hugo/RuDS
1
0
Fork 0
Code Issues 1 Pull Requests Actions Packages Projects Releases Wiki Activity

chore: tended the spec against the bDS app to fix contradictions

Browse Source
This commit is contained in:
Chili Palmer
2026-04-05 08:06:04 +02:00
parent ee61ad56ea
commit f72d87eafe
9 changed files with 213 additions and 124 deletions
Download Patch File Download Diff File Expand all files Collapse all files

235
specs/media_processing.allium
View File

@@ -15,75 +15,111 @@ use "./search.allium" as search
-- ============================================================================
value MediaFileLayout {
-- Binary assets stored in: media/{filename}
-- Sidecar metadata in: media/{id}.md
-- Thumbnails in: thumbnails/{id}.webp
-- Thumbnail source filename: thumbnails/{id}_source.{ext}
binary_path: String -- media/{filename}
sidecar_path: String -- media/{id}.md
thumbnail_path: String -- thumbnails/{id}.webp
thumbnail_source_path: String -- thumbnails/{id}_source.{ext}
-- Binary assets stored in: media/{YYYY}/{MM}/{uuid}.{ext}
-- Sidecar metadata in: {binary_path}.meta
-- Thumbnails in: thumbnails/{id[0:2]}/{id}-{size}.webp
-- (ai thumbnail is JPEG: thumbnails/{id[0:2]}/{id}-ai.jpg)
binary_path: String -- media/{YYYY}/{MM}/{uuid}.{ext}
sidecar_path: String -- {binary_path}.meta
thumbnail_small: String -- thumbnails/{prefix}/{id}-small.webp
thumbnail_medium: String -- thumbnails/{prefix}/{id}-medium.webp
thumbnail_large: String -- thumbnails/{prefix}/{id}-large.webp
thumbnail_ai: String -- thumbnails/{prefix}/{id}-ai.jpg
}
invariant MediaFileNaming {
-- Original filename is preserved in original_name field
-- Stored filename is generated: {timestamp}_{random}.{ext}
-- Example: 1710512345678_abc123.jpg
-- Stored filename uses UUID v4: {uuid}.{ext}
-- Example: a1b2c3d4-e5f6-7890-abcd-ef1234567890.jpg
for m in Media:
m.filename = format("{timestamp}_{random}.{ext}",
timestamp: m.created_at.milliseconds,
random: generate_random_string(6),
m.filename = format("{uuid}.{ext}",
uuid: generate_uuid_v4(),
ext: file_extension(m.original_name))
}
invariant ThumbnailPathBucketing {
-- Thumbnails are bucketed by first 2 chars of media ID
-- This avoids filesystem slowdowns from too many files in one directory
for m in Media:
let prefix = m.id[0:2]
m.thumbnails.small = format("thumbnails/{prefix}/{id}-small.webp",
prefix: prefix, id: m.id)
m.thumbnails.medium = format("thumbnails/{prefix}/{id}-medium.webp",
prefix: prefix, id: m.id)
m.thumbnails.large = format("thumbnails/{prefix}/{id}-large.webp",
prefix: prefix, id: m.id)
m.thumbnails.ai = format("thumbnails/{prefix}/{id}-ai.jpg",
prefix: prefix, id: m.id)
}
-- ============================================================================
-- THUMBNAIL GENERATION
-- ============================================================================
config {
-- Thumbnail configuration
thumbnail_width: Integer = 400
thumbnail_height: Integer = 300
thumbnail_fit: String = "cover" -- "cover", "contain", "fill"
thumbnail_quality: Integer = 80 -- WEBP quality 1-100
thumbnail_format: String = "webp" -- Output format
-- Four thumbnail sizes generated per image
thumbnail_small_width: Integer = 150
thumbnail_medium_width: Integer = 400
thumbnail_large_width: Integer = 800
thumbnail_ai_size: Integer = 448 -- 448x448 square crop, JPEG
thumbnail_quality: Integer = 80 -- WebP quality, hardcoded
thumbnail_format: String = "webp" -- All sizes except AI
thumbnail_ai_format: String = "jpeg" -- AI thumbnail only
}
rule GenerateThumbnail {
rule GenerateThumbnails {
when: MediaImported(media)
requires: is_image(media.mime_type)
-- Generate thumbnail from source image
-- Generate all four thumbnail sizes
ensures: ThumbnailGenerated(
source: media.file_path,
destination: media.thumbnail_path,
width: config.thumbnail_width,
height: config.thumbnail_height,
fit: config.thumbnail_fit,
destination: media.thumbnails.small,
width: config.thumbnail_small_width,
quality: config.thumbnail_quality,
format: config.thumbnail_format
)
ensures: ThumbnailSourceSaved(
ensures: ThumbnailGenerated(
source: media.file_path,
destination: media.thumbnail_source_path
destination: media.thumbnails.medium,
width: config.thumbnail_medium_width,
quality: config.thumbnail_quality,
format: config.thumbnail_format
)
ensures: ThumbnailGenerated(
source: media.file_path,
destination: media.thumbnails.large,
width: config.thumbnail_large_width,
quality: config.thumbnail_quality,
format: config.thumbnail_format
)
ensures: ThumbnailGenerated(
source: media.file_path,
destination: media.thumbnails.ai,
size: config.thumbnail_ai_size,
format: config.thumbnail_ai_format
)
}
-- Thumbnail generation algorithm
value ThumbnailGeneration {
-- 1. Load source image with EXIF orientation
-- 2. Auto-rotate based on EXIF orientation tag
-- 3. Resize using "cover" fit (crop to fill target dimensions)
-- 4. Encode as WEBP with specified quality
-- 5. Write to thumbnails/{id}.webp
-- Preserve original for source thumbnail
-- This allows regenerating thumbnails at different sizes later
-- 1. Load source image
-- 2. Read raw image dimensions from header (not EXIF-orientation-aware)
-- 3. Resize: small/medium/large preserve aspect ratio (width-constrained)
-- AI thumbnail is a 448x448 center crop
-- 4. Encode as WebP (quality 80) for small/medium/large
-- Encode as JPEG for AI thumbnail
-- 5. Write to bucketed thumbnail path: thumbnails/{id[0:2]}/{id}-{size}.{ext}
--
-- No _source copy is made. Thumbnails regenerated from the original binary.
}
invariant ThumbnailExifHandling {
-- EXIF orientation must be respected during thumbnail generation
-- Photos taken with phone cameras often have incorrect orientation
-- without proper EXIF handling
-- TypeScript app reads raw image header dimensions
-- It does NOT apply EXIF orientation correction during thumbnail generation
-- Width/height stored in DB are the raw header values
@guidance
-- The Rust implementation should match this: raw header dimensions
-- EXIF auto-rotation can be added later as an enhancement
}
-- ============================================================================
@@ -97,51 +133,53 @@ value ImageProcessing {
"image/webp", "image/tiff", "image/bmp",
"image/heic", "image/heif"
}
-- Output formats:
output_formats: Set<String> = {
"image/webp", -- Primary output
"image/jpeg", -- Fallback for non-WEBP browsers (not used in Rust version)
"image/png" -- For images with transparency
"image/webp", -- Primary output for thumbnails
"image/jpeg" -- AI thumbnail only
}
-- Processing rules:
-- 1. All images are converted to WEBP for thumbnails
-- 2. Original format is preserved for full-size assets
-- 3. EXIF data is stripped from thumbnails (privacy)
-- 4. EXIF orientation is applied during processing
-- 1. All thumbnails (except AI) are encoded as WebP quality 80
-- 2. AI thumbnail is encoded as JPEG (for vision model compatibility)
-- 3. Original format is preserved for full-size assets (no conversion)
-- 4. EXIF data is not stripped (thumbnails are re-encoded, so EXIF is naturally absent)
}
rule ProcessImageMetadata {
when: MediaImported(media)
-- Extract image metadata
ensures: media.width = extract_width(source_file)
ensures: media.height = extract_height(source_file)
ensures: media.mime_type = detect_mime_type(source_file)
-- Extract image metadata from raw file header
ensures: media.width = extract_width_from_header(source_file)
ensures: media.height = extract_height_from_header(source_file)
ensures: media.mime_type = detect_mime_from_extension(source_file)
ensures: media.size = file_size(source_file)
}
invariant MimeDetection {
-- MIME type is detected from file extension, not from file content/magic bytes
-- Extension mapping: .jpg/.jpeg -> image/jpeg, .png -> image/png, etc.
}
-- ============================================================================
-- MEDIA TRANSLATION FILES
-- ============================================================================
value MediaTranslationFile {
-- File path: media/{id}/{language}.md
-- Format: YAML frontmatter with localized metadata
-- File path: {binary_path}.{language}.meta
-- Format: YAML-like key-value sidecar (same as canonical sidecar)
translation_for: String -- Canonical media ID
language: String -- ISO 639-1 code
title: String?
alt: String?
caption: String?
created_at: Timestamp
updated_at: Timestamp
}
invariant MediaTranslationFileLayout {
for t in MediaTranslations:
-- Stored in subdirectory per media item
t.file_path = format("media/{id}/{lang}.md",
id: t.translation_for,
-- Translation sidecars sit next to the binary, with language suffix
t.file_path = format("{binary_path}.{lang}.meta",
binary_path: t.media.file_path,
lang: t.language)
}
@@ -151,25 +189,24 @@ invariant MediaTranslationFileLayout {
rule ImportMedia {
when: ImportMediaRequested(source_path, project)
-- 1. Validate file type (must be image)
-- 2. Generate unique filename
-- 3. Copy to media/{filename}
-- 4. Generate thumbnail
-- 5. Create sidecar file with metadata
-- 1. Validate file type (must be supported image)
-- 2. Generate UUID v4 filename
-- 3. Copy to media/{YYYY}/{MM}/{uuid}.{ext}
-- 4. Write sidecar {binary_path}.meta
-- 5. Generate four thumbnail sizes
-- 6. Index for search (FTS5)
-- 7. Generate embedding (if enabled)
ensures: media/Media.created(
filename: generate_filename(),
filename: generate_uuid_v4_filename(source_path),
original_name: basename(source_path),
mime_type: detect_mime_type(source_path),
mime_type: detect_mime_from_extension(source_path),
size: file_size(source_path),
width: extract_width(source_path),
height: extract_height(source_path),
file_path: format("media/{filename}", filename: generated_filename),
sidecar_path: format("media/{id}.md", id: media_id),
width: extract_width_from_header(source_path),
height: extract_height_from_header(source_path),
file_path: format("media/{yyyy}/{mm}/{uuid}.{ext}"),
sidecar_path: format("media/{yyyy}/{mm}/{uuid}.{ext}.meta"),
checksum: sha256(source_path)
)
ensures: ThumbnailGenerated(media_id)
ensures: ThumbnailsGenerated(media_id)
ensures: SearchIndexUpdated(media_id)
}
@@ -198,14 +235,17 @@ rule DeleteMedia {
when: DeleteMediaRequested(media)
-- 1. Remove from database
-- 2. Delete binary file
-- 3. Delete sidecar file
-- 4. Delete thumbnail files
-- 5. Remove from search index
-- 6. Remove from all post links
-- 3. Delete sidecar file ({binary_path}.meta)
-- 4. Delete all four thumbnail files
-- 5. Delete translation sidecars ({binary_path}.{lang}.meta)
-- 6. Remove from search index
-- 7. Remove from all post links
ensures: FileDeleted(media.file_path)
ensures: FileDeleted(media.sidecar_path)
ensures: FileDeleted(media.thumbnail_path)
ensures: FileDeleted(media.thumbnail_source_path)
ensures: FileDeleted(media.thumbnails.small)
ensures: FileDeleted(media.thumbnails.medium)
ensures: FileDeleted(media.thumbnails.large)
ensures: FileDeleted(media.thumbnails.ai)
ensures: SearchIndexRemoved(media)
ensures:
for p in media.linked_posts:
@@ -221,7 +261,7 @@ rule ValidateMedia {
-- Check for:
-- 1. Missing binary files
-- 2. Missing sidecar files
-- 3. Missing thumbnails
-- 3. Missing thumbnails (all 4 sizes)
-- 4. Corrupted image files
-- 5. Orphan media (not linked to any post)
for m in project.media:
@@ -229,8 +269,14 @@ rule ValidateMedia {
ensures: ValidationIssueReported(m, "missing_binary")
if not file_exists(m.sidecar_path):
ensures: ValidationIssueReported(m, "missing_sidecar")
if not file_exists(m.thumbnail_path):
ensures: ValidationIssueReported(m, "missing_thumbnail")
if not file_exists(m.thumbnails.small):
ensures: ValidationIssueReported(m, "missing_thumbnail_small")
if not file_exists(m.thumbnails.medium):
ensures: ValidationIssueReported(m, "missing_thumbnail_medium")
if not file_exists(m.thumbnails.large):
ensures: ValidationIssueReported(m, "missing_thumbnail_large")
if not file_exists(m.thumbnails.ai):
ensures: ValidationIssueReported(m, "missing_thumbnail_ai")
if not is_valid_image(m.file_path):
ensures: ValidationIssueReported(m, "corrupted")
if not exists (p in Posts where m in p.linked_media):
@@ -238,13 +284,15 @@ rule ValidateMedia {
}
-- ============================================================================
-- MEDIA EXPORT/EXPORT RULES
-- MEDIA SIDECAR FORMAT
-- ============================================================================
invariant MediaSidecarFormat {
-- Sidecar files use YAML frontmatter
-- Same structure as Media entity fields
-- Sidecar files use YAML-like key-value format (hand-built, not gray-matter)
-- Path: {binary_path}.meta
-- Only truthy fields are written (except required fields)
-- Fields: title, alt, caption, author, tags, language, linkedPostIds
-- Note: 'filename' is NOT written to sidecar (it is the binary filename itself)
}
-- ============================================================================
@@ -252,22 +300,10 @@ invariant MediaSidecarFormat {
-- ============================================================================
config {
-- Image optimization settings
max_original_size: Integer = 20971520 -- 20 MB in bytes
max_width: Integer = 4096
max_height: Integer = 4096
strip_exif: Boolean = true -- Strip EXIF from thumbnails
preserve_color_profile: Boolean = false -- Don't preserve ICC profiles
}
rule OptimizeMedia {
when: ImportMediaRequested(source_path, project)
requires: file_size(source_path) <= config.max_original_size
-- Original files are stored as-is (no compression)
-- Only thumbnails are optimized
@guidance
-- Media entity is created by ImportMedia rule
-- This rule documents the size constraint
-- No file size limit on import (TypeScript app has no max_original_size check)
-- Original files are stored as-is (no compression, no resize)
-- Only thumbnails are generated from the original
strip_exif: Boolean = false -- Not explicitly stripped; re-encoding naturally omits it
}
-- ============================================================================
@@ -278,7 +314,6 @@ rule IndexMediaForSearch {
when: SearchIndexUpdated(media: media/Media)
-- Index fields: title, alt, caption, original_name, tags
-- Plus all translation titles, alts, and captions
-- Note: media.translations not available in this context
let all_text = concat(
media.title,
media.alt,