micro-reader 0.2.0__tar.gz → 0.3.0__tar.gz

{micro_reader-0.2.0 → micro_reader-0.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: micro-reader
-Version: 0.2.0
+Version: 0.3.0
 Summary: Low-level, memory-bounded pixel reading for microscopy formats: read exactly the rectangle you ask for.
 Project-URL: Homepage, https://github.com/bugraoezdemir/micro-reader
 Project-URL: Source, https://github.com/bugraoezdemir/micro-reader
@@ -56,7 +56,7 @@ Description-Content-Type: text/markdown
 
 **Memory-bounded pixel reader for microscopy images.**
 
-`micro-reader` reads regions of large microscopy files without loading the entire image into memory. It provides a common interface across TIFF, CZI, ND2, Leica, OIR, OIB / OIF, Olympus VSI, Imaris, DeltaVision, MRC, common raster formats, OME companion file sets, and high-content screening plates (Harmony, MetaXpress, IN Cell, CellVoyager).
+`micro-reader` reads regions of large microscopy files without loading the entire image into memory. It provides a common interface across TIFF, CZI, ND2, Leica, OIR, OIB / OIF, Olympus VSI, Imaris, DeltaVision, MRC, common raster formats, OME companion file sets, and high-content screening plates (Harmony, Columbus, MetaXpress, ImageXpress HCS.ai, IN Cell, CellVoyager, CQ1, Opera Flex, scanR, Cellomics).
 
 > **Status: pre-alpha (0.0.x)**
 >
@@ -233,7 +233,7 @@ f.path     # the main file: "cells_R3D.dv"
 f.opened   # what was opened: "cells_R3D.dv.log"
 ```
 
-You can also open the folder. A folder opens the image it belongs to (an NDTiff dataset folder, a Harmony or CellVoyager measurement folder), or the one image it holds. A folder with several images raises an error that lists them.
+You can also open the folder. A folder opens the image it belongs to (an NDTiff dataset folder, a Harmony or CellVoyager measurement folder, a folder of Opera `.flex` files), or the one image it holds. An Opera Flex plate has no index file: any of its `.flex` files opens the whole folder as the plate (`multifile=False` reads only that file). A folder with several images raises an error that lists them.
 
 To sort many files at once, for example everything under a folder you want to convert, `group_files` says which files make one image:
 
@@ -246,7 +246,7 @@ for group in g.groups:
 g.singles           # every other file, as given
 ```
 
-A file that a plate or companion index names is listed only in that index's group. When the main file of a multi-file image is missing, its group has `main=None`.
+A file that a plate or companion index names is listed only in that index's group. When the main file of a multi-file image is missing, its group has `main=None`. Pass `plates=True` to group as `open(..., plates=True)` reads: a folder of ND2 files from one NIS-Elements well-plate run is then one group.
 
 ## Axes
 
@@ -324,18 +324,18 @@ Only axes with size 1 can be removed. `y` and `x` are never removed.
 
 ### RGB samples
 
-By default, RGB samples are represented by the `s` axis.
+`micro_reader.open` turns RGB samples into channels: an RGB image has 3 channels, named "<name> R", "G" and "B" and coloured red, green and blue (a file without a channel name gives three unnamed channels in those colours). An image that has channels already gets c × s of them: "<name> R" ... for each.
 
-You can instead represent RGB as channels:
+To keep RGB samples as their own `s` axis, the way the file stores them:
 
 ```python
 f = micro_reader.open(
     "image.tif",
-    samples="channels",
+    samples="keep",
 )
 ```
 
-If there are existing channels already, RGB samples will fold with them, leading to a total of c × s channels.
+`image.source` is then not needed: with the default, it is the image as stored, with its `s` axis. `image.as_axes(...)` keeps `s` unless asked for `samples="channels"`.
 
 
 ## Ambiguous dimensions
@@ -475,15 +475,32 @@ The plate layout is available through:
 plate = f.plates[0]
 plate.wells[0].name
 # "B2"
+plate.row_names, plate.column_names     # ('A', 'B', ...), ('1', '2', ...)
+part = plate.rows[2:5].columns[3:7]     # a part of the plate: still a plate
 ```
 
+Every level knows what it is and what holds it: `.kind` is `"file"`, `"plate"`, `"well"`
+or `"field"`, and `.parent` is the level above. Accessors go down only: a file holds
+`.plates`, a plate `.wells`, a well `.fields`, a field `.images`. `plate.rows` and
+`plate.columns` slice a plate by position, name or a list of them, as `f.select(row=...,
+column=...)` does; rows and columns are a well's address, not a level.
+
 Images can be reached through the wells, or selected by well:
 
 ```python
 image = plate.wells[0].fields[0].image
-f.select(well=plate.well_index("B2"))
+f.select(well=plate.well_index("B2"))       # by its place in plate.wells
+f.select(well_name="B2")                    # by name
+f.select(row_name="B", column=range(1, 4))  # rows and columns: names, or 0-based numbers
+f.select(well=["B2", "C5"], field=[0, 1])   # a list, tuple, range or set: any of them
+f.select(row=slice(2, None))                # a slice of positions: row C onward
 ```
 
+A plate's places have a position and a name, a keyword each: `plate` / `plate_name`,
+`acquisition` / `acquisition_name`, `well` / `well_name`, `row` / `row_name`, `column` /
+`column_name` (the columns of ome-zarr-pyramid's table). A name given to a position keyword
+is its name (`well="B2"` is `well_name="B2"`).
+
 `f.outline()` shows the plates, wells and fields of a file.
 
 Zeiss CZI plates are read as plates with `plates=True`, and as series otherwise:
@@ -581,7 +598,7 @@ The position is omitted when it cannot be determined reliably.
 | **Micro-Manager 1.4**| image-per-file acquisitions (`metadata.txt` + one TIFF per plane) |
 | **DeltaVision**      | `.dv`, `.r3d` (softWoRx, OMX, Priism)                 |
 | **MRC**              | Common MRC modes and extended headers                 |
-| **Plates (HCS)**     | Harmony, MetaXpress, IN Cell and CellVoyager plates   |
+| **Plates (HCS)**     | Harmony, Columbus exports, MetaXpress, ImageXpress HCS.ai (`.jdce`), IN Cell, CellVoyager, Opera Flex, scanR and Cellomics (`.C01`) plates; with `plates=True`, CZI plates, OME companion and OME-TIFF plates (Yokogawa CQ1), ND2 files whose XY points are named after wells, and folders of ND2 files from one NIS-Elements well-plate run |
 | **JPEG / PNG / BMP** | Common raster image formats                           |
 
 All formats use the same `File`, `Image`, axes, indexing, metadata, and reading interface wherever possible.
@@ -645,6 +662,8 @@ f = micro_reader.open("data_t001.ome.tif", multifile=False)
 
 If the companion is missing or does not list the file, opening the whole set raises an error that names `multifile=False`. With `multifile=False` the file is then read without pixel sizes, with a `CompanionWarning`. A TIFF whose metadata lives in another OME-TIFF (instead of a `.companion.ome`) works the same way, with that OME-TIFF in the companion's place.
 
+A companion that describes a plate (`Plate` / `Well` / `WellSample`) reads as one with `plates=True`, as a CZI plate does: `f.plates` lists its plates and wells, and its images are indexed by plate, acquisition, well and field. RGB images in a companion have a sample axis (`t c z y x s`), as RGB OME-TIFFs do.
+
 Some OME-TIFF sets keep the full OME-XML in their TIFF files instead of a companion. Opening one of those files also reads the whole set, and `multifile=False` reads only the planes stored in that file, with the set's metadata.
 
 Micro-Manager stacks always open whole, also with `multifile=False`: they split their files by size, so one file does not hold a complete part of the image.

{micro_reader-0.2.0 → micro_reader-0.3.0}/README.md

@@ -2,7 +2,7 @@
 
 **Memory-bounded pixel reader for microscopy images.**
 
-`micro-reader` reads regions of large microscopy files without loading the entire image into memory. It provides a common interface across TIFF, CZI, ND2, Leica, OIR, OIB / OIF, Olympus VSI, Imaris, DeltaVision, MRC, common raster formats, OME companion file sets, and high-content screening plates (Harmony, MetaXpress, IN Cell, CellVoyager).
+`micro-reader` reads regions of large microscopy files without loading the entire image into memory. It provides a common interface across TIFF, CZI, ND2, Leica, OIR, OIB / OIF, Olympus VSI, Imaris, DeltaVision, MRC, common raster formats, OME companion file sets, and high-content screening plates (Harmony, Columbus, MetaXpress, ImageXpress HCS.ai, IN Cell, CellVoyager, CQ1, Opera Flex, scanR, Cellomics).
 
 > **Status: pre-alpha (0.0.x)**
 >
@@ -179,7 +179,7 @@ f.path     # the main file: "cells_R3D.dv"
 f.opened   # what was opened: "cells_R3D.dv.log"
 ```
 
-You can also open the folder. A folder opens the image it belongs to (an NDTiff dataset folder, a Harmony or CellVoyager measurement folder), or the one image it holds. A folder with several images raises an error that lists them.
+You can also open the folder. A folder opens the image it belongs to (an NDTiff dataset folder, a Harmony or CellVoyager measurement folder, a folder of Opera `.flex` files), or the one image it holds. An Opera Flex plate has no index file: any of its `.flex` files opens the whole folder as the plate (`multifile=False` reads only that file). A folder with several images raises an error that lists them.
 
 To sort many files at once, for example everything under a folder you want to convert, `group_files` says which files make one image:
 
@@ -192,7 +192,7 @@ for group in g.groups:
 g.singles           # every other file, as given
 ```
 
-A file that a plate or companion index names is listed only in that index's group. When the main file of a multi-file image is missing, its group has `main=None`.
+A file that a plate or companion index names is listed only in that index's group. When the main file of a multi-file image is missing, its group has `main=None`. Pass `plates=True` to group as `open(..., plates=True)` reads: a folder of ND2 files from one NIS-Elements well-plate run is then one group.
 
 ## Axes
 
@@ -270,18 +270,18 @@ Only axes with size 1 can be removed. `y` and `x` are never removed.
 
 ### RGB samples
 
-By default, RGB samples are represented by the `s` axis.
+`micro_reader.open` turns RGB samples into channels: an RGB image has 3 channels, named "<name> R", "G" and "B" and coloured red, green and blue (a file without a channel name gives three unnamed channels in those colours). An image that has channels already gets c × s of them: "<name> R" ... for each.
 
-You can instead represent RGB as channels:
+To keep RGB samples as their own `s` axis, the way the file stores them:
 
 ```python
 f = micro_reader.open(
     "image.tif",
-    samples="channels",
+    samples="keep",
 )
 ```
 
-If there are existing channels already, RGB samples will fold with them, leading to a total of c × s channels.
+`image.source` is then not needed: with the default, it is the image as stored, with its `s` axis. `image.as_axes(...)` keeps `s` unless asked for `samples="channels"`.
 
 
 ## Ambiguous dimensions
@@ -421,15 +421,32 @@ The plate layout is available through:
 plate = f.plates[0]
 plate.wells[0].name
 # "B2"
+plate.row_names, plate.column_names     # ('A', 'B', ...), ('1', '2', ...)
+part = plate.rows[2:5].columns[3:7]     # a part of the plate: still a plate
 ```
 
+Every level knows what it is and what holds it: `.kind` is `"file"`, `"plate"`, `"well"`
+or `"field"`, and `.parent` is the level above. Accessors go down only: a file holds
+`.plates`, a plate `.wells`, a well `.fields`, a field `.images`. `plate.rows` and
+`plate.columns` slice a plate by position, name or a list of them, as `f.select(row=...,
+column=...)` does; rows and columns are a well's address, not a level.
+
 Images can be reached through the wells, or selected by well:
 
 ```python
 image = plate.wells[0].fields[0].image
-f.select(well=plate.well_index("B2"))
+f.select(well=plate.well_index("B2"))       # by its place in plate.wells
+f.select(well_name="B2")                    # by name
+f.select(row_name="B", column=range(1, 4))  # rows and columns: names, or 0-based numbers
+f.select(well=["B2", "C5"], field=[0, 1])   # a list, tuple, range or set: any of them
+f.select(row=slice(2, None))                # a slice of positions: row C onward
 ```
 
+A plate's places have a position and a name, a keyword each: `plate` / `plate_name`,
+`acquisition` / `acquisition_name`, `well` / `well_name`, `row` / `row_name`, `column` /
+`column_name` (the columns of ome-zarr-pyramid's table). A name given to a position keyword
+is its name (`well="B2"` is `well_name="B2"`).
+
 `f.outline()` shows the plates, wells and fields of a file.
 
 Zeiss CZI plates are read as plates with `plates=True`, and as series otherwise:
@@ -527,7 +544,7 @@ The position is omitted when it cannot be determined reliably.
 | **Micro-Manager 1.4**| image-per-file acquisitions (`metadata.txt` + one TIFF per plane) |
 | **DeltaVision**      | `.dv`, `.r3d` (softWoRx, OMX, Priism)                 |
 | **MRC**              | Common MRC modes and extended headers                 |
-| **Plates (HCS)**     | Harmony, MetaXpress, IN Cell and CellVoyager plates   |
+| **Plates (HCS)**     | Harmony, Columbus exports, MetaXpress, ImageXpress HCS.ai (`.jdce`), IN Cell, CellVoyager, Opera Flex, scanR and Cellomics (`.C01`) plates; with `plates=True`, CZI plates, OME companion and OME-TIFF plates (Yokogawa CQ1), ND2 files whose XY points are named after wells, and folders of ND2 files from one NIS-Elements well-plate run |
 | **JPEG / PNG / BMP** | Common raster image formats                           |
 
 All formats use the same `File`, `Image`, axes, indexing, metadata, and reading interface wherever possible.
@@ -591,6 +608,8 @@ f = micro_reader.open("data_t001.ome.tif", multifile=False)
 
 If the companion is missing or does not list the file, opening the whole set raises an error that names `multifile=False`. With `multifile=False` the file is then read without pixel sizes, with a `CompanionWarning`. A TIFF whose metadata lives in another OME-TIFF (instead of a `.companion.ome`) works the same way, with that OME-TIFF in the companion's place.
 
+A companion that describes a plate (`Plate` / `Well` / `WellSample`) reads as one with `plates=True`, as a CZI plate does: `f.plates` lists its plates and wells, and its images are indexed by plate, acquisition, well and field. RGB images in a companion have a sample axis (`t c z y x s`), as RGB OME-TIFFs do.
+
 Some OME-TIFF sets keep the full OME-XML in their TIFF files instead of a companion. Opening one of those files also reads the whole set, and `multifile=False` reads only the planes stored in that file, with the set's metadata.
 
 Micro-Manager stacks always open whole, also with `multifile=False`: they split their files by size, so one file does not hold a complete part of the image.

{micro_reader-0.2.0 → micro_reader-0.3.0}/pyproject.toml

@@ -6,7 +6,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "micro-reader"
-version = "0.2.0"
+version = "0.3.0"
 description = "Low-level, memory-bounded pixel reading for microscopy formats: read exactly the rectangle you ask for."
 readme = "README.md"
 requires-python = ">=3.11"

{micro_reader-0.2.0 → micro_reader-0.3.0}/src/micro_reader/__init__.py

@@ -13,7 +13,7 @@ from ._plate import Acquisition, Field, Plate, Well
 from ._spatial import Axis, Collection, CoordinateSystem, Transformation
 from ._errors import (CompanionWarning, CompositeWarning, CorruptFileError, InvariantError,
                       MicroReaderError, MissingImagesWarning, PlateLayoutWarning,
-                      SkewedGridWarning, UnsupportedFormatError)
+                      SkewedGridWarning, StoredValuesWarning, UnsupportedFormatError)
 from ._memory import MemoryLimitError, memory_limit, set_memory_limit
 from ._io import cache_limit, cached_bytes, set_cache_limit
 from ._pool import decode_threads, set_decode_threads
@@ -30,5 +30,5 @@ __all__ = [
     "MicroReaderError", "UnsupportedFormatError", "CorruptFileError", "InvariantError",
     "MemoryLimitError", "DamagedFileWarning",
     "MissingImagesWarning", "CompanionWarning", "CompositeWarning", "PlateLayoutWarning",
-    "SkewedGridWarning", "__version__",
+    "SkewedGridWarning", "StoredValuesWarning", "__version__",
 ]

{micro_reader-0.2.0 → micro_reader-0.3.0}/src/micro_reader/_axes.py

@@ -12,7 +12,11 @@ guessing:
 - axes the file lacks are inserted with size 1;
 - RGB samples ``s`` stay a trailing ``s`` when asked for, or are folded into
   the channels with ``samples="channels"`` (channel ``k`` = sample ``k % S``
-  of channel ``k // S``, as eubi-bridge stores them);
+  of channel ``k // S``, as eubi-bridge stores them; ``open``'s default).
+  The metadata then lists the folded channels: "<name> R" / "G" / "B" ("A"),
+  tinted red / green / blue (white for A); a single unnamed channel gives
+  unnamed, tinted channels; other sample counts are numbered, untinted -- the
+  names eubi-bridge gives (``sample_channels.sample_channel_labels``);
 - ``squeeze=True`` drops every size-1 axis except ``y`` and ``x``;
   ``squeeze="tz"`` drops just those (each must have size 1);
 - an axis the requested ones cannot hold is an error that names it.
@@ -29,6 +33,10 @@ import numpy as np
 from ._base import Image
 
 SAMPLE_POLICIES = ("keep", "channels")
+#: labels and colours of the channels that samples become (as eubi-bridge's)
+SAMPLE_LABELS = {3: ("R", "G", "B"), 4: ("R", "G", "B", "A")}
+SAMPLE_COLORS = {3: ("#FF0000", "#00FF00", "#0000FF"),
+                 4: ("#FF0000", "#00FF00", "#0000FF", "#FFFFFF")}
 #: the array axes, outermost first; ``s`` (RGB samples) after them
 STANDARD = "tczyx"
 CANONICAL = STANDARD + "s"
@@ -36,6 +44,33 @@ CANONICAL = STANDARD + "s"
 KEPT_BY_SQUEEZE = "yx"
 
 
+def folded_channels(channels, n: int, samples: int) -> tuple:
+    """The channels of *n* channels of *samples* samples each, folded into
+    ``n * samples`` channels, channel-major.  A file that lists them already
+    (Bio-Formats-style: one channel per sample) is kept; otherwise each
+    channel becomes "<name> R" / "G" / "B" ("A"), tinted from ``SAMPLE_COLORS``,
+    keeping its wavelengths; an unnamed one among several is "Channel {c} R";
+    a single unnamed channel gives unnamed, tinted channels; sample counts
+    without labels are numbered ("<name> 0") and untinted."""
+    from ._metadata import Channel
+    channels = tuple(channels)
+    if len(channels) == n * samples:
+        return channels
+    if len(channels) != n:
+        channels = (Channel(),) * n
+    labels = SAMPLE_LABELS.get(samples, tuple(str(i) for i in range(samples)))
+    colors = SAMPLE_COLORS.get(samples) or (None,) * samples
+    single_unnamed = n == 1 and not channels[0].name
+    out = []
+    for c, ch in enumerate(channels):
+        for i in range(samples):
+            name = None if single_unnamed else f"{ch.name or f'Channel {c}'} {labels[i]}"
+            out.append(Channel(name=name, color=colors[i],
+                               emission_wavelength=ch.emission_wavelength,
+                               excitation_wavelength=ch.excitation_wavelength))
+    return tuple(out)
+
+
 class AxesImage(Image):
     """*source* presented with *axes* (see the module docstring)."""
 
@@ -165,8 +200,14 @@ class AxesImage(Image):
 
     @property
     def metadata(self):
-        """The native image's metadata (channels follow its own c axis)."""
-        return self.source.metadata
+        """The native image's metadata; with samples folded into the channels,
+        one channel per folded channel (``folded_channels``)."""
+        meta = self.source.metadata
+        if not self._fold:
+            return meta
+        from dataclasses import replace
+        n = self.source.shape[self._names.index("c")] if "c" in self._names else 1
+        return replace(meta, channels=folded_channels(meta.channels, n, self._samples))
 
     @property
     def resolutions(self) -> list:
@@ -192,5 +233,14 @@ class AxesImage(Image):
     def close(self) -> None:
         self.source.close()
 
+    def __getattr__(self, name):
+        """The native image's own public attributes (a CZI series' ``origin``, ...):
+        an RGB image folded by ``open``'s default stays the backend's image to read
+        them from.  Private ones are not passed on."""
+        source = self.__dict__.get("source")
+        if source is None or name.startswith("_"):
+            raise AttributeError(f"{type(self).__name__!r} object has no attribute {name!r}")
+        return getattr(source, name)
+
     def __repr__(self) -> str:
         return super().__repr__()[:-1] + f" of {self.source.axes!r}>"

{micro_reader-0.2.0 → micro_reader-0.3.0}/src/micro_reader/_base.py

@@ -4,6 +4,8 @@ Each image is one spatial frame of the file; its index says which (series,
 tile, rotation, view: see docs/AXES.md)."""
 from __future__ import annotations
 
+import os
+
 import threading
 from dataclasses import dataclass, field
 from typing import Optional, Sequence
@@ -367,6 +369,19 @@ class File:
     backend's own class is an implementation detail."""
 
     backend: str = "base"
+    #: what the object is in the containment tree (decision C-19): a file holds
+    #: plates (``plates``), a plate wells, a well fields, a field images
+    kind = "file"
+    #: what holds it: nothing, for a file (a plate's ``parent`` is its file)
+    parent = None
+
+    def __getattr__(self, name):
+        # an accessor of a lower level (f.wells): name the way down
+        if name in ("wells", "fields", "rows", "columns"):
+            raise AttributeError(f"a file holds plates: f.plates[p].{name}"
+                                 if name != "fields" else
+                                 "a file holds plates: f.plates[p].wells[w].fields")
+        raise AttributeError(f"{type(self).__name__!r} object has no attribute {name!r}")
     #: (coordinate_systems, coordinate_transformations) once declared and checked
     _graph: Optional[tuple] = None
     #: (the images it was built from, collection) once built (``collection``)
@@ -393,6 +408,8 @@ class File:
         if self.plates:
             from ._plate import navigate
             self.plates = navigate(self.plates, self.images)
+            for plate in self.plates:
+                object.__setattr__(plate, "parent", self)
 
     def _index_resolutions(self) -> None:
         """``resolution`` is in every index when the file stores a pyramid of
@@ -441,17 +458,48 @@ class File:
         return {name: len(seen[name]) for name in INDEX_NAMES if name in seen}
 
     def select(self, **index) -> list:
-        """The images whose index matches, e.g. ``select(series=1, view=0)``."""
-        unknown = set(index) - set(INDEX_NAMES)
+        """The images whose index matches, e.g. ``select(series=1, view=0)``.
+
+        - A value may be a list, tuple, range or set: any of them
+          (``select(illumination=[2, 3])``).
+        - A plate's places have a position and a name, a keyword each (the
+          columns of ozp's table): ``plate`` / ``plate_name``, ``acquisition``
+          / ``acquisition_name``, ``well`` / ``well_name``, ``row`` /
+          ``row_name``, ``column`` / ``column_name``.  Positions count from 0
+          (``well`` is a place in ``plates[p].wells``, ``column=2`` the column
+          named "3"); a name given to a position keyword is its name
+          (``well="C3"`` is ``well_name="C3"``).  ``row`` and ``column`` are
+          another address of the well, so they do not go with ``well``.
+        - A ``Plate``, ``Well``, ``Field`` or ``Acquisition`` object is no
+          value (TypeError).
+
+        An unknown index name raises ``KeyError``; a known one no image has
+        selects nothing."""
+        from ._plate import (PLACE_NAMES, WELL_ADDRESS, index_values, place_conditions,
+                             place_matches, places_of)
+        unknown = set(index) - set(INDEX_NAMES) - set(WELL_ADDRESS) - set(PLACE_NAMES)
         if unknown:
             if _OLD_SERIES in unknown:
                 raise KeyError(_RENAMED)
-            raise KeyError(f"unknown indices {sorted(unknown)}; the indices are {INDEX_NAMES}")
-        get, wanted, out = dict.get, tuple(index.items()), []
+            raise KeyError(f"unknown indices {sorted(unknown)}; the indices are {INDEX_NAMES}, "
+                           f"and a plate's {', '.join(WELL_ADDRESS + PLACE_NAMES)}")
+        wanted = {name: index_values(name, value) for name, value in index.items()}
+        conditions = place_conditions(wanted)
+        plates = self.plates if conditions else ()
+        if conditions and not plates:
+            raise ValueError("select: plate names, well names, rows and columns are a "
+                             "plate's places; the file has no plates (open it with "
+                             "plates=True)")
+        get, out = dict.get, []
+        items = tuple(wanted.items())
         for image in self.images:              # dict.get: not ImageIndex's checks
             key = image.index
-            for name, value in wanted:
-                if not get(key, name) == value:
+            if conditions:
+                place = places_of(plates, key)
+                if place is None or not place_matches(conditions, *place):
+                    continue
+            for name, allowed in items:
+                if get(key, name) not in allowed:
                     break
             else:
                 out.append(image)
@@ -536,8 +584,10 @@ class File:
                 paths = getattr(each, "_member_paths", None)
                 if paths is not None:
                     out.update(paths())
-        out.discard(self.path)
-        return out
+        # the file itself (a .flex or .C01 file opens its folder's set, itself
+        # among them), however its path was spelled (relative, other case)
+        own = os.path.normcase(os.path.abspath(self.path))
+        return {p for p in out if os.path.normcase(os.path.abspath(p)) != own}
 
     def _side_files(self) -> set:
         """Files that belong to this file set without being read as its images

{micro_reader-0.2.0 → micro_reader-0.3.0}/src/micro_reader/_errors.py

@@ -45,8 +45,9 @@ class CompositeWarning(UserWarning):
 
 class CompanionWarning(UserWarning):
     """An OME companion set's metadata came from somewhere other than the file
-    opened, or a member path had to be resolved differently than written
-    (docs/COMPANION.md)."""
+    opened, a member path had to be resolved differently than written, or the
+    companion's bytes had to be read in another encoding than its XML
+    declares (docs/COMPANION.md)."""
 
 
 class SkewedGridWarning(UserWarning):
@@ -57,6 +58,12 @@ class SkewedGridWarning(UserWarning):
     the skew."""
 
 
+class StoredValuesWarning(UserWarning):
+    """The file stores its pixels in a form that is not the measured values, and
+    micro-reader reads the measured ones; the warning says what was changed
+    (e.g. an older scanR plate's 12-bit values stored with the top bit set)."""
+
+
 class PlateLayoutWarning(UserWarning):
     """A plate's wells lie outside the layout its file describes (e.g. a slide
     holder whose records address the wells of the slide in it); the layout