feat: show similar keys first on KeyInFile error

uproot4/__init__.py

@@ -102,6 +102,8 @@
 from uproot4.reading import ReadOnlyFile
 from uproot4.reading import ReadOnlyDirectory
 
+from uproot4.exceptions import KeyInFileError
+
 from uproot4.model import Model
 from uproot4.model import classname_decode
 from uproot4.model import classname_encode
@@ -209,79 +211,4 @@ def behavior_of(classname):
 
 del pkgutil
 
-
-class KeyInFileError(KeyError):
-    """
-    Exception raised by attempts to find ROOT objects in ``TDirectories``
-    or ``TBranches`` in :py:class:`~uproot4.behaviors.TBranch.HasBranches`, which
-    both have a Python ``Mapping`` interface (square bracket syntax to extract
-    items).
-
-    This exception descends from Python's ``KeyError``, so it can be used in
-    the normal way by interfaces that expect a missing item in a ``Mapping``
-    to raise ``KeyError``, but it provides more information, depending on
-    availability:
-
-    * ``because``: an explanatory message
-    * ``cycle``: the ROOT cycle number requested, if any
-    * ``keys``: a list or partial list of keys that *are* in the object, in case
-      of misspelling
-    * ``file_path``: a path (or URL) to the file
-    * ``object_path``: a path to the object within the ROOT file.
-    """
-
-    def __init__(
-        self, key, because="", cycle=None, keys=None, file_path=None, object_path=None
-    ):
-        super(KeyInFileError, self).__init__(key)
-        self.key = key
-        self.because = because
-        self.cycle = cycle
-        self.keys = keys
-        self.file_path = file_path
-        self.object_path = object_path
-
-    def __str__(self):
-        if self.because == "":
-            because = ""
-        else:
-            because = " because " + self.because
-
-        with_keys = ""
-        if self.keys is not None:
-            to_show = None
-            for key in self.keys:
-                if to_show is None:
-                    to_show = repr(key)
-                else:
-                    to_show += ", " + repr(key)
-                if len(to_show) > 200:
-                    to_show += "..."
-                    break
-            if to_show is None:
-                to_show = "(none!)"
-            with_keys = "\n\n    Known keys: {0}\n".format(to_show)
-
-        in_file = ""
-        if self.file_path is not None:
-            in_file = "\nin file {0}".format(self.file_path)
-
-        in_object = ""
-        if self.object_path is not None:
-            in_object = "\nin object {0}".format(self.object_path)
-
-        if self.cycle == "any":
-            return """not found: {0} (with any cycle number){1}{2}{3}{4}""".format(
-                repr(self.key), because, with_keys, in_file, in_object
-            )
-        elif self.cycle is None:
-            return """not found: {0}{1}{2}{3}{4}""".format(
-                repr(self.key), because, with_keys, in_file, in_object
-            )
-        else:
-            return """not found: {0} with cycle {1}{2}{3}{4}{5}""".format(
-                repr(self.key), self.cycle, because, with_keys, in_file, in_object
-            )
-
-
 from uproot4._util import no_filter

uproot4/_util.py

@@ -635,3 +635,53 @@ def awkward_form_of_iter(awkward1, form):
         )
     else:
         raise RuntimeError("unrecognized form: {0}".format(type(form)))
+
+def damerau_levenshtein(a, b, ratio=False):
+    """Modified Damerau-Levenshtein distance. Adds a middling penalty
+    for capitalization.
+
+    Parameters
+    ----------
+    a : str
+    b : str
+    ratio : bool, optional, default False
+        Return ratio (len(a)+len(b) - DLH)/(len(a)+len(b))
+
+    Returns
+    -------
+    float
+        Modified Damerau Levenshtein distance
+    """
+    M = [[0]*(len(b)+1) for i in range(len(a)+1)]
+
+    for i in range(len(a)+1):
+        M[i][0] = i
+    for j in range(len(b)+1):
+        M[0][j] = j
+
+    for i in range(1, len(a)+1):
+        for j in range(1, len(b)+1):
+            if a[i-1] == b[j-1]:  # Same char
+                cost = 0
+            elif a[i-1].lower() == b[j-1].lower(): # Same if lowered
+                cost = 0.5
+            else:  # Different char
+                cost = 2
+            M[i][j] = min(M[i-1][j] + 1,  # Addition
+                          M[i][j-1] + 1,  # Removal
+                          M[i-1][j-1] + cost  # Substitution
+                             )
+
+            # Transposition
+            if i > 1 and j > 1 and a[i-1].lower() == b[j-2].lower() and a[i-2].lower() == b[j-2].lower():
+                if a[i-1] == b[j-2] and a[i-2] == b[j-1]:
+                    # Transpose only
+                    M[i][j] = min(M[i][j], M[i-2][j-2] + 1)
+                else:
+                    # Traspose and capitalization
+                    M[i][j] = min(M[i][j], M[i-2][j-2] + 1.5)
+
+    if not ratio:
+        return M[len(a)][len(b)]
+    else:
+        return  ((len(a)+len(b)) - M[len(a)][len(b)] / (len(a)+len(b)))

uproot4/exceptions.py

@@ -0,0 +1,79 @@
+import numpy as np
+from uproot4._util import damerau_levenshtein
+
+class KeyInFileError(KeyError):
+    """
+    Exception raised by attempts to find ROOT objects in ``TDirectories``
+    or ``TBranches`` in :py:class:`~uproot4.behaviors.TBranch.HasBranches`, which
+    both have a Python ``Mapping`` interface (square bracket syntax to extract
+    items).
+
+    This exception descends from Python's ``KeyError``, so it can be used in
+    the normal way by interfaces that expect a missing item in a ``Mapping``
+    to raise ``KeyError``, but it provides more information, depending on
+    availability:
+
+    * ``because``: an explanatory message
+    * ``cycle``: the ROOT cycle number requested, if any
+    * ``keys``: a list or partial list of keys that *are* in the object, in case
+      of misspelling
+    * ``file_path``: a path (or URL) to the file
+    * ``object_path``: a path to the object within the ROOT file.
+    """
+
+    def __init__(
+        self, key, because="", cycle=None, keys=None, file_path=None, object_path=None
+    ):
+        super(KeyInFileError, self).__init__(key)
+        self.key = key
+        self.because = because
+        self.cycle = cycle
+        self.keys = keys
+        self.file_path = file_path
+        self.object_path = object_path
+
+    def __str__(self):
+        if self.because == "":
+            because = ""
+        else:
+            because = " because " + self.because
+
+        with_keys = ""
+        if self.keys is not None:
+            to_show = None
+            distance = []
+            for key in self.keys:
+              distance.append(damerau_levenshtein(self.key, key))
+            self.keys = np.array(self.keys)[np.argsort(distance)]
+            for key in self.keys:
+                if to_show is None:
+                    to_show = repr(key)
+                else:
+                    to_show += ", " + repr(key)
+                if len(to_show) > 200:
+                    to_show += "..."
+                    break
+            if to_show is None:
+                to_show = "(none!)"
+            with_keys = "\n\n    Available keys: {0}\n".format(to_show)
+
+        in_file = ""
+        if self.file_path is not None:
+            in_file = "\nin file {0}".format(self.file_path)
+
+        in_object = ""
+        if self.object_path is not None:
+            in_object = "\nin object {0}".format(self.object_path)
+
+        if self.cycle == "any":
+            return """not found: {0} (with any cycle number){1}{2}{3}{4}""".format(
+                repr(self.key), because, with_keys, in_file, in_object
+            )
+        elif self.cycle is None:
+            return """not found: {0}{1}{2}{3}{4}""".format(
+                repr(self.key), because, with_keys, in_file, in_object
+            )
+        else:
+            return """not found: {0} with cycle {1}{2}{3}{4}{5}""".format(
+                repr(self.key), self.cycle, because, with_keys, in_file, in_object
+            )