Bug Showcase

Check out what we got for you in HasBugs. When you are at one Bug, you can navigate to previous and next with the arrows.

hls-4

HLS can suggest import of modules for the user. These generated imports can be placed incorrect, as it may be inserted before the 'where' of a module declaration, resulting in an invalid file. This has been addressed by rewriting the import-placement.

[integration-test] [rewrite] [ordering issue]
hls-5

Many HSL Plugins deal with text-alternation. One ('stylish-Haskell'), that does formatting on a file level, removes the last line of the file. The fix is to guard the formatting against different types of file-endings.

[unit-test] [whitespace] [string-manipulation] [missing-check]
pandoc-1

Docbook Files support linked figures, which should be convertable to other formats. DocBooks xref elements can link to figures, but the text inserted for the title of the link would be 'figure_title'(default placeholder) rather than the title of the figure. The fix is to add the 'title' attribute and use the title if available.

[integration-test] [missed case] [default value]
pandoc-2

When a file was rendered into DOCX with two tables in them, the space between them would be deleted when there was some comment block in-between.The comment should be ignored and a paragraph should be inserted for spacing.

[integration-test] [missed case] [feature-collision]
pandoc-4

Formatting in code blocks is not allowed by pandoc, causing bold code to be possible but a code block with parts bold in it impossible for LaTeX. The fix is to make sure formatting of a code block happens outside of the codeblock.

[formatting] [missing-check] [integration-test]
pandoc-5

When a code-block starts with a period, ROFF MS interpreted this as a command rather than a part of the code-block. The solution is to escape the periods.

[formatting] [escaping] [integration-test] [close-to-unit-test]
pandoc-6

Inline Markdown code-blocks had a slight parsing/priority problem. Code blocks with '(1)' in them would be wrongly delimited. The fix is to reorder parsing steps to account for lists inner-nested.

[parsing] [unit-test]
pandoc-7

Empty table cells were filtered out when reading LaTeX tables without handling multirows and multicols properly. By adding code for recognising multicols and rows correctly, empty cells can now be filtered without shifting other cells out of place.

[parsing] [data-loss] [unit-test]
pandoc-8

Pandoc is able to convert footnotes from the markdown-based 'Big-Note' to LaTeX. However, it produced a crash for multi-line footnotes when they were in other modulation (e.g., italics). The fix consists of re-ordering the elements and their styles so that their conversion does not clash.

[integration-test] [missed case] [data-representation]