|
| 1 | +# 12. Output Markdown Format |
| 2 | + |
| 3 | +Date: 2021.09.09 |
| 4 | + |
| 5 | +## Status |
| 6 | + |
| 7 | +proposed |
| 8 | + |
| 9 | +## Context |
| 10 | + |
| 11 | +Stencil has an output target that is capable of generating markdown documentation for the components that it compiles. |
| 12 | +Markdown does not have a unifying standard, which has led to multiple different implementations of how the markup is |
| 13 | +parsed. This has resulted in various markdown renderers whose implementations yield different renderings. Depending on |
| 14 | +where an individual views a markdown file, the results may look incorrect. |
| 15 | + |
| 16 | +There was a [recent pull request](https://github.com/ionic-team/stencil/pull/3042/) that the team ultimately decided to |
| 17 | +close for not following CommonMark, which is the proposed/intended solution. Stencil should formally decide on a |
| 18 | +proposed standard and use that for guidance as to how markdown should be formatted. |
| 19 | + |
| 20 | +## Options |
| 21 | + |
| 22 | +1. [Describe the many options, their pros/cons, side effects, and any perspective a given team member has] |
| 23 | +2. Adopt [CommonMark](https://commonmark.org/) |
| 24 | + 1. Pros: |
| 25 | + 1. It has industry support from the likes of GitHub, GitLab, Qt, and more |
| 26 | + 2. It seems to be the leading effort to create an unambiguous specification for Markdown |
| 27 | + 3. Using a well adopted standard may lead to less chance of Markdown renderings not working as intended |
| 28 | + 2. Cons |
| 29 | + 1. Not everyone may be familiar with CommonMark, leading to confusion when reviewing pull requests |
| 30 | + 2. CommonMark has yet to produce a v1.0 specification |
| 31 | + 3. CommonMark does not define a syntax for tables, requiring they be written as |
| 32 | + [HTML](https://spec.commonmark.org/0.30/#html-blocks) |
| 33 | +3. Adopt [GitHub Flavored Markdown (GFM)](https://github.github.com/gfm/) |
| 34 | + 1. Pros: |
| 35 | + 1. Individuals may be more familiar with GFM, given the popularity of the platform |
| 36 | + 2. Using a well adopted standard may lead to less chance of Markdown renderings not working as intended |
| 37 | + 2. Cons: |
| 38 | + 1. GFM is a superset of CommonMark, which may lead to subtle incompatabilities in other renderers that do not |
| 39 | + support GFM |
| 40 | + |
| 41 | +## Decision |
| 42 | + |
| 43 | +The team shall use CommonMark to guide all Markdown formatting decisions for the documentation that is generated by the |
| 44 | +`docs-readme` output target |
| 45 | + |
| 46 | +## Consequences |
| 47 | + |
| 48 | +- There may be changes in the CommonMark specification, which the team will need to adapt to |
| 49 | + - Said changes will likely come as _other_ entities also adopt the changes, e.g. GitHub |
| 50 | +- Meaningful changes to certain audiences in the Stencil community (e.g. users of a specific IDE) may be turned away |
| 51 | + for not following the spec |
| 52 | +- The team should look into tooling around style guides for CommonMark and how they may be integrated into existing |
| 53 | + tests |
| 54 | +- There may be places where the product does not adhere to CommonMark. Cases of these divergences must be documented |
| 55 | + and rectified in the output target |
| 56 | + |
| 57 | +## Links |
| 58 | + |
| 59 | +- [Closed Stencil PR - feat(markdown): improved export to markdown #3042](https://github.com/ionic-team/stencil/pull/3042/) |
| 60 | +- [CommonMark Homepage](https://commonmark.org/) |
| 61 | +- [GitHub Flavored Markdown (GFM)](https://github.github.com/gfm/) |
0 commit comments