steid

feat: the languages this repository actually uses

syntect's default set is Sublime's, and it has no TOML and no Dockerfile — two
file types sitting in this repository's root, rendering as flat text on the one
page whose whole job is to make code legible. TypeScript was missing too.

`two_face::syntax::extra_newlines()` is that set plus the grammars `bat` curates,
so it is a drop-in for `load_defaults_newlines` and the change is one line in the
`OnceLock`. The `hl-` class output and the CSS mapping are untouched: the extra
grammars emit the same TextMate scopes the mapping was already written against,
which is the payoff for having mapped scopes rather than a theme.

`default-features = false` is load-bearing rather than tidy — two-face defaults to
`syntect-onig`, and `syntect-fancy` keeps it on the pure-Rust engine so no C
dependency reaches the Debian release build. The cost is 0.59 MiB of embedded
grammar dump: 13.65 MB to 14.27 MB.

A test now asserts the sixteen languages that highlight rather than describing
them, so a future syntax-set swap cannot quietly reopen the hole. `.jsx` is the
one extension still plain, and the handover says why.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01JnmRPpETeB9y6RSowpigye

JamesPatrickGill authored 15 hours agoparent30018a8Browse files8331d0cc1f2dcc2daf2f7d69345e50c730285035

4 files changed+76 −24

Cargo.lock+12 −0View file

		@@ -2015,6 +2015,7 @@ dependencies = [
2015	2015	"tokio",
2016	2016	"tokio-util",
2017	2017	"topcoat",
	2018	+ "two-face",
2018	2019	"uuid",
2019	2020	]
2020	2021
		@@ -2764,6 +2765,17 @@ dependencies = [
2764	2765	"thiserror",
2765	2766	]
2766	2767
	2768	+[[package]]
	2769	+name = "two-face"
	2770	+version = "0.5.2+bat-0.26.1"
	2771	+source = "registry+https://github.com/rust-lang/crates.io-index"
	2772	+checksum = "915be7adc2ff6f4338acbf71f3eda0a146f46003b992a1669c674308610efdac"
	2773	+dependencies = [
	2774	+ "serde",
	2775	+ "serde_derive",
	2776	+ "syntect",
	2777	+]
	2778	+
2767	2779	[[package]]
2768	2780	name = "typenum"
2769	2781	version = "1.20.1"

Cargo.toml+1 −0View file

		@@ -21,6 +21,7 @@ syntect = { version = "5.3.0", default-features = false, features = ["default-fa
21	21	tokio = { version = "1.53.1", features = ["rt-multi-thread", "macros", "process", "fs", "io-util", "time"] }
22	22	tokio-util = { version = "0.7", features = ["io"] }
23	23	topcoat = { version = "0.5.0", features = ["icon-iconify", "tailwind", "ui"] }
	24	+two-face = { version = "0.5.2", default-features = false, features = ["syntect-fancy"] }
24	25	uuid = { version = "1.24.0", features = ["v4"] }
25	26
26	27	[build-dependencies]

plans/handover-highlight.md+30 −10View file

		@@ -20,6 +20,10 @@ highlighting. Nothing else changed: no new route, no new `GitQuery` method, **no
20	20	`hl-` classes at the bottom of the file.
21	21	- `syntect 5.3` with `default-features = false, features = ["default-fancy"]`: the pure
22	22	Rust regex engine, so the Debian release build gains no C dependency.
	23	+- `two-face 0.5` with `default-features = false, features = ["syntect-fancy"]`: the
	24	+ grammars syntect's defaults lack. Its default feature is `syntect-onig`, so
	25	+ `default-features = false` is not tidiness — it is what keeps `onig`, and therefore a
	26	+ C dependency, out of the release build.
23	27
24	28	## progress.md material — decisions worth not rediscovering
25	29
		@@ -62,17 +66,15 @@ highlighting. Nothing else changed: no new route, no new `GitQuery` method, **no
62	66
63	67	## current.md material — holes opened and shortcuts taken
64	68
65		−- The default syntax set is missing four languages this project itself uses. Checked
66		− by test (`which_common_languages_the_default_set_knows`, prints its findings):
	69	+- `.jsx` is the one extension still unhighlighted. `two_face` closed the gap that
	70	+ mattered — TOML, Dockerfile and TypeScript all highlight now, and the first two are in
	71	+ this repository — but its JavaScript grammar claims only `js` and `htc`, while React
	72	+ is covered on the TypeScript side by `TypeScriptReact` for `tsx`. So a `.jsx` file
	73	+ renders plain. Nothing here uses one; it is recorded rather than fixed, and the fix if
	74	+ it is ever wanted is to alias `jsx` onto the JavaScript syntax in `syntax_for`, not
	75	+ another crate. `the_languages_this_repository_uses_are_all_highlighted` asserts the
	76	+ sixteen that do work, so a future syntax-set swap cannot quietly reopen the hole.
67	77
68		− \| Highlighted \| Plain \|
69		− \|---\|---\|
70		− \| Rust, shell, Markdown, Python, JavaScript, Go, YAML, JSON, CSS, SQL, Makefile, C \| TOML, Dockerfile, TypeScript, JSX/TSX \|
71		−
72		− `Cargo.toml` and `Dockerfile` render unhighlighted in this repository today — the
73		− `highlight-toml-dark.png` screenshot is exactly that. `two-face` bundles the missing
74		− grammars and is the fix, deliberately not added here: it is another dependency and
75		− a decision for the orchestrator. As briefed, it was not attempted.
76	78	- Highlighting costs ~87 ms per 1,255-line file in a release build — measured on
77	79	`browse.rs` with `cargo test --release`, and it is paid on every view because nothing
78	80	caches. In the `topcoat dev` (debug) build the same page takes ~0.7 s to serve versus
		@@ -91,6 +93,24 @@ highlighting. Nothing else changed: no new route, no new `GitQuery` method, **no
91	93	- The blob view's own header has no "this is Rust" indicator. The language is
92	94	detected and then only ever visible as colour.
93	95
	96	+## The `two-face` decision, and what it cost
	97	+
	98	+Taken by the user after the first pass reported the gap.
	99	+
	100	+- The syntax set is now `two_face::syntax::extra_newlines()`, which is syntect's
	101	+ defaults plus the grammars `bat` curates. It is a drop-in for
	102	+ `SyntaxSet::load_defaults_newlines` — one line in the `OnceLock`. Nothing else moved:
	103	+ the `hl-` class output and every rule in `styles.css` are unchanged, because the extra
	104	+ grammars emit the same TextMate scopes the mapping was already written against.
	105	+- The release binary grew 0.59 MiB, from 13.65 MB to 14.27 MB (`cargo build
	106	+ --release`, measured on this branch with and without the dependency). That is the
	107	+ embedded grammar dump; it is data, not code. Worth knowing because the artifact is
	108	+ downloaded by `install.sh` over whatever connection an operator has, and because
	109	+ 5b's release notes quote a size.
	110	+- No new C dependency. `syntect-fancy` resolves to `syntect/regex-fancy`, the same
	111	+ pure-Rust engine already in use, so the musl/glibc reasoning recorded in `current.md`
	112	+ is untouched.
	113	+
94	114	## ui.md material
95	115
96	116	- The blob's code cell is the only place colour carries information rather than

src/infrastructure/highlight.rs+33 −14View file

		@@ -104,14 +104,21 @@ pub fn source_lines(file_name: &str, text: &str) -> Source {
104	104	Source { lines, too_large }
105	105	}
106	106
107		−/// The default syntax set, loaded once.
	107	+/// The syntax set, loaded once.
	108	+///
	109	+/// `two_face` rather than `SyntaxSet::load_defaults_newlines`: syntect's own defaults
	110	+/// are Sublime's, which have no TOML, no Dockerfile and no TypeScript — the first two
	111	+/// of which are in this repository. `two_face` is the same set plus the grammars `bat`
	112	+/// curates, so the gap closes without teaching this module anything about individual
	113	+/// languages. Its `syntect-fancy` feature pins it to the same pure-Rust regex engine,
	114	+/// so no `onig` — and therefore no C — comes in behind it.
108	115	///
109	116	/// Loading is a few megabytes of deserialization, so it happens on the first blob
110	117	/// viewed and never again. `_newlines` is the variant whose rules expect a trailing
111	118	/// newline on each line, which is what [`LinesWithEndings`] hands it.
112	119	fn syntaxes() -> &'static SyntaxSet {
113	120	static SYNTAXES: OnceLock<SyntaxSet> = OnceLock::new();
114		− SYNTAXES.get_or_init(SyntaxSet::load_defaults_newlines)
	121	+ SYNTAXES.get_or_init(two_face::syntax::extra_newlines)
115	122	}
116	123
117	124	/// The language, from the file name and then the first line.
		@@ -301,30 +308,42 @@ mod tests {
301	308	assert_eq!(source.lines, vec![SourceLine::Plain("hello".to_owned())]);
302	309	}
303	310
304		− /// The default syntax set is Sublime's, not a forge's, so the languages it lacks
305		− /// are worth knowing rather than discovering on a page.
	311	+ /// The gap `two_face` was added to close, asserted rather than described: TOML,
	312	+ /// Dockerfile and TypeScript rendered plain until it landed, and the first two are
	313	+ /// in this repository. `.jsx` is still absent — see the handover.
306	314	#[test]
307		− fn which_common_languages_the_default_set_knows() {
	315	+ fn the_languages_this_repository_uses_are_all_highlighted() {
308	316	for name in [
309	317	"main.rs",
	318	+ "install.sh",
310	319	"Cargo.toml",
311	320	"Dockerfile",
312		− "install.sh",
	321	+ "app.ts",
	322	+ "App.tsx",
313	323	"README.md",
	324	+ "styles.css",
	325	+ "config.yaml",
	326	+ "data.json",
	327	+ "Makefile",
314	328	"app.py",
315	329	"index.js",
316		− "app.ts",
317	330	"main.go",
318		− "config.yaml",
319		− "data.json",
320		− "styles.css",
321	331	"query.sql",
322		− "Makefile",
323	332	"main.c",
324		− "App.jsx",
325	333	] {
326		− let known = syntax_for(syntaxes(), name, "").is_some();
327		− println!("{name}: {}", if known { "highlighted" } else { "PLAIN" });
	334	+ assert!(
	335	+ syntax_for(syntaxes(), name, "").is_some(),
	336	+ "{name} has no syntax"
	337	+ );
328	338	}
329	339	}
	340	+
	341	+ #[test]
	342	+ fn toml_is_highlighted_and_not_merely_recognised() {
	343	+ let source = source_lines("Cargo.toml", "[package]\nname = \"steid\"\n");
	344	+
	345	+ assert!(matches!(source.lines[1], SourceLine::Classed(_)));
	346	+ let html = classes(&source.lines[1]);
	347	+ assert!(html.contains("hl-string"), "{html}");
	348	+ }
330	349	}