Place html table formatting in util.py

Allows it to be used by other tabular outputting routines.

Also put the "limit" functionality into a single util function

Author: hyanwong

python/tests/test_util.py

@@ -489,12 +489,12 @@ def test_unicode_table():
     )
 
 
-def test_unicode_table_alignments():
+def test_unicode_table_column_alignments():
     assert (
         util.unicode_table(
             [["5", "6", "7", "8"], ["90", "10", "11", "12"]],
             header=["1", "2", "3", "4"],
-            alignments="<>><",
+            column_alignments="<>><",
         )
         == textwrap.dedent(
             """

python/tskit/tables.py

@@ -27,7 +27,6 @@
 import collections.abc
 import dataclasses
 import datetime
-import itertools
 import json
 import numbers
 import warnings
@@ -657,41 +656,10 @@ def __str__(self):
         return util.unicode_table(rows, header=headers, row_separator=False)
 
     def _repr_html_(self):
-        """
-        Called by jupyter notebooks to render tables
-        """
         headers, rows = self._text_header_and_rows(
             limit=tskit._print_options["max_lines"]
         )
-        headers = "".join(f"<th>{header}</th>" for header in headers)
-        rows = (
-            f'<td style="text-align: center;" colspan="{len(headers)}"><em>{row[11:]}'
-            f" rows skipped (tskit.set_print_options)</em></td>"
-            if "__skipped__" in row
-            else "".join(f"<td>{cell}</td>" for cell in row)
-            for row in rows
-        )
-        rows = "".join(f"<tr>{row}</tr>\n" for row in rows)
-        return f"""
-            <div>
-                <style scoped="">
-                    .tskit-table tbody tr th:only-of-type {{vertical-align: middle;}}
-                    .tskit-table tbody tr th {{vertical-align: top;}}
-                    .tskit-table tbody td {{text-align: right;padding: 0.5em 0.5em;}}
-                    .tskit-table tbody th {{padding: 0.5em 0.5em;}}
-                </style>
-                <table border="1" class="tskit-table">
-                    <thead>
-                        <tr>
-                            {headers}
-                        </tr>
-                    </thead>
-                    <tbody>
-                        {rows}
-                    </tbody>
-                </table>
-            </div>
-        """
+        return util.html_table(rows, header=headers)
 
     def _columns_all_integer(self, *colnames):
         # For displaying floating point values without loads of decimal places
@@ -852,15 +820,8 @@ def __init__(self, max_rows_increment=0, ll_table=None):
     def _text_header_and_rows(self, limit=None):
         headers = ("id", "flags", "location", "parents", "metadata")
         rows = []
-        if limit is None or self.num_rows <= limit:
-            indexes = range(self.num_rows)
-        else:
-            indexes = itertools.chain(
-                range(limit // 2),
-                [-1],
-                range(self.num_rows - (limit - (limit // 2)), self.num_rows),
-            )
-        for j in indexes:
+        row_indexes = util.truncate_rows(self.num_rows, limit)
+        for j in row_indexes:
             if j == -1:
                 rows.append(f"__skipped__{self.num_rows-limit}")
             else:
@@ -1105,16 +1066,9 @@ def __init__(self, max_rows_increment=0, ll_table=None):
     def _text_header_and_rows(self, limit=None):
         headers = ("id", "flags", "population", "individual", "time", "metadata")
         rows = []
-        if limit is None or self.num_rows <= limit:
-            indexes = range(self.num_rows)
-        else:
-            indexes = itertools.chain(
-                range(limit // 2),
-                [-1],
-                range(self.num_rows - (limit - (limit // 2)), self.num_rows),
-            )
+        row_indexes = util.truncate_rows(self.num_rows, limit)
         decimal_places_times = 0 if self._columns_all_integer("time") else 8
-        for j in indexes:
+        for j in row_indexes:
             row = self[j]
             if j == -1:
                 rows.append(f"__skipped__{self.num_rows-limit}")
@@ -1306,16 +1260,9 @@ def __init__(self, max_rows_increment=0, ll_table=None):
     def _text_header_and_rows(self, limit=None):
         headers = ("id", "left", "right", "parent", "child", "metadata")
         rows = []
-        if limit is None or self.num_rows <= limit:
-            indexes = range(self.num_rows)
-        else:
-            indexes = itertools.chain(
-                range(limit // 2),
-                [-1],
-                range(self.num_rows - (limit - (limit // 2)), self.num_rows),
-            )
+        row_indexes = util.truncate_rows(self.num_rows, limit)
         decimal_places = 0 if self._columns_all_integer("left", "right") else 8
-        for j in indexes:
+        for j in row_indexes:
             if j == -1:
                 rows.append(f"__skipped__{self.num_rows-limit}")
             else:
@@ -1528,17 +1475,10 @@ def __init__(self, max_rows_increment=0, ll_table=None):
     def _text_header_and_rows(self, limit=None):
         headers = ("id", "left", "right", "node", "source", "dest", "time", "metadata")
         rows = []
-        if limit is None or self.num_rows <= limit:
-            indexes = range(self.num_rows)
-        else:
-            indexes = itertools.chain(
-                range(limit // 2),
-                [-1],
-                range(self.num_rows - (limit - (limit // 2)), self.num_rows),
-            )
+        row_indexes = util.truncate_rows(self.num_rows, limit)
         decimal_places_coords = 0 if self._columns_all_integer("left", "right") else 8
         decimal_places_times = 0 if self._columns_all_integer("time") else 8
-        for j in indexes:
+        for j in row_indexes:
             if j == -1:
                 rows.append(f"__skipped__{self.num_rows-limit}")
             else:
@@ -1748,16 +1688,9 @@ def __init__(self, max_rows_increment=0, ll_table=None):
     def _text_header_and_rows(self, limit=None):
         headers = ("id", "position", "ancestral_state", "metadata")
         rows = []
-        if limit is None or self.num_rows <= limit:
-            indexes = range(self.num_rows)
-        else:
-            indexes = itertools.chain(
-                range(limit // 2),
-                [-1],
-                range(self.num_rows - (limit - (limit // 2)), self.num_rows),
-            )
+        row_indexes = util.truncate_rows(self.num_rows, limit)
         decimal_places = 0 if self._columns_all_integer("position") else 8
-        for j in indexes:
+        for j in row_indexes:
             if j == -1:
                 rows.append(f"__skipped__{self.num_rows-limit}")
             else:
@@ -1971,17 +1904,10 @@ def __init__(self, max_rows_increment=0, ll_table=None):
     def _text_header_and_rows(self, limit=None):
         headers = ("id", "site", "node", "time", "derived_state", "parent", "metadata")
         rows = []
-        if limit is None or self.num_rows <= limit:
-            indexes = range(self.num_rows)
-        else:
-            indexes = itertools.chain(
-                range(limit // 2),
-                [-1],
-                range(self.num_rows - (limit - (limit // 2)), self.num_rows),
-            )
+        row_indexes = util.truncate_rows(self.num_rows, limit)
         # Currently mutations do not have discretised times: this for consistency
         decimal_places_times = 0 if self._columns_all_integer("time") else 8
-        for j in indexes:
+        for j in row_indexes:
             if j == -1:
                 rows.append(f"__skipped__{self.num_rows-limit}")
             else:
@@ -2232,15 +2158,8 @@ def add_row(self, metadata=None):
     def _text_header_and_rows(self, limit=None):
         headers = ("id", "metadata")
         rows = []
-        if limit is None or self.num_rows <= limit:
-            indexes = range(self.num_rows)
-        else:
-            indexes = itertools.chain(
-                range(limit // 2),
-                [-1],
-                range(self.num_rows - (limit - (limit // 2)), self.num_rows),
-            )
-        for j in indexes:
+        row_indexes = util.truncate_rows(self.num_rows, limit)
+        for j in row_indexes:
             if j == -1:
                 rows.append(f"__skipped__{self.num_rows-limit}")
             else:
@@ -2490,15 +2409,8 @@ def append_columns(
     def _text_header_and_rows(self, limit=None):
         headers = ("id", "timestamp", "record")
         rows = []
-        if limit is None or self.num_rows <= limit:
-            indexes = range(self.num_rows)
-        else:
-            indexes = itertools.chain(
-                range(limit // 2),
-                [-1],
-                range(self.num_rows - (limit - (limit // 2)), self.num_rows),
-            )
-        for j in indexes:
+        row_indexes = util.truncate_rows(self.num_rows, limit)
+        for j in row_indexes:
             if j == -1:
                 rows.append(f"__skipped__{self.num_rows-limit}")
             else:

python/tskit/util.py

@@ -23,6 +23,7 @@
 Module responsible for various utility functions used in other modules.
 """
 import dataclasses
+import itertools
 import json
 import numbers
 import os
@@ -370,7 +371,7 @@ def render_metadata(md, length=40):
 
 
 def unicode_table(
-    rows, title=None, header=None, row_separator=True, column_alignments=None
+    rows, *, title=None, header=None, row_separator=True, column_alignments=None
 ):
     """
     Convert a table (list of lists) of strings to a unicode table. If a row contains
@@ -445,6 +446,41 @@ def unicode_table(
     return "".join(out)
 
 
+def html_table(rows, *, header):
+    """
+    Called by jupyter notebooks to render tables
+    """
+    headers = "".join(f"<th>{h}</th>" for h in header)
+    rows = (
+        f'<td style="text-align: center;" colspan="{len(headers)}"><em>{row[11:]}'
+        f" rows skipped (tskit.set_print_options)</em></td>"
+        if "__skipped__" in row
+        else "".join(f"<td>{cell}</td>" for cell in row)
+        for row in rows
+    )
+    rows = "".join(f"<tr>{row}</tr>\n" for row in rows)
+    return f"""
+        <div>
+            <style scoped="">
+                .tskit-table tbody tr th:only-of-type {{vertical-align: middle;}}
+                .tskit-table tbody tr th {{vertical-align: top;}}
+                .tskit-table tbody td {{text-align: right;padding: 0.5em 0.5em;}}
+                .tskit-table tbody th {{padding: 0.5em 0.5em;}}
+            </style>
+            <table border="1" class="tskit-table">
+                <thead>
+                    <tr>
+                        {headers}
+                    </tr>
+                </thead>
+                <tbody>
+                    {rows}
+                </tbody>
+            </table>
+        </div>
+    """
+
+
 def tree_sequence_html(ts):
     table_rows = "".join(
         f"""
@@ -686,6 +722,20 @@ def set_print_options(*, max_lines=40):
     tskit._print_options = {"max_lines": max_lines}
 
 
+def truncate_rows(num_rows, limit=None):
+    """
+    Return a list of indexes into a set of rows, but is limit is set, truncate the
+    number of rows and place a `-1` instead of the intermediate indexes
+    """
+    if limit is None or num_rows <= limit:
+        return range(num_rows)
+    return itertools.chain(
+        range(limit // 2),
+        [-1],
+        range(num_rows - (limit - (limit // 2)), num_rows),
+    )
+
+
 def random_nucleotides(length: numbers.Number, *, seed: Union[int, None] = None) -> str:
     """
     Returns a random string of nucleotides of the specified length. Characters