modularizing

This commit is contained in:
Michael Freno
2025-10-31 17:42:30 -04:00
parent 747382614b
commit d947bc04e4
16 changed files with 6785 additions and 6593 deletions

234
flexlove/ImageRenderer.lua Normal file
View File

@@ -0,0 +1,234 @@
--[[
ImageRenderer.lua - Image rendering utilities for FlexLove
Provides object-fit modes and object-position support for image rendering
]]
-- ====================
-- Error Handling Utilities
-- ====================
--- Standardized error message formatter
---@param module string -- Module name (e.g., "Color", "Theme", "Units")
---@param message string -- Error message
---@return string -- Formatted error message
local function formatError(module, message)
return string.format("[FlexLove.%s] %s", module, message)
end
-- ====================
-- ImageRenderer
-- ====================
---@class ImageRenderer
local ImageRenderer = {}
--- Calculate rendering parameters for object-fit modes
--- Returns source and destination rectangles for rendering
---@param imageWidth number -- Natural width of the image
---@param imageHeight number -- Natural height of the image
---@param boundsWidth number -- Width of the bounds to fit within
---@param boundsHeight number -- Height of the bounds to fit within
---@param fitMode string? -- One of: "fill", "contain", "cover", "scale-down", "none" (default: "fill")
---@param objectPosition string? -- Position like "center center", "top left", "50% 50%" (default: "center center")
---@return {sx: number, sy: number, sw: number, sh: number, dx: number, dy: number, dw: number, dh: number, scaleX: number, scaleY: number}
function ImageRenderer.calculateFit(imageWidth, imageHeight, boundsWidth, boundsHeight, fitMode, objectPosition)
fitMode = fitMode or "fill"
objectPosition = objectPosition or "center center"
-- Validate inputs
if imageWidth <= 0 or imageHeight <= 0 or boundsWidth <= 0 or boundsHeight <= 0 then
error(formatError("ImageRenderer", "Dimensions must be positive"))
end
local result = {
sx = 0, -- Source X
sy = 0, -- Source Y
sw = imageWidth, -- Source width
sh = imageHeight, -- Source height
dx = 0, -- Destination X
dy = 0, -- Destination Y
dw = boundsWidth, -- Destination width
dh = boundsHeight, -- Destination height
scaleX = 1, -- Scale factor X
scaleY = 1, -- Scale factor Y
}
-- Calculate based on fit mode
if fitMode == "fill" then
-- Stretch to fill bounds (may distort)
result.scaleX = boundsWidth / imageWidth
result.scaleY = boundsHeight / imageHeight
result.dw = boundsWidth
result.dh = boundsHeight
elseif fitMode == "contain" then
-- Scale to fit within bounds (preserves aspect ratio)
local scale = math.min(boundsWidth / imageWidth, boundsHeight / imageHeight)
result.scaleX = scale
result.scaleY = scale
result.dw = imageWidth * scale
result.dh = imageHeight * scale
-- Apply object-position for letterbox alignment
local posX, posY = ImageRenderer._parsePosition(objectPosition)
result.dx = (boundsWidth - result.dw) * posX
result.dy = (boundsHeight - result.dh) * posY
elseif fitMode == "cover" then
-- Scale to cover bounds (preserves aspect ratio, may crop)
local scale = math.max(boundsWidth / imageWidth, boundsHeight / imageHeight)
result.scaleX = scale
result.scaleY = scale
local scaledWidth = imageWidth * scale
local scaledHeight = imageHeight * scale
-- Apply object-position for crop alignment
local posX, posY = ImageRenderer._parsePosition(objectPosition)
-- Calculate which part of the scaled image to show
local cropX = (scaledWidth - boundsWidth) * posX
local cropY = (scaledHeight - boundsHeight) * posY
-- Convert back to source coordinates
result.sx = cropX / scale
result.sy = cropY / scale
result.sw = boundsWidth / scale
result.sh = boundsHeight / scale
result.dx = 0
result.dy = 0
result.dw = boundsWidth
result.dh = boundsHeight
elseif fitMode == "none" then
-- Use natural size (no scaling)
result.scaleX = 1
result.scaleY = 1
result.dw = imageWidth
result.dh = imageHeight
-- Apply object-position
local posX, posY = ImageRenderer._parsePosition(objectPosition)
result.dx = (boundsWidth - imageWidth) * posX
result.dy = (boundsHeight - imageHeight) * posY
elseif fitMode == "scale-down" then
-- Use none or contain, whichever is smaller
if imageWidth <= boundsWidth and imageHeight <= boundsHeight then
-- Image fits naturally, use "none"
return ImageRenderer.calculateFit(imageWidth, imageHeight, boundsWidth, boundsHeight, "none", objectPosition)
else
-- Image too large, use "contain"
return ImageRenderer.calculateFit(imageWidth, imageHeight, boundsWidth, boundsHeight, "contain", objectPosition)
end
else
error(formatError("ImageRenderer", string.format("Invalid fit mode: '%s'. Must be one of: fill, contain, cover, scale-down, none", tostring(fitMode))))
end
return result
end
--- Parse object-position string into normalized coordinates (0-1)
--- Supports keywords (center, top, bottom, left, right) and percentages
---@param position string -- Position string like "center center", "top left", "50% 50%"
---@return number, number -- Normalized X and Y positions (0-1)
function ImageRenderer._parsePosition(position)
if not position or type(position) ~= "string" then
return 0.5, 0.5 -- Default to center
end
-- Split into X and Y components
local parts = {}
for part in position:gmatch("%S+") do
table.insert(parts, part:lower())
end
-- If only one value, use it for both axes (with special handling)
if #parts == 1 then
local val = parts[1]
if val == "left" or val == "right" then
parts = { val, "center" }
elseif val == "top" or val == "bottom" then
parts = { "center", val }
else
parts = { val, val }
end
elseif #parts == 0 then
return 0.5, 0.5 -- Default to center
end
local function parseValue(val)
-- Handle keywords
if val == "center" then
return 0.5
elseif val == "left" or val == "top" then
return 0
elseif val == "right" or val == "bottom" then
return 1
end
-- Handle percentages
local percent = val:match("^([%d%.]+)%%$")
if percent then
return tonumber(percent) / 100
end
-- Handle plain numbers (treat as percentage)
local num = tonumber(val)
if num then
return num / 100
end
-- Invalid value, default to center
return 0.5
end
local x = parseValue(parts[1])
local y = parseValue(parts[2] or parts[1])
-- Clamp to 0-1 range
x = math.max(0, math.min(1, x))
y = math.max(0, math.min(1, y))
return x, y
end
--- Draw an image with specified object-fit mode
---@param image love.Image -- Image to draw
---@param x number -- X position of bounds
---@param y number -- Y position of bounds
---@param width number -- Width of bounds
---@param height number -- Height of bounds
---@param fitMode string? -- Object-fit mode (default: "fill")
---@param objectPosition string? -- Object-position (default: "center center")
---@param opacity number? -- Opacity 0-1 (default: 1)
function ImageRenderer.draw(image, x, y, width, height, fitMode, objectPosition, opacity)
if not image then
return -- Nothing to draw
end
opacity = opacity or 1
fitMode = fitMode or "fill"
objectPosition = objectPosition or "center center"
local imgWidth, imgHeight = image:getDimensions()
local params = ImageRenderer.calculateFit(imgWidth, imgHeight, width, height, fitMode, objectPosition)
-- Save current color
local r, g, b, a = love.graphics.getColor()
-- Apply opacity
love.graphics.setColor(1, 1, 1, opacity)
-- Draw image
if params.sx ~= 0 or params.sy ~= 0 or params.sw ~= imgWidth or params.sh ~= imgHeight then
-- Need to use a quad for cropping
local quad = love.graphics.newQuad(params.sx, params.sy, params.sw, params.sh, imgWidth, imgHeight)
love.graphics.draw(image, quad, x + params.dx, y + params.dy, 0, params.dw / params.sw, params.dh / params.sh)
else
-- Simple draw with scaling
love.graphics.draw(image, x + params.dx, y + params.dy, 0, params.scaleX, params.scaleY)
end
-- Restore color
love.graphics.setColor(r, g, b, a)
end
return ImageRenderer