kl

API/Tidal.py

@@ -3,6 +3,7 @@ from __future__ import annotations
 from typing import Any, Dict, List, Optional, Set
 
 from .base import API, ApiError
+from SYS.logger import debug
 
 DEFAULT_BASE_URL = "https://tidal-api.binimum.org"
 
@@ -241,21 +242,24 @@ class Tidal(API):
 
         # 1. Fetch info (metadata) - fetch raw to ensure all fields are available for merging
         info_resp = self._get_json("info/", params={"id": track_int})
+        debug(f"[API.Tidal] info_resp (len={len(str(info_resp))}): {info_resp}")
         info_data = info_resp.get("data") if isinstance(info_resp, dict) else info_resp
         if not isinstance(info_data, dict) or "id" not in info_data:
             info_data = info_resp if isinstance(info_resp, dict) and "id" in info_resp else {}
 
         # 2. Fetch track (manifest/bit depth)
         track_resp = self.track(track_id)
+        debug(f"[API.Tidal] track_resp (len={len(str(track_resp))}): {track_resp}")
         # Note: track() method in this class currently returns raw JSON, so we handle it similarly.
         track_data = track_resp.get("data") if isinstance(track_resp, dict) else track_resp
-        if not isinstance(track_data, dict) or "id" not in track_data:
+        if not isinstance(track_data, dict):
-            track_data = track_resp if isinstance(track_resp, dict) and "id" in track_resp else {}
+            track_data = track_resp if isinstance(track_resp, dict) else {}
 
         # 3. Fetch lyrics
         lyrics_data = {}
         try:
             lyr_resp = self.lyrics(track_id)
+            debug(f"[API.Tidal] lyrics_resp (len={len(str(lyr_resp))}): {lyr_resp}")
             lyrics_data = lyr_resp.get("lyrics") or lyr_resp if isinstance(lyr_resp, dict) else {}
         except Exception:
             pass
@@ -267,18 +271,24 @@ class Tidal(API):
         if isinstance(track_data, dict):
             merged_md.update(track_data)
 
+        debug(f"[API.Tidal] merged_md keys: {list(merged_md.keys())}")
+
         # Derived tags and normalized/parsed info
         tags = build_track_tags(merged_md)
+        debug(f"[API.Tidal] generated tags: {tags}")
         parsed_info = parse_track_item(merged_md)
 
         # Structure for return
-        return {
+        res = {
             "metadata": merged_md,
             "parsed": parsed_info,
             "tags": list(tags),
             "lyrics": lyrics_data,
         }
+        debug(f"[API.Tidal] returning full_track_metadata keys: {list(res.keys())}")
+        return res
 
 
 # Legacy alias for TidalApiClient
 TidalApiClient = Tidal
+HifiApiClient = Tidal

CLI.py

@@ -3731,18 +3731,32 @@ class PipelineExecutor:
                             if emits:
                                 try:
                                     from cmdlet import _shared as sh
+                                    from SYS import models
 
+                                    # 1. Apply -path persistence (moves temp files to final destination)
                                     emits = sh.apply_output_path_from_pipeobjects(
                                         cmd_name=cmd_name,
                                         args=list(stage_args),
                                         emits=emits,
                                     )
+
+                                    # 2. METADATA STICKINESS / PROPAGATION
+                                    # We normalize all emitted items and merge metadata/tags from the previous stage.
+                                    # This ensures info like track titles/lyrics survive downloads/conversions.
+                                    # See cmdlet._shared.propagate_metadata for the merge logic.
+                                    prev_items = piped_result
+                                    if not isinstance(prev_items, (list, tuple)):
+                                        prev_items = [prev_items] if prev_items else []
+
+                                    emits = sh.propagate_metadata(prev_items, emits)
+
                                     try:
                                         pipeline_ctx.emits = list(emits)
                                     except Exception:
                                         pass
                                 except Exception:
                                     pass
+
                             if emits:
                                 # If the cmdlet already installed an overlay table (e.g. get-tag),
                                 # don't overwrite it: set_last_result_items_only() would clear the

Provider/HIFI.py

@@ -13,13 +13,15 @@ from typing import Any, Dict, Iterable, List, Optional, Tuple
 from urllib.parse import urlparse
 
 from API.Tidal import (
-    HifiApiClient,
+    Tidal as TidalApiClient,
     build_track_tags,
     coerce_duration_seconds,
     extract_artists,
     stringify,
 )
 from ProviderCore.base import Provider, SearchResult, parse_inline_query_arguments
+from ProviderCore.inline_utils import collect_choice
+from cmdlet._shared import get_field
 from SYS import pipeline as pipeline_context
 from SYS.logger import debug, log
 
@@ -64,7 +66,9 @@ def _format_total_seconds(seconds: Any) -> str:
     return f"{mins}:{secs:02d}"
 
 
-class Tidal(Provider):
+class HIFI(Provider):
+
+    PROVIDER_NAME = "hifi"
 
     TABLE_AUTO_STAGES = {
         "hifi.track": ["download-file"],
@@ -97,7 +101,7 @@ class Tidal(Provider):
             self.api_timeout = float(self.config.get("timeout", 10.0))
         except Exception:
             self.api_timeout = 10.0
-        self.api_clients = [HifiApiClient(base_url=url, timeout=self.api_timeout) for url in self.api_urls]
+        self.api_clients = [TidalApiClient(base_url=url, timeout=self.api_timeout) for url in self.api_urls]
 
     def extract_query_arguments(self, query: str) -> Tuple[str, Dict[str, Any]]:
         normalized, parsed = parse_inline_query_arguments(query)
@@ -281,7 +285,7 @@ class Tidal(Provider):
         if isinstance(detail, dict):
             title = self._stringify(detail.get("title")) or title
 
-        return SearchResult(
+        res = SearchResult(
             table="hifi.track",
             title=title,
             path=f"hifi://track/{track_id}",
@@ -291,6 +295,12 @@ class Tidal(Provider):
             full_metadata=dict(detail) if isinstance(detail, dict) else {},
             selection_args=["-url", f"hifi://track/{track_id}"],
         )
+        if isinstance(detail, dict):
+            try:
+                res.tag = self._build_track_tags(detail)
+            except Exception:
+                pass
+        return res
 
     def _extract_artist_selection_context(self, selected_items: List[Any]) -> List[Tuple[int, str]]:
         contexts: List[Tuple[int, str]] = []
@@ -1130,25 +1140,36 @@ class Tidal(Provider):
             md = dict(getattr(result, "full_metadata") or {})
 
         track_id = self._extract_track_id_from_result(result)
-        if track_id:
+        debug(f"[hifi] download: track_id={track_id}, manifest_present={bool(md.get('manifest'))}, tag_count={len(result.tag) if result.tag else 0}")
+        
+        # Enrichment: fetch full metadata if manifest or detailed info (like tags/lyrics) is missing.
+        # We check for 'manifest' because it's required for DASH playback.
+        # We also check for lyrics/subtitles to ensure they are available for add-file.
+        has_lyrics = bool(md.get("_tidal_lyrics_subtitles")) or bool(md.get("lyrics"))
+
+        if track_id and (not md.get("manifest") or not md.get("artist") or len(result.tag or []) <= 1 or not has_lyrics):
+            debug(f"[hifi] Enriching track data (reason: manifest={not md.get('manifest')}, lyrics={not has_lyrics}, tags={len(result.tag or [])})")
             # Multi-part enrichment from API: metadata, tags, and lyrics.
             full_data = self._fetch_all_track_data(track_id)
+            debug(f"[hifi] download: enrichment full_data present={bool(full_data)}")
             if isinstance(full_data, dict):
                 # 1. Update metadata
                 api_md = full_data.get("metadata")
                 if isinstance(api_md, dict):
+                    debug(f"[hifi] download: updating metadata with {len(api_md)} keys")
                     md.update(api_md)
                 
                 # 2. Update tags (re-sync result.tag so cmdlet sees them)
                 api_tags = full_data.get("tags")
+                debug(f"[hifi] download: enrichment tags={api_tags}")
                 if isinstance(api_tags, list) and api_tags:
                     result.tag = set(api_tags)
 
                 # 3. Handle lyrics
-                lyrics = full_data.get("lyrics")
+                lyrics_dict = full_data.get("lyrics")
-                if isinstance(lyrics, dict) and lyrics:
+                if isinstance(lyrics_dict, dict) and lyrics_dict:
-                    md.setdefault("lyrics", lyrics)
+                    md.setdefault("lyrics", lyrics_dict)
-                    subtitles = lyrics.get("subtitles")
+                    subtitles = lyrics_dict.get("subtitles")
                     if isinstance(subtitles, str) and subtitles.strip():
                         md["_tidal_lyrics_subtitles"] = subtitles.strip()
 
@@ -1328,7 +1349,7 @@ class Tidal(Provider):
 
         return False, None
 
-    def _get_api_client_for_base(self, base_url: str) -> Optional[HifiApiClient]:
+    def _get_api_client_for_base(self, base_url: str) -> Optional[TidalApiClient]:
         base = base_url.rstrip("/")
         for client in self.api_clients:
             if getattr(client, "base_url", "").rstrip("/") == base:
@@ -1739,6 +1760,10 @@ class Tidal(Provider):
                 or payload.get("path")
                 or payload.get("url")
             )
+            # Guard against method binding (e.g. str.title) being returned by getattr(str, "title")
+            if callable(title):
+                title = None
+
             if not title:
                 title = f"Track {track_id}"
             path = (
@@ -1983,12 +2008,6 @@ class Tidal(Provider):
 
             return True
 
-        # Optimization: If we are selecting tracks, do NOT force a "Detail View" (resolving manifest) here.
-        # This allows batch selection to flow immediately to `download-file` (via TABLE_AUTO_STAGES)
-        # or other downstream cmdlets. The download logic (HIFI.download) handles manifest resolution locally.
-        if table_type == "hifi.track" or (is_generic_hifi and any(str(get_field(i, "path")).startswith("hifi://track/") for i in selected_items)):
-             return False
-
         contexts = self._extract_track_selection_context(selected_items)
         try:
             debug(f"[hifi.selector] track contexts={len(contexts)}")

Provider/internetarchive.py

@@ -501,6 +501,26 @@ class InternetArchive(Provider):
         "internetarchive.formats": ["download-file"],
     }
 
+    def maybe_show_picker(
+        self,
+        *,
+        url: str,
+        item: Optional[Any] = None,
+        parsed: Dict[str, Any],
+        config: Dict[str, Any],
+        quiet_mode: bool,
+    ) -> Optional[int]:
+        """Generic hook for download-file to show a selection table for IA items."""
+        from cmdlet._shared import get_field as sh_get_field
+        return maybe_show_formats_table(
+            raw_urls=[url] if url else [],
+            piped_items=[item] if item else [],
+            parsed=parsed,
+            config=config,
+            quiet_mode=quiet_mode,
+            get_field=sh_get_field,
+        )
+
     def __init__(self, config: Optional[Dict[str, Any]] = None):
         super().__init__(config)
         conf = _pick_provider_config(self.config)

cmdlet/_shared.py

@@ -11,11 +11,16 @@ import sys
 import tempfile
 from collections.abc import Iterable as IterableABC
 
-from SYS.logger import log
+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
@@ -2405,6 +2410,117 @@ def coerce_to_pipe_object(
     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,
@@ -2518,12 +2634,12 @@ def resolve_tidal_manifest_path(item: Any) -> Optional[str]:
 
             if candidate_path:
                 m = re.search(
-                    r"tidal:(?://)?track[\\/](\d+)",
+                    r"(tidal|hifi):(?://)?track[\\/](\d+)",
                     str(candidate_path),
                     flags=re.IGNORECASE,
                 )
                 if m:
-                    track_id = m.group(1)
+                    track_id = m.group(2)
 
         if (not already) and track_id is not None:
             try:
@@ -2706,3 +2822,327 @@ def resolve_tidal_manifest_path(item: Any) -> Optional[str]:
         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
+

cmdlet/add_note.py

@@ -209,11 +209,8 @@ class Add_Note(Cmdlet):
         note_name = str(note_name or "").strip()
         note_text = str(note_text or "").strip()
         if not note_name or not note_text:
-            log(
+            pass # We now support implicit pipeline notes if -query is missing
-                "[add_note] Error: -query must include title:<title> and text:<text>",
+            # But if explicit targeting (store+hash) is used, we still demand args below.
-                file=sys.stderr,
-            )
-            return 1
 
         if hash_override and not store_override:
             log(
@@ -224,6 +221,14 @@ class Add_Note(Cmdlet):
 
         explicit_target = bool(hash_override and store_override)
         results = normalize_result_input(result)
+
+        if explicit_target and (not note_name or not note_text):
+            log(
+                "[add_note] Error: Explicit target (store+hash) requires -query with title/text",
+                file=sys.stderr,
+            )
+            return 1
+
         if results and explicit_target:
             # Direct targeting mode: apply note once to the explicit target and
             # pass through any piped items unchanged.
@@ -287,7 +292,36 @@ class Add_Note(Cmdlet):
                 ctx.emit(res)
                 continue
 
-            item_note_text = note_text
+            # Determine notes to write for this item
+            notes_to_write: List[Tuple[str, str]] = []
+            
+            # 1. Explicit arguments always take precedence
+            if note_name and note_text:
+                notes_to_write.append((note_name, note_text))
+            
+            # 2. Pipeline notes auto-ingestion
+            # Look for 'notes' dictionary in the item (propagated by pipeline/download-file)
+            # Structure: {'notes': {'lyric': '...', 'sub': '...'}}
+            # Check both root and nested 'extra'
+            
+            # Check root 'notes' (dict or extra.notes)
+            pipeline_notes = res.get("notes")
+            if not isinstance(pipeline_notes, dict):
+                extra = res.get("extra")
+                if isinstance(extra, dict):
+                    pipeline_notes = extra.get("notes")
+                    
+            if isinstance(pipeline_notes, dict):
+                for k, v in pipeline_notes.items():
+                    # If arg-provided note conflicts effectively with pipeline note? 
+                    # We just append both.
+                    if v and str(v).strip():
+                        notes_to_write.append((str(k), str(v)))
+
+            if not notes_to_write:
+                # Pass through items that have nothing to add
+                ctx.emit(res)
+                continue
 
             store_name = str(store_override or res.get("store") or "").strip()
             raw_hash = res.get("hash")
@@ -298,7 +332,7 @@ class Add_Note(Cmdlet):
                     "[add_note] Error: Missing -store and item has no store field",
                     file=sys.stderr
                 )
-                return 1
+                continue
 
             resolved_hash = self._resolve_hash(
                 raw_hash=str(raw_hash) if raw_hash else None,
@@ -313,79 +347,42 @@ class Add_Note(Cmdlet):
                 ctx.emit(res)
                 continue
             
-            try:
+            # Queue operations
-                backend = store_registry[store_name]
+            if store_name not in note_ops:
-            except Exception as exc:
+                note_ops[store_name] = []
-                log(
-                    f"[add_note] Error: Unknown store '{store_name}': {exc}",
-                    file=sys.stderr
-                )
-                return 1
             
-            # Queue for bulk write per store. We still emit items immediately;
+            for (n_name, n_text) in notes_to_write:
-            # the pipeline only advances after this cmdlet returns.
+                note_ops[store_name].append((resolved_hash, n_name, n_text))
-            note_ops.setdefault(store_name,
+                planned_ops += 1
-                                []).append((resolved_hash,
-                                            note_name,
-                                            item_note_text))
-            planned_ops += 1
                 
             ctx.emit(res)
 
-        # Execute bulk writes per store.
+
-        successful_writes = 0
+        # Execute batch operations
+        success_count = 0
         for store_name, ops in note_ops.items():
-            if not ops:
-                continue
             try:
                 backend = store_registry[store_name]
-            except Exception:
+                if not hasattr(backend, "set_note"):
-                continue
+                     log(f"[add_note] Store '{store_name}' does not support notes", file=sys.stderr)
+                     continue
                      
-            store_success = 0
+                for (h, name, text) in ops:
-            bulk_fn = getattr(backend, "set_note_bulk", None)
+                    try:
-            if callable(bulk_fn):
+                        if backend.set_note(h, name, text, config=config):
-                try:
+                            success_count += 1
-                    ok = bool(bulk_fn(list(ops), config=config))
+                    except Exception as e:
-                    if ok:
+                        log(f"[add_note] Write failed {store_name}:{h} ({name}): {e}", file=sys.stderr)
-                        store_success += len(ops)
-                        ctx.print_if_visible(
-                            f"✓ add-note: {len(ops)} item(s) in '{store_name}'",
-                            file=sys.stderr
-                        )
-                        successful_writes += store_success
-                        continue
-                    log(
-                        f"[add_note] Warning: bulk set_note returned False for '{store_name}'",
-                        file=sys.stderr,
-                    )
-                except Exception as exc:
-                    log(
-                        f"[add_note] Warning: bulk set_note failed for '{store_name}': {exc}; falling back",
-                        file=sys.stderr,
-                    )
                         
-            # Fallback: per-item writes
+            except Exception as e:
-            for file_hash, name, text in ops:
+                 log(f"[add_note] Store access failed '{store_name}': {e}", file=sys.stderr)
-                try:
-                    ok = bool(backend.set_note(file_hash, name, text, config=config))
-                    if ok:
-                        store_success += 1
-                except Exception:
-                    continue
 
-            if store_success:
+        if planned_ops > 0:
-                successful_writes += store_success
+            msg = f"✓ add-note: Updated {success_count}/{planned_ops} notes across {len(note_ops)} stores"
-                ctx.print_if_visible(
+            ctx.print_if_visible(msg, file=sys.stderr)
-                    f"✓ add-note: {store_success} item(s) in '{store_name}'",
-                    file=sys.stderr
-                )
             
-        log(
+        return 0
-            f"[add_note] Updated {successful_writes}/{planned_ops} item(s)",
-            file=sys.stderr
-        )
-        return 0 if successful_writes > 0 else 1
 
 
 CMDLET = Add_Note()
+