Cyclic0007/python-markdownify
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add a new configuration option to control tabler header row inference (#161)

Browse Source 
Add option to infer first table row as table header (defaults to false)
This commit is contained in:
SomeBottle
2025-01-19 21:13:24 +08:00
committed by GitHub
parent 1783995cb2
commit 3bf0b527a4
4 changed files with 81 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
README.rst
View File

@@ -139,6 +139,11 @@ keep_inline_images_in
that should be allowed to contain inline images, for example ``['td']``.
Defaults to an empty list.
table_infer_header
Controls handling of tables with no header row (as indicated by ``<thead>``
or ``<th>``). When set to ``True``, the first body row is used as the header row.
Defaults to ``False``, which leaves the header row empty.
wrap, wrap_width
If ``wrap`` is set to ``True``, all text paragraphs are wrapped at
``wrap_width`` characters. Defaults to ``False`` and ``80``.

33
markdownify/__init__.py
View File

@@ -102,6 +102,7 @@ class MarkdownConverter(object):
strong_em_symbol = ASTERISK
sub_symbol = ''
sup_symbol = ''
table_infer_header = False
wrap = False
wrap_width = 80
@@ -518,13 +519,24 @@ class MarkdownConverter(object):
cells = el.find_all(['td', 'th'])
is_headrow = (
all([cell.name == 'th' for cell in cells])
or (not el.previous_sibling and not el.parent.name == 'tbody')
or (el.parent.name == 'thead'
# avoid multiple tr in thead
and len(el.parent.find_all('tr')) == 1)
)
is_head_row_missing = (
(not el.previous_sibling and not el.parent.name == 'tbody')
or (not el.previous_sibling and el.parent.name == 'tbody' and len(el.parent.parent.find_all(['thead'])) < 1)
)
overline = ''
underline = ''
if is_headrow and not el.previous_sibling:
# first row and is headline: print headline underline
if ((is_headrow
or (is_head_row_missing
and self.options['table_infer_header']))
and not el.previous_sibling):
# first row and:
# - is headline or
# - headline is missing and header inference is enabled
# print headline underline
full_colspan = 0
for cell in cells:
if 'colspan' in cell.attrs and cell['colspan'].isdigit():
@@ -532,13 +544,16 @@ class MarkdownConverter(object):
else:
full_colspan += 1
underline += '| ' + ' | '.join(['---'] * full_colspan) + ' |' + '\n'
elif (not el.previous_sibling
and (el.parent.name == 'table'
or (el.parent.name == 'tbody'
and not el.parent.previous_sibling))):
elif ((is_head_row_missing
and not self.options['table_infer_header'])
or (not el.previous_sibling
and (el.parent.name == 'table'
or (el.parent.name == 'tbody'
and not el.parent.previous_sibling)))):
# headline is missing and header inference is disabled or:
# first row, not headline, and:
# - the parent is table or
# - the parent is tbody at the beginning of a table.
# - the parent is table or
# - the parent is tbody at the beginning of a table.
# print empty headline above this row
overline += '| ' + ' | '.join([''] * len(cells)) + ' |' + '\n'
overline += '| ' + ' | '.join(['---'] * len(cells)) + ' |' + '\n'

4
markdownify/main.py
View File

@@ -61,6 +61,10 @@ def main(argv=sys.argv[1:]):
"should be converted to markdown images instead, this option can "
"be set to a list of parent tags that should be allowed to "
"contain inline images.")
parser.add_argument('--table-infer-header', dest='table_infer_header',
action='store_true',
help="When a table has no header row (as indicated by '<thead>' "
"or '<th>'), use the first body row as the header row.")
parser.add_argument('-w', '--wrap', action='store_true',
help="Wrap all text paragraphs at --wrap-width characters.")
parser.add_argument('--wrap-width', type=int, default=80)

51
tests/test_tables.py
View File

@@ -141,6 +141,33 @@ table_head_body_missing_head = """<table>
</tbody>
</table>"""
table_head_body_multiple_head = """<table>
<thead>
<tr>
<td>Creator</td>
<td>Editor</td>
<td>Server</td>
</tr>
<tr>
<td>Operator</td>
<td>Manager</td>
<td>Engineer</td>
</tr>
</thead>
<tbody>
<tr>
<td>Bob</td>
<td>Oliver</td>
<td>Tom</td>
</tr>
<tr>
<td>Thomas</td>
<td>Lucas</td>
<td>Ethan</td>
</tr>
</tbody>
</table>"""
table_missing_text = """<table>
<thead>
<tr>
@@ -245,10 +272,28 @@ def test_table():
assert md(table_with_linebreaks) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith Jackson | 50 |\n| Eve | Jackson Smith | 94 |\n\n'
assert md(table_with_header_column) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_head_body) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_head_body_multiple_head) == '\n\n| | | |\n| --- | --- | --- |\n| Creator | Editor | Server |\n| Operator | Manager | Engineer |\n| Bob | Oliver | Tom |\n| Thomas | Lucas | Ethan |\n\n'
assert md(table_head_body_missing_head) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_missing_text) == '\n\n| | Lastname | Age |\n| --- | --- | --- |\n| Jill | | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_missing_head) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_body) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_with_caption) == 'TEXT\n\nCaption\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n\n'
assert md(table_missing_head) == '\n\n| | | |\n| --- | --- | --- |\n| Firstname | Lastname | Age |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_body) == '\n\n| | | |\n| --- | --- | --- |\n| Firstname | Lastname | Age |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_with_caption) == 'TEXT\n\nCaption\n\n| | | |\n| --- | --- | --- |\n| Firstname | Lastname | Age |\n\n'
assert md(table_with_colspan) == '\n\n| Name | | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_with_undefined_colspan) == '\n\n| Name | Age |\n| --- | --- |\n| Jill | Smith |\n\n'
def test_table_infer_header():
assert md(table, table_infer_header=True) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_with_html_content, table_infer_header=True) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| **Jill** | *Smith* | [50](#) |\n| Eve | Jackson | 94 |\n\n'
assert md(table_with_paragraphs, table_infer_header=True) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_with_linebreaks, table_infer_header=True) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith Jackson | 50 |\n| Eve | Jackson Smith | 94 |\n\n'
assert md(table_with_header_column, table_infer_header=True) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_head_body, table_infer_header=True) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_head_body_multiple_head, table_infer_header=True) == '\n\n| Creator | Editor | Server |\n| --- | --- | --- |\n| Operator | Manager | Engineer |\n| Bob | Oliver | Tom |\n| Thomas | Lucas | Ethan |\n\n'
assert md(table_head_body_missing_head, table_infer_header=True) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_missing_text, table_infer_header=True) == '\n\n| | Lastname | Age |\n| --- | --- | --- |\n| Jill | | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_missing_head, table_infer_header=True) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_body, table_infer_header=True) == '\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_with_caption, table_infer_header=True) == 'TEXT\n\nCaption\n\n| Firstname | Lastname | Age |\n| --- | --- | --- |\n\n'
assert md(table_with_colspan, table_infer_header=True) == '\n\n| Name | | Age |\n| --- | --- | --- |\n| Jill | Smith | 50 |\n| Eve | Jackson | 94 |\n\n'
assert md(table_with_undefined_colspan, table_infer_header=True) == '\n\n| Name | Age |\n| --- | --- |\n| Jill | Smith |\n\n'