PERF: styler uuid control and security

doc/source/whatsnew/v1.2.0.rst

@@ -208,6 +208,7 @@ Performance improvements
 - Performance improvements when creating Series with dtype `str` or :class:`StringDtype` from array with many string elements (:issue:`36304`, :issue:`36317`)
 - Performance improvement in :meth:`GroupBy.agg` with the ``numba`` engine (:issue:`35759`)
 - Performance improvement in :meth:`GroupBy.transform` with the ``numba`` engine (:issue:`36240`)
+- ``Styler`` uuid method altered to compress data transmission over web whilst maintaining reasonably low table collision probability (:issue:`36345`)
 
 .. ---------------------------------------------------------------------------
 

pandas/io/formats/style.py

@@ -18,7 +18,7 @@
     Tuple,
     Union,
 )
-from uuid import uuid1
+from uuid import uuid4
 
 import numpy as np
 
@@ -73,6 +73,9 @@ class Styler:
         List of {selector: (attr, value)} dicts; see Notes.
     uuid : str, default None
         A unique identifier to avoid CSS collisions; generated automatically.
+    uuid_len : int, default 5
+        If ``uuid`` is not specified, the length of the ``uuid`` to randomly generate
+        in characters, max is 32.
     caption : str, default None
         Caption to attach to the table.
     table_attributes : str, default None
@@ -128,6 +131,7 @@ class Styler:
 
     * Blank cells include ``blank``
     * Data cells include ``data``
+
     """
 
     loader = jinja2.PackageLoader("pandas", "io/formats/templates")
@@ -140,6 +144,7 @@ def __init__(
         precision: Optional[int] = None,
         table_styles: Optional[List[Dict[str, List[Tuple[str, str]]]]] = None,
         uuid: Optional[str] = None,
+        uuid_len: int = 5,
         caption: Optional[str] = None,
         table_attributes: Optional[str] = None,
         cell_ids: bool = True,
@@ -159,7 +164,8 @@ def __init__(
         self.index = data.index
         self.columns = data.columns
 
-        self.uuid = uuid
+        self.uuid_len = uuid_len if uuid_len < 33 else 32
+        self.uuid = (uuid or uuid4().hex[: self.uuid_len]) + "_"
         self.table_styles = table_styles
         self.caption = caption
         if precision is None:
@@ -248,7 +254,7 @@ def _translate(self):
         precision = self.precision
         hidden_index = self.hidden_index
         hidden_columns = self.hidden_columns
-        uuid = self.uuid or str(uuid1()).replace("-", "_")
+        uuid = self.uuid
         ROW_HEADING_CLASS = "row_heading"
         COL_HEADING_CLASS = "col_heading"
         INDEX_NAME_CLASS = "index_name"

pandas/tests/io/formats/test_style.py

@@ -1718,6 +1718,24 @@ def test_colspan_w3(self):
         s = Styler(df, uuid="_", cell_ids=False)
         assert '<th class="col_heading level0 col0" colspan="2">l0</th>' in s.render()
 
+    @pytest.mark.parametrize("len_", [1, 5, 32])
+    def test_uuid_len(self, len_):
+        # GH 36345
+        df = pd.DataFrame(data=[["A"]])
+        s = Styler(df, uuid_len=len_, cell_ids=False).render()
+        strt = s.find('id="T_')
+        end = s[strt + 6 :].find('"')
+        assert end == len_ + 1
+
+    @pytest.mark.parametrize("len_", [33, 100])
+    def test_uuid_max_len(self, len_):
+        # GH 36345
+        df = pd.DataFrame(data=[["A"]])
+        s = Styler(df, uuid_len=len_, cell_ids=False).render()
+        strt = s.find('id="T_')
+        end = s[strt + 6 :].find('"')
+        assert end == 32 + 1
+
 
 @td.skip_if_no_mpl
 class TestStylerMatplotlibDep: