Add option to specify asset paths for SPAs #53
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Rewritten on top of the |
|
Rebased on top of the |
Allows asset path prefixes to be specified by the CLI when running in single-page mode. If a path is requested that doesn't exist, and that path matches one of the asset path prefixes, the request will 404 instead of retuning a 200 containing the root index. Fixes #44
Codecov Report
@@ Coverage Diff @@
## next #53 +/- ##
=======================================
Coverage 98.48% 98.48%
=======================================
Files 2 2
Lines 66 66
=======================================
Hits 65 65
Misses 1 1
Continue to review full report at Codecov.
|
lukeed
approved these changes
May 31, 2020
There was a problem hiding this comment.
Thank you for this :) Sorry it took some time to get to – slowly but surely!
Part of my delay here was that I was debating whether or not all asset requests (read, URLs that ended in an extension) should not be given an index.html fallback by default. If you don't want this, you can provide assets: false or --no-assets to stop. Otherwise, you can provide additional patterns (no longer just prefixes) that you want to opt out of the fallback behavior.
For example: "/static/" (like in your example) will match any URL that contains "/static/" within it, but providing "^/static/" will ignore pathnames that start with "/static/" ... which is to say, any String value is turned into a RegExp as given. If you provide a RegExp, it's left as it is.
Let me know if the changes look okay to you.
lukeed
added a commit
that referenced
this pull request
Jun 6, 2020
- ran before extensions could be applied - effectively reverts to #53
lukeed
added a commit
that referenced
this pull request
Jun 10, 2020
* chore(sirv): bump “@polka/url” for better decoding * feat(sirv): inline `opts.single` behavior * fix(sirv): unknown Content-Type as empty string * feat(sirv-cli): run HTTP/2 server behind flags * chore(sirv-cli): update readme * feat: run `sade` in single-command mode * feat(sirv): allow `--single` to accept path; - Closes #35 * fix(sirv): respect existing `Content-Type` header; - Closes #38 - Closes #39 * fix(sirv-cli): use "-K" for `--http2` alias * feat: allow precompiled gzip & brotli files; - Closes #3 * refactor(sirv): consolidate "dev & prod" modes; - still optimized - adds ETag support to "dev" mode - adds brotli/gzip support to "dev" mode - Related: #3 * chore: bump devdeps * v1.0.0-next.0 * chore(docs): fix Polka demo usage; - Closes #42 * fix(sirv): set "no-store" during `dev` mode; - Closes #45 * v1.0.0-next.1 * fix(sirv): extract `list` into package * fix(sirv): ignore deeply-nested dotfiles too * chore(sirv): update `polka/url` dep; lukeed/polka#119 * v1.0.0-next.2 * feat(sirv-cli): add `--pass` flag for passphrase * v1.0.0-next.3 * fix(sirv): compare `ETag` against `If-None-Match` header (#56) * fix(sirv-cli): use `host` for port availability * Add support for ETag-based caching When providing an ETag header on responses, the client automatically sends that header value on subsequent requests for the matching resource as an `If-None-Match` request header. If the ETag value received in `If-None-Match` is the same as the ETag computed by the disk crawl, a 304 "Not Modified" status can be returned along with an empty body to indicate that the requestor's cached value should be used. * chore: use existing `isEtag` var Co-authored-by: Luke Edwards <[email protected]> * chore(sirv): squash declaration * v1.0.0-next.4 * break: require Node 10.x minimum; - Node 8.x is already End of Life (2019-12-31) * v1.0.0-next.5 * feat(sirv): add type definitions; - Closes #61 * chore: add `notice` to base readme re: website; - Closes #62 * feat(sirv): add `filter` option to determine file list; - supplants default `dotfiles` matcher, responsible for own logic - Closes #50 * fix(sirv): replace `filter` w/ built-in ".well-known" allowance; - Closes #50 * chore: rename files * chore(sirv-cli): rename `boot` file * chore(sirv): produce `build` files w/ "module" entry * chore: run tests with esm loader * chore: re-extract `middleware` helper file * chore: begin integration test suites; - include fixtures directory * chore: use `pnpm` again for lerna; This reverts commit a56872b. # Conflicts: # lerna.json * chore(action): install `pnpm` for ci * chore: add `single` tests * chore(action): use `curl` installer for pnpm * chore: add "extensions" tests * chore: add `dotfiles` test suite * chore: add `dev` test suite * chore: add `etag` test suite * chore: add `brotli` and `gzip` suites * chore: add `maxAge` and `immutable` test suites * chore: add `ranges` test suite * fix(test): remove `only` on dotfiles suite * chore: import `setHeaders` and `onNoMatch` suites * chore: cleanup test helpers * chore: include "build" step in CI * chore: initial CLI tests * revert: allow normal resolution * chore: try `actions/cache` recipe * chore: add loose CLI tests * fix(test): terminate child procs correctly * chore: boot `http2` server within test * fix(sirv): normalize encoded/accented characters for cache * debug: server spawner really annoying * feat: add "assets" option (#53) * Add option to specify asset paths for SPAs Allows asset path prefixes to be specified by the CLI when running in single-page mode. If a path is requested that doesn't exist, and that path matches one of the asset path prefixes, the request will 404 instead of retuning a 200 containing the root index. Fixes #44 * fix: ensure `assets` array & preload prep work * chore: add `assets` tests * chore: update options definition * update `--single` & `--assets` text Co-authored-by: Luke Edwards <[email protected]> * refactor: rename "assets" to "ignores" option * fix: include dotfiles & well-known in default ignores * fix: move `ignores` check into `toAssume` builder * refactor: hoist bound `lookup` function * fix(sirv): do not fill cache when `dev` true * v1.0.0-next.6 * fix(sirv): extension regexp * v1.0.0-next.7 * fix(sirv): remove premature ignore check; - ran before extensions could be applied - effectively reverts to #53 * v1.0.0-next.8 * fix(sirv): allow definition to export type members * v1.0.0-next.9 * fix(sirv): dedupe extensions list w/ brotli + gzip * chore: update documentation * chore: update bench figures; - include dummy apps directly * chore: fix spacing Co-authored-by: Jason Miller <[email protected]> Co-authored-by: Carey Metcalfe <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Allows asset path prefixes to be specified by the CLI when running in single-page mode.
If a path is requested that doesn't exist, and that path matches one of the asset path prefixes, the request will 404 instead of retuning a 200 containing the root index.
Fixes #44