Complain about trailing underscores, find closest key

Tests in tests/test_core/test_errors/test_dict_path_errors.py
not yet updated though.

Author: nicholas-esterer

packages/python/plotly/_plotly_utils/utils.py

@@ -364,10 +364,10 @@ def display_string_positions(p, i=None, offset=0, length=1, char="^", trim=True)
     return ret
 
 
-def chomp_empty_strings(strings, c):
+def chomp_empty_strings(strings, c, reverse=False):
     """
     Given a list of strings, some of which are the empty string "", replace the
-    empty strings with "_" and combine them with the closest non-empty string on
+    empty strings with c and combine them with the closest non-empty string on
     the left or "" if it is the first string.
     Examples:
     for c="_"
@@ -379,7 +379,15 @@ def chomp_empty_strings(strings, c):
     [''] -> ['']
     ['', ''] -> ['_']
     ['', '', '', ''] -> ['___']
+    If reverse is true, empty strings are combined with closest non-empty string
+    on the right or "" if it is the last string.
     """
+
+    def _rev(l):
+        return [s[::-1] for s in l][::-1]
+
+    if reverse:
+        return _rev(chomp_empty_strings(_rev(strings), c))
     if not len(strings):
         return strings
     if sum(map(len, strings)) == 0:
@@ -400,3 +408,31 @@ def __call__(self, x, y):
                 return x + [y]
 
     return list(filter(len, reduce(_Chomper(c), strings, [""])))
+
+
+# taken from
+# https://en.wikibooks.org/wiki/Algorithm_Implementation/Strings/Levenshtein_distance#Python
+def levenshtein(s1, s2):
+    if len(s1) < len(s2):
+        return levenshtein(s2, s1)  # len(s1) >= len(s2)
+    if len(s2) == 0:
+        return len(s1)
+    previous_row = range(len(s2) + 1)
+    for i, c1 in enumerate(s1):
+        current_row = [i + 1]
+        for j, c2 in enumerate(s2):
+            # j+1 instead of j since previous_row and current_row are one character longer
+            # than s2
+            insertions = previous_row[j + 1] + 1
+            deletions = current_row[j] + 1
+            substitutions = previous_row[j] + (c1 != c2)
+            current_row.append(min(insertions, deletions, substitutions))
+        previous_row = current_row
+    return previous_row[-1]
+
+
+def find_closest_string(string, strings):
+    def _key(s):
+        return levenshtein(s, string)
+
+    return sorted(strings, key=_key)[0]

packages/python/plotly/plotly/basedatatypes.py

@@ -18,6 +18,7 @@
     split_string_positions,
     display_string_positions,
     chomp_empty_strings,
+    find_closest_string,
 )
 from _plotly_utils.exceptions import PlotlyKeyError
 from .optional_imports import get_module
@@ -95,9 +96,9 @@ def _split_and_chomp(s):
                 return s
             s_split = split_multichar([s], list("_"))
             # handle key paths like "a_path_", "_another_path", or
-            # "yet__another_path" by joining extra "_" to the string to the left or
-            # the empty string if at the beginning
-            s_chomped = chomp_empty_strings(s_split, "_")
+            # "yet__another_path" by joining extra "_" to the string to the right or
+            # the empty string if at the end
+            s_chomped = chomp_empty_strings(s_split, "_", reverse=True)
             return s_chomped
 
         # after running _split_and_chomp on key_path2b, it will be a list
@@ -204,14 +205,25 @@ def _check_path_in_prop_tree(obj, path, error_cast=None):
                 # In case i is 0, the best we can do is indicate the first
                 # property in the string as having caused the error
                 disp_i = max(i - 1, 0)
+                dict_item_len = _len_dict_item(prop[disp_i])
+                # if the path has trailing underscores, the prop string will start with "_"
+                trailing_underscores = ""
+                if prop[i][0] == "_":
+                    trailing_underscores = " and path has trailing underscores"
+                # if the path has trailing underscores and the display index is
+                # one less than the prop index (see above), then we can also
+                # indicate the offending underscores
+                if (trailing_underscores != "") and (disp_i != i):
+                    dict_item_len += _len_dict_item(prop[i])
                 arg += """
 
-Property does not support subscripting:
+Property does not support subscripting%s:
 %s
 %s""" % (
+                    trailing_underscores,
                     path,
                     display_string_positions(
-                        prop_idcs, disp_i, length=_len_dict_item(prop[disp_i]), char="^"
+                        prop_idcs, disp_i, length=dict_item_len, char="^"
                     ),
                 )
             else:
@@ -226,6 +238,17 @@ def _check_path_in_prop_tree(obj, path, error_cast=None):
                         prop_idcs, i, length=_len_dict_item(prop[i]), char="^"
                     ),
                 )
+                guessed_prop = None
+                # If obj has _valid_props then we can try and guess what key was intended
+                try:
+                    guessed_prop = find_closest_string(prop[i], obj._valid_props)
+                except Exception:
+                    pass
+                if guessed_prop is not None:
+                    arg += """
+Did you mean "%s"?""" % (
+                        guessed_prop,
+                    )
             # Make KeyError more pretty by changing it to a PlotlyKeyError,
             # because the Python interpreter has a special way of printing
             # KeyError

packages/python/plotly/plotly/tests/test_core/test_errors/test_dict_path_errors.py

@@ -49,7 +49,8 @@ def test_raises_on_bad_dot_property(some_fig):
             e.args[0].find(
                 """Bad property path:
 layout.shapes[1].x2000
-                 ^^^^^"""
+                 ^^^^^
+Did you mean "x0"?"""
             )
             >= 0
         )
@@ -68,7 +69,8 @@ def test_raises_on_bad_ancestor_dot_property(some_fig):
             e.args[0].find(
                 """Bad property path:
 layout.shapa[1].x2000
-       ^^^^^"""
+       ^^^^^
+Did you mean "shapes"?"""
             )
             >= 0
         )