From d98cd4fdce027dd0ad00badb590dc0c63a0b89ce Mon Sep 17 00:00:00 2001 From: Peter Massey-Plantinga Date: Thu, 30 Apr 2020 11:05:19 -0400 Subject: [PATCH] (yaml) Add support for inline sequences and mappings (#2513) * Use containers to match inline sequences and mappings * Add string type for inside inline elements * Handle nested inline sequences and mappings * Disallow all braces brackets and commas from strings inside inline mappings or sequences * clean up implementation * feed the linter Co-authored-by: Josh Goebel --- src/languages/yaml.js | 197 ++++++++++++++++++----------- test/markup/yaml/inline.expect.txt | 3 + test/markup/yaml/inline.txt | 3 + 3 files changed, 126 insertions(+), 77 deletions(-) create mode 100644 test/markup/yaml/inline.expect.txt create mode 100644 test/markup/yaml/inline.txt diff --git a/src/languages/yaml.js b/src/languages/yaml.js index 4fc4ba0a28..02a1d9cee2 100644 --- a/src/languages/yaml.js +++ b/src/languages/yaml.js @@ -11,7 +11,7 @@ export default function(hljs) { var LITERALS = 'true false yes no null'; // YAML spec allows non-reserved URI characters in tags. - var URI_CHARACTERS = '[\\w#;/?:@&=+$,.~*\\\'()[\\]]+' + var URI_CHARACTERS = '[\\w#;/?:@&=+$,.~*\\\'()[\\]]+'; // Define keys as starting with a word character // ...containing word chars, spaces, colons, forward-slashes, hyphens and periods @@ -21,25 +21,25 @@ export default function(hljs) { className: 'attr', variants: [ { begin: '\\w[\\w :\\/.-]*:(?=[ \t]|$)' }, - { begin: '"\\w[\\w :\\/.-]*":(?=[ \t]|$)' }, //double quoted keys - { begin: '\'\\w[\\w :\\/.-]*\':(?=[ \t]|$)' } //single quoted keys + { begin: '"\\w[\\w :\\/.-]*":(?=[ \t]|$)' }, // double quoted keys + { begin: '\'\\w[\\w :\\/.-]*\':(?=[ \t]|$)' } // single quoted keys ] }; var TEMPLATE_VARIABLES = { className: 'template-variable', variants: [ - { begin: '\{\{', end: '\}\}' }, // jinja templates Ansible - { begin: '%\{', end: '\}' } // Ruby i18n + { begin: '{{', end: '}}' }, // jinja templates Ansible + { begin: '%{', end: '}' } // Ruby i18n ] }; var STRING = { className: 'string', relevance: 0, variants: [ - {begin: /'/, end: /'/}, - {begin: /"/, end: /"/}, - {begin: /\S+/} + { begin: /'/, end: /'/ }, + { begin: /"/, end: /"/ }, + { begin: /\S+/ } ], contains: [ hljs.BACKSLASH_ESCAPE, @@ -47,85 +47,128 @@ export default function(hljs) { ] }; + // Strings inside of value containers (objects) can't contain braces, + // brackets, or commas + var CONTAINER_STRING = hljs.inherit(STRING, { + variants: [ + { begin: /'/, end: /'/ }, + { begin: /"/, end: /"/ }, + { begin: /[^\s,{}[\]]+/ } + ] + }); + var DATE_RE = '[0-9]{4}(-[0-9][0-9]){0,2}'; var TIME_RE = '([Tt \\t][0-9][0-9]?(:[0-9][0-9]){2})?'; var FRACTION_RE = '(\\.[0-9]*)?'; var ZONE_RE = '([ \\t])*(Z|[-+][0-9][0-9]?(:[0-9][0-9])?)?'; var TIMESTAMP = { className: 'number', - begin: '\\b' + DATE_RE + TIME_RE + FRACTION_RE + ZONE_RE + '\\b', - } + begin: '\\b' + DATE_RE + TIME_RE + FRACTION_RE + ZONE_RE + '\\b' + }; + + var VALUE_CONTAINER = { + end: ',', + endsWithParent: true, + excludeEnd: true, + contains: [], + keywords: LITERALS, + relevance: 0 + }; + var OBJECT = { + begin: '{', + end: '}', + contains: [VALUE_CONTAINER], + illegal: '\\n', + relevance: 0 + }; + var ARRAY = { + begin: '\\[', + end: '\\]', + contains: [VALUE_CONTAINER], + illegal: '\\n', + relevance: 0 + }; + + var MODES = [ + KEY, + { + className: 'meta', + begin: '^---\s*$', + relevance: 10 + }, + { // multi line string + // Blocks start with a | or > followed by a newline + // + // Indentation of subsequent lines must be the same to + // be considered part of the block + className: 'string', + begin: '[\\|>]([0-9]?[+-])?[ ]*\\n( *)[\\S ]+\\n(\\2[\\S ]+\\n?)*' + }, + { // Ruby/Rails erb + begin: '<%[%=-]?', + end: '[%-]?%>', + subLanguage: 'ruby', + excludeBegin: true, + excludeEnd: true, + relevance: 0 + }, + { // named tags + className: 'type', + begin: '!\\w+!' + URI_CHARACTERS + }, + // https://yaml.org/spec/1.2/spec.html#id2784064 + { // verbatim tags + className: 'type', + begin: '!<' + URI_CHARACTERS + ">" + }, + { // primary tags + className: 'type', + begin: '!' + URI_CHARACTERS + }, + { // secondary tags + className: 'type', + begin: '!!' + URI_CHARACTERS + }, + { // fragment id &ref + className: 'meta', + begin: '&' + hljs.UNDERSCORE_IDENT_RE + '$' + }, + { // fragment reference *ref + className: 'meta', + begin: '\\*' + hljs.UNDERSCORE_IDENT_RE + '$' + }, + { // array listing + className: 'bullet', + // TODO: remove |$ hack when we have proper look-ahead support + begin: '\\-(?=[ ]|$)', + relevance: 0 + }, + hljs.HASH_COMMENT_MODE, + { + beginKeywords: LITERALS, + keywords: { literal: LITERALS } + }, + TIMESTAMP, + // numbers are any valid C-style number that + // sit isolated from other words + { + className: 'number', + begin: hljs.C_NUMBER_RE + '\\b' + }, + OBJECT, + ARRAY, + STRING + ]; + + var VALUE_MODES = [...MODES]; + VALUE_MODES.pop(); + VALUE_MODES.push(CONTAINER_STRING); + VALUE_CONTAINER.contains = VALUE_MODES; return { name: 'YAML', case_insensitive: true, aliases: ['yml', 'YAML'], - contains: [ - KEY, - { - className: 'meta', - begin: '^---\s*$', - relevance: 10 - }, - { // multi line string - // Blocks start with a | or > followed by a newline - // - // Indentation of subsequent lines must be the same to - // be considered part of the block - className: 'string', - begin: '[\\|>]([0-9]?[+-])?[ ]*\\n( *)[\\S ]+\\n(\\2[\\S ]+\\n?)*', - }, - { // Ruby/Rails erb - begin: '<%[%=-]?', end: '[%-]?%>', - subLanguage: 'ruby', - excludeBegin: true, - excludeEnd: true, - relevance: 0 - }, - { // named tags - className: 'type', - begin: '!\\w+!' + URI_CHARACTERS, - }, - // https://yaml.org/spec/1.2/spec.html#id2784064 - { // verbatim tags - className: 'type', - begin: '!<' + URI_CHARACTERS + ">", - }, - { // primary tags - className: 'type', - begin: '!' + URI_CHARACTERS, - }, - { // secondary tags - className: 'type', - begin: '!!' + URI_CHARACTERS, - }, - { // fragment id &ref - className: 'meta', - begin: '&' + hljs.UNDERSCORE_IDENT_RE + '$', - }, - { // fragment reference *ref - className: 'meta', - begin: '\\*' + hljs.UNDERSCORE_IDENT_RE + '$' - }, - { // array listing - className: 'bullet', - // TODO: remove |$ hack when we have proper look-ahead support - begin: '\\-(?=[ ]|$)', - relevance: 0 - }, - hljs.HASH_COMMENT_MODE, - { - beginKeywords: LITERALS, - keywords: {literal: LITERALS} - }, - TIMESTAMP, - // numbers are any valid C-style number that - // sit isolated from other words - { - className: 'number', - begin: hljs.C_NUMBER_RE + '\\b' - }, - STRING - ] + contains: MODES }; } diff --git a/test/markup/yaml/inline.expect.txt b/test/markup/yaml/inline.expect.txt new file mode 100644 index 0000000000..d12626f0ae --- /dev/null +++ b/test/markup/yaml/inline.expect.txt @@ -0,0 +1,3 @@ +foo: [bar, bar2, [1, 2], 3] +foo: {bar: [1, 2], baz: {inside: 3}} +foo: ba{}r,ba[]z diff --git a/test/markup/yaml/inline.txt b/test/markup/yaml/inline.txt new file mode 100644 index 0000000000..8dfa15b2ea --- /dev/null +++ b/test/markup/yaml/inline.txt @@ -0,0 +1,3 @@ +foo: [bar, bar2, [1, 2], 3] +foo: {bar: [1, 2], baz: {inside: 3}} +foo: ba{}r,ba[]z