Clean commit: only Nuke-related changes

Author: Dallin Clark

pipeline/software/nuke/tools/BobukeTools/menu.py

@@ -21,6 +21,11 @@ def make_bobo_fx_read_node():
 
     # run the FX read now
     bobo_read_node.auto_read_latest_fx_exr()
+def make_bobo_cfx_read_node():
+    import bobo_read_node  # type: ignore[import-not-found]
+
+    # run the FX read now
+    bobo_read_node.auto_read_latest_cfx_exr()
 
 
 def make_bobo_write_node():
@@ -111,7 +116,7 @@ def set_frameRange_and_aspectRatio():
 m.addCommand("Bobo Read Node", "make_bobo_read_node()", icon="BobukeIcon.png")
 
 m.addCommand("Bobo FX Read", "make_bobo_fx_read_node()", icon="BobukeIcon.png")
-
+m.addCommand("Bobo CFX Read", "make_bobo_cfx_read_node()", icon="BobukeIcon.png")
 
 ################################### Nungeon Shelf Tool Buttons ###################################
 menu = nuke.menu("Nuke")

pipeline/software/nuke/tools/BobukeTools/scripts/bobo_read_node.py

@@ -2,21 +2,83 @@
 import os
 import re
 import glob
+import math
+from functools import reduce
 from datetime import datetime
 
 
-def get_latest_exr_sequence(render_root):
+def _find_images_dir(base):
+    """Return images dir under base, preferring images_dn over images."""
+    for sub in ("images_dn", "images"):
+        p = os.path.join(base, sub)
+        if os.path.isdir(p):
+            return p
+    return None
+
+
+def _gcd_list(values):
+    """Greatest common divisor for a list of positive integers."""
+    if not values:
+        return 1
+    return reduce(math.gcd, values)
+
+
+def _scan_exr_sequence(images_dir):
     """
-    Scan render_root for MM-DD-YYYY_V_### folders and, inside the newest one,
-    discover all .exr files, extract their frame numbers & padding, and return:
-      (pattern, first_frame, last_frame)
-    where pattern is something like '/…/images/%04d.exr'
+    Return (pattern, first_frame, last_frame, pad, cadence_step)
+    cadence_step is the GCD of frame gaps (1 = full, 2 = on 2s, 4 = on 4s, etc.)
+    """
+    files = glob.glob(os.path.join(images_dir, "*.exr"))
+    if not files:
+        return None
+
+    nums, pad = [], 0
+    for f in files:
+        fname = os.path.basename(f)
+        m = re.match(r"^(.*?)(\d+)\.exr$", fname)
+        if not m:
+            continue
+        n = int(m.group(2))
+        nums.append(n)
+        pad = max(pad, len(m.group(2)))
+
+    if not nums:
+        return None
+
+    nums = sorted(set(nums))
+    first_frame = nums[0]
+    last_frame = nums[-1]
+    diffs = [b - a for a, b in zip(nums, nums[1:]) if b > a]
+    step = _gcd_list(diffs) if diffs else 1
+    if step <= 0:
+        step = 1
+
+    pattern = os.path.join(images_dir, "%%0%dd.exr" % pad)
+    return pattern, first_frame, last_frame, pad, step
+
+
+def get_latest_exr_sequences(render_root):
+    """
+    Discover EXR sequences in the newest MM-DD-YY_V_### directory.
+
+    - If subfolders exist under the newest date/version dir:
+        scan each subfolder; in each, prefer images_dn/ over images/
+    - If no subfolders:
+        look directly under the newest date/version dir for images_dn/ or images/
+
+    Returns a list of dicts:
+      [{
+        'label': <subfolder or 'root'>,
+        'pattern': '/…/%04d.exr',
+        'first': int,
+        'last': int,
+        'step': int   # 1 (full), 2 (on 2s), 4 (on 4s), etc.
+      }, ...]
     """
     if not os.path.isdir(render_root):
         nuke.message(f"[Auto Read] Render folder does not exist:\n  {render_root}")
-        return None, None, None
+        return []
 
-    # find latest date/version subfolder
     folder_pattern = re.compile(r"^(\d{2}-\d{2}-\d{2})_V_(\d{3})$")
     candidates = []
     for name in os.listdir(render_root):
@@ -28,77 +90,169 @@ def get_latest_exr_sequence(render_root):
             candidates.append((date_obj, version, path))
     if not candidates:
         nuke.message("[Auto Read] No valid render subfolders found.")
-        return None, None, None
+        return []
 
-    # pick newest, then check for images_dn/images
     _, _, newest = sorted(candidates, key=lambda x: (x[0], x[1]))[-1]
-    for sub in ("images_dn", "images"):
-        subdir = os.path.join(newest, sub)
-        if os.path.isdir(subdir):
-            newest = subdir
-            break
 
-    # glob all .exr files
-    files = glob.glob(os.path.join(newest, "*.exr"))
-    if not files:
-        nuke.message(f"[Auto Read] No .exr files found in:\n  {newest}")
-        return None, None, None
+    subfolders = [d for d in os.listdir(newest)
+                  if os.path.isdir(os.path.join(newest, d))]
 
-    # extract frame numbers and detect padding
-    nums, pad = [], 0
-    for f in files:
-        fname = os.path.basename(f)
-        m = re.match(r"^(.*?)(\d+)\.exr$", fname)
-        if not m:
-            continue
-        nums.append(int(m.group(2)))
-        pad = max(pad, len(m.group(2)))
+    sequences = []
 
-    if not nums:
-        nuke.message(f"[Auto Read] Couldn't parse frame numbers in:\n  {newest}")
-        return None, None, None
+    if subfolders:
+        for sub in sorted(subfolders):
+            base = os.path.join(newest, sub)
+            images_dir = _find_images_dir(base)
+            if not images_dir:
+                continue
+            seq = _scan_exr_sequence(images_dir)
+            if not seq:
+                continue
+            pattern, first, last, _pad, step = seq
+            sequences.append({
+                "label": sub,
+                "pattern": pattern,
+                "first": first,
+                "last": last,
+                "step": step
+            })
+        if not sequences:
+            nuke.message(f"[Auto Read] No .exr sequences found under any subfolder in:\n  {newest}")
+            return []
+    else:
+        images_dir = _find_images_dir(newest)
+        if not images_dir:
+            nuke.message(f"[Auto Read] Neither images_dn nor images found in:\n  {newest}")
+            return []
+        seq = _scan_exr_sequence(images_dir)
+        if not seq:
+            nuke.message(f"[Auto Read] No .exr files found in:\n  {images_dir}")
+            return []
+        pattern, first, last, _pad, step = seq
+        sequences.append({
+            "label": "root",
+            "pattern": pattern,
+            "first": first,
+            "last": last,
+            "step": step
+        })
+
+    return sequences
 
-    first_frame = min(nums)
-    last_frame = max(nums)
-    pattern = os.path.join(newest, "%%0%dd.exr" % pad)
 
-    return pattern, first_frame, last_frame
+def _sanitize_for_nuke(name):
+    """Keep node names safe for Nuke."""
+    safe = re.sub(r"[^0-9a-zA-Z_]", "_", name)
+    if safe and safe[0].isdigit():
+        safe = "_" + safe
+    return safe or "seq"
 
 
-def make_read_node(render_subdir="render", node_name="EXR_read"):
+def _nearest_hold_expr(first, last, step):
     """
-    Create a Read node pointing at the full sequence (from first_frame to last_frame).
+    Expression that maps timeline frame -> nearest existing frame in a stepped sequence.
+    Uses floor(x + 0.5) to emulate round() for stability.
+    """
+    if step <= 1:
+        # identity mapping (no missing frames cadence)
+        return f"clamp(frame, {first}, {last})"
+    # nearest multiple of 'step' from 'first', then clamp to [first,last]
+    return f"clamp({first}+{step}*floor((frame-{first})/{step}+0.5), {first}, {last})"
+
+
+def make_read_nodes(render_subdir="render", node_name_prefix="EXR_read"):
+    """
+    Create one Read node per discovered sequence inside the newest date/version folder.
+
+    - Nodes are named: <prefix>_<label> (e.g., EXR_read_beauty)
+    - Project frame range is set to the union [min(first), max(last)] across all sequences.
+    - For sequences detected as rendered on 2s/4s (or any N-s cadence), the Read node's
+      'frame' knob is set to hold the nearest available frame so playback never errors.
+    - If ALL sequences share the same cadence of 2 or 4, project FPS is divided by that cadence.
     """
     script_path = nuke.root()["name"].value()
     if not script_path or script_path == "Root":
         nuke.message("Open your shot before using Auto Read.")
-        return None
+        return []
 
     shot_dir = os.path.dirname(os.path.dirname(script_path))
     render_dir = os.path.join(shot_dir, render_subdir)
 
-    seq_pattern, first, last = get_latest_exr_sequence(render_dir)
-    if not seq_pattern:
-        return None
+    sequences = get_latest_exr_sequences(render_dir)
+    if not sequences:
+        return []
+
+    # remove any existing nodes created by this tool (by prefix)
     for n in nuke.allNodes("Read"):
-        if n.name() == node_name:
-            nuke.delete(n)
-    read = nuke.nodes.Read(name=node_name, file=seq_pattern, on_error="black")
-    # set both the sequence's native range and the playback range
-    read["origfirst"].setValue(first)
-    read["origlast"].setValue(last)
-    read["first"].setValue(first)
-    read["last"].setValue(last)
+        try:
+            if n.name().startswith(node_name_prefix):
+                nuke.delete(n)
+        except Exception:
+            pass
+
+    created = []
+    global_first = min(s["first"] for s in sequences)
+    global_last = max(s["last"] for s in sequences)
+
+    for s in sequences:
+        label = _sanitize_for_nuke(s["label"])
+        node_name = f"{node_name_prefix}_{label}" if label != "root" else node_name_prefix
+        read = nuke.nodes.Read(name=node_name, file=s["pattern"], on_error="black")
+        # native sequence range
+        read["origfirst"].setValue(s["first"])
+        read["origlast"].setValue(s["last"])
+        read["first"].setValue(s["first"])
+        read["last"].setValue(s["last"])
+
+        # Time-map to handle sparse cadence (2s/4s/etc.) by holding nearest available frame
+        expr = _nearest_hold_expr(s["first"], s["last"], s["step"])
+        try:
+            read["frame"].setExpression(expr)
+        except Exception:
+            # Fallback: if 'frame' knob is unavailable for some reason, do nothing
+            pass
+
+        # Optional label so it's obvious what's happening
+        try:
+            read["label"].setValue(f"{s['label']}  step:{s['step']}")
+        except Exception:
+            pass
+
+        created.append(read)
+
+    # Set the project frame range to cover all created sequences
+    nuke.root()["first_frame"].setValue(global_first)
+    nuke.root()["last_frame"].setValue(global_last)
 
-    nuke.root()["first_frame"].setValue(first)
-    nuke.root()["last_frame"].setValue(last)
+    # If ALL sequences share cadence 2 or 4, adjust project FPS accordingly
+    steps = set(s["step"] for s in sequences)
+    if len(steps) == 1:
+        only_step = steps.pop()
+        if only_step in (2, 4):
+            try:
+                current_fps = float(nuke.root()["fps"].value())
+                new_fps = current_fps / float(only_step)
+                nuke.root()["fps"].setValue(new_fps)
+                nuke.tprint(f"[Auto Read] Detected cadence {only_step}s; FPS set to {new_fps:.3f}")
+            except Exception as e:
+                nuke.tprint(f"[Auto Read] Could not adjust FPS: {e}")
 
-    return read
+    return created
 
 
 def auto_read_latest_fx_exr():
-    read_node = make_read_node("FX/render", node_name="Bobo_FX_read")
-    if not read_node:
+    nodes = make_read_nodes("fx/render", node_name_prefix="Bobo_FX_read")
+    if not nodes:
+        return
+    try:
+        viewer = nuke.activeViewer().node()
+        nuke.zoom(1, [viewer["xpos"].value(), viewer["ypos"].value()])
+    except Exception as e:
+        nuke.tprint(f"[Auto Read] Viewer zoom error: {e}")
+
+def auto_read_latest_cfx_exr():
+    nodes = make_read_nodes("cfx/render", node_name_prefix="Bobo_CFX_read")
+    if not nodes:
         return
     try:
         viewer = nuke.activeViewer().node()
@@ -109,10 +263,10 @@ def auto_read_latest_fx_exr():
 
 def auto_read_latest_exr():
     """
-    Callback: build the Read node and zoom the Viewer.
+    Callback: build the Read nodes and zoom the Viewer.
     """
-    read_node = make_read_node()
-    if not read_node:
+    nodes = make_read_nodes()
+    if not nodes:
         return
     try:
         viewer = nuke.activeViewer().node()
@@ -137,4 +291,4 @@ def main():
     main()
     nuke.tprint("[Auto Read] Module loaded and command registered.")
 except Exception as e:
-    nuke.tprint(f"[Auto Read] Failed to initialize: {e}")
+    nuke.tprint(f"[Auto Read] Failed to initialize: {e}")