rxiv-maker 1.18.1__tar.gz → 1.18.2__tar.gz

{rxiv_maker-1.18.1 → rxiv_maker-1.18.2}/CHANGELOG.md

@@ -7,6 +7,16 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [1.18.2] - 2026-01-06
+
+### Fixed
+
+- **Table Caption Parser Enhancement**: Improved robustness of table caption detection
+  - Flexible whitespace handling: Parser now skips blank lines and HTML comments between tables and captions
+  - Cross-reference support: Table captions can now include references to figures, tables, and equations
+  - Dynamic line-skipping logic: Intelligently determines how many lines to skip based on actual content
+  - Fixes issues where captions separated from tables by comments or extra whitespace were not detected
+
 ## [1.18.1] - 2024-12-24
 
 ### Added

{rxiv_maker-1.18.1 → rxiv_maker-1.18.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: rxiv-maker
-Version: 1.18.1
+Version: 1.18.2
 Summary: Write scientific preprints in Markdown. Generate publication-ready PDFs efficiently.
 Project-URL: Homepage, https://github.com/HenriquesLab/rxiv-maker
 Project-URL: Documentation, https://github.com/HenriquesLab/rxiv-maker#readme

{rxiv_maker-1.18.1 → rxiv_maker-1.18.2}/src/rxiv_maker/__version__.py

@@ -1,3 +1,3 @@
 """Version information."""
 
-__version__ = "1.18.1"
+__version__ = "1.18.2"

{rxiv_maker-1.18.1 → rxiv_maker-1.18.2}/src/rxiv_maker/converters/table_processor.py

@@ -139,9 +139,9 @@ def convert_tables_to_latex(
                     result_lines.pop()  # Remove caption line
 
             # Check for new format table caption after the table
-            new_format_caption, table_id, rotation_angle = _parse_table_caption(lines, i)
-            if new_format_caption:
-                i += 2  # Skip blank line and caption line
+            new_format_caption, table_id, rotation_angle, lines_to_skip = _parse_table_caption(lines, i)
+            if new_format_caption and lines_to_skip:
+                i += lines_to_skip  # Skip blank lines, comments, and caption line
 
             # Generate LaTeX table with the processed caption
             latex_table = generate_latex_table(
@@ -844,39 +844,68 @@ def _is_table_row(line: str) -> bool:
     return "|" in line and line.startswith("|") and line.endswith("|")
 
 
-def _parse_table_caption(lines: list[str], i: int) -> tuple[str | None, str | None, int | None]:
-    """Parse table caption in new format after table."""
+def _parse_table_caption(lines: list[str], i: int) -> tuple[str | None, str | None, int | None, int | None]:
+    """Parse table caption in new format after table.
+
+    Returns:
+        tuple of (caption, table_id, rotation_angle, lines_to_skip)
+        lines_to_skip indicates how many lines after position i should be skipped
+    """
     new_format_caption = None
     table_id = None
     rotation_angle = None
+    lines_to_skip = None
 
-    if (
-        i < len(lines)
-        and lines[i].strip() == ""
-        and i + 1 < len(lines)
-        and re.match(r"^\{#[a-zA-Z0-9_:-]+.*\}\s*\*\*.*\*\*", lines[i + 1].strip())
-    ):
-        # Found new format caption, parse it
-        caption_line = lines[i + 1].strip()
-
-        # Parse caption with optional attributes like rotate=90
-        caption_match = re.match(r"^\{#([a-zA-Z0-9_:-]+)([^}]*)\}\s*(.+)$", caption_line)
-        if caption_match:
-            table_id = caption_match.group(1)
-            attributes_str = caption_match.group(2).strip()
-            caption_text = caption_match.group(3)
-
-            # Extract rotation attribute if present
-            if attributes_str:
-                rotation_match = re.search(r"rotate=(\d+)", attributes_str)
-                if rotation_match:
-                    rotation_angle = int(rotation_match.group(1))
-
-            # Process caption text to handle markdown formatting
-            new_format_caption = re.sub(r"\*\*([^*]+)\*\*", r"\\textbf{\1}", caption_text)
-            new_format_caption = re.sub(r"\*([^*]+)\*", r"\\textit{\1}", new_format_caption)
-
-    return new_format_caption, table_id, rotation_angle
+    # Look ahead for caption, skipping blank lines and HTML comments
+    # Search up to 5 lines ahead to be flexible
+    search_limit = min(i + 5, len(lines))
+
+    for offset in range(search_limit - i):
+        check_line = lines[i + offset].strip()
+
+        # Skip blank lines and HTML comments
+        if check_line == "" or (check_line.startswith("<!--") and check_line.endswith("-->")):
+            continue
+
+        # Check if this is a new-format caption
+        if re.match(r"^\{#[a-zA-Z0-9_:-]+.*\}\s*\*\*.*\*\*", check_line):
+            caption_line = check_line
+
+            # Parse caption with optional attributes like rotate=90
+            caption_match = re.match(r"^\{#([a-zA-Z0-9_:-]+)([^}]*)\}\s*(.+)$", caption_line)
+            if caption_match:
+                table_id = caption_match.group(1)
+                attributes_str = caption_match.group(2).strip()
+                caption_text = caption_match.group(3)
+
+                # Extract rotation attribute if present
+                if attributes_str:
+                    rotation_match = re.search(r"rotate=(\d+)", attributes_str)
+                    if rotation_match:
+                        rotation_angle = int(rotation_match.group(1))
+
+                # Process caption text to handle markdown formatting
+                new_format_caption = re.sub(r"\*\*([^*]+)\*\*", r"\\textbf{\1}", caption_text)
+                new_format_caption = re.sub(r"\*([^*]+)\*", r"\\textit{\1}", new_format_caption)
+
+                # Convert cross-references (figures, tables, equations) in caption
+                from .figure_processor import (
+                    convert_equation_references_to_latex,
+                    convert_figure_references_to_latex,
+                )
+
+                new_format_caption = convert_figure_references_to_latex(new_format_caption)
+                new_format_caption = convert_table_references_to_latex(new_format_caption)
+                new_format_caption = convert_equation_references_to_latex(new_format_caption)
+
+                # Return how many lines to skip (all lines from i to caption, inclusive)
+                lines_to_skip = offset + 1
+                break
+        else:
+            # Found a non-blank, non-comment, non-caption line - stop searching
+            break
+
+    return new_format_caption, table_id, rotation_angle, lines_to_skip
 
 
 def _split_table_row_respecting_backticks(row: str) -> list[str]:

{rxiv_maker-1.18.1 → rxiv_maker-1.18.2}/tests/cli/test_changelog_command.py

@@ -27,6 +27,12 @@ def sample_changelog():
     # Include current version to match __version__
     return """# Changelog
 
+## [v1.18.2] - 2026-01-06
+
+### Fixed
+- Table caption parser enhancement with flexible whitespace handling
+- Cross-reference support in table captions
+
 ## [v1.18.0] - 2025-12-23
 
 ### Added
@@ -144,7 +150,7 @@ class TestChangelogCommand:
         """Test --breaking-only flag."""
         mock_fetch.return_value = sample_changelog
 
-        result = runner.invoke(changelog, ["--recent", "7", "--breaking-only"])
+        result = runner.invoke(changelog, ["--recent", "8", "--breaking-only"])
 
         assert result.exit_code == 0
         # Should show v1.12.0 which has breaking changes
@@ -199,7 +205,7 @@ class TestChangelogCommand:
         mock_fetch.return_value = sample_changelog
 
         # Use the latest version to ensure no versions after it
-        result = runner.invoke(changelog, ["--since", "v1.18.0"])
+        result = runner.invoke(changelog, ["--since", "v1.18.2"])
 
         assert result.exit_code == 0
         output = strip_ansi(result.output)