Project

Language Inventory

TanStack Docs Language Inventory

Generated from local markdown and MDX files in:

  • Sibling repositories under ../*/docs
  • The sibling ../tanstack.com repository

The scan parses fenced code blocks statefully, so closing fences are not counted as plaintext blocks. Hidden directories and this highlight package are ignored.

Scanned files: 2940

Generated real-code fixtures:

shell
pnpm run extract:fixtures

The current fixture set stores up to twenty sampled code fences per normalized language target in test/generated/tanstack-doc-fixtures.json.

Raw Fence Tags

CountTag
8173ts
4078tsx
904js
877typescript
561bash
226plaintext
211json
136sh
107vue
96jsx
86svelte
51angular-ts
43text
38txt
33mermaid
33shell
28html
26js-vue
21yaml
15toml
14javascript
11css
11markdown
11md
11sql
9angular-html
8http
6env
4dockerfile
4nginx
4python
3diff
3xml
2apache
2dotenv
2ejs
1-->
1jsonc
1scheme

Normalized Support Targets

--> is malformed markdown and is normalized to plaintext.

TargetAliases
tsxtsx
tsts, typescript, angular-ts
jsxjsx
jsjs, javascript, js-vue
shellbash, sh, shell
jsonjson, jsonc
plaintextplaintext, text, txt, -->
vuevue
sveltesvelte
htmlhtml, angular-html, xml
markdownmd, markdown
yamlyaml
csscss
diffdiff
mermaidmermaid
tomltoml
sqlsql
httphttp
envenv, dotenv
dockerfiledockerfile
nginxnginx
pythonpython
apacheapache
ejsejs
schemescheme

Contract

The highlighter output should be one HTML tree with stable semantic classes. Theme switching should happen in CSS, not by duplicating HTML, embedding inline colors, or re-tokenizing for light and dark themes.