fix(expr): fix formatting of table info using tabulate

ibis/backends/impala/tests/test_exprs.py

@@ -1,5 +1,4 @@
 from decimal import Decimal
-from io import StringIO
 
 import pandas as pd
 import pandas.testing as tm
@@ -20,13 +19,6 @@ def test_embedded_identifier_quoting(alltypes):
     expr.execute()
 
 
-def test_table_info(alltypes):
-    buf = StringIO()
-    alltypes.info(buf=buf)
-
-    assert buf.getvalue() is not None
-
-
 def test_summary_execute(alltypes):
     table = alltypes
 

ibis/backends/tests/test_generic.py

@@ -1,4 +1,5 @@
 import decimal
+import io
 
 import numpy as np
 import pandas as pd
@@ -402,3 +403,13 @@ def test_dropna_table(backend, alltypes, how, subset):
 def test_select_sort_sort(alltypes):
     query = alltypes[alltypes.year, alltypes.bool_col]
     query = query.sort_by(query.year).sort_by(query.bool_col)
+
+
+def test_table_info(alltypes):
+    buf = io.StringIO()
+    alltypes.info(buf=buf)
+
+    info_str = buf.getvalue()
+    assert info_str is not None
+    assert "Nulls" in info_str
+    assert all(str(type) in info_str for type in alltypes.schema().types)

ibis/expr/types/relations.py

@@ -7,6 +7,7 @@
 from typing import IO, TYPE_CHECKING, Any, Iterable, Literal, Mapping, Sequence
 
 import numpy as np
+import tabulate
 from cached_property import cached_property
 from public import public
 
@@ -850,19 +851,29 @@ def info(self, buf: IO[str] | None = None) -> None:
         buf
             A writable buffer, defaults to stdout
         """
-        metrics = [self.count().name("nrows")]
-        for col in self.columns:
-            metrics.append(self[col].count().name(col))
+        metrics = [self[col].count().name(col) for col in self.columns]
+        metrics.append(self.count().name("nrows"))
 
-        metrics = self.aggregate(metrics).execute().loc[0]
+        schema = self.schema()
 
-        names = ["Column", "------"] + self.columns
-        types = ["Type", "----"] + [repr(x) for x in self.schema().types]
-        counts = ["Non-null #", "----------"] + [str(x) for x in metrics[1:]]
-        col_metrics = util.adjoin(2, names, types, counts)
-        result = f"Table rows: {metrics[0]}\n\n{col_metrics}"
+        *items, (_, n) = self.aggregate(metrics).execute().squeeze().items()
 
-        print(result, file=buf)
+        tabulated = tabulate.tabulate(
+            [
+                (
+                    column,
+                    schema[column],
+                    f"{n - non_nulls} ({100 * (1.0 - non_nulls / n):>3.3g}%)",
+                )
+                for column, non_nulls in items
+            ],
+            headers=["Column", "Type", "Nulls (%)"],
+            colalign=("left", "left", "right"),
+        )
+        width = tabulated[tabulated.index("\n") + 1 :].index("\n")
+        row_count = f"Rows: {n}".center(width)
+        footer_line = "-" * width
+        print("\n".join([tabulated, footer_line, row_count]), file=buf)
 
     def set_column(self, name: str, expr: ir.ValueExpr) -> TableExpr:
         """Replace an existing column with a new expression.

ibis/util.py

@@ -16,7 +16,6 @@
     TYPE_CHECKING,
     Any,
     Hashable,
-    Iterable,
     Iterator,
     Mapping,
     Sequence,
@@ -159,33 +158,6 @@ def is_function(v: Any) -> bool:
     return isinstance(v, (types.FunctionType, types.LambdaType))
 
 
-def adjoin(space: int, *lists: Iterable[str]) -> str:
-    """Glue together two sets of strings using `space`.
-
-    Parameters
-    ----------
-    space : int
-    lists : list or tuple
-
-    Returns
-    -------
-    str
-    """
-    lengths = [max(map(len, x)) + space for x in lists[:-1]]
-
-    # not the last one
-    lengths.append(max(map(len, lists[-1])))
-    max_len = max(map(len, lists))
-    chains = (
-        itertools.chain(
-            (x.ljust(length) for x in lst),
-            itertools.repeat(' ' * length, max_len - len(lst)),
-        )
-        for lst, length in zip(lists, lengths)
-    )
-    return '\n'.join(map(''.join, zip(*chains)))
-
-
 def log(msg: str) -> None:
     """Log `msg` using ``options.verbose_log`` if set, otherwise ``print``.
 

poetry-overrides.nix

@@ -78,4 +78,8 @@ self: super:
       self.poetry-core
     ];
   });
+
+  tabulate = super.tabulate.overridePythonAttrs (_: {
+    TABULATE_INSTALL = "lib-only";
+  });
 }

pyproject.toml

@@ -34,8 +34,10 @@ multipledispatch = ">=0.6,<0.7"
 numpy = ">=1,<2"
 pandas = ">=1.2.5,<2"
 parsy = ">=1.3.0,<2"
+poetry-dynamic-versioning = ">=0.14.0,<1"
 pydantic = ">=1.9.0,<2"
 regex = ">=2021.7.6"
+tabulate = ">=0.8.9,<1"
 toolz = ">=0.11,<0.12"
 clickhouse-cityhash = { version = ">=1.0.2,<2", optional = true }
 clickhouse-driver = { version = ">=0.1,<0.3", optional = true }
@@ -59,7 +61,6 @@ pyspark = { version = ">=3,<4", optional = true }
 requests = { version = ">=2,<3", optional = true }
 Shapely = { version = ">=1.6,<1.8.1", optional = true }
 sqlalchemy = { version = ">=1.4,<2.0", optional = true }
-poetry-dynamic-versioning = "^0.14.0"
 
 [tool.poetry.dev-dependencies]
 black = ">=22.1.0,<23"