Merge pull request #2146 from midichef/html_captions

[html] display title/aria-label/caption/summary
saulpw · Dec 7, 2023 · 6bf20c8 · 6bf20c8
2 parents 9fd728b + 243db37
commit 6bf20c8
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 3 deletions.
diff --git a/tests/golden/pull2140.tsv b/tests/golden/pull2140.tsv
@@ -1,3 +1,3 @@
-name	rows	cols	id	classes
-table_0	0	0		test_empty
-links	0	5		
+name	rows	cols	id	classes	title	aria_label	caption	summary	heading
+table_0	0	0		test_empty					
+links	0	5							
diff --git a/visidata/loaders/html.py b/visidata/loaders/html.py
@@ -33,6 +33,11 @@ class HtmlTablesSheet(IndexSheet):
         Column('tag', width=0, getter=lambda col,row: row.html.tag),
         Column('id', getter=lambda col,row: row.html.attrib.get('id')),
         Column('classes', getter=lambda col,row: row.html.attrib.get('class')),
+        Column('title', getter=lambda col,row: row.html.attrib.get('title')),
+        Column('aria_label', getter=lambda col,row: row.html.attrib.get('aria-label')),
+        Column('caption', getter=lambda col,row: row.html.xpath('normalize-space(./caption)') if row.html.xpath('./caption') else None, cache=True),
+        Column('summary', getter=lambda col,row: row.html.attrib.get('summary')),
+        Column('heading', getter=lambda col,row: row.html.xpath('normalize-space(./preceding-sibling::*[self::h1 or self::h2 or self::h3 or self::h4 or self::h5 or self::h6][1])') or None, cache=True),
     ]
     def iterload(self):
         lxml = vd.importExternal('lxml')