Fix XRD legend labels to preserve distinguishing characters

Author: BenjaminCharmes

pydatalab/src/pydatalab/bokeh_plots.py

@@ -28,7 +28,7 @@
 from bokeh.themes import Theme
 from scipy.signal import find_peaks
 
-from .utils import shrink_label
+from .utils import generate_unique_labels
 
 FONTSIZE = "12pt"
 TYPEFACE = "Helvetica, sans-serif"
@@ -335,7 +335,6 @@ def selectable_axes_plot(
     if isinstance(df, dict):
         labels = list(df.keys())
 
-    label_counts: dict[str, int] = {}
     original_labels_list = []
 
     for ind, df_ in enumerate(df):
@@ -353,10 +352,10 @@ def selectable_axes_plot(
                 orig = df_temp.index.name if len(df) > 1 else ""
 
         original_labels_list.append(orig)
-        shrunk = shrink_label(orig)
-        label_counts[shrunk] = label_counts.get(shrunk, 0) + 1
 
-    label_counter: dict[str, int] = {}
+    legend_labels = (
+        generate_unique_labels(original_labels_list) if len(df) > 1 else original_labels_list
+    )
     plot_columns = []
 
     for ind, df_ in enumerate(df):
@@ -366,23 +365,14 @@ def selectable_axes_plot(
         if isinstance(df, dict):
             df_ = df[df_]
 
-        df_with_metadata = df_.copy()
-
-        original_label = original_labels_list[ind]
-        label = shrink_label(original_label)
-
-        if label and len(df) > 1 and label_counts.get(label, 0) > 1:
-            if label not in label_counter:
-                label_counter[label] = 0
-            label_counter[label] += 1
-            label = f"{label} [{label_counter[label]:02d}]"
+        label = legend_labels[ind] if legend_labels else ""
 
         if hasattr(df_, "attrs"):
             for attr in ["item_id", "original_filename", "wavelength"]:
                 if attr in df_.attrs:
-                    df_with_metadata[attr] = df_.attrs[attr]
+                    df_[attr] = df_.attrs[attr]
 
-        source = ColumnDataSource(df_with_metadata)
+        source = ColumnDataSource(df_)
 
         if color_options:
             color = {"field": color_options[0], "transform": color_mapper}

pydatalab/src/pydatalab/utils.py

@@ -4,6 +4,7 @@
 """
 
 import datetime
+import re
 from json import JSONEncoder
 from math import ceil
 
@@ -59,13 +60,33 @@ def default(o):
 
 
 def shrink_label(label: str | None, max_length: int = 15) -> str:
-    """Shrink label to fit within max_length, preserving file extension when possible."""
+    """Shrink label to fit within max_length, preserving file extension."""
     if not label or len(label) <= max_length:
         return label or ""
 
     if "." in label:
         name, ext = label.rsplit(".", 1)
         if len(ext) < 6:
+            pattern = r"(\d+)"
+            match = re.search(pattern, name)
+            if match:
+                number = match.group(1)
+                if len(number) > 4 and number.startswith("0"):
+                    number_stripped = number.lstrip("0") or "0"
+                    name_shortened = name[: match.start()] + number_stripped + name[match.end() :]
+
+                    if len(name_shortened) + len(ext) + 1 <= max_length:
+                        return f"{name_shortened}.{ext}"
+
+                    number_with_ext_length = len(number_stripped) + len(ext) + 1
+                    available_for_prefix = max_length - number_with_ext_length - 3
+                    if available_for_prefix >= 2:
+                        prefix = name[: match.start()][:available_for_prefix]
+                        return f"{prefix}...{number_stripped}.{ext}"
+
+                    if number_with_ext_length <= max_length:
+                        return f"{number_stripped}.{ext}"
+
             available = max_length - len(ext) - 4
             if available > 3:
                 return f"{name[:available]}...{ext}"
@@ -75,3 +96,167 @@ def shrink_label(label: str | None, max_length: int = 15) -> str:
             return f"{label[:12]}..."
     else:
         return f"{label[:12]}..."
+
+
+def generate_unique_labels(
+    filenames: list[str],
+    max_length: int = 15,
+) -> list[str]:
+    if not filenames or len(filenames) == 1:
+        return filenames if filenames else []
+
+    common_prefix = _find_common_prefix_smart(filenames)
+    common_suffix = _find_common_suffix_smart(filenames)
+
+    extension = ""
+    if all("." in f for f in filenames):
+        extensions = [f.rsplit(".", 1)[1] for f in filenames]
+        if len(set(extensions)) == 1:
+            extension = f".{extensions[0]}"
+
+    unique_parts = []
+    for filename in filenames:
+        start_idx = len(common_prefix)
+        end_idx = len(filename) - len(common_suffix)
+        unique_part = filename[start_idx:end_idx] if start_idx < end_idx else filename
+
+        if not unique_part.strip():
+            unique_part = filename
+
+        if extension and unique_part.endswith(extension):
+            unique_part = unique_part[: -len(extension)]
+
+        unique_part_without_ext = unique_part
+
+        if len(unique_part_without_ext) < 5 and len(filename) <= max_length:
+            unique_parts.append(filename)
+        else:
+            if extension:
+                unique_part = unique_part + extension
+            unique_parts.append(unique_part)
+
+    labels = []
+    for i, part in enumerate(unique_parts):
+        shrunken = shrink_label(part, max_length)
+
+        if "." in shrunken:
+            name_part, ext_part = shrunken.rsplit(".", 1)
+            if name_part.replace("0", "").isdigit() and name_part.startswith("0"):
+                stripped = name_part.lstrip("0") or "0"
+                if len(stripped) < 6 and common_prefix:
+                    available = max_length - len(stripped) - len(ext_part) - 4
+                    prefix_length = min(available, 4)
+                    if prefix_length >= 2:
+                        prefix_to_add = common_prefix[:prefix_length].rstrip("-_. /\\")
+                        if prefix_to_add:
+                            shrunken = f"{prefix_to_add}...{stripped}.{ext_part}"
+
+        if len(shrunken) < 8 and common_prefix:
+            available = max_length - len(shrunken) - 3
+            prefix_length = min(available, 4)
+            if prefix_length >= 2:
+                prefix_to_add = common_prefix[:prefix_length].rstrip("-_. /\\")
+                if prefix_to_add and "..." not in shrunken:
+                    shrunken = f"{prefix_to_add}...{shrunken}"
+
+        labels.append(shrunken)
+
+    return _add_numbering_for_duplicates(labels)
+
+
+def _find_common_prefix_smart(strings: list[str]) -> str:
+    if not strings or len(strings) < 2:
+        return ""
+
+    prefix = _find_common_prefix(strings)
+
+    if not prefix:
+        return ""
+
+    if prefix[-1] in ("-", "_", " ", "/", "\\"):
+        return prefix
+
+    last_sep = max(
+        prefix.rfind("-"),
+        prefix.rfind("_"),
+        prefix.rfind(" "),
+        prefix.rfind("/"),
+        prefix.rfind("\\"),
+    )
+
+    if last_sep > 0:
+        return prefix[: last_sep + 1]
+
+    return ""
+
+
+def _find_common_suffix_smart(strings: list[str]) -> str:
+    if not strings or len(strings) < 2:
+        return ""
+
+    suffix = _find_common_suffix(strings)
+
+    if not suffix:
+        return ""
+
+    if suffix.startswith("."):
+        return ""
+
+    if suffix[0] in ("-", "_", " ", "/", "\\"):
+        return suffix
+
+    first_sep = len(suffix)
+    for sep in ("-", "_", " ", "/", "\\"):
+        pos = suffix.find(sep)
+        if pos != -1 and pos < first_sep:
+            first_sep = pos
+
+    if first_sep < len(suffix):
+        return suffix[first_sep:]
+
+    return ""
+
+
+def _find_common_prefix(strings: list[str]) -> str:
+    if not strings or len(strings) < 2:
+        return ""
+
+    min_str = min(strings)
+    max_str = max(strings)
+
+    for i, char in enumerate(min_str):
+        if char != max_str[i]:
+            return min_str[:i]
+
+    return min_str
+
+
+def _find_common_suffix(strings: list[str]) -> str:
+    if not strings or len(strings) < 2:
+        return ""
+
+    reversed_strings = [s[::-1] for s in strings]
+    common_reversed_prefix = _find_common_prefix(reversed_strings)
+
+    return common_reversed_prefix[::-1]
+
+
+def _add_numbering_for_duplicates(labels: list[str]) -> list[str]:
+    label_counts: dict[str, int] = {}
+    for label in labels:
+        label_counts[label] = label_counts.get(label, 0) + 1
+
+    if all(count == 1 for count in label_counts.values()):
+        return labels
+
+    label_counter: dict[str, int] = {}
+    numbered_labels = []
+
+    for label in labels:
+        if label_counts[label] > 1:
+            label_counter[label] = label_counter.get(label, 0) + 1
+            numbered_labels.append(f"{label} [{label_counter[label]:02d}]")
+        else:
+            numbered_labels.append(label)
+
+    return numbered_labels

pydatalab/tests/test_utils.py

@@ -0,0 +1,64 @@
+from pydatalab.utils import generate_unique_labels
+
+
+def test_generate_unique_labels_single_file():
+    result = generate_unique_labels(["sample_xrd_pattern.cif"])
+    assert result == ["sample_xrd_pattern.cif"]
+
+
+def test_generate_unique_labels_empty():
+    result = generate_unique_labels([])
+    assert result == []
+
+
+def test_generate_unique_labels_common_suffix():
+    filenames = ["sample1-xrd.xrdml", "sample2-xrd.xrdml"]
+    result = generate_unique_labels(filenames)
+    assert result == ["sample1.xrdml", "sample2.xrdml"]
+
+
+def test_generate_unique_labels_prefix_and_suffix():
+    filenames = [
+        "experiment_run1_final.dat",
+        "experiment_run2_final.dat",
+        "experiment_run3_final.dat",
+    ]
+    result = generate_unique_labels(filenames)
+    assert result == ["run1.dat", "run2.dat", "run3.dat"]
+
+
+def test_generate_unique_labels_long_unique_parts():
+    filenames = [
+        "very_long_sample_name_with_many_characters_001.cif",
+        "very_long_sample_name_with_many_characters_002.cif",
+    ]
+    result = generate_unique_labels(filenames, max_length=10)
+    assert all(len(label) <= 15 for label in result)
+    assert result[0] != result[1]
+
+
+def test_generate_unique_labels_duplicates_after_shortening():
+    filenames = [
+        "CIF_0000000000000001.xrdml",
+        "CIF_0000000000000002.xrdml",
+    ]
+    result = generate_unique_labels(filenames, max_length=15)
+    assert result == ["CIF...1.xrdml", "CIF...2.xrdml"]
+
+
+def test_generate_unique_labels_common_prefix():
+    filenames = ["ICSDCollCode-000002.cif", "ICSDCollCode-000003.cif"]
+    result = generate_unique_labels(filenames)
+    assert result == ["ICSD...2.cif", "ICSD...3.cif"]
+
+
+def test_generate_unique_labels_same_extension():
+    filenames = ["sample_A.cif", "sample_B.cif", "sample_C.cif"]
+    result = generate_unique_labels(filenames)
+    assert result == ["sample_A.cif", "sample_B.cif", "sample_C.cif"]
+
+
+def test_generate_unique_labels_cif_pattern():
+    filenames = ["CIF_00000001.cif", "CIF_00000002.cif"]
+    result = generate_unique_labels(filenames)
+    assert result == ["CIF...1.cif", "CIF...2.cif"]