This pull request aims to fix the issue of excessive nesting brought up in the following comment: https://github.com/prettier/prettier/issues/737#issuecomment-392541493
This "fix" is very simple. It adds a new possible way to activate JSX mode for ternary operators. If the first non-conditional parent of the expression is a JSX expression container, then JSX mode gets activated.
This branch also contains a test case for said scenario.
- preserve trailing spaces in html, excluding html comment
- `doc-printer` is now unrelated to `options.parser`
- fix some cases that two trailing spaces are mis-considered as `break`
Fixes#4605
The AST here is actually two `html`s and it did follow the [CommonMark spec](https://spec.commonmark.org/0.28/#html-blocks), so I ended up with respecting the blank line between block `html`s.
* Add various tests for ts type annotated function properties.
* Improve format when passing ts arrow function with return type
* Add test for case from #2581.
* Add a single test suite for blank files in all parsers
The parsers that don’t yet support blank files have been filtered out of the list until they do.
* Don’t directly `require()` Prettier
* Update `require`
* Add missing Babylon parser
* Update snapshots
We landed a change that added a new `InterpreterDirective` AST node type for hashbangs, and no longer add it as a comment/leadingComment.
Ref: https://github.com/babel/babel/pull/7928
I mimicked what we do in `@babel/generator` here, since I found it better than trying to add comments to the ast in `parser-include-shebang.js`). Definitely open to a better/cleaner option though!
Note: I'll follow this up with enabling tests for https://github.com/prettier/prettier/pull/4543 and https://github.com/prettier/prettier/pull/4540 once they land too.
* Support new node types for optionals
* Update babylon
* Fix decorator tests
* fix manual errors
* Update flow
* pretty print
* enable option for flow to support ??
* fix AST_COMPARE=1
* fix lint and explicitly test 1_2_3 syntax for flow
* Don't default parser to babylon
* Different default values according to version
* Fix error logging
* Add tests specifically for parser inference
* Add another test case
* Add API test
* Break link definitions onto multiple lines when needed
* Simplify the conditional for breaking the title
* Don’t print the title if it’s blank
* Revert "Don’t print the title if it’s blank"
This reverts commit 2a8d0dd9995a2e5ab8e6d888c07b88dda058637f.
* Add another test
* Fix title printing
* Second time’s the charm
* Third time’s the charm?
* Fix snapshot
* Prettify
* Don’t break unless `proseWrap` is `always`
* Test `proseWrap: never` on link references
* Added support for the graph ql comment tag
As described in the issue https://github.com/prettier/prettier/issues/4360
Prettier still adds a space inbetween the comment tag and the template literal.
* Added comment explaining the random spaces around GraphQL
* Fixed failing AST_COMPARE tests
* Add tests showing existing behavior of various TS casts.
* Improve formatting of TS casts with generics and unions.
By adding a soft break between the cast and the expression. Also
includes wrapping parentheses for clarity as suggested in #4171.
Avoids changing behavior at all though if casting an array or object
literal because those already have good behavior where the array or
object literal breaks before the cast does so including them would just
result in a pointless extra layer of parentheses that would add no
clarity.
* Update tests in prepation for updating long cast format.
* Update so nested breaks inside type cast are indented.
* Add tests where expression being cast is too long itself.
* Update cast formatting to only sometimes break after cast.
Specifically it will break after the cast if that makes the cast itself
fit on a single line. If the cast itself won't fit on a single line then
the expression being cast will be placed directly after the `>` at the
end of the cast.
* Add heuristic to format functional composition more nicely
* Regex -> array
* Use object instead of array for faster lookup
* Remove empty lines
* Add redux connect
Also add comments for libraries
* handle cusor position relative to comments correctly
* use more robust method of cursor offset tracking
* add support for cursorOffset with rangeStart/End
* add tests for cursorOffset
* use old JS syntax for node4
* add another cursor offset test case
* remove trailing whitespace in cursor tests
* fix null findNodeAtOffset handling
* remove dead code
* add more test cases
* use indexOf for node4
* revert comments.js
* add more comment tests
* change tests infra
* Fix AST_COMPARE
* move CURSOR to top level of file
* mutate result
* use es6 property shorthand
* simplify diff offset calculation
* put remaining test cases in individual files
* Fix: Always preserve frontmatter, remove dependency
* Move correct test to correct filename
* Add test for malformed frontmatter, finetune regex
* Use other malformed test case
* Update yarn.lock
* fix(glimmer): preserve leading/trailing spaces in text nodes underneath attribute nodes when next to mustache statements.
* fix(glimmer): format printer-glimmer.js
* fix(glimmer): lint fix
* fix(glimmer): removed decomposition assignments to be compatible with Node 4
* fix(glimmer): Array.includes() -> indexOf() for Node 4
* feat(typescript): support generic JSX element
* fix(playground): `buffer is not defined` in `--parser typescript`
* chore(deps): update typescript-eslint-parser
* d3 counts as a constructor
* adds test
* Tweak wording
* docs: fix linting
* updates test
* no $ special case
* renames function
* shorter function name
* removes trailing white space
* feat(markdown): support hasPragma/insertPragma
* test: add require-pragma tests
* feat: support pragma with front matter
* feat: support pragma in multiline comment
* fix: don't lowercase custom selectors variable name #4254
* proper fix for camelCased custom selector
* Move customs-selectors case snapshots to correct place
* Allow new interface style for GraphQL.
This is a breaking change since it will upgrade the old style to the new one.
Closes#3600.
* Allow old interface style.
When in present of mixed style, it updates to the new one.
* feat(typescript): support for conditonal types
* refactor(js): reuse conditional expression logic
* chore(typescript): update snapshot for typescript conditional type test suite
* chore(js): make code support Node 4
* chore(js): rename utility functions
* chore(js): add comments for formatTernaryOperator
* fix(ts): support infer keyword
* chore(js): new line
* chore(js): improve readablity a little bit
* Upgrade flow to 0.64
This also enables spread operator on jsx on the flow parser:
```js
<div>{...list}</div>
```
I had no idea it was even a thing.
I had to comment out a flow test that no longer parses, but it's no longer in the flow repo and the comment said that it shouldn't be working anyway so I don't think it's going to be a problem.
* Fix build script
* Treat define calls like require
See issue #3829. This commit keeps define calls as a unit when possible,
to prevent an extra indent in the body of an AMD-style define.
Rather than adding the "define" check in the same case as the existing
"require" check, I added a separate condition to explicitly call out the
fact that this checks for both AMD and CommonJS modules.
`yarn test -u` yields no changes, and `yarn lint` passes successfully.
* Added tests for define and require amd formatting
This tests the solution for issue #3829.
* Add babylon interpreter to require-amd test
* Merge require and define checks
* test: add test cases
* test: add test cases
* fix(javascript): ternary with `--use-tabs`
* chore: add istanbul ignore comment
* fix: remove unnecessary condition
* docs(commands): update `align`
* fix: do not transform the middle part
* refactor: markAsRoot
* fix: print tabs in sub-ternaries
* docs(commands): update
* test: add a cool test case
* test: add test cases
* test: add test cases
* fix(javascript): indentation for sub-ternaries
* test: update test cases
* fix: no extra tab for `tabWidth: 4`
* Added prettier layout for class extends and implements
* Added prettier layout for class extends and implements, fix with keeping it in online if it fits
* Added prettier layout for class extends and implements, fixed tests as a side change
* Extends implements comments in same line fix
* Extends implements comments in same line without ugly workaround
* added empty line to test typescript_class/extends_implements.ts
* If only one extends keep old behaviour
* Removed console logs
* Update GraphQL parser
Fixes#3601
This adds support for
- description as strings
- extending all the possible types
- block strings
- allow removing {} if there's no implementation for all graphql types
This is a breaking change but shouldn't be a big issue.
- Empty types are no longer allowed (there's an option to enable it but it hasn't been released yet). The fix is to remove `{}`
Something that hasn't been changed because not released:
- Doesn't support the new `&` separator for implementing multiple interfaces
A bug has been fixed:
- Now properly prints @directives for unions.
* Properly handle triple quotes
* feat(glimmer): initial HTML/Handlebars/Glimmer support
* feat(glimmer): support un-escaped mustache expressions
* fix(glimmer): add 'as |foo|' syntax
* fix(glimmer): support {{foo bar}} syntax
* feat(glimmer): implement all AST types
* feat(glimmer): implement if/else if/else
* test(glimmer): ignore leading/trailing whitespace from AST_COMPARE
* fix(glimmer): do not use trailing slash for void self-closing elements
* chore(build): disable html tests until we can sort out the glimmer dependency
* feat(html): complete rebase on parse5 addition
* Prettier Support For Glimmer VM/Handlebars
Polished Support for Sub Expressions
Added test cases for glimmer primitives
Added support for concat statements
Attempted to make element nodes work
Attempted block statements
Element Nodes are OK
Added support for block elements that are not else-if related
Added support for Else/If
Cleaning up
Rebase
Switch node 4 syntax
Update build
* Removed dead code/partials
* Added new lines to end of test files.
* Rebase after plugin change
* Added ignores to code coverage
There's a lot of demand for vue sfc (#2097). This introduces partial support for them: all the html is printed as is, except for the script and style tags which are printed using prettier. I believe that this should cover a lot of the use cases while being simple to support and if we want we can extend to more in the future.
I copy pasted the html parser used by vue (it's just a single 400 lines file) so that we don't run the chancesof conflicts. I'm also very conservative: I only print the style and script at the top level and for the lang attributes we support.
I expect this to be landable as is and provide value, review welcome :)
* Add an option to utils.printString to use the passed string as-is
* Conform the quotes in Markdown link titles to the singleQuote option
* Don’t use destructuring
* Add some tests
* Format util.js
* Use custom string-printing logic
* Add some more tests
* Fix tests
* Address review comments
* Revert "Format util.js"
This reverts commit d2533ef7fb9231e2f733ece7decac9b923cf6941.
* Revert "Don’t use destructuring"
This reverts commit a5d3b0ac76a076e9a89cffc8cec396cabeacd1ee.
* Revert "Add an option to utils.printString to use the passed string as-is"
This reverts commit ef8e8322609ca3c2808885f3fadcd03f0790078c.
* Even more tests
Based on https://github.com/prettier/prettier/pull/2676
* Thread `path` and `options` through helpers so we don't need to add `needsParens` onto the AST node anymore (mutation)
* Pull test call detection logic out into helper method so it can be re-used for arrow function parens
* Add arrow function parens option implementation (avoid/always)
* Don't break arrow function parens around (done) in test call
* Do not prepend / append with a semicolon the only JSX element in a program
Fixes#3196
* Limit single JSX element without semicolon to Markdown only
* Fix tests
* Don't lowercase element names and attribute names in selectors
https://www.w3.org/TR/css3-selectors/#casesens
> All Selectors syntax is case-insensitive within the ASCII range (i.e.
> [a-z] and [A-Z] are equivalent), except for parts that are not under the
> control of Selectors. The case sensitivity of document language element
> names, attribute names, and attribute values in selectors depends on the
> document language. For example, in HTML, element names are
> case-insensitive, but in XML, they are case-sensitive.
Fixes#3304.
* Fix number normalization in selector interpolation
* Adding --stdin-filepath option in order to enable filetype inference from
stdin
* Fixing typo on the documentation
* Fixing typos - fixing some typos
* Restoring static lib files back since they are artifacts
* fix(jsx): Break opening elements with single text attribute if there's a comment
* Single text attribute with a comment on the JSXElement name
* Consolidate rules of comments in JSX
This was only accounting for a single element in the first group. Now it handles arbitrary elements. Only the last element being capitalized matters.
This fixes one issue in #3107
* Split source elements relative to their language.
Colliding node types which are not source elements in every language.
Example ObjectExpression in JSON / JS
* Node4's Array.prototype.includes
* Ensure there is no confusion
ObjectExpression vs BlockStatement & LabeledStatement
* fixing issue #2984
* removed typescript from test #2984
* prettier
* removed obsolete test copied from another
* added verification parsers
* added new line
* Account for empty lines in member chain
* Add tests
* Account for parens
* Improve tests
* Add some comments
* Remove an outdated comment
* Fix lint errors
* Refactor a line of code
* Add one more comment for clarification
* refactor: extract `splitText`
* fix: respcet CJK width in table
* test: add failing test
* fix: support emoji
* test: add failing test
* feat: support CJK character
* feat: enable adding whitespace between non-CJK/CJK-character by default
* fix: do not print softline at node that is sensitive to its identifier
* fix: treat full-width whitespace as CJK punctuation
* disallow leading/trailing full-width whitespace
* feat: remove `--split-cjk-text` option and enable it by default
* refactor: simplify regex and remove unnecessary `g` flag
* feat(markdown): inital implementation
* feat(markdown): support strong
* fix: add missing default value
* feat(markdown): support inlineCode
* feat: support delete
* feat: support link
* feat: support image
* feat: support blockquote
* feat: support heading
* feat: support code
* feat: support yaml
* feat: support html
* feat: support list
* feat: support thematicBreak
* feat: support table
* feat: support linkReference
* feat: support imageReference
* feat: support definition
* feat: support footnote
* feat: support footnoteReference
* feat: support footnoteDefinition
* test(cli): update snapshots
* refactor: extract SINGLE_LINE_NODE_TYPES
* refactor: printChildren
* fix: correct newlines
* test: add trailing newline
* fix: blockquote formatting
* fix: node types
* fix: break line correctly
* fix: remove unnecessary properties to make AST_COMPARE happy
* fix: escape `|` in tableCell content
* fix: unexpected line break
* fix: ast difference from loose list
* fix: html break lines
* refactor: fix linting
* fix: normalize ast
* fix: escape specific chars
* test: add more tests
* fix: build markdown parser
* chore: remove unnecessary *.log
* fix: escape html entity
* feat: support prettier-ignore
* fix: line break for non-loose listItem
* feat: support formatting `code` based on `lang`
* fix: add `jsx` and `tsx`
* fix: use multiparser
* refactor: fix linting
* test: update test case 😉
* feat: switch to `_` style emphasis
* fix: sequence list should use different prefix
* test: add tests
* fix: do not print additional new line after `prettier-ignore`
* fix(list): enforce `1.` to avoid unnecessary git diff
* feat: enable `commonmark` option
* feat: support `break`
* fix: escape backslash
* refactor: escape html entity using backslash
* fix: respect autolink-style link
* feat: support md`...` and markdown`...`
* docs: replace ands with commas
* fix: respect indented code block
* fix: respect html entity
* docs: add docs for modified MDAST
* fix: inlineCode is breakline-able
* feat: support backtick in inlineCode
* feat: support a-lot-of-backtick in fenced code block
* feat: use `~~~`-style code block in js template
* fix: respect escaped chars
* fix: use `*`-style emphasis for invalid `_`-style output
* test: add test cases
* fix: use `- - -`-style thematicBreak to avoid conflict with yaml
* fix: remain the same content for linkReference identifier
* fix: `inlineCode` gap can be a line break
* fix: `html` should not print trailing spaces if it's in root
* refactor: fix typo
* fix: wrap `definition`'s url if there's whitespace
* fix: remove unnecessary whitespace at the end of paragraph
* fix: fix: remove unnecessary whitespace at the start of paragraph
* fix: setence children length is possible 0
* fix: support continuous ordered list
* fix: do not print addtional hardline after loose list
* fix: use double-backtick style for single-backtick value in inlineCode
* fix: support nested emphasis
* fix: support space-url in link/image
* fix: escape `)` in link/image url
* fix: support single-quote in link/image/definition title
* fix: respect alt in image/imageReference
* fix: use `*`-style thematicBreak in list
* fix: loose/tight list linebreaks
* fix: print third linebreak before indented code block with a tight list in the previous
* test: move bug cases
* fix: remove unnecessary linebreaks
* refactor: fix typo
* Fix TypeScript comment snapshot
* Add failing test for TypeScript method comments
* Remove trailing space in comment
* Add snapshot for comment with many lines
* Add TODO marking potential location for comment re-indentation
* Update snapshot
* Indent comments based on the parent node
* Only adjust indentation if the comment looks like a JSDoc comment
* Only strip leading spaces when adjusting comment indentation
* Update snapshot
* Add test to show incorrect indentation based on parent node
* Update snapshot
* Update snapshot
* Update comment indentation to not rely on the parent node
* Extract code for getting the indent size
* Add test where the parent indentation changes as well
* Fix lint warning regarding lexical declarations within case block
* Update lockfile
* Update snapshot
* Add test case for block comment inside of a method
* Replace spread operator with `concat`
* Use `align` instead of `addAlignmentToDoc`
* Add test for mismatched indentation within JSDoc comment
* Always add a leading space to subsequent lines in a JSDoc comment
* Refactor JSDoc comment printing into a separate function
* Remove unneeded `align`
* Replace `forEach` with `map` for better style
* Only `trimLeft` when the end comment token is on the last line of the comment
* Remove unneeded `docBuilders`
* Add tests for JSDoc comments in JSX
* Update snapshots