Run this with cargo test --features gen-tests suite::table.

Adapted from original by replacing td elements by th elements inside thead and disabling the third test.

False match

Test header
-----------

<h2>Test header</h2>

True match

Test|Table
----|-----

<table><thead><tr><th>Test</th><th>Table</th></tr></thead><tbody></tbody>
</table>

Actual rows in it

Test|Table
----|-----
Test row
Test|2

Test ending
.
<table><thead><tr><th>Test</th><th>Table</th></tr></thead><tbody>
<tr><td>Test row</td></tr>
<tr><td>Test</td><td>2</td></tr>
</tbody></table>
<p>Test ending</p>

Test with quote

> Test  | Table
> ------|------
> Row 1 | Every
> Row 2 | Day
>
> Paragraph

<blockquote>
<table><thead><tr><th>Test</th><th>Table</th></tr></thead><tbody>
<tr><td>Row 1</td><td>Every</td></tr>
<tr><td>Row 2</td><td>Day</td></tr>
</tbody></table>
<p>Paragraph</p>
</blockquote>

Test with list

 1. First entry
 2. Second entry

    Col 1|Col 2
    -|-
    Row 1|Part 2
    Row 2|Part 2

<ol>
<li>
<p>First entry</p>
</li>
<li>
<p>Second entry</p>
<table><thead><tr><th>Col 1</th><th>Col 2</th></tr></thead><tbody>
<tr><td>Row 1</td><td>Part 2</td></tr>
<tr><td>Row 2</td><td>Part 2</td></tr>
</tbody></table>
</li>
</ol>

Test with border

|Col 1|Col 2|
|-----|-----|
|R1C1 |R1C2 |
|R2C1 |R2C2 |

<table><thead><tr><th>Col 1</th><th>Col 2</th></tr></thead><tbody>
<tr><td>R1C1</td><td>R1C2</td></tr>
<tr><td>R2C1</td><td>R2C2</td></tr>
</tbody></table>

Test with empty cells

Empty cells should work.

| Col 1 | Col 2 |
|-------|-------|
|       |       |
|       |       |

<table><thead><tr><th>Col 1</th><th>Col 2</th></tr></thead><tbody>
<tr><td></td><td></td></tr>
<tr><td></td><td></td></tr>
</tbody></table>

... and properly mix with filled cells.

| Col 1 | Col 2 |
|-------|-------|
|   x   |       |
|       |    x  |

<table><thead><tr><th>Col 1</th><th>Col 2</th></tr></thead><tbody>
<tr><td>x</td><td></td></tr>
<tr><td></td><td>x</td></tr>
</tbody></table>

Table with UTF-8

Basic example.

|Col 1|Col 2|
|-----|-----|
|✓    |✓    |
|✓    |✓    |

<table><thead><tr><th>Col 1</th><th>Col 2</th></tr></thead><tbody>
<tr><td>✓</td><td>✓</td></tr>
<tr><td>✓</td><td>✓</td></tr>
</tbody></table>

More advanced example.

|  Target                       | std |rustc|cargo| notes                      |
|-------------------------------|-----|-----|-----|----------------------------|
| `x86_64-unknown-linux-musl`   |  ✓  |     |     | 64-bit Linux with MUSL     |
| `arm-linux-androideabi`       |  ✓  |     |     | ARM Android                |
| `arm-unknown-linux-gnueabi`   |  ✓  |  ✓  |     | ARM Linux (2.6.18+)        |
| `arm-unknown-linux-gnueabihf` |  ✓  |  ✓  |     | ARM Linux (2.6.18+)        |
| `aarch64-unknown-linux-gnu`   |  ✓  |     |     | ARM64 Linux (2.6.18+)      |
| `mips-unknown-linux-gnu`      |  ✓  |     |     | MIPS Linux (2.6.18+)       |
| `mipsel-unknown-linux-gnu`    |  ✓  |     |     | MIPS (LE) Linux (2.6.18+)  |

<table><thead><tr><th>Target</th><th>std</th><th>rustc</th><th>cargo</th><th>notes</th></tr></thead><tbody>
<tr><td><code>x86_64-unknown-linux-musl</code></td><td>✓</td><td></td><td></td><td>64-bit Linux with MUSL</td></tr>
<tr><td><code>arm-linux-androideabi</code></td><td>✓</td><td></td><td></td><td>ARM Android</td></tr>
<tr><td><code>arm-unknown-linux-gnueabi</code></td><td>✓</td><td>✓</td><td></td><td>ARM Linux (2.6.18+)</td></tr>
<tr><td><code>arm-unknown-linux-gnueabihf</code></td><td>✓</td><td>✓</td><td></td><td>ARM Linux (2.6.18+)</td></tr>
<tr><td><code>aarch64-unknown-linux-gnu</code></td><td>✓</td><td></td><td></td><td>ARM64 Linux (2.6.18+)</td></tr>
<tr><td><code>mips-unknown-linux-gnu</code></td><td>✓</td><td></td><td></td><td>MIPS Linux (2.6.18+)</td></tr>
<tr><td><code>mipsel-unknown-linux-gnu</code></td><td>✓</td><td></td><td></td><td>MIPS (LE) Linux (2.6.18+)</td></tr>
</tbody></table>

Hiragana-containing pseudo-table.

|-|-|
|ぃ|い|

<p>|-|-|
|ぃ|い|</p>

Hiragana-containing actual table.

|ぁ|ぃ|
|-|-|
|ぃ|ぃ|

<table><thead><tr><th>ぁ</th><th>ぃ</th></tr></thead><tbody>
<tr><td>ぃ</td><td>ぃ</td></tr>
</tbody></table>

Test russian symbols.

|Колонка 1|Колонка 2|
|---------|---------|
|Ячейка 1 |Ячейка 2 |

<table><thead><tr><th>Колонка 1</th><th>Колонка 2</th></tr></thead><tbody>
<tr><td>Ячейка 1</td><td>Ячейка 2</td></tr>
</tbody></table>

Test cases based on https://github.com/github/cmark-gfm/issues/180 and https://github.com/raphlinus/pulldown-cmark/issues/540.

Pulldown-cmark gives "heavy" tables (with a leading |) higher priority than "light" tables for paragraph interruption. This is compatible with older versions of GitHub, but 2023-06-21 GitHub treats them the same.

A quick run through https://babelmark.github.io shows, of the implementations that support pipe tables at all¹:

Implementation ->	a	b	c	d	e	f	g	h	i	j
pulldown-cmark	✓	x	x	x	x	✓	x	x	✓	x
github/cmark-gfm	✓	x	✓	x	✓	✓	✓	✓	✓	x
pycmarkgfm	✓	x	✓	x	✓	✓	✓	✓	✓	x
DFM²	✓	✓	✓	✓	✓	✓	✓	-	-	x
league/commonmark GFM	✓	✓	✓	✓	✓	✓	✓	x	✓	x
s9e/TextFormatter	✓	✓	✓	✓	✓	✓	✓	x	x	x
markdown-it[^3]³	✓	x	✓	x	✓	✓	✓	x	-	-
php-markdown-extra	✓	x	✓	x	✓	✓	✓	x	x	x
markdig	x	x	x	x	x	x	x	x	x	x
MD4C	x	x	x	x	x	x	x	x	x	x
parsedown	x	x	x	x	x	x	x	x	x	x
maruku⁴	x	x	x	x	x	x	x	x	x	-
cebe	x	x	x	x	x	x	x	x	x	x
pandoc	x	x	x	x	x	x	x	x	x	x
multimarkdown	x	x	x	x	x	x	x	x	x	x

While pulldown-cmark could probably afford to align itself with GFM, tables interrupting paragraphs is not a portable feature, and anyone trying to write widely-compatible markdown syntax can't rely on it.

Implementations that "support pipe tables at all" were identified by including "table a" without a preceding paragraph and seeing if it produced an HTML table for that.

On "table h" and "table i", DFM sees a table with no cells, followed by a paragraph with b in it.

⁴

On "table j", maruku and markdown-it see a table with no cells, followed by a paragraph with b in it.

On "table i", markdown-it sees a table with no cells, followed by a paragraph with b in it.

table a
|  a  |  b  |
| --- | --- |
|  c  |  d  |


table b
    |  a  |  b  |
    | --- | --- |
    |  c  |  d  |


table c
 a  |  b
--- | ---
 c  |  d


table d
    a | b
    --|--
    c | d


table e
a | b
--|--
c | d

table f
  |  a  |  b  |
  | --- | --- |
  |  c  |  d  |


table g
   a  |  b
  --- | ---
   c  |  d

table h
a
|-|
b

table i
| a
|-
b

table j
| a
-
b

<p>table a</p>
<table><thead><tr><th>a</th><th>b</th></tr></thead><tbody>
<tr><td>c</td><td>d</td></tr>
</tbody></table>
<p>table b
|  a  |  b  |
| --- | --- |
|  c  |  d  |</p>
<p>table c
a  |  b
--- | ---
c  |  d</p>
<p>table d
a | b
--|--
c | d</p>
<p>table e
a | b
--|--
c | d</p>
<p>table f</p>
<table><thead><tr><th>a</th><th>b</th></tr></thead><tbody>
<tr><td>c</td><td>d</td></tr>
</tbody></table>
<p>table g
a  |  b
--- | ---
c  |  d</p>
<p>table h
a
|-|
b</p>
<p>table i</p>
<table><thead><tr><th>a</th></tr></thead><tbody>
<tr><td>b</td></tr>
</tbody></table>
<h2>table j
| a</h2>
<p>b</p>

Test case based on https://github.com/github/cmark-gfm/issues/333. Normally, we'd follow GFM's lead, but this one seems like a bug, and Pandoc, the other big, authoritative Extended Markdown parser, treats it as a table.

Like the example above showing interruption, this ambiguity is tested:

Implementation ->	parsed as
pulldown-cmark	table
github/cmark-gfm	list
pycmarkgfm	list
markdown-it	table
php-markdown-extra	table
DFM	table
league/commonmark GFM	list
s9e/TextFormatter	table
markdig	table
MD4C	list
parsedown	table
maruku	table
cebe	table
pandoc	table
multimarkdown	table

a | b
- | -
1 | 2

<table><thead><tr><th>a</th><th>b</th></tr></thead><tbody>
<tr><td>1</td><td>2</td></tr>
</tbody></table>

Hard line breaks are not allowed at the end of table header lines. It's about a 50:50 split on babelmark, but this change makes pulldown-cmark consistent with GFM and Pandoc.

Implementation ->	parsed as
pulldown-cmark	paragraph
github/cmark-gfm	paragraph
pycmarkgfm	paragraph
markdown-it	table
php-markdown-extra	table
DFM	table
league/commonmark GFM	paragraph
s9e/TextFormatter	table
markdig	paragraph
MD4C	paragraph
parsedown	table
maruku	table
cebe	paragraph
pandoc	paragraph
multimarkdown	table

a | b\
- | -
1 | 2

<p>a | b\</p>
<ul>
<li>| -
1 | 2</li>
</ul>

As a block structure, table parsing has a higher priority than backslashes. This is consistent with how other paragraph-interrupting structures work, and is what GitHub does.

a\
| b | c |
|---|---|
| d | e |

<p>a\</p>
<table><thead><tr><th>b</th><th>c</th></tr></thead><tbody>
<tr><td>d</td><td>e</td></tr>
</tbody></table>

As a special case, pipes in inline code in tables are escaped with backslashes.

The parsing rule for CommonMark is that block structures are parsed before inline structures are. Normally, this means backslashes aren't allowed to have any effect on block structures at all. Tables do consider backslashes, but not the same way inline syntax does: they have higher precedence than anything else, as-if they were parsed in a completely separate pass.

This rule should be identical to GitHub's. See https://gist.github.com/notriddle/c027512ee849f12098fec3a3256c89d3 for what they do. This is totally different from Pandoc's commonmark_x, for example, and from many other Markdown parsers, because of some weird corner cases it creates.

| Description | Test case |
|-------------|-----------|
| Single      | `\`       |
| Double      | `\\`      |
| Basic test  | `\|`      |
| Basic test 2| `\|\|\`   |
| Basic test 3| `x\|y\|z\`|
| Not pipe    | `\.`      |
| Combo       | `\.\|\`   |
| Extra       | `\\\.`    |
| Wait, what? | `\\|`     |
| Wait, what? | `\\\|`    |
| Wait, what? | `\\\\|`   |
| Wait, what? | `\\\\\|`  |
| Wait, what? |          \|
| Wait, what? |         \\|
| Wait, what? |        \\\|
| Wait, what?x|          \|x
| Wait, what?x|         \\|x
| Wait, what?x|        \\\|x
| Direct trail|         \.|x

<table><thead><tr><th>Description</th><th>Test case</th></tr></thead><tbody>
<tr><td>Single</td><td><code>\</code></td></tr>
<tr><td>Double</td><td><code>\\</code></td></tr>
<tr><td>Basic test</td><td><code>|</code></td></tr>
<tr><td>Basic test 2</td><td><code>||\</code></td></tr>
<tr><td>Basic test 3</td><td><code>x|y|z\</code></td></tr>
<tr><td>Not pipe</td><td><code>\.</code></td></tr>
<tr><td>Combo</td><td><code>\.|\</code></td></tr>
<tr><td>Extra</td><td><code>\\\.</code></td></tr>
<tr><td>Wait, what?</td><td><code>\|</code></td></tr>
<tr><td>Wait, what?</td><td><code>\\|</code></td></tr>
<tr><td>Wait, what?</td><td><code>\\\|</code></td></tr>
<tr><td>Wait, what?</td><td><code>\\\\|</code></td></tr>
<tr><td>Wait, what?</td><td>|</td></tr>
<tr><td>Wait, what?</td><td>|</td></tr>
<tr><td>Wait, what?</td><td>\|</td></tr>
<tr><td>Wait, what?x</td><td>|x</td></tr>
<tr><td>Wait, what?x</td><td>|x</td></tr>
<tr><td>Wait, what?x</td><td>\|x</td></tr>
<tr><td>Direct trail</td><td>.</td></tr>
</tbody></table>

Example with code blocks in the table's header.

| Single | `\|` |
|--|--|
| Single | `\|` |


| Double | `\\|` |
|--|--|
| Double | `\\|` |


| Double Twice | `\\|\\|` |
|--|--|
| Double Twice | `\\|\\|` |


| Triple | `\\\|` |
|--|--|
| Triple | `\\\|` |

<table><thead><tr><th>Single</th><th><code>|</code></th></tr></thead><tbody>
<tr><td>Single</td><td><code>|</code></td></tr>
</tbody></table>
<table><thead><tr><th>Double</th><th><code>\|</code></th></tr></thead><tbody>
<tr><td>Double</td><td><code>\|</code></td></tr>
</tbody></table>
<table><thead><tr><th>Double Twice</th><th><code>\|\|</code></th></tr></thead><tbody>
<tr><td>Double Twice</td><td><code>\|\|</code></td></tr>
</tbody></table>
<table><thead><tr><th>Triple</th><th><code>\\|</code></th></tr></thead><tbody>
<tr><td>Triple</td><td><code>\\|</code></td></tr>
</tbody></table>

Table rows must have at least one cell. A single pipe, on its own, doesn't count.

The behavior on babelmark is pretty diverse, but this behavior is chosen to align with GFM.

✓ means it treats the zero cell lines with only a pipe as not being valid rows, but otherwise parses as a table.

x means it treats the zero cell lines with only a pipe as a valid row.

? means it does something else.

Implementation ->	first	second	third
pulldown-cmark	✓	✓	✓
github/cmark-gfm	✓	✓	✓
pycmarkgfm	✓	✓	✓
pandoc gfm³	✓	✓	✓
parsedown	x	x	✓
php-markdown-extra	x	x	✓
markdown-it	x	x	✓⁴
DFM	x	x	x
league/commonmark GFM	x	x	x
s9e/TextFormatter	x	x	x
MD4C	x	x	x
cebe	x	x	x
pandoc	x	x	x
multimarkdown	?¹	?¹	?¹
markdig	?¹	?¹	?¹
maruku	?²	?²	?²

babelmark test cases

Multimarkdown and Markdig don't recognize these at tables at all, even though I know they support pipe tables.

Maruku swallows much of the text entirely in these test cases.

⁴

Markdown-it changed some time between the version on their playground https://markdown-it.github.io/#md3=%7B%22source%22%3A%22%7C%20Table%20%7C%20Header%20%7C%5Cn%7C-------%7C--------%7C%5Cn%7C%20Table%20%7C%20Body%20%20%20%7C%5Cn%7C%5Cn%7C%20Not%20%20%20%7C%20Enough%20%7C%5Cn%5Cn%7C%20Table%20%7C%20Header%20%7C%5Cn%7C-------%7C--------%7C%5Cn%7C%5Cn%5Cn%7C%5Cn%7C-------%7C--------%7C%5Cn%7C%20Table%20%7C%20Body%20%20%20%7C%5Cn%22%2C%22defaults%22%3A%7B%22html%22%3Afalse%2C%22xhtmlOut%22%3Afalse%2C%22breaks%22%3Afalse%2C%22langPrefix%22%3A%22language-%22%2C%22linkify%22%3Atrue%2C%22typographer%22%3Atrue%2C%22_highlight%22%3Atrue%2C%22_strict%22%3Afalse%2C%22_view%22%3A%22html%22%7D%7D and the version Babelmark 3 uses.

Not on babelmark 3. I ran it on https://pandoc.org/try/?params=%7B%22text%22%3A%22%7C+Table+%7C+Header+%7C%5Cn%7C-------%7C--------%7C%5Cn%7C+Table+%7C+Body+++%7C%5Cn%7C%5Cn%7C+Not+++%7C+Enough+%7C%5Cn%5Cn%7C+Table+%7C+Header+%7C%5Cn%7C-------%7C--------%7C%5Cn%7C%5Cn%5Cn%7C%5Cn%7C-------%7C--------%7C%5Cn%7C+Table+%7C+Body+++%7C%5Cn%22%2C%22to%22%3A%22html5%22%2C%22from%22%3A%22gfm%22%2C%22standalone%22%3Afalse%2C%22embed-resources%22%3Afalse%2C%22table-of-contents%22%3Afalse%2C%22number-sections%22%3Afalse%2C%22citeproc%22%3Afalse%2C%22html-math-method%22%3A%22plain%22%2C%22wrap%22%3A%22auto%22%2C%22highlight-style%22%3Anull%2C%22files%22%3A%7B%7D%2C%22template%22%3Anull%7D

| Table | Header |
|-------|--------|
| Table | Body   |
|
| Not   | Enough |


| Table | Header |
|-------|--------|
| Table | Body   |
|→
| Not   | Enough |

<table><thead><tr><th>Table</th><th>Header</th></tr></thead><tbody>
<tr><td>Table</td><td>Body</td></tr>
</tbody></table>
<p>|
| Not   | Enough |</p>
<table><thead><tr><th>Table</th><th>Header</th></tr></thead><tbody>
<tr><td>Table</td><td>Body</td></tr>
</tbody></table>
<p>|
| Not   | Enough |</p>

| Table | Header |
|-------|--------|
|

<table><thead><tr><th>Table</th><th>Header</th></tr></thead><tbody>
</tbody></table>
<p>|</p>

|
|-------|--------|
| Table | Body   |

<p>|
|-------|--------|
| Table | Body   |</p>

Double escaping in link URLs

https://markdown-it.github.io/#md3=%7B%22source%22%3A%22%7C%20Single%20%7C%20%5Btest%5D%28first%5C%5C%7Csecond%29%20%7C%5Cn%7C--%7C--%7C%5Cn%5Cn%7C%20Double%20%7C%20%5Btest%5D%28first%5C%5C%5C%5C%7Csecond%29%20%7C%5Cn%7C--%7C--%7C%5Cn%5Cn%7C%20Triple%20%7C%20%5Btest%5D%28first%5C%5C%5C%5C%5C%5C%7Csecond%29%20%7C%5Cn%7C--%7C--%7C%5Cn%22%2C%22defaults%22%3A%7B%22html%22%3Afalse%2C%22xhtmlOut%22%3Afalse%2C%22breaks%22%3Afalse%2C%22langPrefix%22%3A%22language-%22%2C%22linkify%22%3Atrue%2C%22typographer%22%3Atrue%2C%22_highlight%22%3Atrue%2C%22_strict%22%3Afalse%2C%22_view%22%3A%22src%22%7D%7D

https://gist.github.com/notriddle/a789116ccec90d261242f3e5fd3bf0ee

| Single | [test](first\|second) |
|--|--|

| Double | [test](first\\|second) |
|--|--|

| Triple | [test](first\\\|second) |
|--|--|

<table><thead><tr><th>Single</th><th><a href="first%7Csecond">test</a></th></tr></thead><tbody>
</tbody></table>
<table><thead><tr><th>Double</th><th><a href="first%7Csecond">test</a></th></tr></thead><tbody>
</tbody></table>
<table><thead><tr><th>Triple</th><th><a href="first%5C%7Csecond">test</a></th></tr></thead><tbody>
</tbody></table>

Double escaping in link references

https://gist.github.com/notriddle/02d6dfe2559a1a05fa9308e0bf4a9c39

https://markdown-it.github.io/#md3=%7B%22source%22%3A%22%7C%20Single%20%7C%20%5Bfirst%5C%5C%7Csecond%5D%20%7C%5Cn%7C--%7C--%7C%5Cn%5Cn%7C%20Double%20%7C%20%5Bfirst%5C%5C%5C%5C%7Csecond%5D%20%7C%5Cn%7C--%7C--%7C%5Cn%5Cn%7C%20Triple%20%7C%20%5Bfirst%5C%5C%5C%5C%5C%5C%7Csecond%5D%20%7C%5Cn%7C--%7C--%7C%5Cn%5Cn%5Bfirst%5C%5C%7Csecond%5D%3A%20https%3A%2F%2Frust-lang.org%5Cn%5Cn%5Bfirst%5C%5C%5C%5C%7Csecond%5D%3A%20https%3A%2F%2Fdocs.rs%22%2C%22defaults%22%3A%7B%22html%22%3Afalse%2C%22xhtmlOut%22%3Afalse%2C%22breaks%22%3Afalse%2C%22langPrefix%22%3A%22language-%22%2C%22linkify%22%3Atrue%2C%22typographer%22%3Atrue%2C%22_highlight%22%3Atrue%2C%22_strict%22%3Afalse%2C%22_view%22%3A%22html%22%7D%7D

| Single | [first\|second] |
|--|--|

| Double | [first\\|second] |
|--|--|

| Triple | [first\\\|second] |
|--|--|

[first\|second]: https://rust-lang.org

[first\\|second]: https://docs.rs

<table><thead><tr><th>Single</th><th>[first|second]</th></tr></thead><tbody>
</tbody></table>
<table><thead><tr><th>Double</th><th><a href="https://rust-lang.org">first|second</a></th></tr></thead><tbody>
</tbody></table>
<table><thead><tr><th>Triple</th><th><a href="https://docs.rs">first\|second</a></th></tr></thead><tbody>
</tbody></table>

Double escaping in table interruption

Q: Knock knock.
A: Who's there.
Q: Interrupting cow.
A: Interrupting —?
| `Moo\\|ooo` |
|-------------|
| `ooo\\|ooo` |

<p>Q: Knock knock.
A: Who's there.
Q: Interrupting cow.
A: Interrupting —?</p>
<table><thead><tr><th><code>Moo\|ooo</code></th></tr></thead><tbody>
<tr><td><code>ooo\|ooo</code></td></tr>
</tbody></table>

Double escaping in image alt text

| ![Moo\\|Moo](image.png) |
|-------------|
| ![Moo\\\|Moo](image.png) |

<table><thead><tr><th><img src="image.png" alt="Moo|Moo" /></th></tr></thead><tbody>
<tr><td><img src="image.png" alt="Moo\|Moo" /></td></tr>
</tbody></table>

Double escaping in link title

| [Moo](https://example.org "Example\\|Link") |
|---------------------------------------------|
| [Moo](https://example.org "Example\\\|Link") |

<table><thead><tr><th><a href="https://example.org" title="Example|Link">Moo</a></th></tr></thead><tbody>
<tr><td><a href="https://example.org" title="Example\|Link">Moo</a></td></tr>
</tbody></table>

Blank list items and lists that don't start at one can interrupt tables.

moo | moo
----|----
moo | moo
*

<table><thead><tr><th>moo</th><th>moo</th></tr></thead><tbody>
<tr><td>moo</td><td>moo</td></tr>
</tbody></table>
<ul>
<li></li>
</ul>

moo | moo
----|----
moo | moo
2.

<table><thead><tr><th>moo</th><th>moo</th></tr></thead><tbody>
<tr><td>moo</td><td>moo</td></tr>
</tbody></table>
<ol start="2">
<li></li>
</ol>

pulldown-cmark guide