The cook log SQLite DB was created at recipeDir/cook-log.db. In the Docker
deployment /recipes is mounted read-only, so SQLite could never create the
file there — the cook log has never worked in the container. The writable
/data volume was mounted but unused.
- initDb now creates the DB's parent directory if missing, so startup
initialization is self-contained and idempotent.
- app takes a dataDir and places cook-log.db there.
- Add optional --data-dir flag (defaults to the recipe dir, preserving
local-dev behavior); Docker CMD passes --data-dir /data.
Verified end-to-end as root against a read-only recipe dir and a
non-existent data dir: the dir and DB are created and cook-log POST/GET
round-trips.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
groupEntries was sorting month groups by reverse-alphabetical name
(e.g. 'June' before 'July'), producing oldest-first ordering.
Now sorts by the actual entry dates within each group (newest first).
Combined with the earlier DESC fix in fetchAllEntries, the cook
history page now shows newest months at top, newest entries within
each month at top.
- Add Import link to all page headers (landing, recipe, cook history, import)
- Match ingredient list font size to recipe method text (both 15px)
- Show recipe titles instead of filenames on Cook History page
- Make clicking anywhere on recipe hero image/placeholder trigger file upload dialog
- New landing page (/) based on roux-landing-mockup.html layout
- Shuffle card that suggests random recipes
- Recently cooked shelf (driven by cook log DB)
- Recently added shelf (by file modification time)
- Existing recipe index moved to /recipes
- Updated nav links throughout to point to /recipes
- All hlint warnings fixed
- Use newtype instead of data for single-field types (RouxConfig, ImportError)
- Replace maybe "" id with fromMaybe "" throughout
- Replace not (x `elem` ys) with x `notElem` ys
- Remove redundant brackets, redundant $, and eta-reduce lookupRecipe
- Use numeric underscore for 1_000_000
Previously images were saved directly in the recipe directory
alongside .cook files. Now they go into <recipes-dir>/recipe-images/,
keeping the recipe directory clean and making the purpose of the
file clear from its path.
Users can now upload an photo for any recipe directly from the
view page. The upload button appears as either a 'Replace' overlay
on existing images or an empty dashed placeholder for recipes
without one.
Upload flow:
1. Click upload button -> file picker opens (accepts image/*)
2. JavaScript reads the file as base64, POSTs to /upload-image
with form fields: filename, file-b64, file-name
3. Server decodes the base64, saves to <recipe-dir>/<basename>.<ext>
4. Server updates the .cook file's YAML front matter, adding or
updating the 'image:' key with a /recipe-images/ URL
5. Server returns JSON success, page reloads to show the image
Server routes added:
- POST /upload-image — accepts base64-encoded image upload
- GET /recipe-images/<filename> — serves saved recipe images
(extension-whitelisted to .jpg/.jpeg/.png/.gif/.webp)
Image files are stored alongside .cook files in the recipe directory
and served via the /recipe-images/ path. The fsnotify watcher detects
the .cook metadata change and triggers an SSE reload.
When importing a recipe, the raw JSON output from scrape-recipe is now
saved as <recipe-name>.json in the recipe directory alongside the .cook
file. This lets users inspect the intermediate schema.org representation
to debug step breakdown or ingredient parsing issues.
The lazy I/O in LB.hGetContents returns a thunk immediately without
reading any data. If waitForProcess is called before the ByteString
is forced, the parent blocks on the process while the child blocks
on a full pipe buffer (no one is reading). Fix by computing LB.length
(which forces full traversal) before calling waitForProcess.
createProcess returns (stdin, stdout, stderr, process), not
(stdout, stderr, _, process). Since stdin is Inherited (not CreatePipe),
it returns Nothing, not Just. The pattern (Just outH, Just errH, _, _)
was matching on the stdin and stdout fields, so it always fell through
to the error branch.
Also adds type-safety conventions to AGENTS.md: no partial patterns on
IO results, no partial functions (head/tail/fromJust), always use case
with explicit branches.
readProcessWithExitCode returns stdout/stderr as String, which forces GHC
to decode raw bytes using the current locale encoding. In a Docker
container with C locale (default), non-ASCII UTF-8 bytes (like 0xE2)
cause 'invalid argument' errors.
Replace with readProcessBytes which uses createProcess + hSetBinaryMode
to capture stdout/stderr as raw lazy ByteStrings, bypassing locale
decoding entirely. JSON is decoded directly from raw bytes via A.decode.
The urldecode helper in parseFormBody was using Html.urlDecode which only
handled %%20, %%23, and %%25. Full URLs encode colons, slashes, and dots
as %%3A, %%2F, and %%2E, so they were never decoded. The scraper received
the still-encoded URL and failed to fetch it.
Replace Html.urlDecode with a proper percent-decode that handles all %%XX
hex sequences using digitToInt.
Logs each stage of the import with [roux] prefix: request URL, script
path, exit codes, stderr output, JSON parse status, conversion status,
and file write path. Error messages now include the attempted URL for
easier debugging.
- Strip .cook from map keys to match lookupRecipe behavior
- Use makeRelative for event paths to match relative map keys
- Filter watcher to only .cook files
- Log watcher thread exceptions
- Use imported isSuffixOf from Data.List
- Add SortMode type (AlphaSort | TagSort | CourseSort) to Roux.Html
- Routes: / (alpha), /sorted/tags, /sorted/course
- AlphaSort: recipes sorted by title (from metadata) or filename
- TagSort: recipes grouped under tag headings, empty state when none
- CourseSort: recipes grouped by course/category metadata
- Navigation bar in index page with active-highlighted buttons
- Add metaCourse field to Metadata model (+ emptyMetadata update)
- Add nubOrd helper for deduplicating tag/course lists
- All 26 tests passing, hlint clean
- Html.hs: replace lambdas with section/composition where possible
- Html.hs: replace case/fromMaybe with fromMaybe
- Html.hs: add fromMaybe to imports, use unqualified
- Parser.hs: replace span with break (2 locations)
- Parser.hs: replace if/then/else with list comprehension
- Parser.hs: remove redundant catch-all patterns (break is exhaustive)
- Server.hs: eta reduce htmlResponse
- Add .hlint.yaml to suppress 'Avoid lambda' (suggested fix has
precedence issues with blaze-html's infix (!) operator)
- All 26 tests passing, hlint: No hints, exit code 0
- Add Roux.Html.recipePage: full recipe detail page with:
- Back navigation link
- Title and metadata (servings, times, difficulty, etc.)
- Ingredients summary list with quantities
- Sections with named headings
- Steps with inline element rendering:
- Ingredients highlighted in pumpkin (with quantity badge)
- Cookware in violet italic
- Timers in azure with ⏱ icon
- Comments in sand italic
- Recipe refs as links
- Line breaks
- Update Roux.Server: route /recipes/FILENAME to recipe page
- Case-insensitive filename lookup, .cook extension optional
- URL decoding for filenames
- Fix ingredient name parsing: restrict name characters to
alphaNum + spaces/hyphens/apostrophes, preventing greedy
consumption across special chars like ) and #
- Split name chars into pMultiNameChar (with spaces, before braces)
and pSingleNameChar (without spaces, fallback)
- Suppress -Wname-shadowing in Html.hs (blaze-html exports clash
with common variable names)
- All 26 tests passing, server smoke-tested with examples
- Add Roux.Server: WAI application using wai+warp directly
- Scans .cook files from recipe directory on startup
- GET / → HTML index page listing all recipes
- Other routes → 404
- Add Roux.RecipeIndex: directory scanning and recipe parsing
- scanRecipes: lists .cook files, parses each, extracts titles
- Falls back to filename when metadata title is absent
- Add Roux.Html: blaze-html rendering with Pico CSS
- indexPage: renders recipe list with links
- urlEncode: percent-encodes filenames for URLs
- Update Roux.hs: re-exports Server.app and Types
- Add filepath to dependencies
jbrechtel