Compare

+# My git viewer

+

+### Why I did this

+

+So, the premise behind this project was a few different things:

+

+- I wanted a git viewer stack that, while modular, remained light.

+- I liked the idea of having reusable "templates" rather than generating the HTML dynamically in the code. This is much closer to how things like Hugo operate.

+- It needs to be easy to extend (modularity!). A core is responsible for loading a git repo's context and other request state, but all "real" functionality is implemented via modules. I could add/remove things without needing to worry (too much...) about breaking unrelated things.

+- I wanted something that was modern, but didn't require javascript to operate. All features on this site work without javascript, though your experience is improved with it.

+- I wanted to have a way to view private git repositories using systems I already have.

+ - I don't want to/have the capability to SSH clone my repos to show something off.

+ - Could I have done this in cgit? Perhaps, but it wasn't built for it.

+- And, probably most visibly, this was a trial in "vibe coding". While the idea around the design of this software was mine, I wrote very little of the code. Especially the UI, and this is for the better. I am not a graphics designer and whatever slop the LLM pumps out (with my input here and there) will be leagues better than whatever I do; for better or worse.

+

+The name "gsg" doesn't mean much (I don't name things). It're pretty much "git static generator". It was the placeholder when I was thinking of a system that was "kind of a combination of cgit's lightness and hugo's composability and templating".

+

+### The design

+

+First and foremost, the AI writeup is in ./llm_readme.md. Read that if you want all of the verbose details.

+

+Also, as alluded to above, no javascript is strictly required. The few places it is used it is just used to make the experience nicer, and a fallback always exists. For example, the JSON/YAML codepath has a JS module that allows you to search for values/keys, perform folding, etc. This is a core design tenant: JS always hydrates what exists already.

+

+(Not) Broadly speaking, as stated above, this project is split up into:

+

+- The core

+ - gitx - the module that actually interfaces with git. This transforms a context into a re-usable struct (should I ever want to use a real implementation)

+ - access - Consolidated checks for whether someone can see/not see a repo

+ - session - Owns all privileged access logic. It generates/verifies tokens that are used across the site. Everything is stateless once authenticated.

+ - authn - Route owner that defines login/logout flows. Uses session internally

+ - httpx - Shared helpers used when generating responses/evaling a req

+ - notices - Logic for me to post notices to the site or repos

+ - share - Logic for authenticated users to mint links to private repos they are allowed to. Uses same auth logic as users under the hood.

+ - api - Exposes repo details as a JSON endpoint, will be used by my main site later on

+ - core/site - All the main things needed to make this tick. Build req context, define routes and where they live, define module registration structs.

+- The modules

+ - blob - Responsible for returning files (raw or templated)

+ - commit - Render a commit

+ - compare - Render a diff between two commits/refs

+ - graph - Render commit log graph

+ - image - Image preview, defines compatible file types and max size

+ - markdown - Render markdown

+ - readme - Inherits markdown to render a readme (what you see here!)

+ - notices - Handles cache invalidation for repos that need one rendered

+ - log - View the git log

+ - repos - Aggregates the repos you may view based on session

+ - structured - Whether a file is a YAML or JSON. If so, it prepares data for JS in the frontend to perform folding and suchA

+ - tree - the directory-structure viewer

+

+As time goes on, I am sure I will find more things to add to this. This is really meant to be a project for me, not necessarily building "the next git viewer for everyone".

+

+I figured I would upload this here in case my tokens are useful for others. Enjoy!

+go 1.23.0

+ golang.org/x/sync v0.14.0

+golang.org/x/sync v0.14.0 h1:woo0S4Yywslg6hp4eUFjTVOyKt0RookbpAHG4c1HmhQ=

+golang.org/x/sync v0.14.0/go.mod h1:1dzgHSNfp02xaA81J2MS99Qcpr2w7fw1gpm99rleRqA=

+// cache-max-age = 2m

+

+ // CacheMaxAge is gsg.cache-max-age: how long an anonymous (no session,

+ // no share grant) response for this repo may be reused -- both by the

+ // requesting browser (Cache-Control: public, max-age=...) and by gsg's

+ // own in-memory response cache (Server.respCache) -- before either has

+ // to ask again. Falls back to defaultAnonCacheMaxAge when unset.

+ CacheMaxAge time.Duration

+ CacheMaxAge: parseCacheMaxAge(r.Config("gsg.cache-max-age")),

+// defaultAnonCacheMaxAge is the ceiling/default applied when a repo doesn't

+// set gsg.cache-max-age.

+const defaultAnonCacheMaxAge = 2 * time.Minute

+

+func parseCacheMaxAge(v string) time.Duration {

+ if v == "" {

+ return defaultAnonCacheMaxAge

+ }

+ d, err := time.ParseDuration(v)

+ if err != nil || d <= 0 {

+ return defaultAnonCacheMaxAge

+ }

+ return d

+}

+

+ // StaticTag is Server.StaticTag, threaded through so the `static`

+ // template func can build cache-busting asset URLs without a package

+ // import cycle back to Server.

+ StaticTag string

+

+// cacheMaxAge is how long an anonymous response for this repo may be

+// reused before revalidating. Server.buildResponseHead (the browser-facing

+// Cache-Control header) and Server.fillResponse (the server-side response

+// cache's own TTL) both read this, so the two never disagree about

+// staleness.

+func (c *Ctx) cacheMaxAge() time.Duration {

+ if c.Info.CacheMaxAge > 0 {

+ return c.Info.CacheMaxAge

+ }

+ return defaultAnonCacheMaxAge

+}

+

+ // static builds a cache-busting asset URL, e.g. "css/base.css" ->

+ // "/static/css/base.css?v=<tag>". The `v` param is what lets

+ // main.go's static handler hand out a year-long immutable

+ // Cache-Control instead of a short one -- a URL without it is

+ // still served, just not frozen for a year.

+ "static": func(p string) string {

+ return "/static/" + p + "?v=" + c.StaticTag

+ },

+ return s[:n] + "…"

+ // CacheableAnon opts a route into Server's in-memory response cache for

+ // anonymous, shareless requests (see Server.cacheEligible) -- a full

+ // rendered response is stored and replayed to the next matching

+ // request within Ctx.cacheMaxAge, without touching git at all. Reserved

+ // for the small-HTML page routes most likely to be a spike's hot page;

+ // left off routes whose bodies can be large or unboundedly numerous

+ // (blob, raw, compare -- see routes.go) and the JSON api-* endpoints,

+ // to keep this pass's scope and memory footprint bounded: respCache is

+ // an LRU bounded by entry count, not bytes.

+ CacheableAnon bool

+ "context"

+

+ "golang.org/x/sync/singleflight"

+ // iteration without a restart; leave empty in production. Also disables

+ // the response cache below, so a live-edited template is never masked

+ // by a stale cached response.

+ // StaticTag is a content hash of the embedded static assets (set once

+ // at startup in main.go via HashFS), threaded onto every Ctx so the

+ // `static` template func can build cache-busting asset URLs.

+ StaticTag string

+

+

+ // respCache holds full rendered responses for CacheableAnon routes,

+ // keyed on the request (route+path+query, see cacheEligible) rather

+ // than content identity. sf collapses every request that arrives while

+ // a key is cold into one fill instead of one per request -- the thing

+ // that actually matters during a traffic spike is many distinct

+ // anonymous visitors hitting the same hot page, not just requests that

+ // happen to land in the same instant.

+ respCache *LRU

+ sf singleflight.Group

+ Log: slog.Default(),

+ Timeout: 10 * time.Second,

+ MaxGit: 16 << 20, // headroom above modules.maxPreviewSize (10 MiB)

+ cache: NewLRU(cacheSize),

+ respCache: NewLRU(cacheSize),

+ tag, err := HashFS(fsys)

+// HashFS digests every file in fsys, path and contents both, into a short

+// stable tag. Used for the template set (so editing the theme invalidates

+// every ETag at once -- without this a restyled page keeps serving 304s)

+// and, via main.go, for the embedded static assets, so a CSS/JS edit mints

+// a new cache-busting URL instead of being masked by a year-long immutable

+// Cache-Control.

+func HashFS(fsys fs.FS) (string, error) {

+// cachedResponse is a full rendered response, header and body both, as

+// stored in respCache. expiresAt is only meaningful for entries actually

+// stored there.

+type cachedResponse struct {

+ header http.Header

+ body []byte

+ expiresAt time.Time

+}

+

+ if key, ok := s.cacheEligible(rt, r); ok {

+ if cr, ok := s.respCacheGet(key); ok {

+ writeResponse(w, r, cr)

+ return

+ }

+ // singleflight collapses every request that arrives while this key

+ // is cold into one fill. That fill runs on a context detached from

+ // whichever request happened to trigger it: without that, the

+ // first caller to disconnect or time out would kill the shared

+ // git/render work and fail every other request that coalesced onto

+ // it too, even ones with their own deadline still well in the

+ // future. Bounded by s.Timeout so a genuinely stuck fill still

+ // gets reaped.

+ //

+ // The closure below captures r, the *triggering* (leader) request

+ // -- every follower that coalesces onto this Do gets a response

+ // built from the leader's request, not their own. That's only safe

+ // because cacheEligible already pinned every request-varying input

+ // that matters (user == "", no share cookie) and the key covers

+ // path+query, so leader and followers are asking an identical

+ // question. A future CacheableAnon route whose render depends on

+ // anything else request-specific (a header, say) would silently

+ // break this.

+ v, err, _ := s.sf.Do(key, func() (any, error) {

+ fillCtx, cancel := context.WithTimeout(context.Background(), s.Timeout)

+ defer cancel()

+ return s.fillResponse(reqCtx{fillCtx}, rt, r)

+ })

+ if err != nil {

+ s.fail(w, r, rt, err)

+ return

+ }

+ cr := v.(*cachedResponse)

+ s.respCache.Put(key, cr)

+ writeResponse(w, r, cr)

+ return

+ }

+

+ c, header, err := s.buildResponseHead(ctx, rt, r)

+ copyHeader(w.Header(), header)

+ if match := r.Header.Get("If-None-Match"); match != "" && strings.Contains(match, header.Get("ETag")) {

+ w.WriteHeader(http.StatusNotModified)

+ return

+ }

+

+ s.mu.RLock()

+ ts := s.templates

+ s.mu.RUnlock()

+ body, err := rt.Render.Render(c, ts)

+ if err != nil {

+ s.fail(w, r, rt, err)

+ return

+ }

+ w.Header().Set("Content-Type", rt.Render.ContentType())

+ if hs, ok := rt.Render.(HeaderSetter); ok {

+ hs.Headers(c, w.Header())

+ }

+ w.Write(body)

+}

+

+// cacheEligible reports whether rt/r may be served from, and stored in,

+// the shared response cache, and if so, the key to use. Eligibility is

+// decided entirely from cheap, git-free signals -- cookies and

+// already-loaded repo config -- so a cache hit costs nothing beyond a map

+// lookup, and a miss doesn't cost anything extra either before falling

+// through to the real render.

+//

+// Two independent conditions gate it, deliberately redundant: the request

+// must carry neither a session nor a share cookie (the same

+// anonymous-and-shareless condition the Cache-Control split in

+// buildResponseHead already trusts), AND, for a repo-scoped route, that

+// repo must not be Hidden -- checked directly from RepoInfo, with no

+// reference to any cookie at all. A cache entry is shared state replayed

+// to whoever asks next, so a false negative here has a bigger blast

+// radius than a single mis-issued header would: the Hidden check means a

+// bug in cookie detection alone can never leak a hidden repo's content,

+// since that repo's responses are never eligible for the shared cache in

+// the first place.

+func (s *Server) cacheEligible(rt Route, r *http.Request) (string, bool) {

+ if !rt.CacheableAnon || s.TemplateDir != "" {

+ return "", false

+ }

+ if s.Sessions.Verify(r) != "" {

+ return "", false

+ }

+ if _, ok := s.Sessions.VerifyShareCookie(r); ok {

+ return "", false

+ }

+ if rt.Scope == ScopeRepo {

+ h, ok := s.Site.Lookup(r.PathValue("repo"))

+ if !ok || h.Info.Hidden {

+ return "", false

+ }

+ }

+ return rt.Name + "\x00" + r.URL.Path + "\x00" + r.URL.RawQuery, true

+}

+

+func (s *Server) respCacheGet(key string) (*cachedResponse, bool) {

+ v, ok := s.respCache.Get(key)

+ if !ok {

+ return nil, false

+ }

+ cr := v.(*cachedResponse)

+ if time.Now().After(cr.expiresAt) {

+ return nil, false

+ }

+ return cr, true

+}

+

+// fillResponse builds a full response for storage in respCache: header and

+// body both, since an entry has to be complete before anyone else can be

+// served from it.

+func (s *Server) fillResponse(ctx contexter, rt Route, r *http.Request) (*cachedResponse, error) {

+ c, header, err := s.buildResponseHead(ctx, rt, r)

+ if err != nil {

+ return nil, err

+ }

+ s.mu.RLock()

+ ts := s.templates

+ s.mu.RUnlock()

+ body, err := rt.Render.Render(c, ts)

+ if err != nil {

+ return nil, err

+ }

+ header.Set("Content-Type", rt.Render.ContentType())

+ if hs, ok := rt.Render.(HeaderSetter); ok {

+ hs.Headers(c, header)

+ }

+ return &cachedResponse{header: header, body: body, expiresAt: time.Now().Add(c.cacheMaxAge())}, nil

+}

+

+// buildResponseHead resolves the Ctx and computes the ETag/Cache-Control

+// header, without executing the template -- split out so a matching

+// If-None-Match can skip rendering the body entirely.

+//

+// ETag is derived from the resolved commit SHA and every other input that

+// can change what's rendered, not the response body, so a conditional

+// request costs one ref lookup instead of a render.

+func (s *Server) buildResponseHead(ctx contexter, rt Route, r *http.Request) (*Ctx, http.Header, error) {

+ c, err := s.buildCtx(ctx, rt, r)

+ if err != nil {

+ return nil, nil, err

+ }

+ //

+ // s.StaticTag is folded in too: a CSS/JS-only edit changes every

+ // rendered `static "..."` URL, so page ETags have to move with it.

+ etag := `"` + shortHash(rt.Name, r.URL.Path, r.URL.RawQuery, tag, s.StaticTag, c.contentIdentity(), c.User, shareCookie) + `"`

+

+ header := http.Header{}

+ header.Set("ETag", etag)

+ header.Set("Cache-Control", "public, max-age=31536000, immutable")

+ header.Add("Vary", "Cookie")

+ // No shared/intermediary cache (proxy, CDN, or gsg's own

+ // respCache) may ever store this response, only the requesting

+ // browser -- otherwise a misbehaving shared cache could serve

+ // one user's hidden-repo listing to a different visitor. Vary:

+ // Cookie above is defense in depth for any cache that honors

+ // Vary but not Cache-Control: private.

+ header.Set("Cache-Control", "private, no-cache")

+ // public, with a real max-age rather than no-cache: the ETag

+ // already folds in c.User and the share cookie (above), so the

+ // only failure mode for a shared cache that mishandles Vary is

+ // serving this anonymous view to a logged-in or share-holding

+ // visitor within the TTL -- they see fewer repos, or a 404 on a

+ // hidden one. A UX papercut, not a disclosure: the dangerous

+ // direction (private content served to an anonymous visitor)

+ // can't happen here, since that's the separate branch above.

+ header.Set("Cache-Control", fmt.Sprintf("public, max-age=%d", int(c.cacheMaxAge().Seconds())))

+ return c, header, nil

+}

+

+func copyHeader(dst, src http.Header) {

+ for k, vs := range src {

+ for _, v := range vs {

+ dst.Add(k, v)

+ }

+}

+func writeResponse(w http.ResponseWriter, r *http.Request, cr *cachedResponse) {

+ copyHeader(w.Header(), cr.header)

+ if match := r.Header.Get("If-None-Match"); match != "" && strings.Contains(match, cr.header.Get("ETag")) {

+ w.WriteHeader(http.StatusNotModified)

+ w.Write(cr.body)

+ c.StaticTag = s.StaticTag

+ c.StaticTag = s.StaticTag

+ // buildResponseHead reads the same field, so page chrome that must not

+ // be frozen into a year-long response is suppressed by exactly the

+ // condition that makes the response year-long.

+ // The cached-response and cache-fill paths in serve() only ever reach

+ // fail() before anything is written to w, so for those there is

+ // nothing to strip. But the plain (non-cacheable) path copies

+ // buildResponseHead's ETag/Cache-Control onto w *before* calling

+ // Render, on the assumption that Render will succeed -- if it doesn't,

+ // this Del/Set is still what prevents a failed response from going out

+ // with a stale ETag or, worse, the year-long immutable Cache-Control

+ // from a commit route.

+ {Name: "index", Scope: core.ScopeSite, CacheableAnon: true,

+ {Name: "repo", CacheableAnon: true,

+ {Name: "tree", NavLabel: "files", CacheableAnon: true,

+ {Name: "log", NavLabel: "log", CacheableAnon: true,

+ {Name: "graph", NavLabel: "graph", CacheableAnon: true,

+ {Name: "refs", NavLabel: "refs", CacheableAnon: true,

+ // Arbitrary two-commit diff: /demo/compare/main...feature. Not

+ // CacheableAnon: a diff body is bounded only by MaxGit (16 MiB,

+ // larger than a blob preview), and every valid ref pair is a

+ // distinct key -- a repo with N refs has ~N^2 reachable ones. The

+ // response cache has no size-aware eviction, so this route would

+ // risk filling it with a few huge entries for no real hit-rate

+ // benefit (see internal/core/route.go's CacheableAnon doc).

+# gsg — a modular git browser

+

+cgit's weight, Hugo's composability, served dynamically. A **module**

+computes one piece of a view; a **route** binds a renderer to a URL pattern.

+Adding a feature means adding a module and calling it from a template —

+`internal/core` never changes.

+

+One dependency (`goldmark`, which itself has none) plus the `git` binary. Go 1.22+ (uses `net/http`

+pattern wildcards, so there is no router dependency either).

+

+```

+go build -o gsg .

+./gsg -root /srv/git -addr :8080 -name "rschanz.org"

+./gsg -root /srv/git -templates ./templates # re-parse the theme per request

+./gsg -root /srv/git -notices /etc/gsg/notices.yaml -admin-addr 127.0.0.1:8081

+```

+

+## The idea

+

+```go

+type Module func(c *Ctx, args ...string) (any, error)

+```

+

+Templates reach modules through one funcmap entry:

+

+```gotmpl

+{{ with mod "meta" }}<h1>{{ .Title }}</h1>{{ end }}

+{{ $d := mod "diff" }}

+{{ with modOr "readme" }}{{ .HTML }}{{ end }}

+```

+

+- **Lazy.** A template that calls `mod "authors"` is what *causes* authors to

+ be computed. Views that don't ask don't pay.

+- **Memoized twice.** Once per request (`Ctx.cache`), and once process-wide

+ against the resolved commit SHA (`core.LRU`).

+- **No manifest.** Dependencies are declared by use. There is no module list

+ to keep in sync with the templates.

+

+`internal/modules/example_authors.go` is a live demonstration: 30 lines, an

+`init()`, and one `{{ with mod "authors" }}` block. Nothing else moved.

+

+## Pages and endpoints are the same mechanism

+

+A route is `{pattern, renderer}`. `core.View` renders a template,

+`core.JSON` serializes an endpoint. Both take a `*Ctx`, both call modules.

+

+```go

+{Name: "commit", Immutable: true,

+ Pattern: "GET /{repo}/commit/{rev}",

+ Render: core.View{Template: "commit.html"}},

+

+{Name: "api-commit", Immutable: true,

+ Pattern: "GET /{repo}/api/commit/{rev}",

+ Render: core.JSON{Build: api.CommitEndpoint}},

+```

+

+### Endpoints use DTOs, not module values

+

+`internal/api` defines explicit structs and never marshals a module's return

+value. Go marshals every exported field by default, so serializing a module

+type means the day someone adds `Body` to `gitx.Commit`, commit messages are

+public. The allowlist is one readable file.

+

+```

+GET /api/repos.json site index

+GET /{repo}/api/repo.json slug, title, description, owner, section,

+ clone_url, default_branch, last_commit

+GET /{repo}/api/commit/{sha} sha, short, authored, parents, and

+ file/line counts — no subject, no body

+```

+

+## Routes

+

+| Route | URL |

+|---|---|

+| index | `/` |

+| repo | `/{repo}/` — default branch, last update, latest commit, README |

+| files | `/{repo}/tree/{ref}/{path}` |

+| file | `/{repo}/blob/{ref}/{path}` |

+| **raw bytes** | `/{repo}/raw/{ref}/{path}` |

+| log | `/{repo}/log/{ref}` |

+| refs | `/{repo}/refs/` |

+| commit | `/{repo}/commit/{sha}` |

+| **compare** | `/{repo}/compare/{base}..{head}` or `{base}...{head}` |

+| **branch graph** | `/{repo}/graph/` |

+

+### Compare

+

+`a..b` is the literal endpoint difference. `a...b` diffs from the merge

+base, which is what you want when reviewing a branch — it excludes changes

+that landed on the base since the branch forked. Verified against

+`git merge-base` on a diverged fixture: `topic..main` lists `topic.txt` as

+removed, `topic...main` does not.

+

+### Raw files

+

+A repository can contain an `.html` file with a `<script>` tag. Serving that

+as `text/html` on your own origin turns any push into stored XSS against the

+whole site — which is why GitHub puts raw content on

+`raw.githubusercontent.com`, a separate origin. Same-host serving is only

+safe if the response cannot be interpreted as a document, so `/raw/` sends

+`text/plain` (or `application/octet-stream` for binaries, plus

+`Content-Disposition: attachment`), with `X-Content-Type-Options: nosniff`

+and `Content-Security-Policy: default-src \'none\'; sandbox` as a backstop.

+

+The exception is a validated raster image (PNG, JPEG, GIF, WebP, BMP, ICO,

+AVIF): when the extension and the magic bytes agree, `/raw/` declares the

+real `image/*` type inline so the blob page can preview it. An extension by

+itself is not enough — an `.png` whose contents are HTML still downloads as

+`application/octet-stream`. SVG is deliberately excluded: it is an XML

+document that can execute script, so it stays `text/plain`.

+

+If you would rather have real isolation, serve `/raw/` from a second

+hostname and the CSP becomes belt-and-braces rather than the only line.

+

+### Images and structured config files

+

+An image blob shows an inline preview with a `download` link to the same

+`/raw/` URL; images over 10 MiB skip the preview and offer the download only.

+

+JSON and YAML blobs get per-level indent guides, fold controls, key/value

+search, and YAML anchor/alias links. The search box also accepts a value

+path: typing `obj1.property3.items[4].val` resolves it against the parsed

+document and highlights that line and, for a container, its subtree. A

+query containing `.` or `[` is tried as a path first and falls back to

+plain substring search when it does not resolve, so dotted text stays

+searchable. Paths quote keys that need it (`obj1["a.b"]`) and attach indices

+directly; SVG is never previewed, and files past 512 KiB skip the parsing

+tooling entirely.

+

+### Branch graph

+

+`modules/graph.go` lays out the DAG in lanes in one pass over

+`git log --all --topo-order`, which guarantees a child is emitted before its

+parents. `active` holds, per lane, the SHA that lane is waiting to draw; a

+commit claims the lane expecting it (or a fresh one), hands that lane to its

+first parent, and opens a new lane per additional parent — which is what

+makes a merge fork visually.

+

+The module emits the per-row SVG because the geometry *is* the data, but

+every stroke refers to a CSS custom property (`--lane-0` … `--lane-5`), so

+the theme still owns the palette. Row height in the SVG must stay in sync

+with `table.graph td { height }` in the stylesheet.

+

+## Caching, which is the whole reason this works dynamically

+

+Every module result is a pure function of `(repo, commit SHA, path, args)`.

+Refs move; SHAs do not. So the process-wide cache is keyed on the resolved

+SHA and needs **no invalidation logic at all** — a stale entry is impossible

+by construction.

+

+`Ctx.Vars` are part of the cache key, not just the commit. They were not

+originally, and the bug was exactly the kind this cache should be immune to:

+`/compare/a..b` and `/compare/a...b` resolve to the same head commit, so the

+second request was served the first one\'s diff. Anything that changes what a

+module computes has to be in the key.

+

+ETags are derived from the resolved SHA plus a digest of every template

+file, *not* from the response body. A conditional request costs one ref

+lookup instead of a render, and restyling the theme invalidates every ETag

+at once (without the template digest, a restyled page keeps serving 304s).

+

+Two URL classes, two policies:

+

+| URL shape | Example | Cache-Control |

+|---|---|---|

+| SHA-addressed | `/demo/commit/64b90fff…` | `max-age=31536000, immutable` |

+| ref-addressed | `/demo/tree/main` | `no-cache` (revalidate via ETag) |

+

+## Ref-vs-path ambiguity

+

+`/{repo}/tree/{rest...}` fuses them, and `gitx.SplitRefPath` separates them

+by **longest matching prefix against the repo's actual ref list**, so

+`feature/x/src/main.go` resolves correctly. Splitting at the first `/` is

+wrong because branch names contain slashes; cgit dodges this with `?h=`

+query params, GitHub does what this does.

+

+## Serving is an attack surface; static generation was not

+

+- **Ref input never reaches `git rev-parse`.** rev-parse accepts an

+ expression language — `:/regex` searches history, `@{...}` reads the

+ reflog — and git subcommands treat a leading `-` as an option.

+ `gitx.Resolve` accepts either a hex SHA that already exists or an exact

+ match against the repo's own ref list. Verified: `--upload-pack=…`,

+ `:/secret`, `HEAD@{0}` and `-n` all return 404 with no side effects.

+- **Every git call is `exec.CommandContext`** with a per-request deadline

+ (`-timeout`, default 10s) and a scrubbed environment

+ (`GIT_CONFIG_NOSYSTEM=1`, no `HOME`, no terminal prompt).

+- **Output is capped** at `MaxGit` (8 MiB) per command; a diff that exceeds

+ it renders truncated rather than eating the heap.

+

+## Layout

+

+```

+main.go flags, repo discovery, embedded assets

+internal/gitx/ the ONLY package that knows git exists

+internal/core/core.go Ctx, RepoInfo, module registry, Store

+internal/core/route.go Route, View, JSON, route request parsing

+internal/core/server.go request -> Ctx, ETags, cache policy

+internal/core/cache.go LRU keyed on commit SHAs

+internal/core/funcs.go template funcmap (mod, modOr, dict, date, bytes)

+internal/access/ visibility and share-grant policy (imports no core)

+internal/authn/ login/logout endpoints

+internal/httpx/ shared admin/share HTTP security helpers

+internal/modules/ meta refs log tree readme repos commit diff blob notices

+internal/notices/ notices YAML store + admin UI (imports no core)

+internal/api/ JSON DTOs — the serialization allowlist

+internal/site/routes.go the route table (the file you edit most)

+internal/site/request.go route-specific path parsers (compare ranges)

+templates/ layouts/ + views/

+```

+

+`gitx` shells out to plumbing, one process per query, `git log` in a single

+pass. Swapping in go-git means rewriting that one package.

+

+## Sections

+

+`gsg.section` groups repositories on the index. The `sections` module builds

+on `repos` rather than re-walking the repo set, so the index pays for its log

+lookups once. When no repo sets a section, it returns a single unnamed group

+and the template renders a flat list — no config, no headings.

+

+```sh

+git --git-dir=/srv/git/demo.git config gsg.section "Infrastructure"

+```

+

+## Notices

+

+Operator-authored banners at the top of every page, site-wide and per

+repository. They live in one YAML file that is watched for changes, so

+editing it never needs a restart or a rebuild:

+

+```yaml

+global:

+ - type: critical # critical | warning | info | fix

+ title: Maintenance window

+ body: |

+ Saturday 02:00-04:00 UTC.

+ The site is read-only during the window.

+ link: https://status.example.org

+ link_text: status page

+repos:

+ demo:

+ - type: warning

+ title: Migrating to a new remote

+```

+

+```sh

+./gsg -root /srv/git -notices /etc/gsg/notices.yaml

+```

+

+Global notices render above every repository's; on the site index only the

+global list appears. The file is re-read on write (atomic-rename saves from

+an editor included); a parse error keeps the last good content rather than

+blanking the banners.

+

+### Editing from the web

+

+`-admin-addr` starts a second listener with a small HTML form to add, edit,

+and delete notices. It is deliberately separate from the read-only server:

+the admin routes are not reachable on the public port even if the reverse

+proxy is misconfigured.

+

+```sh

+./gsg -root /srv/git -notices /etc/gsg/notices.yaml -admin-addr 127.0.0.1:8081

+```

+

+Protect it with Caddy `forward_auth` and proxy it to the admin listener. gsg

+trusts the network in front of it, so the admin port must not be published

+directly. In Compose, put 8081 on the internal network only and pass

+`-admin-allow-remote`, because a container address is not loopback. Note

+that gsg performs no authentication itself -- the `forward_auth` block below

+is the only thing between the internet and the admin form.

+

+```caddyfile

+example.com {

+ # Admin: authenticate first, then proxy to the separate admin listener.

+ # handle, not handle_path: the admin mux matches the full /admin/...

+ # path, so stripping the prefix would 404 every request.

+ handle /admin/* {

+ forward_auth authentik:9000 {

+ uri /outpost.goauthentik.io/auth/caddy

+ copy_headers X-Authentik-Username

+ }

+ reverse_proxy gsg:8081

+ }

+

+ # Public, read-only.

+ handle {

+ reverse_proxy gsg:8080

+ }

+}

+```

+

+The `X-Authentik-Username` header is used only to attribute the mutation in

+the log; it is never consulted for authorization. By default Authentik

+decides who may reach the handler at all -- anyone forward_auth lets through

+`/admin/*` can edit notices.

+

+Mutations are POST-only, CSRF-protected (double-submit cookie plus

+`Origin`/`Sec-Fetch-Site`), and logged with the `X-Authentik-Username`

+header, which is used for audit only, never for authorization.

+

+### Gating notices editing to a group

+

+`/admin/*` is also where [Login](#login) mints the `gsg_user` cookie, so

+forward_auth for the whole prefix is usually left open to everyone you want

+able to log in (for share links, hidden-repo visibility, etc). To let more

+people log in without also handing them notices editing, add `-admin-group`:

+

+```sh

+./gsg -root /srv/git -notices /etc/gsg/notices.yaml \

+ -admin-addr 127.0.0.1:8081 -admin-group gsg-admins

+```

+

+This checks the `X-Authentik-Groups` header (override with

+`-admin-group-header`) for `gsg-admins` among its `|`-delimited entries,

+Authentik's own format for that header, and answers `403` to any request

+under `/admin/notices` from a caller that isn't a member -- before the

+handler does anything else, including setting the CSRF cookie. Add the group

+header to the same `forward_auth` block used above:

+

+```caddyfile

+ forward_auth authentik:9000 {

+ uri /outpost.goauthentik.io/auth/caddy

+ copy_headers X-Authentik-Username X-Authentik-Groups

+ }

+```

+

+Leaving `-admin-group` unset keeps today's behavior: anyone forward_auth

+admits to `/admin/*` can edit notices.

+

+**Immutable commit pages carry no notices.** `/{repo}/commit/{rev}` is

+served `Cache-Control: public, max-age=31536000, immutable`, so a banner

+rendered into it would stay frozen in the reader's browser. Notices are

+suppressed there by design; every other page picks them up on the next

+request.

+

+## Login

+

+Logging in reveals hidden repos whose `gsg.visible-to` list includes your

+username; everyone else, logged in or not, sees exactly what an anonymous

+visitor sees. It reuses the same Authentik `forward_auth` setup as the

+notices admin UI above, and needs nothing extra beyond `-admin-addr`

+(notices editing is not required — `-admin-addr` alone is enough to enable

+login).

+

+```sh

+./gsg -root /srv/git -admin-addr 127.0.0.1:8081

+```

+

+```caddyfile

+example.com {

+ # Login: same trust boundary as /admin/*, one hop mints the cookie

+ # and redirects back. This must be the exact path, not /auth/*: gsg

+ # also serves /auth/logout, but on the *public* listener, since

+ # logging out only clears gsg's own cookie and needs no Authentik

+ # round trip at all. Routing /auth/* here would send logout to the

+ # admin listener too, where no such handler exists, and gate it

+ # behind an Authentik check it doesn't need.

+ handle /auth/login {

+ forward_auth authentik:9000 {

+ uri /outpost.goauthentik.io/auth/caddy

+ copy_headers X-Authentik-Username

+ }

+ reverse_proxy gsg:8081

+ }

+

+ # Public, read-only (also handles /auth/logout, /share/new, /share/view/*).

+ handle {

+ reverse_proxy gsg:8080

+ }

+}

+```

+

+Visiting `/auth/login` (through Caddy, so `X-Authentik-Username` is set and

+trustworthy by the time gsg sees the request) mints a signed, stateless

+`gsg_user` cookie and redirects back to `/`. The cookie is verified on the

+public listener with an in-memory key — nothing is written to disk, and a

+process restart signs everyone out. `/auth/logout` clears it and needs no

+round trip through Authentik, since it only has to forget gsg's own cookie.

+

+**There is no "log in" link anywhere in the UI.** `/auth/login` is not

+discoverable from the site; people who need it are told the URL out of

+band. Once logged in, the nav bar shows who you are and a log-out link.

+

+## Share links

+

+Once logged in, `/share/new` lets you pick from your own hidden repos and

+mint a link that grants view access to exactly those repos to anyone who

+has it — no login, no Authentik involvement. It's entirely stateless: the

+link is a signed, self-contained token, and gsg keeps no record of what's

+been shared. An optional password can gate it.

+

+```

+git --git-dir=/srv/git/demo.git config gsg.share-max-age "720h"

+```

+

+`gsg.share-max-age` (Go duration syntax, default 30 days if unset) is both

+the ceiling no link's expiry may exceed and the default when the minting

+form doesn't specify one — you can always ask for less, never more. Sharing

+several repos at once uses the smallest ceiling among them.

+

+**Passwords are never stored, not even hashed.** The password is folded

+directly into the link's signature instead of being embedded in it: nobody

+holding just the URL has anything to brute-force offline, and checking a

+guess requires asking the server, which rate-limits attempts. This is a

+soft gate appropriate for a small self-hosted tool, not a cryptographic

+secret protecting high-value data.

+

+**Revocation.** A stateless token can't be individually revoked — there's

+no list of what's been issued. The one lever:

+

+```

+git --git-dir=/srv/git/demo.git config gsg.share-revoked-after "$(date +%s)"

+```

+

+Any share grant for that repo issued at or before this timestamp stops

+working immediately, including in a browser that already redeemed the

+link — not just future redemptions. Two things worth knowing before you

+rely on this:

+

+- There's no per-recipient state, so revoking means invalidating *every*

+ outstanding share of that repo, then re-sharing with whoever should still

+ have it.

+- Removing a username from `gsg.visible-to` does **not** revoke a share

+ link that user already minted for that repo. The two mechanisms are

+ independent; only `gsg.share-revoked-after` invalidates outstanding

+ shares.

+

+## README rendering

+

+`modules/markdown.go` uses goldmark with GFM, Typographer, and automatic

+heading IDs — Hugo\'s own goldmark defaults, so a README renders identically

+to the same file on your blog.

+

+`WithUnsafe` is deliberately not set, which is also Hugo\'s default. Anyone

+who can push can write a README, so raw HTML in markdown is untrusted input;

+goldmark drops it. A `<script>` committed to a README comes out as

+`<!-- raw HTML omitted -->`.

+

+Relative links are rewritten against the file\'s own directory: links become

+the blob view, images become `/raw/` (an `<img>` needs bytes, not an HTML

+page). Absolute URLs, site-absolute paths, fragments and `mailto:` are left

+alone, and anything resolving above the repo root is not rewritten.

+

+## Per-repository config

+

+Read from each repo's own git config, cgit-style, with `gitweb.*` and the

+legacy `description` file as fallbacks, so an existing cgit tree works as-is:

+

+```sh

+git --git-dir=/srv/git/demo.git config gsg.title "Demo Project"

+git --git-dir=/srv/git/demo.git config gsg.description "A demo repository"

+git --git-dir=/srv/git/demo.git config gsg.clone "git://rschanz.org/demo"

+git --git-dir=/srv/git/demo.git config gsg.hide true

+git --git-dir=/srv/git/demo.git config gsg.visible-to "alice,bob"

+git --git-dir=/srv/git/demo.git config gsg.share-max-age "720h"

+git --git-dir=/srv/git/demo.git config gsg.share-revoked-after "$(date +%s)"

+```

+

+Add a key by adding a field to `core.RepoInfo` and a line in `LoadRepoInfo`.

+

+A repo's slug is its directory name (minus `.git`), and slugs share the

+public listener's URL namespace with a few reserved top-level paths:

+`static`, `auth`, and `share`. A repo named e.g. `share.git` would collide

+with the share-link endpoints below and is not supported.

+

+`gsg.visible-to` is a comma-separated allowlist of usernames (matched

+case-insensitively against a logged-in session, see "Login" below) who may

+see this repo even though `gsg.hide` is set. It's a separate key from

+`gsg.owner` on purpose: `Owner` is a free-text display field that falls

+back to `gitweb.owner` for cgit-migrated repos, so treating it as an access

+list could silently grant visibility nobody intended.

+

+## Theme

+

+Monospace-first: every value on these pages is an identifier, hash, path or

+count, so the interface type is the same face as the code, and the sans face

+appears only inside rendered README prose. A cool slate ramp, one indigo

+accent, and a six-colour lane ramp shared by the commit graph and its branch

+chips. Light and dark both defined via `prefers-color-scheme`.

+

+Diffs mark the gutter rather than flooding whole rows with green and red —

+the signal stays readable when a hunk is mostly additions.

+

+No webfonts; system stacks only, so there is no third-party request. All

+colour lives in `:root` custom properties in `static/style.css`.

+

+## Known gaps

+

+- **No syntax highlighting.** `blob` returns numbered plain lines; Chroma

+ drops into the same module.

+- **No pagination on `log` or `graph`.** They cap at 200 and 150 commits.

+- **The graph loads all refs into memory** before laying out. Fine to a few

+ thousand commits; beyond that it needs windowing.

+- **The LRU is bounded by entry count, not bytes.** A giant diff and a ref

+ list cost the same slot. The `Store` interface is the only thing a

+ size-aware replacement has to satisfy.

+- **Repo discovery is one level deep** and happens once at startup; a repo

+ created later needs a restart.

+- **Module errors surface at render time**, not load time. Add a CI pass

+ that renders every view against a fixture repo.

+ flag.IntVar(&o.cacheSize, "cache", 4096, "entries per cache (module results and the anonymous response cache each get their own, sized identically)")

+ // Content hash of the embedded assets, not a build timestamp, so it's

+ // stable across restarts that don't change any file and only changes

+ // when one actually does -- templates use it (via the `static` funcmap

+ // entry) to mint cache-busting URLs.

+ staticTag, err := core.HashFS(static)

+ if err != nil {

+ return err

+ }

+ srv.StaticTag = staticTag

+ mux.Handle("/static/", cachedStatic(static))

+ amux.Handle("/static/", cachedStatic(static))

+// cachedStatic serves fsys with a Cache-Control that depends on whether the

+// request carries the `v` cache-busting query param the `static` template

+// func mints (see internal/core/funcs.go). Requests that carry it are

+// addressed by content hash, so they're frozen for a year; everything else

+// -- share/notices/error-page assets, which don't go through that funcmap

+// and so never get a `v` -- still gets a real improvement over no

+// Cache-Control at all, just a short one, in case a template edit changes

+// what a bare URL should serve.

+func cachedStatic(fsys fs.FS) http.Handler {

+ fileServer := http.FileServer(http.FS(fsys))

+ h := http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {

+ if r.URL.Query().Get("v") != "" {

+ w.Header().Set("Cache-Control", "public, max-age=31536000, immutable")

+ } else {

+ w.Header().Set("Cache-Control", "public, max-age=600")

+ }

+ fileServer.ServeHTTP(w, r)

+ })

+ return http.StripPrefix("/static/", h)

+}

+

+<link rel="stylesheet" href="{{ static "css/base.css" }}">

+ <span>{{ .Site.Name }} v1.4</span>

+{{ define "styles" }}<link rel="stylesheet" href="{{ static "css/blob.css" }}">{{ end }}

+{{ define "scripts" }}{{ if (mod "blob").Structured }}<script src="{{ static "js/structured.js" }}" defer></script>{{ end }}{{ end }}

+<link rel="stylesheet" href="{{ static "css/diff.css" }}">

+<link rel="stylesheet" href="{{ static "css/readme.css" }}">

+{{ define "styles" }}<link rel="stylesheet" href="{{ static "css/diff.css" }}">{{ end }}

+{{ define "styles" }}<link rel="stylesheet" href="{{ static "css/graph.css" }}">{{ end }}

+{{ define "styles" }}<link rel="stylesheet" href="{{ static "css/repo-list.css" }}">{{ end }}

+<link rel="stylesheet" href="{{ static "css/repo.css" }}">

+<link rel="stylesheet" href="{{ static "css/readme.css" }}">