item, "extra", None)
if isinstance(extra, dict):
extra["_tidal_manifest_path"] = str(target_path)
except Exception:
return None
return str(target_path)
def check_url_exists_in_storage(
urls: Sequence[str],
storage: Any,
hydrus_available: bool,
final_output_dir: Optional[Path] = None,
) -> bool:
"""Pre-flight check to see if URLs already exist in storage.
Args:
urls: List of URLs to check
storage: The storage interface
hydrus_available: Whether Hydrus is available
final_output_dir: Final output directory (to skip if same as storage)
Returns:
True if check passed (user said yes or no dups), False if user said no (stop).
"""
if storage is None:
debug("Bulk URL preflight skipped: storage unavailable")
return True
try:
current_cmd_text = pipeline_context.get_current_command_text("")
except Exception:
current_cmd_text = ""
try:
stage_ctx = pipeline_context.get_stage_context()
except Exception:
stage_ctx = None
in_pipeline = bool(stage_ctx is not None or ("|" in str(current_cmd_text or "")))
if in_pipeline:
try:
cached_cmd = pipeline_context.load_value("preflight.url_duplicates.command", default="")
cached_decision = pipeline_context.load_value("preflight.url_duplicates.continue", default=None)
except Exception:
cached_cmd = ""
cached_decision = None
if cached_decision is not None and str(cached_cmd or "") == str(current_cmd_text or ""):
if bool(cached_decision):
return True
try:
pipeline_context.request_pipeline_stop(reason="duplicate-url declined", exit_code=0)
except Exception:
pass
return False
unique_urls: List[str] = []
for u in urls or []:
s = str(u or "").strip()
if s and s not in unique_urls:
unique_urls.append(s)
if len(unique_urls) == 0:
return True
try:
from SYS.metadata import normalize_urls
except Exception:
normalize_urls = None # type: ignore[assignment]
def _httpish(value: str) -> bool:
try:
return bool(value) and (value.startswith("http://") or value.startswith("https://"))
except Exception:
return False
url_needles: Dict[str, List[str]] = {}
for u in unique_urls:
needles: List[str] = []
if normalize_urls is not None:
try:
needles.extend([n for n in (normalize_urls(u) or []) if isinstance(n, str)])
except Exception:
needles = []
if not needles:
needles = [u]
filtered: List[str] = []
for n in needles:
n2 = str(n or "").strip()
if not n2:
continue
if not _httpish(n2):
continue
if n2 not in filtered:
filtered.append(n2)
url_needles[u] = filtered if filtered else [u]
backend_names: List[str] = []
try:
backend_names_all = storage.list_searchable_backends()
except Exception:
backend_names_all = []
for backend_name in backend_names_all:
try:
backend = storage[backend_name]
except Exception:
continue
try:
if str(backend_name).strip().lower() == "temp":
continue
except Exception:
pass
try:
backend_location = getattr(backend, "_location", None)
if backend_location and final_output_dir:
backend_path = Path(str(backend_location)).expanduser().resolve()
temp_path = Path(str(final_output_dir)).expanduser().resolve()
if backend_path == temp_path:
continue
except Exception:
pass
backend_names.append(backend_name)
if not backend_names:
debug("Bulk URL preflight skipped: no searchable backends")
return True
seen_pairs: set[tuple[str, str]] = set()
matched_urls: set[str] = set()
match_rows: List[Dict[str, Any]] = []
max_rows = 200
try:
from Store.HydrusNetwork import HydrusNetwork
except Exception:
HydrusNetwork = None # type: ignore
for backend_name in backend_names:
if len(match_rows) >= max_rows:
break
try:
backend = storage[backend_name]
except Exception:
continue
if HydrusNetwork is not None and isinstance(backend, HydrusNetwork):
if not hydrus_available:
continue
client = getattr(backend, "_client", None)
if client is None:
continue
for original_url, needles in url_needles.items():
if len(match_rows) >= max_rows:
break
if (original_url, str(backend_name)) in seen_pairs:
continue
found_hash: Optional[str] = None
found = False
for needle in (needles or [])[:3]:
if not _httpish(needle):
continue
try:
from API.HydrusNetwork import HydrusRequestSpec
spec = HydrusRequestSpec(
method="GET",
endpoint="/add_urls/get_url_files",
query={"url": needle},
)
# Access internal client safely if possible, else skip check
if hasattr(client, "_perform_request"):
response = client._perform_request(spec)
raw_hashes = None
if isinstance(response, dict):
raw_hashes = response.get("hashes") or response.get("file_hashes")
raw_ids = response.get("file_ids")
has_ids = isinstance(raw_ids, list) and len(raw_ids) > 0
has_hashes = isinstance(raw_hashes, list) and len(raw_hashes) > 0
if has_hashes:
try:
found_hash = str(raw_hashes[0]).strip()
except Exception:
found_hash = None
if has_ids or has_hashes:
found = True
break
except Exception:
continue
if not found:
continue
seen_pairs.add((original_url, str(backend_name)))
matched_urls.add(original_url)
display_row = {
"title": "(exists)",
"store": str(backend_name),
"hash": found_hash or "",
"url": original_url,
"columns": [
("Title", "(exists)"),
("Store", str(backend_name)),
("Hash", found_hash or ""),
("URL", original_url),
],
}
match_rows.append(display_row)
continue
for original_url, needles in url_needles.items():
if len(match_rows) >= max_rows:
break
if (original_url, str(backend_name)) in seen_pairs:
continue
backend_hits: List[Dict[str, Any]] = []
for needle in (needles or [])[:3]:
try:
backend_hits = backend.search(f"url:{needle}", limit=1) or []
if backend_hits:
break
except Exception:
continue
if not backend_hits:
continue
seen_pairs.add((original_url, str(backend_name)))
matched_urls.add(original_url)
hit = backend_hits[0]
title = hit.get("title") or hit.get("name") or hit.get("target") or hit.get("path") or "(exists)"
file_hash = hit.get("hash") or hit.get("file_hash") or hit.get("sha256") or ""
try:
from SYS.result_table import build_display_row
extracted = build_display_row(hit, keys=["title", "store", "hash", "ext", "size"])
except Exception:
extracted = {}
extracted["title"] = str(title)
extracted["store"] = str(hit.get("store") or backend_name)
extracted["hash"] = str(file_hash or "")
ext = extracted.get("ext")
size_val = extracted.get("size")
display_row = {
"title": str(title),
"store": str(hit.get("store") or backend_name),
"hash": str(file_hash or ""),
"ext": str(ext or ""),
"size": size_val,
"url": original_url,
"columns": [
("Title", str(title)),
("Store", str(hit.get("store") or backend_name)),
("Hash", str(file_hash or "")),
("Ext", str(ext or "")),
("Size", size_val),
("URL", original_url),
],
}
match_rows.append(display_row)
if not match_rows:
debug("Bulk URL preflight: no matches")
return True
table = ResultTable(f"URL already exists ({len(matched_urls)} url(s))", max_columns=10)
table.set_no_choice(True)
try:
table.set_preserve_order(True)
except Exception:
pass
for row in match_rows:
table.add_result(row)
try:
pipeline_context.set_last_result_table_overlay(table, match_rows)
except Exception:
pass
suspend = getattr(pipeline_context, "suspend_live_progress", None)
cm: AbstractContextManager[Any] = nullcontext()
if callable(suspend):
try:
maybe_cm = suspend()
if maybe_cm is not None:
cm = maybe_cm # type: ignore[assignment]
except Exception:
cm = nullcontext()
with cm:
get_stderr_console().print(table)
setattr(table, "_rendered_by_cmdlet", True)
answered_yes = bool(Confirm.ask("Continue anyway?", default=False, console=get_stderr_console()))
if in_pipeline:
try:
existing = pipeline_context.load_value("preflight", default=None)
except Exception:
existing = None
preflight_cache: Dict[str, Any] = existing if isinstance(existing, dict) else {}
url_dup_cache = preflight_cache.get("url_duplicates")
if not isinstance(url_dup_cache, dict):
url_dup_cache = {}
url_dup_cache["command"] = str(current_cmd_text or "")
url_dup_cache["continue"] = bool(answered_yes)
preflight_cache["url_duplicates"] = url_dup_cache
try:
pipeline_context.store_value("preflight", preflight_cache)
except Exception:
pass
if not answered_yes:
if in_pipeline:
try:
pipeline_context.request_pipeline_stop(reason="duplicate-url declined", exit_code=0)
except Exception:
pass
return False
return True
Reference in New Issue View Git Blame Copy Permalink