WIP: Rendering/UI refactor.

Input line rendering logic in width-constrained situations.
WIP: vi-like input system.
2025-04-02 06:13:27 -05:00 · 2025-03-15 20:42:29 -05:00 · 2025-03-15 07:19:30 -05:00 · 2025-03-04 15:51:02 -06:00 · 2025-03-04 15:49:54 -06:00 · 2025-02-23 08:23:24 -06:00
53 changed files with 2121 additions and 3844 deletions
--- a/.gitignore
+++ b/.gitignore
@ -1,3 +1,4 @@
 *.sw?
-.gradle/
-build/
+/wdiwtlt
+nimble.develop
+nimble.paths
--- a/README.md
+++ b/README.md
@ -1,76 +0,0 @@
-# What Do I Want To Listen To
-
-## A simple, tag-based music library manager.
-
-This project is born out of a frustration I had managing my music library.
-I have found playlists, genres, and other ways of organizing my music too
-restrictive and cumbersome to keep up with. Here are the main features I want
-out of music player, in order of priority:
-
-### Implemented
-
-* Song tagging.
-* Playlists.
-* Bookmarks (temporary song tag marking location in a playlist/album)
-* Read meta-data from ID3.
-
-### TODO
-
-* Web-based interface.
-* Mobile interface (could implement the Subsonic API on the back-end to be able
-  to use pre-made mobile apps)
-* Transcoding on the fly.
-* Read songs from Amazon S3.
-* Stream from Amazon Cloudfront.
-
-I have not found a music manager that gives me all of the above, so I'm
-writing my own.
-
-## Overview
-
-WDIWTLT is currently made up of two subprojects:
-
-* `core`
-* `cli`
-
-### `core`
-
-`core` contains the data layer implementation, built with a lightweight ORM
-layer over JDBC, and common functionality for managing a media library.
-
-### `cli`
-
-`cli` is a command-line interface built using the WDIWTLT core and VLC for
-media playback.
-
-## Install
-
-The current version, 0.1.0, is an alpha version. The CLI client depends on
-[VLC](http://www.videolan.org/vlc/) and expects it to already be installed on
-the system. 
-
-To install the CLI client:
-
-    $ wget http://mvn.jdb-labs.com/repo/com/jdbernard/wdiwtlt-cli/0.1.0/wdiwtlt-cli-0.1.0.zip
-    $ unzip wdiwtlt-cli-0.1.0.zip
-
-The `wdiwtlt-cli` binary is located in `wdiwtlt-cli/bin`, which you can add to
-you `PATH` environment variable.
-
-## Building From Source
-
-The `wdiwtlt` project is written in [Groovy](http://www.groovy-lang.org/) uses
-the [Gradle](http://gradle.org) build tool. If you do not already have a Groovy
-environment installed you can do:
-
-    $ curl -s https://get.sdkman.io | bash
-    $ source "$HOME/.sdkman/bin/sdkman-init.sh"
-    $ yes | sdkman install groovy
-    $ yes | sdkman install gradle
-
-Once you have Groovy and Gradle, the `wdiwtlt` project can be built by invoking
-`gradle assembleDist`.
-
-    $ git clone https://git.jdb-labs.com/jdb/wdiwtlt.git
-    $ cd wdiwtlt
-    $ gradle assembleDist
--- a/build.gradle
+++ b/build.gradle
@ -1,25 +0,0 @@
-allprojects {
-
-    buildscript {
-        repositories {
-            mavenLocal()
-            mavenCentral()
-        }
-
-        dependencies {
-            classpath 'ch.raffael.pegdown-doclet:pegdown-doclet:1.2'
-        }
-    }
-
-    group = 'com.jdbernard'
-    version = '0.1.3'
-
-    repositories {
-        mavenLocal()
-        mavenCentral()
-        maven { url "https://dl.bintray.com/ijabz/maven" }
-        maven { url "http://mvn.jdb-labs.com/repo" }
-    }
-
-    apply plugin: 'maven'
-}
--- a/cli/build.gradle
+++ b/cli/build.gradle
@ -1,23 +0,0 @@
-apply plugin: 'java'
-apply plugin: 'groovy'
-apply plugin: 'ch.raffael.pegdown-doclet'
-apply plugin: 'application'
-
-mainClassName = 'com.jdbernard.wdiwtlt.cli.CommandLineInterface'
-
-dependencies {
-  compile localGroovy()
-  compile 'ch.qos.logback:logback-classic:1.1.3'
-  compile 'ch.qos.logback:logback-core:1.1.3'
-  compile 'org.slf4j:slf4j-api:1.7.14'
-  compile 'com.offbytwo:docopt:0.6.0.20150202'
-  compile 'com.jdbernard:jdb-util:4.4'
-  compile 'jline:jline:2.12'
-  compile project(":wdiwtlt-core")
-  compile 'uk.co.caprica:vlcj:3.10.1'
-
-  testCompile 'junit:junit:4.12'
-
-  runtime 'net.java.dev.jna:jna:4.2.1'
-  runtime 'net.java.dev.jna:jna-platform:4.2.1'
-}
--- a/cli/interface.txt
+++ b/cli/interface.txt
@ -1,76 +0,0 @@
-# Selection
-
-A selection can be a set of Artists, Albums, or songs. It is only ever one of
-these things. So you can select multiple artists, or multiple albums, or
-multiple songs.
-
-select {album, artist, file, playlist, tag} <id | name>
-select {albums, artists, files, playlists, tags} where <criteria>
-select files tagged as <tag>... and not as <tag>..
-select playing {album, artist, file, playlist}
-
-# Play Queue Management
-
-enqueue selection
-enqueue <selection-criteria>
-
-remove selection from queue
-remove <selection-criteria> from queue
-
-play selection
-play bookmark <id | name>
-play <selection-criteria>
-
-clear queue
-
-# Tagging
-
-tag <tag>...
-tag selection as <tag>...
-tag <selection-criteria> as <tag>...
-
-# List
-
-list <selection-criteria>
-list selected {albums, artists, files, playlists, tags}
-list playing {albums, artists, files, playlists, tags}
-
-# Playlist management
-
-create playlist named <playlist name>
-create playlist named <playlist name> from {selection, queue}
-copy playlist <id | name> as <new name>
-rename playlist <old name> to <new name>
-
-add selection to playlist <id | name>
-add <selection-criteria> to playlist <id | name>
-
-remove selection from playlist <id | name>
-remove <selection-criteria> from playlist <id | name>
-
-clear playlist <id | name>
-
-delete playlist <id | name>
-
-# Bookmarking
-
-create bookmark named <name>
-create bookmark named <name> on <playlist id | name> at <media file id | name>
-rename bookmark <old name> to <new name>
-delete bookmark <name | id>
-
-# Transport Functions
-
-play
-pause
-stop
-next <count>
-prev <count>
-ff <amount> <unit>
-rw <amount> <unit>
-jump <media file id | name>
-
-# Misc
-
-scan
-clear
--- a/cli/src/main/groovy/com/jdbernard/wdiwtlt/cli/CliErr.groovy
+++ b/cli/src/main/groovy/com/jdbernard/wdiwtlt/cli/CliErr.groovy
@ -1,9 +0,0 @@
-package com.jdbernard.wdiwtlt.cli
-
-public class CliErr extends Exception {
-    public CliErr(String message) { super(message) }
-    public CliErr(String message, Throwable t) { super(message, t) }
-    public CliErr(Throwable t) { super(t) }
-
-    public static err(String msg) { throw new CliErr(msg) }
-}
--- a/cli/src/main/groovy/com/jdbernard/wdiwtlt/cli/CommandLineInterface.groovy
+++ b/cli/src/main/groovy/com/jdbernard/wdiwtlt/cli/CommandLineInterface.groovy
--- a/cli/src/main/groovy/com/jdbernard/wdiwtlt/cli/ScrollText.groovy
+++ b/cli/src/main/groovy/com/jdbernard/wdiwtlt/cli/ScrollText.groovy
@ -1,69 +0,0 @@
-package com.jdbernard.wdiwtlt.cli
-
-import com.jdbernard.util.AnsiEscapeCodeSequence as ANSI
-
-public class ScrollText {
-
-    public String text = ""
-    public int maxWidth
-
-    private int scrollIdx
-    private String curAnsiPrefix = ""
-
-    public void setMaxWidth(int max) {
-        this.maxWidth = Math.max(max, 1) }
-
-    public void setText(String t) {
-        this.text = t
-        this.scrollIdx = Math.max(0, text.size() - 10) }
-
-    public String getNextScroll() {
-        if (ANSI.strip(text).size() < maxWidth) return text
-
-        scrollIdx = (scrollIdx + 1) % text.size()
-
-        // If we're on the start of an ANSI escape sequence, skip past the end
-        // of it.
-        while (text[scrollIdx] == '\u001b') {
-          def endIdx = text.findIndexOf(scrollIdx) {
-            Character.isLetter(it.charAt(0)) }
-          curAnsiPrefix = text[scrollIdx..endIdx]
-          scrollIdx = (endIdx + 1) % text.size() }
-
-        int toWalk = maxWidth
-        int endIdx = scrollIdx + 1
-        int wrapIdx = 0
-
-        boolean inAnsiSeq = false
-        while (toWalk > 0 && wrapIdx < scrollIdx) {
-            def cur
-            if (endIdx == text.size()) {
-                cur = text[wrapIdx]
-
-                if (cur == '\u001b') inAnsiSeq = true
-
-                if (inAnsiSeq && Character.isLetter(cur.charAt(0)))
-                    inAnsiSeq = false
-
-                if (!inAnsiSeq) toWalk--
-                wrapIdx++ }
-
-            else {
-                cur = text[endIdx]
-
-                if (cur == '\u001b') inAnsiSeq = true
-
-                if (inAnsiSeq && Character.isLetter(cur.charAt(0)))
-                    inAnsiSeq = false
-
-                if (!inAnsiSeq) toWalk--
-                endIdx++ } }
-
-        if (wrapIdx == 0) return curAnsiPrefix + text[scrollIdx..<endIdx]
-        return new StringBuilder()
-            .append(curAnsiPrefix)
-            .append(text[scrollIdx..<endIdx])
-            .append(" ").append(text[0..<wrapIdx]) }
-
-    public String toString() { return getNextScroll() }
-}
--- a/cli/src/main/groovy/com/jdbernard/wdiwtlt/cli/UniversalNoopImplementation.groovy
+++ b/cli/src/main/groovy/com/jdbernard/wdiwtlt/cli/UniversalNoopImplementation.groovy
@ -1,7 +0,0 @@
-package com.jdbernard.wdiwtlt.cli
-
-public class UniversalNoopImplementation {
-  Map methods = [:]
-
-  public def methodMissing(String name, def args) {
-    if (methods[name]) return methods[name](*args) } }
--- a/cli/src/main/resources/logback.groovy
+++ b/cli/src/main/resources/logback.groovy
@ -1,19 +0,0 @@
-import ch.qos.logback.core.*;
-import ch.qos.logback.core.encoder.*;
-import ch.qos.logback.core.read.*;
-import ch.qos.logback.core.rolling.*;
-import ch.qos.logback.core.status.*;
-import ch.qos.logback.classic.net.*;
-import ch.qos.logback.classic.encoder.PatternLayoutEncoder;
-
-
-appender("FILE", FileAppender) {
-  file = "wdiwtlt.log"
-  append = true
-  encoder(PatternLayoutEncoder) {
-    pattern = "%level %logger - %msg%n"
-  }
-}
-
-root(OFF)
-logger("com.jdbernard.wdiwtlt", WARN, ["FILE"])
--- a/config.nims
+++ b/config.nims
@ -0,0 +1,4 @@
+# begin Nimble config (version 2)
+when withDir(thisDir(), system.fileExists("nimble.paths")):
+  include "nimble.paths"
+# end Nimble config
--- a/core/README.md
+++ b/core/README.md
@ -1,10 +0,0 @@
-## Database
-
-Uses [db-migrate][http://db-migrate.readthedocs.org/en/latest/] to manage
-database migrations. Migration scripts live if `src/main/db`.
-
-Database environment configuration lives in `database.json`. 
-
-To initialize a new database do:
-
-    db-migrate -m src/main/db up
--- a/core/build.gradle
+++ b/core/build.gradle
@ -1,30 +0,0 @@
-buildscript {
-    repositories {
-        mavenLocal()
-        mavenCentral()
-    }
-
-    dependencies {
-        classpath 'ch.raffael.pegdown-doclet:pegdown-doclet:1.2'
-    }
-}
-
-apply plugin: 'java'
-apply plugin: 'groovy'
-apply plugin: 'ch.raffael.pegdown-doclet'
-
-dependencies {
-  compile localGroovy()
-  compile 'ch.qos.logback:logback-classic:1.1.3'
-  compile 'ch.qos.logback:logback-core:1.1.3'
-  compile 'org.slf4j:slf4j-api:1.7.14'
-  compile 'com.zaxxer:HikariCP:2.4.3'
-  compile 'net.jthink:jaudiotagger:2.2.3'
-  compile 'commons-codec:commons-codec:1.10'
-  compile 'javax.persistence:persistence-api:1.0.2'
-
-  testCompile 'junit:junit:4.12'
-
-  runtime 'com.h2database:h2:1.4.192'
-  runtime 'org.postgresql:postgresql:9.4.1207.jre7'
-}
--- a/core/database.properties
+++ b/core/database.properties
@ -1,5 +0,0 @@
-dataSourceClassName=org.h2.jdbcx.JdbcDataSource
-dataSource.url=jdbc:h2:/Users/jonathan/programs/wdiwtlt/db;DATABASE_TO_UPPER=FALSE;AUTO_SERVER=TRUE
-dataSource.user=sa
-dataSource.password=
-migrations.dir=src/main/sql/migrations
--- a/core/src/main/groovy/com/jdbernard/wdiwtlt/ConfigWrapper.groovy
+++ b/core/src/main/groovy/com/jdbernard/wdiwtlt/ConfigWrapper.groovy
@ -1,93 +0,0 @@
-package com.jdbernard.wdiwtlt
-
-import com.zaxxer.hikari.HikariConfig
-
-public class ConfigWrapper {
-
-    public static final String LIBRARY_DIR_KEY = "library.dir"
-    public static final String DB_CONFIG_FILE_KEY = "database.config.file"
-
-    Properties configProps
-    Map<String, String> env = System.getenv()
-
-    public ConfigWrapper() {
-        if (env.WDIWTLT_CONFIG_FILE) {
-            if (tryLoadConfigFile(env.WDIWTLT_CONFIG_FILE)) {
-                return } }
-
-        if (env.HOME &&
-            tryLoadConfigFile(new File(env.HOME, ".wdiwtlt.properties"))) return
-
-        String userHome = System.getProperty('user.home')
-        if (userHome &&
-            tryLoadConfigFile(new File(userHome, '.wdiwtlt.properties'))) return
-
-        try {
-            ConfigWrapper.getResourceAsStream("/com/jdbernard/wdiwtlt/db/default.properties")
-                .withStream { tryLoadConfig(is) } }
-        catch (Exception e) {} }
-
-    public ConfigWrapper(File propertiesFile) {
-        this.configPropertiesFile = configPropertiesFile }
-
-    public String getLibraryRootPath() {
-        // 1. Check config properties
-        if (configProps && configProps[LIBRARY_DIR_KEY])
-            return configProps[LIBRARY_DIR_KEY]
-
-        // 2. Check environment variable 
-        if (env.WDIWTLT_LIBRARY_DIR)
-            return env.WDIWTLT_LIBRARY_DIR
-
-        return null }
-
-    public HikariConfig getHikariConfig() {
-        Properties props = new Properties()
-
-        // 1. Check config properties for database config file
-        if (configProps) {
-
-            // 1.1 Look for a reference to a dedicated config file
-            if (configProps[DB_CONFIG_FILE_KEY]) {
-                File cf = new File(configProps[DB_CONFIG_FILE_KEY])
-                if (cf.exists() && cf.isFile())
-                    cf.withInputStream { props.load(it) } }
-
-            // 1.2 Look for prefixed properties in this config file
-            else if (configProps["database.config.dataSourceClassName"]) {
-                props.putAll(configProps
-                    .findAll { it.key.startsWith("database.config.") }
-                    .collectEntries { [it.key[16..-1], it.value] } ) } }
-
-        // 2. Look for environment variables
-        if (!props) {
-            if (env.WDIWTLT_DATASOURCE_CLASSNAME)
-                props["dataSourceClassName"] = env.WDIWTLT_DATABASE_DATASOURCECLASSNAME
-
-            if (env.WDIWTLT_DATASOURCE_PROPERTIES) {
-                props.putAll(env.WDIWTLT_DATASOURCE_PROPERTIES.split(":")
-                    .collect { it.split("=") }
-                    .collectEntries { ["dataSource.${it[0]}", it[1]] }) } }
-
-        // If properties exist, create and return HikariConfig
-        if (props) return new HikariConfig(props)
-        else return null }
-
-    private boolean tryLoadConfigFile(String configFilePath) {
-      if (!configFilePath) return false
-      return tryLoadConfigFile(new File(configFilePath)) }
-
-    private boolean tryLoadConfigFile(File configFile) {
-        if (!configFile.exists() || !configFile.isFile()) return false
-
-        return configFile.withInputStream { tryLoadConfig(it) } }
-
-    private boolean tryLoadConfig(InputStream is) {
-        try {
-            Properties props = new Properties()
-            props.load(is)
-            this.configProps = props }
-        catch (Exception e) { return false }
-
-        return true }
-}
--- a/core/src/main/groovy/com/jdbernard/wdiwtlt/MediaLibrary.groovy
+++ b/core/src/main/groovy/com/jdbernard/wdiwtlt/MediaLibrary.groovy
@ -1,299 +0,0 @@
-package com.jdbernard.wdiwtlt
-
-import com.jdbernard.wdiwtlt.db.DbApi
-import com.jdbernard.wdiwtlt.db.models.*
-
-import java.sql.Timestamp
-import java.util.regex.Pattern
-
-import org.apache.commons.codec.digest.DigestUtils
-import org.jaudiotagger.audio.AudioFile
-import org.jaudiotagger.audio.AudioFileIO
-import org.jaudiotagger.tag.Tag as JATag
-import org.jaudiotagger.tag.FieldKey
-import org.slf4j.Logger
-import org.slf4j.LoggerFactory
-
-import static org.jaudiotagger.tag.FieldKey.*
-
-public class MediaLibrary {
-
-    private static Logger logger = LoggerFactory.getLogger(MediaLibrary)
-
-    private static UPPERCASE_PATTERN = Pattern.compile(/(.)(\p{javaUpperCase})/)
-
-    @Delegate DbApi dbapi
-    private File libraryRoot
-
-    public long autoDeletePeriod = 1000 * 60 * 60 * 24 * 7 // one week
-
-    public MediaLibrary(DbApi dbapi, File rootDir) {
-        logger.debug("Creating a MediaLibrary rooted at: {}",
-            rootDir.canonicalPath)
-
-        this.dbapi = dbapi
-        this.libraryRoot = rootDir }
-
-    public void clean() {
-        dbapi.removeEmptyAlbums()
-        dbapi.removeEmptyArtists()
-        dbapi.removeEmptyPlaylists()
-
-        Timestamp staleTS = new Timestamp(new Date().time - autoDeletePeriod)
-        dbapi.withTransaction {
-            def stalePlaylists = dbapi.getPlaylistsWhere(
-                userCreated: false, lastUsedBefore: staleTS)
-            stalePlaylists.each { dbapi.delete(it) } }
-        dbapi.withTransaction {
-            def staleBookmarks = dbapi.getBookmarksWhere(
-                userCreated: false, lastUsedBefore: staleTS)
-            staleBookmarks.each { dbapi.delete(it) } } }
-
-    public def rescanLibrary() {
-        def results = [ total: 0, ignored: 0, new: 0, present: 0, absent: 0]
-
-        List<MediaFile> missingFiles = dbapi.getMediaFiles()
-        List<MediaFile> foundFiles = []
-
-        Date startDate = new Date()
-        libraryRoot.eachFileRecurse { file ->
-
-            if (!file.isFile()) return
-            def mf = addFile(file)
-
-            results.total++
-            if (!mf) results.ignored++
-            else {
-                foundFiles << mf
-                if (missingFiles.contains(mf)) missingFiles.remove(mf)
-                if (mf.dateAdded > startDate) results.new++ } }
-
-        foundFiles.each { mf ->
-            if (!mf.presentLocally) {
-                mf.presentLocally = true
-                dbapi.update(mf) } }
-
-        missingFiles.each { mf ->
-            if (mf.presentLocally) {
-                mf.presentLocally = false
-                dbapi.update(mf) } }
-
-        results.present = foundFiles.size()
-        results.absent = missingFiles.size()
-
-        return results }
-
-    public MediaFile addFile(File f) {
-        if (!f.exists() || !f.isFile()) {
-            logger.info("Ignoring non-existant file: {}", f.canonicalPath)
-            return null }
-
-        def relPath = getRelativePath(libraryRoot, f)
-        MediaFile found = dbapi.getMediaFileByFilePath(relPath)
-
-        if (found) {
-            logger.info(
-                "Ignoring a media file I already know about: {}", relPath)
-            return found }
-
-        MediaFile mf = new MediaFile()
-        mf.filePath = relPath
-
-        // Hash the file
-        mf.fileHash = f.withInputStream { DigestUtils.md5Hex(it) }
-
-        // Look for an entry for that hash
-        found = dbapi.getMediaFileByFileHash(mf.fileHash)
-
-        if (found) {
-            logger.info('Found a media file by hash in a new location. ' +
-                "I'm updating my relative path to this file:\n\t{}\n\t\t--> {}.",
-                found.filePath, mf.filePath)
-            found.filePath = mf.filePath
-            return dbapi.update(found) }
-
-        // Read in the media's tags
-        def af
-        try { af = AudioFileIO.read(f) }
-        catch (Exception e) {
-            logger.info("Ignoring a file because I can't " +
-                "read the media tag info:\n\t{}\n\t{}",
-                f.canonicalPath, e.localizedMessage)
-            return null }
-
-        def fileTag = af.tag
-
-        mf.name = fileTag?.getFirst(TITLE)?.trim() ?: f.name
-        mf.comment = fileTag?.getAll(COMMENT)?.collect { it.trim() }?.join('\n\n')
-        mf.discNumber = fileTag?.getFirst(DISC_NO) ?: '1'
-        mf.trackNumber = safeToInteger(fileTag?.getFirst(TRACK))
-
-        def folderParts = mf.filePath.split("[\\\\/]")[1..<-1] as LinkedList
-
-        // Find artist and album names (if any)
-        def artistNames = fileTag?.getAll(ARTIST)?.collect { it.trim() }
-        def albumNames = fileTag?.getAll(ALBUM)?.collect { it.trim() }
-
-        if (!artistNames) {
-            mf.metaInfoSource = MediaFile.FILE_LOCATION
-            artistNames = folderParts.size() >= 2 ?  [folderParts[0]] : [] }
-
-        if (!albumNames) {
-            mf.metaInfoSource = MediaFile.FILE_LOCATION
-            albumNames = [folderParts.peekLast()] }
-
-        dbapi.withTransaction {
-            dbapi.create(mf)
-            associateWithArtistsAndAlbums(mf, artistNames, albumNames,
-                safeToInteger(fileTag?.getFirst(YEAR))) }
-
-        return mf
-    }
-
-    public def getByIdOrName(Class modelClass, String input) {
-        def match
-
-        if (safeToUUID(input)) {
-            match = dbapi.getById(modelClass, safeToUUID(input))
-            if (match) match = [match] }
-
-        else {
-            match = dbapi.getByIdLike(modelClass, input)
-            if (!match) match = dbapi.getByName(modelClass, input)
-            if (!match) match = dbapi.getLike(modelClass, ["name"], [input]) }
-        return match }
-
-    public List<Artist> getArtistsByName(String name) {
-        return [dbapi.getArtistByName(name)] ?: dbapi.getArtistsLikeName(name) }
-
-    public List<Album> getAlbumsByName(String name) {
-        return dbapi.getAlbumsWhere(name: name) ?: dbapi.getAlbumsLikeName(name) }
-
-    public List<MediaFile> collectMediaFiles(List<Model> models) {
-        if (!models) return []
-
-        return models.collectMany { m ->
-            if (m.class == MediaFile) return [m]
-            return dbapi.getMediaFilesWhere((idKeyFor(m.class)): m.id) }.findAll() }
-
-    public List<Artist> splitArtist(Artist toSplit, Pattern splitPattern) {
-        return splitArtist(toSplit, pattern.split(toSplit.name)) }
-
-    public List<Artist> splitArtist(Artist toSplit, List<String> newNames) {
-        def albums = dbapi.getAlbumsWhere(artistId: toSplit.id)
-        def mediaFiles = dbapi.getMediaFilesByArtistId(toSplit.id)
-
-        toSplit.name = newNames[0]
-        List<Artist> newArtists = newNames[1..-1].collect { new Artist(it) }
-        newArtists.each { newArtist ->
-            albums.each { dbapi.associate(newArtist, it) }
-            mediaFiles.each { dbapi.associate(newArtist, it) } }
-
-        return [toSplit] + newArtists }
-
-    private void associateWithArtistsAndAlbums(MediaFile mf,
-    List<String> artistNames, List<String> albumNames, Integer albumYear) {
-
-        // Find or create artists.
-        def artists = artistNames.collect { artistName ->
-            Artist artist = dbapi.getArtistByName(artistName)
-            if (!artist) artist = dbapi.create(new Artist(name: artistName))
-            return artist }
-
-        // Associate file with artists.
-        artists.each { dbapi.associate(it, mf) }
-
-        // Find or create albums
-        def albums = albumNames.collect { albumName ->
-            Album album
-
-            // If we know what year the album was released we can use that to
-            // narrow down the list of matching albums.
-            if (albumYear != null) {
-                // Look first to see if we already know about this album
-                // associated with one of the artists for this piece.
-                album = artists.inject(null, { foundAlbum, artist ->
-                    if (foundAlbum) return foundAlbum
-                    def cur = dbapi.getAlbumsWhere(name: albumName,
-                        year: albumYear, artistId: artist.id)
-                    return cur ? cur[0] : null })
-
-                // If we don't have it with one of the artists, see if we have
-                // one that matches the name and year.
-                if (!album) {
-                    def cur = dbapi.getAlbumsWhere(
-                        name: albumName, year: albumYear)
-                    album = cur ? cur[0] : null } }
-
-            else {
-                album = artists.inject(null, { foundAlbum, artist ->
-                    if (foundAlbum) return foundAlbum
-                    def cur = dbapi.getAlbumsWhere(
-                        name: albumName, artistId: artist.id)
-                    return cur ? cur[0] : null })
-
-                if (!album) {
-                    def cur = dbapi.getAlbumsWhere(name: albumName)
-                    album = cur ? cur[0] : null } }
-
-            // We still can't find the album at all. We'll need to create it
-            if (!album)
-                album = dbapi.create(new Album(name: albumName, year: albumYear))
-
-            return album }
-
-        // Associate file with albums
-        albums.each { dbapi.associate(it, mf) }
-
-        // Make sure we have association between all of the artists and albums.
-        artists.each { artist ->
-            def albumsForArtist = dbapi.getAlbumsWhere(artistId: artist.id)
-            def albumsMissing = albums - albumsForArtist
-            albumsMissing.each { album -> dbapi.associate(artist, album) } } }
-
-    /** #### `getRelativePath`
-      * Given a parent path and a child path, assuming the child path is
-      * contained within the parent path, return the relative path from the
-      * parent to the child. */
-    public static String getRelativePath(File parent, File child) {
-        def parentPath = parent.canonicalPath.split("[\\\\/]")
-        def childPath = child.canonicalPath.split("[\\\\/]")
-
-        /// If the parent path is longer it cannot contain the child path and
-        /// we cannot construct a relative path without backtracking.
-        if (parentPath.length > childPath.length) return ""
-
-        /// Compare the parent and child path up until the end of the parent
-        /// path.
-        int b = 0
-        while (b < parentPath.length && parentPath[b] == childPath[b] ) b++
-
-        /// If we stopped before reaching the end of the parent path it must be
-        /// that the paths do not match. The parent cannot contain the child and
-        /// we cannot build a relative path without backtracking.
-        if (b != parentPath.length) return ""
-        return (['.'] + childPath[b..<childPath.length]).join('/') }
-
-    public static UUID safeToUUID(def val) {
-        if (val == null) return null
-        try { return UUID.fromString(val as String) }
-        catch (IllegalArgumentException iae) { return null } }
-
-    public static Integer safeToInteger(def val) {
-        if (val == null) return null
-        try { return val.trim() as Integer }
-        catch (NumberFormatException nfe) { return null } }
-
-    public static String uncapitalize(String s) {
-        if (s == null) return null;
-        if (s.length() < 2) return s.toLowerCase();
-        return s[0].toLowerCase() + s[1..-1] }
-
-    public static String toEnglish(Class c) {
-        return UPPERCASE_PATTERN.matcher(c.simpleName).
-            replaceAll(/$1 $2/).toLowerCase() }
-
-    public static String idKeyFor(Class c) {
-        return uncapitalize(c.simpleName) + 'Id'; }
-
-}
--- a/core/src/main/groovy/com/jdbernard/wdiwtlt/db/DbApi.groovy
+++ b/core/src/main/groovy/com/jdbernard/wdiwtlt/db/DbApi.groovy
--- a/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Album.java
+++ b/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Album.java
@ -1,11 +0,0 @@
-package com.jdbernard.wdiwtlt.db.models;
-
-public class Album extends Model {
-    public String name;
-    public Integer trackTotal;
-    public Integer year;
-
-    public String toString() {
-        if (year != null) return name + " (" + year + ")";
-        else return name; }
-}
--- a/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Artist.java
+++ b/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Artist.java
@ -1,7 +0,0 @@
-package com.jdbernard.wdiwtlt.db.models;
-
-public class Artist extends Model {
-    public String name;
-
-    public String toString() { return name; }
-}
--- a/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Bookmark.java
+++ b/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Bookmark.java
@ -1,18 +0,0 @@
-package com.jdbernard.wdiwtlt.db.models;
-
-import java.sql.Timestamp;
-import java.util.Date;
-import java.util.UUID;
-
-public class Bookmark extends Model {
-    public String name;
-    public UUID playlistId;
-    public UUID mediaFileId;
-    public int playIndex = 0;
-    public int playTimeMs = 0;
-    public boolean userCreated;
-    public Timestamp createdAt = new Timestamp(new Date().getTime());
-    public Timestamp lastUsed = new Timestamp(new Date().getTime());
-
-    public String toString() { return name; }
-}
--- a/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Image.java
+++ b/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Image.java
@ -1,7 +0,0 @@
-package com.jdbernard.wdiwtlt.db.models;
-
-public class Image extends Model {
-    public String url;
-
-    public String toString() { return url; }
-}
--- a/core/src/main/java/com/jdbernard/wdiwtlt/db/models/MediaFile.java
+++ b/core/src/main/java/com/jdbernard/wdiwtlt/db/models/MediaFile.java
@ -1,25 +0,0 @@
-package com.jdbernard.wdiwtlt.db.models;
-
-import java.util.Date;
-import java.sql.Timestamp;
-
-public class MediaFile extends Model {
-    public static final String TAG_INFO = "tag info";
-    public static final String FILE_LOCATION = "file location";
-
-    public String name;
-    public String discNumber;
-    public Integer trackNumber;
-    public int playCount = 0;
-    public String filePath;
-    public String fileHash;
-    public String metaInfoSource = TAG_INFO;
-    public Timestamp dateAdded = new Timestamp(new Date().getTime());
-    public Timestamp lastPlayed;
-    public boolean presentLocally = true;
-    public String comment;
-
-    public String toString() {
-      if (trackNumber != null) return trackNumber + " - " + name;
-      else return name; }
-}
--- a/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Model.java
+++ b/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Model.java
@ -1,26 +0,0 @@
-package com.jdbernard.wdiwtlt.db.models;
-
-import java.util.UUID;
-import javax.persistence.Entity;
-
-@Entity
-public class Model implements Comparable<Model>, Cloneable {
-    public UUID id;
-
-    public boolean equals(Object thatObj) {
-        if (thatObj == null) return false;
-        if (this.getClass() != thatObj.getClass()) return false;
-
-        Model that = (Model) thatObj;
-
-        if (this.id == null || that.id == null) return false;
-        return this.id.equals(that.id); }
-
-    public int compareTo(Model that) {
-        if (this.id == null) return -1;
-        if (this.getClass() != that.getClass()) {
-            return this.getClass().getSimpleName().compareTo(
-                that.getClass().getSimpleName()); }
-
-        return this.id.compareTo(that.id); }
-}
--- a/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Playlist.java
+++ b/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Playlist.java
@ -1,18 +0,0 @@
-package com.jdbernard.wdiwtlt.db.models;
-
-import java.sql.Timestamp;
-import java.util.Date;
-import java.util.UUID;
-
-public class Playlist extends Model {
-    public boolean userCreated = true;
-    public String name;
-    public int mediaFileCount = 0;
-    public UUID copiedFromId = null;
-    public Timestamp createdAt = new Timestamp(new Date().getTime());
-    public Timestamp lastUsed = new Timestamp(new Date().getTime());
-
-    public String toString() {
-        if (userCreated) return name;
-        return name + " (auto)"; }
-}
--- a/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Tag.java
+++ b/core/src/main/java/com/jdbernard/wdiwtlt/db/models/Tag.java
@ -1,11 +0,0 @@
-package com.jdbernard.wdiwtlt.db.models;
-
-public class Tag extends Model {
-    public String name;
-    public String description = "";
-
-    public String toString() {
-      if (description != null && description.length() > 0)
-        return name + " (" + description + ")";
-      else return name; }
-}
--- a/core/src/main/sql/migrations/20151209054632-initial-schema-up.sql
+++ b/core/src/main/sql/migrations/20151209054632-initial-schema-up.sql
@ -1,112 +0,0 @@
-CREATE TABLE artists (
-  id UUID PRIMARY KEY,
-  name VARCHAR UNIQUE NOT NULL
-);
-
-CREATE INDEX artists_name_idx ON artists(name);
-
-CREATE TABLE albums (
-  id UUID PRIMARY KEY,
-  name VARCHAR NOT NULL,
-  year INTEGER,
-  track_total INTEGER
-);
-
-CREATE INDEX albums_name_idx ON albums(name);
-
-CREATE TABLE media_files (
-  id UUID PRIMARY KEY,
-  name VARCHAR NOT NULL,
-  disc_number VARCHAR NOT NULL DEFAULT '1',
-  track_number INTEGER,
-  play_count INTEGER NOT NULL DEFAULT 0,
-  file_path VARCHAR NOT NULL,
-  file_hash VARCHAR NOT NULL,
-  meta_info_source VARCHAR NOT NULL, -- 'tag' or 'filesystem'
-  date_added TIMESTAMP NOT NULL DEFAULT NOW(),
-  last_played TIMESTAMP,
-  present_locally BOOLEAN NOT NULL DEFAULT TRUE,
-  comment VARCHAR DEFAULT ''
-);
-
-CREATE INDEX media_files_name_idx ON media_files(name);
-
-CREATE TABLE artists_media_files (
-  artist_id UUID NOT NULL REFERENCES artists(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  media_file_id UUID NOT NULL REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  PRIMARY KEY (artist_id, media_file_id)
-);
-
-CREATE TABLE albums_media_files (
-  album_id UUID NOT NULL REFERENCES albums(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  media_file_id UUID NOT NULL REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  PRIMARY KEY (album_id, media_file_id)
-);
-
-CREATE TABLE tags (
-  id UUID PRIMARY KEY,
-  name VARCHAR UNIQUE NOT NULL,
-  description VARCHAR NOT NULL DEFAULT ''
-);
-
-CREATE TABLE playlists (
-  id UUID PRIMARY KEY,
-  user_created BOOLEAN NOT NULL DEFAULT FALSE,
-  name VARCHAR NOT NULL,
-  media_file_count INTEGER NOT NULL DEFAULT 0,
-  copied_from_id UUID DEFAULT NULL,
-  created_at TIMESTAMP NOT NULL DEFAULT NOW(),
-  last_used TIMESTAMP NOT NULL DEFAULT NOW()
-);
-
-CREATE TABLE playlists_media_files (
-  playlist_id UUID NOT NULL REFERENCES playlists(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  media_file_id UUID NOT NULL REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  position INTEGER NOT NULL DEFAULT 0,
-  PRIMARY KEY (playlist_id, media_file_id, position),
-  UNIQUE (playlist_id, position)
-);
-
-CREATE TABLE bookmarks (
-  id UUID PRIMARY KEY,
-  name VARCHAR,
-  user_created BOOLEAN NOT NULL DEFAULT FALSE,
-  playlist_id UUID NOT NULL REFERENCES playlists(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  media_file_id UUID NOT NULL REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  play_index INTEGER NOT NULL,
-  play_time_ms INTEGER NOT NULL DEFAULT 0,
-  created_at TIMESTAMP NOT NULL DEFAULT NOW(),
-  last_used TIMESTAMP NOT NULL DEFAULT NOW()
-);
-
-CREATE INDEX bookmarks_playlist_id_idx ON bookmarks (playlist_id);
-CREATE INDEX bookmarks_media_file_id_idx ON bookmarks (media_file_id);
-
-CREATE TABLE media_files_tags (
-  media_file_id UUID REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  tag_id UUID REFERENCES tags(id) ON DELETE CASCADE ON UPDATE CASCADE,
-  PRIMARY KEY (media_file_id, tag_id)
-);
-
-CREATE TABLE images (
-  id UUID PRIMARY KEY,
-  url VARCHAR
-);
-
-CREATE TABLE artists_images (
-  artist_id UUID REFERENCES artists (id) ON DELETE CASCADE ON UPDATE CASCADE,
-  image_id UUID REFERENCES images (id) ON DELETE CASCADE ON UPDATE CASCADE,
-  PRIMARY KEY (artist_id, image_id)
-);
-
-CREATE TABLE albums_images (
-  album_id UUID REFERENCES albums (id) ON DELETE CASCADE ON UPDATE CASCADE,
-  image_id UUID REFERENCES images (id) ON DELETE CASCADE ON UPDATE CASCADE,
-  PRIMARY KEY (album_id, image_id)
-);
-
-CREATE TABLE artists_albums (
-  artist_id UUID NOT NULL REFERENCES artists (id) ON DELETE CASCADE ON UPDATE CASCADE,
-  album_id UUID NOT NULL REFERENCES albums (id) ON DELETE CASCADE ON UPDATE CASCADE,
-  PRIMARY KEY (artist_id, album_id)
-);
--- a/core/test.groovy
+++ b/core/test.groovy
@ -1,35 +0,0 @@
-import groovy.grape.Grape
-import groovy.sql.Sql
-Grape.grab(group: 'com.zaxxer', module: 'HikariCP', version: '2.4.3')
-Grape.grab(group: 'org.postgresql', module: 'postgresql', version: '9.4.1207.jre7')
-Grape.grab(group: 'commons-codec', module: 'commons-codec', version: '1.10')
-
-import com.jdbernard.wdiwtlt.MediaLibrary
-import com.jdbernard.wdiwtlt.db.ORM
-import com.jdbernard.wdiwtlt.db.models.*
-import com.zaxxer.hikari.*
-import org.jaudiotagger.audio.*
-import org.jaudiotagger.tag.*
-import ch.qos.logback.classic.Level
-import ch.qos.logback.classic.Logger
-import org.slf4j.LoggerFactory
-
-rootLogger = (Logger) LoggerFactory.getLogger(Logger.ROOT_LOGGER_NAME)
-rootLogger.level = Level.INFO
-
-// myLogger = (Logger) LoggerFactory.getLogger("com.jdbernard.wdiwtlt")
-// myLogger.level = Level.DEBUG
-
-config = new Properties()
-config.dataSourceClassName = "org.postgresql.ds.PGSimpleDataSource"
-config."dataSource.databaseName" = "wdiwtlt"
-config."dataSource.user" = "jonathan"
-config."dataSource.password" = ""
-
-hcfg = new HikariConfig(config)
-hds = new HikariDataSource(hcfg)
-
-db = new ORM(hds)
-
-musicDir = new File('/Users/jonathan/Music')
-library = new MediaLibrary(db, musicDir)
--- a/rest-server/build.gradle
+++ b/rest-server/build.gradle
@ -1,24 +0,0 @@
-apply plugin: 'groovy'
-apply plugin: 'war'
-apply plugin: "jetty"
-
-dependencies {
-    compile localGroovy()
-    compile 'ch.qos.logback:logback-classic:1.1.3'
-    compile 'ch.qos.logback:logback-core:1.1.3'
-    compile 'org.slf4j:slf4j-api:1.7.14'
-    compile 'com.impossibl.pgjdbc-ng:pgjdbc-ng:0.3'
-    compile 'com.lambdaworks:scrypt:1.4.0'
-    compile 'com.zaxxer:HikariCP-java6:2.3.2'
-    compile 'javax:javaee-api:7.0'
-    compile 'javax.ws.rs:javax.ws.rs-api:2.0.1'
-    compile project(":wdiwtlt-core")
-
-    runtime 'com.fasterxml.jackson.jaxrs:jackson-jaxrs-json-provider:2.3.2'
-    runtime 'org.glassfish.jersey.containers:jersey-container-servlet:2.16'
-    runtime 'org.glassfish.jersey.media:jersey-media-json-jackson:2.16'
-
-    providedCompile 'javax.servlet:javax.servlet-api:3.1.0'
-
-    testCompile 'junit:junit:4.12'
-    testRuntime 'com.h2database:h2:1.4.186' }
--- a/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/rest/AllowCors.java
+++ b/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/rest/AllowCors.java
@ -1,12 +0,0 @@
-package com.jdbernard.nlsongs.rest;
-
-import javax.ws.rs.NameBinding;
-import java.lang.annotation.ElementType;
-import java.lang.annotation.RetentionPolicy;
-import java.lang.annotation.Retention;
-import java.lang.annotation.Target;
-
-@NameBinding
-@Target({ ElementType.TYPE, ElementType.METHOD })
-@Retention(value = RetentionPolicy.RUNTIME)
-public @interface AllowCors {}
--- a/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/rest/CorsResponseFilter.java
+++ b/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/rest/CorsResponseFilter.java
@ -1,30 +0,0 @@
-package com.jdbernard.nlsongs.rest;
-
-import javax.annotation.Priority;
-import javax.ws.rs.Priorities;
-import javax.ws.rs.core.MultivaluedMap;
-import javax.ws.rs.container.ContainerRequestContext;
-import javax.ws.rs.container.ContainerResponseContext;
-import javax.ws.rs.container.ContainerResponseFilter;
-import javax.ws.rs.ext.Provider;
-
-@Provider @AllowCors @Priority(Priorities.HEADER_DECORATOR)
-public class CorsResponseFilter implements ContainerResponseFilter {
-
-    @Override
-    public void filter(ContainerRequestContext reqCtx,
-                       ContainerResponseContext respCtx) {
-
-        MultivaluedMap<String, Object> headers = respCtx.getHeaders();
-
-        headers.add("Access-Control-Allow-Origin",
-            reqCtx.getHeaderString("Origin"));
-
-        headers.add("Access-Control-Allow-Methods",
-            "GET, POST, PUT, DELETE, OPTIONS");
-
-        headers.add("Access-Control-Allow-Headers",
-            reqCtx.getHeaderString("Access-Control-Request-Headers"));
-
-    }
-}
--- a/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/rest/PingResource.java
+++ b/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/rest/PingResource.java
@ -1,12 +0,0 @@
-package com.jdbernard.nlsongs.rest;
-
-import javax.ws.rs.GET;
-import javax.ws.rs.Path;
-import javax.ws.rs.Produces;
-
-@Path("v1/ping") @AllowCors
-public class PingResource {
-
-    @GET
-    @Produces("text/plain")
-    public String ping() { return "pong"; } }
--- a/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/servlet/WdiwtltContext.groovy
+++ b/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/servlet/WdiwtltContext.groovy
@ -1,8 +0,0 @@
-package com.jdbernard.nlsongs.servlet
-
-import com.jdbernard.wdiwtlt.db.DbApi
-
-public class WdiwtltContext {
-
-    public static DbApi dbapi
-}
--- a/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/servlet/WdiwtltContextListener.groovy
+++ b/rest-server/src/main/groovy/com/jdbernard/wdiwtlt/servlet/WdiwtltContextListener.groovy
@ -1,43 +0,0 @@
-package com.jdbernard.nlsongs.servlet
-
-import javax.servlet.ServletContext
-import javax.servlet.ServletContextEvent
-import javax.servlet.ServletContextListener
-
-import com.jdbernard.wdiwtlt.db.DbApi
-
-import com.zaxxer.hikari.HikariConfig
-import com.zaxxer.hikari.HikariDataSource
-
-public final class WdiwtltContextListener implements ServletContextListener {
-
-    public void contextInitialized(ServletContextEvent event) {
-        def context = event.servletContext
-
-        // Load the context configuration.
-        Properties props = new Properties()
-        WdiwtltContextListener.getResourceAsStream(
-            context.getInitParameter('context.config.file')).withStream { is ->
-                props.load(is) }
-
-        // Create the pooled data source
-        HikariConfig hcfg = new HikariConfig(
-            context.getInitParameter('datasource.config.file'))
-
-        HikariDataSource hds = new HikariDataSource(hcfg)
-
-        // Create the NLSonsDB instance.
-        DbApi dbapi = new DbApi(hds)
-
-        context.setAttribute('dbapi', dbapi)
-        WdiwtltContext.dbapi = dbapi }
-
-    public void contextDestroyed(ServletContextEvent event) {
-        def context = event.servletContext
-
-        // Shutdown the DB API instance (it will shut down the data source).
-        DbApi dbapi = context.getAttribute('dbapi')
-        if (dbapi) dbapi.shutdown()
-
-        context.removeAttribute('dbapi') }
-}
--- a/settings.gradle
+++ b/settings.gradle
@ -1,4 +0,0 @@
-include 'core', 'cli', 'rest-server'
-
-project(":core").name ="wdiwtlt-core"
-project(":cli").name = "wdiwtlt-cli"
--- a/src/main/nim/wdiwtlt.nim
+++ b/src/main/nim/wdiwtlt.nim
@ -0,0 +1,334 @@
+import std/[json, options, os, paths, sequtils, strutils, terminal, times,
+            unicode]
+import cliutils, docopt, namespaced_logging, mpv
+
+import wdiwtlt/[cliconstants, commandbuffer, logging, medialibrary,
+                messagebuffer, models, nonblockingio, scrolltext, usage]
+
+type
+  CliMode {.pure.} = enum Direct, Command
+
+  MessageBuffers* = ref object
+    playing: MessageBuffer[ScrollText]
+    status: MessageBuffer[ScrollText]
+    longform: MessageBuffer[string]
+    input: MessageBuffer[string]
+
+  History = ref object
+    entries: seq[string]
+    idx: int
+
+  CliContext = ref object
+    cfg: CombinedConfig
+    cmd: CommandBuffer
+    curMediaFile: Option[MediaFile]
+    lib: MediaLibrary
+    log: Option[Logger]
+    logs: seq[LogMessage]
+    logThreshold: Level
+    mode: CliMode
+    msgs: MessageBuffers
+    mpv: ptr handle
+    msgDuration: Duration
+    stop*: bool
+    width*: int
+    height*: int
+
+var ctx {.threadvar.}: CliContext
+
+const FRAME_DUR_MS = 200
+
+proc initMpv(): ptr handle =
+  result = mpv.create()
+
+  if result.isNil:
+    raise newException(ValueError, "failed creating mpv context")
+
+  # Enable default key bindings, so the user can actually interact with
+  # the player (and e.g. close the window).
+  result.set_option("terminal", "no")
+  result.set_option("input-default-bindings", "yes")
+  result.set_option("input-vo-keyboard", "yes")
+  result.set_option("osc", false)
+  #result.set_property("volume", "50.0")
+  discard result.request_log_messages("no")
+
+  # Done setting up options.
+  check_error result.initialize()
+
+
+proc cleanup() =
+  terminate_destroy(ctx.mpv)
+  setNonBlockingTty(false)
+  echo ""
+
+
+#[
+proc exitHandlerHook() {.noconv.} =
+  # This is called when the user presses Ctrl+C
+  # or when the program exits normally.
+  terminate_destroy(ctx.mpv)
+  echo ""
+  quit(QuitSuccess)
+]#
+
+proc initContext(args: Table[string, Value]): CliContext =
+  var wdiwtltCfgFilename: string
+
+  let log = getLogger("wdiwtlt")
+
+  let cfgLocations =
+    if args["--config"]: @[$args["--config"]]
+    elif existsEnv("XDG_CONFIG_HOME"): @[getEnv("XDG_CONFIG_HOME") / "wdiwtlt.json"]
+    else: @[$getEnv("HOME") / ".config" / "wdiwtlt.json" ]
+
+  try: wdiwtltCfgFilename = findConfigFile(".wdiwtlt.json", cfgLocations)
+  except ValueError:
+    log.error "could not find wdiwtlt config file: " & wdiwtltCfgFilename
+    if not args["--config"]:
+      wdiwtltCfgFilename = cfgLocations[0]
+
+      var cfgFile: File
+      try:
+        cfgFile = open(wdiwtltCfgFilename, fmWrite)
+        cfgFile.write((%*{
+          "libraryRoot":
+            if args["--library-root"]:
+              $args["--library-root"]
+            else:
+              getEnv("XDG_MUSIC_DIR", getEnv("HOME") / "music"),
+          "dbPath": getEnv("XDG_DATA_HOME", (getEnv("HOME") / ".local/share")) / "wdiwtlt/db.sqlite",
+        }).pretty)
+        log.info "created sample config file at " & wdiwtltCfgFilename
+      except CatchableError:
+        log.error "could not write default .wdiwtlt.json to " &
+          wdiwtltCfgFilename
+      finally: close(cfgFile)
+
+    quit(QuitFailure)
+
+  log.debug("loading config from '$#'" % [wdiwtltCfgFilename])
+  let cfg = initCombinedConfig(wdiwtltCfgFilename, args)
+
+  let width = min(terminalWidth(), cfg.getVal("maxWidth", "80").parseInt)
+  let height = min(terminalHeight(), cfg.getVal("maxHeight", "60").parseInt)
+
+  let msgs = MessageBuffers(
+    playing: initMessageBuffer[ScrollText](
+      dims = BufferDimensions(width: width, height: 1, x: 0, y: 0),
+      initialMsg = "Nothing playing"),
+    status: initMessageBuffer[ScrollText](
+      dims = BufferDimensions(width: width, height: 1, x: 0, y: 1),
+      initialMsg = ""),
+    input: initMessageBuffer[string](
+      dims = BufferDimensions(width: width, height: 1, x: 0, y: 2),
+      initialMsg = ""),
+    longform: initMessageBuffer[string](
+      dims = BufferDimensions(width: width, height: height - 3, x: 0, y: 4),
+      initialMsg = ""))
+
+  result = CliContext(
+    cfg: cfg,
+    cmd: initCommandBuffer(msgs.input),
+    curMediaFile: none(MediaFile),
+    lib: initMediaLibrary(
+      rootDir = getVal(cfg, "libraryRoot").Path,
+      dbPath = getVal(cfg, "dbPath").Path),
+    log: log,
+    mode: CliMode.Direct,
+    msgs: msgs,
+    mpv: initMpv(),
+    msgDuration: initDuration(seconds = 5, milliseconds = 0),
+    stop: false,
+    width: width,
+    height: height)
+
+  if logService.isSome:
+    let customLogAppender = initCustomLogAppender(doLogMessage =
+      proc (msg: LogMessage): void = ctx.logs.add(msg))
+
+    logService.get.clearAppenders()
+    logService.get.addAppender(customLogAppender)
+
+
+proc nextTick(ctx: var CliContext) =
+  ctx.msgs.playing.nextTick
+  ctx.msgs.status.nextTick
+  ctx.msgs.longform.nextTick
+
+
+proc setMode(ctx: var CliContext, mode: CliMode) =
+  ctx.mode = mode
+  case mode
+
+  of CliMode.Direct:
+    hideCursor()
+
+  of CliMode.Command:
+    showCursor()
+    ctx.cmd.setMode(EditMode.Insert)
+
+
+proc statusMsg(ctx: var CliContext, msg: string) =
+  ctx.msgs.status.showMsg(msg, some(ctx.msgDuration))
+
+
+proc longMsg[T: string or seq[string]](ctx: var CliContext, msg: T) =
+  ctx.msgs.longform.showMsg(msg, some(ctx.msgDuration))
+
+
+proc viewLogLoops(ctx: var CliContext) =
+  let logLines = ctx.logs
+    .filterIt(it.level > ctx.logThreshold)
+    .mapIt("$# - [$#]: $#" % [$it.level, $it.scope, it.message])
+    .join("\p")
+
+  ctx.msgs.longform.showMsg(logLines, none[Duration]())
+
+  var viewLogs = true
+  while viewLogs:
+    case getKeyAsync()
+    of Key.Q: viewLogs = false
+    of Key.K, Key.Up: ctx.msgs.longform.scrollLines(-1)
+    of Key.J, Key.Down: ctx.msgs.longform.scrollLines(1)
+    of Key.H, Key.Left: ctx.msgs.longform.scrollColumns(-1)
+    of Key.L, Key.Right: ctx.msgs.longform.scrollColumns(1)
+    of Key.PageUp: ctx.msgs.longform.scrollPages(-1)
+    of Key.PageDown: ctx.msgs.longform.scrollPages(1)
+    else: discard
+
+    sleep(FRAME_DUR_MS)
+
+  ctx.msgs.longform.clear()
+
+proc scanLibrary(ctx: var CliContext) =
+  ctx.statusMsg("Scanning media library...")
+  let counts = ctx.lib.rescanLibrary()
+
+  var scanResults = @[
+    "─".repeat(ctx.width),
+    "Scan complete:",
+    "\t$# files total." % [ $counts.total ]
+  ]
+
+  if counts.new > 0:
+    scanResults.add("\t$# new files added." % [ $counts.new ])
+
+  if counts.ignored > 0:
+    scanResults.add("\t$# files ignored." % [ $counts.ignored ])
+
+  if counts.absent > 0:
+    scanResults.add(
+      "\t$# files in the database but not stored locally." %
+      [ $counts.absent ])
+
+  ctx.longMsg(scanResults.join("\p"))
+
+proc processLogs(ctx: var CliContext, args: seq[string]) =
+  case args[0]
+  of "view": ctx.viewLogLoops()
+  of "clear": ctx.logs = @[]
+  of "set-threshold":
+    if args.len < 2:
+      ctx.longMsg(("missing $# argument to $#" %
+        [ "log-threshold", "logs set-threshold" ]) &
+        usageOf(@["logs", "set-threshold"]))
+
+    # TODO
+  else: ctx.statusMsg("Unrecognized logs command: " & args[0])
+
+
+proc processCommand(ctx: var CliContext, command: string) =
+  let parts = command.strip.split(' ')
+
+  case parts[0]
+  of "scan":                scanLibrary(ctx)
+  of "q", "quit", "exit":   ctx.stop = true
+  of "logs":                ctx.processLogs(parts[1..^1])
+  else: statusMsg(ctx, "Unrecognized command: " & command)
+
+
+proc handleKey(ctx: var CliContext, key: Key) =
+  if ctx.mode == CliMode.Direct:
+    case key
+    of Key.Q:     ctx.stop = true
+    of Key.Colon, Key.I: ctx.setMode(CliMode.Command)
+    else: discard
+
+  elif ctx.mode == CliMode.Command:
+    case key
+    of Key.Enter:
+      let command = $ctx.cmd
+      ctx.cmd.clear
+      ctx.cmd.mode = EditMode.Insert
+      processCommand(ctx, command)
+    of Key.Backspace: ctx.cmd.handleInput(Key.Backspace)
+    of Key.Escape:
+      if ctx.cmd.mode == EditMode.Command:
+        ctx.setMode(CliMode.Direct)
+        ctx.cmd.clear
+      else: ctx.cmd.handleInput(Key.Escape)
+    else: ctx.cmd.handleInput(key)
+
+
+proc mainLoop(ctx: var CliContext) =
+
+  hideCursor()
+  var frame = 0
+
+  while not ctx.stop:
+
+    # This sleep below will be the primary driver of the frame rate of the
+    # application (examples below):
+    #
+    # | frames/sec | sec/frame | ms/frame |
+    # |------------|-----------|----------|
+    # |    5.00    |   0.200   |    200   |
+    # |   10.00    |   0.100   |    100   |
+    # |   12.50    |   0.080   |     80   |
+    # |   20.00    |   0.050   |     50   |
+    # |   25.00    |   0.040   |     40   |
+    # |   31.25    |   0.032   |     32   |
+    # |   50.00    |   0.020   |     20   |
+    # |   62.50    |   0.016   |     16   |
+    #
+    # Previously, when we were using illwill and rendering every frame we
+    # targeted a faster FPS to allow for a responsive feeling to the command
+    # input, etc. In this case, we still had a "logic update" period that was
+    # lower than our rendering FPS. Specifically, we were targeting 60fps
+    # rendering speed, but 5fps for logic updates (scrolling text, expiring
+    # messages, etc.).
+    #
+    # Now that we are directly rendering (components of the UI render updates
+    # themselves immediately) we don't need to render at 60fps to have a
+    # responsive UI. Now we really only care about the frequency of logic
+    # updates, so we're targeting 5fps and triggering our logic updates every
+    # frame.
+    handleKey(ctx, getKeyAsync())
+    sleep(FRAME_DUR_MS)
+    nextTick(ctx) # tick on every frame
+    frame = (frame + 1) mod 1000 # use a max frame count to prevent overflow
+
+
+when isMainModule:
+
+  discard enableLogging()
+  try:
+    let args = docopt(USAGE, version=VERSION)
+
+    ctx = initContext(args)
+    setNonBlockingTty(true)
+
+    mainLoop(ctx)
+
+  except Exception:
+    let ex = getCurrentException()
+    getLogger("wdiwtlt").error(%*{
+      "msg": "Unhandled exception",
+      "error": ex.msg,
+      "trace": ex.getStackTrace()
+    })
+    cleanup()
+    quit(QuitFailure)
+  finally:
+    cleanup()
--- a/src/main/nim/wdiwtlt/cliconstants.nim
+++ b/src/main/nim/wdiwtlt/cliconstants.nim
@ -0,0 +1,18 @@
+const VERSION* = "0.2.0"
+
+const USAGE* = "wdiwtlt v" & VERSION & """
+
+Usage:
+  wdiwtlt [options]
+  wdiwtlt --version
+  wdwtlt --help
+
+Options:
+  -c, --config <cfgPath>          Path to the configuration file. Defaults to:
+                                  ~/.config/wdiwtlt/config.json
+
+  -L, --library-root <root-dir>   The path to a local media library directory.
+
+  -D, --database-config <dbCfg>   Path to the database file (SQLite3). Defaults to:
+                                  ~/.config/wdiwtlt/db.sqlite
+"""
--- a/src/main/nim/wdiwtlt/commandbuffer.nim
+++ b/src/main/nim/wdiwtlt/commandbuffer.nim
@ -0,0 +1,247 @@
+import std/[options, strutils]
+
+import ./[messagebuffer, nonblockingio, terminal, util]
+
+type
+  EditMode* {.pure.} = enum Command, Insert, Overwrite, Visual
+
+  Command = ref object
+    buffer: string
+    idx: int
+    selectionStartIdx: Option[int]
+
+  CommandBuffer* = ref object
+    history: seq[Command]
+    mb: MessageBuffer[string]
+    idx: int
+    visibleIdx: int # first visible character in the cmd buffer
+    mode*: EditMode
+
+func initCommandBuffer*(mb: MessageBuffer[string]): CommandBuffer =
+  result = CommandBuffer(
+    history: @[Command(
+      buffer: "",
+      idx: 0,
+      selectionStartIdx: none(int))],
+    idx: 0,
+    mb: mb,
+    mode: Insert)
+
+proc cur(cb: CommandBuffer): Command = cb.history[cb.idx]
+proc cur(cb: var CommandBuffer): Command = cb.history[cb.idx]
+
+proc prev*(cb: var CommandBuffer) =
+  cb.idx = clamp( cb.idx - 1, 0, cb.history.len - 1)
+
+proc next*(cb: var CommandBuffer) =
+  cb.idx = clamp( cb.idx + 1, 0, cb.history.len - 1)
+
+proc insert*(cb: var CommandBuffer, s: string) =
+  var cmd = cb.history[cb.idx]
+  cmd.buffer = cmd.buffer[0..<cmd.idx] & s & cmd.buffer[cmd.idx..^1]
+  cmd.idx += s.len
+
+proc overwrite*(cb: var CommandBuffer, s: string) =
+  var cmd = cb.history[cb.idx]
+  if cmd.idx + s.len > cmd.buffer.len:
+    cmd.buffer = cmd.buffer[0..<cmd.idx] & s
+  else:
+    cmd.buffer = cmd.buffer[0..<cmd.idx] & s & cmd.buffer[cmd.idx + s.len..^1]
+  cmd.idx += s.len
+
+proc write*(cb: var CommandBuffer, s: string) =
+  if cb.mode == EditMode.Insert: cb.insert(s)
+  elif cb.mode == EditMode.Overwrite: cb.overwrite(s)
+
+proc delete*(cb: var CommandBuffer, backspace = true) =
+  var cmd = cb.history[cb.idx]
+  if backspace and cmd.idx > 0:
+    cmd.buffer = cmd.buffer[0..<cmd.idx - 1] & cmd.buffer[cmd.idx..^1]
+    cmd.idx -= 1
+  elif not backspace and cmd.idx < cmd.buffer.len:
+    cmd.buffer = cmd.buffer[0..<cmd.idx] & cmd.buffer[cmd.idx + 1..^1]
+
+proc clear*(cb: var CommandBuffer) =
+  cb.history.add(Command(
+    buffer: "",
+    idx: 0,
+    selectionStartIdx: none(int)))
+  cb.idx = cb.history.len - 1
+  cb.visibleIdx = 0
+
+proc left*(cb: var CommandBuffer) =
+  cb.cur.idx = clamp(cb.cur.idx - 1, 0, cb.cur.buffer.len)
+
+proc right*(cb: var CommandBuffer) =
+  cb.cur.idx = clamp(cb.cur.idx + 1, 0, cb.cur.buffer.len)
+
+proc toHome*(cb: var CommandBuffer) = cb.cur.idx = 0
+proc toEnd*(cb: var CommandBuffer) = cb.cur.idx = cb.cur.buffer.len
+
+proc backWord*(cb: var CommandBuffer) =
+  var cmd = cb.history[cb.idx]
+  while cmd.idx > 0 and cmd.buffer[cmd.idx - 1] == ' ': cmd.idx -= 1
+  while cmd.idx > 0 and cmd.buffer[cmd.idx - 1] != ' ': cmd.idx -= 1
+
+proc forwardWord*(cb: var CommandBuffer) =
+  var cmd = cb.history[cb.idx]
+  while cmd.idx < cmd.buffer.len and cmd.buffer[cmd.idx] != ' ': cmd.idx += 1
+  while cmd.idx < cmd.buffer.len and cmd.buffer[cmd.idx] == ' ': cmd.idx += 1
+
+proc endOfWord*(cb: var CommandBuffer) =
+  var cmd = cb.history[cb.idx]
+  cmd.idx = clamp(cmd.idx + 1, 0, cmd.buffer.len - 1)
+  while cmd.idx < cmd.buffer.len and cmd.buffer[cmd.idx] == ' ': cmd.idx += 1
+  while cmd.idx < cmd.buffer.len - 1 and
+        cmd.buffer[cmd.idx + 1] != ' ':
+    cmd.idx += 1
+
+func toChar(k: Key): char = cast[char](ord(k))
+
+
+proc setMode*(cb: var CommandBuffer, mode: EditMode) =
+  cb.mode = mode
+  case mode
+  of EditMode.Insert:     setCursorType(ctBarBlink)
+  of EditMode.Overwrite:  setCursorType(ctBlock)
+  else:                   setCursorType(ctBlockBlink)
+
+
+func `$`*(cb: CommandBuffer): string =
+  if cb.cur.selectionStartIdx.isNone:
+    return cb.cur.buffer
+  else:
+    let selIdx = cb.cur.selectionStartIdx.get
+    let curIdx = cb.cur.idx
+
+    if selIdx < curIdx:
+      return cb.cur.buffer[0..<selIdx] &
+        invert(cb.cur.buffer[selIdx..<curIdx]) &
+        cb.cur.buffer[curIdx..^1]
+    else:
+      return cb.cur.buffer[0..<curIdx] &
+        invert(cb.cur.buffer[curIdx..<selIdx]) &
+        cb.cur.buffer[selIdx..^1]
+
+func idx*(cb: CommandBuffer): int = cb.cur.idx
+
+
+proc render*(cb: var CommandBuffer) =
+  write(cb.mb.dims.x, cb.mb.dims.y, ' '.repeat(cb.mb.dims.width))
+
+  let cmd = cb.cur
+  var inputLine = cmd.buffer
+  let spaceAvailable = cb.mb.dims.width - 2
+
+  # Check to see that the line entered fits entirely within the space
+  # available to display it. We need two characters of space: one for the
+  # colon and one for the cursor. If we don't have enough space, we need to
+  # decide which portion of the line to display.
+  if len(cmd.buffer) >= spaceAvailable:
+    # If the user has moved the cursor left of our visible index, follow them
+    # back
+    if cmd.idx < cb.visibleIdx: cb.visibleIdx = cmd.idx
+
+    # If the user has moved right, push our index to the right to follow them
+    elif (cmd.idx - cb.visibleIdx) > spaceAvailable:
+      cb.visibleIdx = cmd.idx - spaceAvailable
+
+    # If the user has started backspacing (we are showing less that we
+    # could), follow them back
+    elif (len(cmd.buffer) - cb.visibleIdx) < spaceAvailable:
+      cb.visibleIdx = len(cmd.buffer) - spaceAvailable
+
+    # Show the portion of the line starting at the visible index and ending
+    # either with the end of the line or the end of the visible space
+    # (whichever comes first). We need to check both because the user may
+    # delete characters from the end of the line, which would make the
+    # portion of the line we're currently showing shorter than the visible
+    # space (don't try to access an index past the end of the string).
+    inputLine = cmd.buffer[
+      cb.visibleIdx ..<
+      min(cb.visibleIdx + spaceAvailable, len(cmd.buffer))]
+
+  elif cb.visibleIdx > 0:
+    # We know the line fits within the space available, but we're still
+    # showing a subset of the line (probably because it was longer and the
+    # user backspaced). Let's just reset and show the whole line now that we
+    # can.
+    cb.visibleIdx = 0
+
+  # TODO: implement VISUAL mode selection highlighting
+  cb.mb.showMsg(":" & inputLine)
+  setCursorPosition(
+    cb.mb.dims.x + (cmd.idx - cb.visibleIdx) + 1,
+    cb.mb.dims.y)
+
+proc handleInput*(cb: var CommandBuffer, key: Key) =
+  case cb.mode
+  of EditMode.Insert, EditMode.Overwrite:
+    case key
+    of Key.Escape: cb.setMode(EditMode.Command)
+    of Key.Backspace: cb.delete()
+    of Key.Delete: cb.delete(backspace = false)
+    of Key.Left: cb.left()
+    of Key.Right: cb.right()
+    of Key.Up: cb.prev()
+    of Key.Down: cb.next()
+    of Key.Home: cb.toHome()
+    of Key.End: cb.toEnd()
+    of Key.CtrlH: cb.backWord()
+    of Key.CtrlL: cb.forwardWord()
+    elif key >= Key.Space and key <= Key.Tilde: cb.write($toChar(key))
+    else: discard
+
+  of EditMode.Command:
+    case key
+    of Key.Backspace: cb.delete()
+    of Key.Delete,  Key.X: cb.delete(backspace = false)
+    of Key.Left,    Key.H: cb.left()
+    of Key.Right,   Key.L: cb.right()
+    of Key.Up,      Key.J: cb.prev()
+    of Key.Down,    Key.K: cb.next()
+    of Key.Home,    Key.Zero:   cb.toHome()
+    of Key.End,     Key.Dollar: cb.toEnd()
+    of Key.B: cb.backWord()
+    of Key.W: cb.forwardWord()
+    of Key.E: cb.endOfWord()
+    of Key.I: cb.setMode(EditMode.Insert)
+    of Key.A:
+      cb.cur.idx = clamp(cb.cur.idx + 1, 0, cb.cur.buffer.len)
+      cb.setMode(EditMode.Insert)
+    of Key.ShiftA:
+      cb.cur.idx = cb.cur.buffer.len
+      cb.setMode(EditMode.Insert)
+    of Key.ShiftI:
+      cb.cur.idx = 0
+      cb.setMode(EditMode.Insert)
+    of Key.ShiftR:
+      cb.setMode(EditMode.Overwrite)
+    of Key.V:
+      cb.setMode(EditMode.Visual)
+      cb.cur.selectionStartIdx = some(cb.cur.idx)
+    else: discard
+
+  of EditMode.Visual:
+    case key
+    of Key.Escape:
+      cb.setMode(EditMode.Command)
+      cb.cur.selectionStartIdx = none(int)
+    of Key.Backspace: cb.left()
+    of Key.Left,   Key.H: cb.left()
+    of Key.Right,  Key.L: cb.right()
+    of Key.Up,     Key.J:
+      cb.cur.selectionStartIdx = none[int]()
+      cb.prev()
+    of Key.Down,   Key.K:
+      cb.cur.selectionStartIdx = none[int]()
+    of Key.Home,   Key.Zero: cb.toHome()
+    of Key.End,    Key.Dollar: cb.toEnd()
+    of Key.B: cb.backWord()
+    of Key.W: cb.forwardWord()
+    of Key.V:
+      cb.setMode(EditMode.Command)
+      cb.cur.selectionStartIdx = none(int)
+    else: discard
+
+  render(cb)
--- a/src/main/nim/wdiwtlt/db.nim
+++ b/src/main/nim/wdiwtlt/db.nim
@ -0,0 +1,213 @@
+import std/[dirs, files, json, options, paths, sequtils, strutils, times]
+import db_connector/db_sqlite
+import waterpark/sqlite
+import fiber_orm, timeutils, uuids
+
+export fiber_orm.NotFoundError
+export fiber_orm.PaginationParams
+export fiber_orm.PagedRecords
+export fiber_orm.enableDbLogging
+export sqlite.close
+
+import ./[logging, models]
+
+const schemaDDL = readFile("src/main/sql/media-library-schema.sql")
+  .split(";")
+  .filterIt(not isEmptyOrWhitespace(it))
+
+type
+  WdiwtltDb* = SqlitePool
+
+func toJsonHook*(dt: DateTime): JsonNode = %(dt.formatIso8601)
+proc fromJsonHook*(dt: var DateTime, n: JsonNode): void =
+  dt = n.getStr.parseIso8601
+
+
+proc createTables*(db: WdiwtltDb) =
+  ## Create the database tables if they don't exist.
+
+  db.withConnection conn:
+    let rows = conn.getRow(sql("SELECT name FROM sqlite_master WHERE type = 'table' AND name = 'artists'"))
+
+    if rows.len == 0 or isEmptyOrWhitespace(rows[0]):
+      for ddl in schemaDDL:
+        try: conn.exec(sql(ddl))
+        except DbError:
+          let ex = getCurrentException()
+          getLogger("wdiwtlt/db").error(%*{
+            "msg": "Failed creating initial schema.",
+            "ddl": ddl,
+            "error": ex.msg})
+          raise ex
+
+proc initDB*(dbPath: Path): SqlitePool =
+  # Create the DB file if it doesn't exist.
+  if not fileExists(dbPath):
+    let dbDir = dbPath.splitFile().dir
+    if not dirExists(dbDir):
+      createDir(dbDir)
+
+  result = newSqlitePool(10, dbPath.string)
+
+  # Create the database tables if they don't exist.
+  result.createTables()
+
+
+generateProcsForModels(WdiwtltDb,
+  [Artist, Album, MediaFile, Tag, Playlist, Bookmark, Image])
+
+generateJoinTableProcs(WdiwtltDb, Artist, Album, "artists_albums")
+generateJoinTableProcs(WdiwtltDb, Artist, MediaFile, "artists_media_files")
+generateJoinTableProcs(WdiwtltDb, Album, MediaFile, "albums_media_files")
+generateJoinTableProcs(WdiwtltDb, Playlist, MediaFile, "playlists_media_files")
+generateJoinTableProcs(WdiwtltDb, MediaFile, Tag, "media_files_tags")
+generateJoinTableProcs(WdiwtltDb, Artist, Image, "artists_images")
+generateJoinTableProcs(WdiwtltDb, Album, Image, "albums_images")
+
+generateLookup(WdiwtltDb, Artist, @["name"])
+generateLookup(WdiwtltDb, Album, @["artistId"])
+generateLookup(WdiwtltDb, Album, @["name"])
+generateLookup(WdiwtltDb, Album, @["name", "year"])
+generateLookup(WdiwtltDb, MediaFile, @["filePath"])
+generateLookup(WdiwtltDb, MediaFile, @["fileHash"])
+
+proc findAlbumsByArtistAndName*(
+    db: WdiwtltDb,
+    artistId: UUID,
+    albumName: string,
+    page: Option[PaginationParams] = none[PaginationParams]()): PagedRecords[Album] =
+  ## Find albums by artist and name.
+
+  db.withConnection conn:
+    var query = """
+      SELECT """ & columnNamesForModel(Album).join(",") & """
+      FROM albums al
+      JOIN artists_albums aral ON al.id = aral.album_id
+      WHERE aral.artist_id = ? AND al.name = ? """
+
+    let countStmt = """
+      SELECT COUNT(*)
+      FROM albums al
+      JOIN artists_albums aral ON al.id = aral.album_id
+      WHERE aral.artist_id = ? AND al.name = ? """
+
+    if page.isSome: query &= getPagingClause(page.get)
+
+    logQuery("findAlbumsByArtistAndName", query,
+      [("artist_id", $artistId), ("album.name", albumName)])
+
+    let values = @[$artistId, albumName]
+    let records = conn.getAllRows(sql(query), values).mapIt(rowToModel(Album, it))
+
+    result = PagedRecords[Album](
+      pagination: page,
+      records: records,
+      totalRecords:
+        if page.isNone: records.len
+        else: conn.getRow(sql(countStmt), values)[0].parseInt)
+
+
+proc findAlbumsByArtistAndName*(
+    db: WdiwtltDb,
+    artist: Artist,
+    albumName: string,
+    page: Option[PaginationParams] = none[PaginationParams]()): PagedRecords[Album] =
+  ## Find albums by artist and name.
+  return db.findAlbumsByArtistAndName(artist.id, albumName, page)
+
+
+proc findAlbumsByArtistNameAndYear*(
+    db: WdiwtltDb,
+    artistId: UUID,
+    albumName: string,
+    year: int,
+    page: Option[PaginationParams] = none[PaginationParams]()): PagedRecords[Album] =
+  ## Find albums by artist name and year.
+
+  db.withConnection conn:
+    var query = """
+      SELECT """ & columnNamesForModel(Album).join(",") & """
+      FROM albums al
+      JOIN artists_albums aral ON al.id = aral.album_id
+      WHERE aral.artist_id = ? AND al.name = ? AND al.year = ? """
+
+    let countStmt = """
+      SELECT COUNT(*)
+      FROM albums al
+      JOIN artists_albums aral ON al.id = aral.album_id
+      WHERE aral.artist_id = ? AND al.name = ? AND al.year = ? """
+
+    if page.isSome: query &= getPagingClause(page.get)
+
+    logQuery("findAlbumsByArtistNameAndYear", query,
+      [("artist_id", $artistId), ("album.name", albumName), ("album.year", $year)])
+
+    let values = @[$artistId, albumName, $year]
+    let records = conn.getAllRows(sql(query), values).mapIt(rowToModel(Album, it))
+
+    result = PagedRecords[Album](
+      pagination: page,
+      records: records,
+      totalRecords:
+        if page.isNone: records.len
+        else: conn.getRow(sql(countStmt), values)[0].parseInt)
+
+proc findAlbumsByArtistNameAndYear*(
+    db: WdiwtltDb,
+    artist: Artist,
+    albumName: string,
+    year: int,
+    page: Option[PaginationParams] = none[PaginationParams]()): PagedRecords[Album] =
+  ## Find albums by artist name and year.
+  return db.findAlbumsByArtistNameAndYear(artist.id, albumName, year, page)
+
+proc removeEmptyAlbums*(db: WdiwtltDb) =
+  ## Remove albums that have no media files.
+
+  db.withConnection conn:
+    let query = """
+        DELETE FROM albums WHERE id IN (
+            SELECT DISTINCT al.id
+            FROM albums al LEFT OUTER JOIN albums_media_files almf ON
+                al.id = almf.album_id
+            WHERE almf.album_id IS NULL)"""
+
+    getLogger("wdiwtlt/db").debug(%*{
+      "msg": "Deleting empty albums.",
+      "query": query})
+
+    conn.exec(sql(query))
+
+proc removeEmptyArtists*(db: WdiwtltDb) =
+  ## Remove artists that have no albums.
+
+  db.withConnection conn:
+    let query = """
+        DELETE FROM artists WHERE id IN (
+            SELECT DISTINCT ar.id
+            FROM artists ar LEFT OUTER JOIN artists_albums aral ON
+                ar.id = aral.artist_id
+            WHERE aral.artist_id IS NULL)"""
+
+    getLogger("wdiwtlt/db").debug(%*{
+      "msg": "Deleting empty artists.",
+      "query": query})
+
+    conn.exec(sql(query))
+
+proc removeEmptyPlaylists*(db: WdiwtltDb) =
+  ## Remove playlists that have no media files.
+
+  db.withConnection conn:
+    let query = """
+        DELETE FROM playlists WHERE id IN (
+            SELECT DISTINCT pl.id
+            FROM playlists pl LEFT OUTER JOIN playlists_media_files plmf ON
+                pl.id = plmf.playlist_id
+            WHERE plmf.playlist_id IS NULL)"""
+
+    getLogger("wdiwtlt/db").debug(%*{
+      "msg": "Deleting empty playlists.",
+      "query": query})
+
+    conn.exec(sql(query))
--- a/src/main/nim/wdiwtlt/incremental_md5.nim
+++ b/src/main/nim/wdiwtlt/incremental_md5.nim
@ -0,0 +1,39 @@
+import std/[paths, streams]
+import checksums/md5
+
+proc fileToMD5*(filename: Path) : string =
+
+  const blockSize: int = 8192 # read files in 8KB chunnks
+  var
+    c: MD5Context
+    d: MD5Digest
+    fs: FileStream
+    buffer: string
+
+  #read chunk of file, calling update until all bytes have been read
+  try:
+    fs = newFileStream(open($filename))
+
+    md5Init(c)
+    buffer = fs.readStr(blockSize)
+
+    while buffer.len > 0:
+      md5Update(c, buffer.cstring, buffer.len)
+      buffer = fs.readStr(blockSize)
+
+    md5Final(c, d)
+
+  finally:
+    if fs != nil:
+      close(fs)
+
+  result = $d
+
+when isMainModule:
+
+  if paramCount() > 0:
+    let arguments = commandLineParams()
+    echo("MD5: ", fileToMD5(arguments[0]))
+  else:
+    echo("Must pass filename.")
+    quit(-1)
--- a/src/main/nim/wdiwtlt/logging.nim
+++ b/src/main/nim/wdiwtlt/logging.nim
@ -0,0 +1,46 @@
+import std/[options, os, strutils, unicode]
+import namespaced_logging, zero_functional
+
+from fiber_orm import enableDbLogging
+
+export namespaced_logging
+
+var logService* {.threadvar.}: Option[LogService]
+
+proc enableLogging*(svc: LogService = initLogService(), debug = false): LogService =
+  if not (svc.cfg.appenders --> exists(it of ConsoleLogAppender)):
+    svc.addAppender(initConsoleLogAppender(threshold = lvlAll))
+
+  enableDbLogging(svc)
+  logService = some(svc)
+  result = svc
+
+
+proc configureLoggingThresholds*(debug = false) =
+  if not logService.isSome: return
+  let logSvc = logService.get
+
+  if debug:
+    logSvc.cfg.rootLevel = Level.lvlDebug
+    logSvc.cfg.loggers.add([
+      LoggerConfig(name: "wdiwtlt", threshold: some(Level.lvlDebug)),
+      LoggerConfig(name: "fiber_orm", threshold: some(Level.lvlDebug)),
+    ])
+  else: logSvc.cfg.rootLevel = Level.lvlInfo
+
+  logSvc.reloadThreadState()
+
+
+proc enableLoggingByEnvVar*(envVar = "DEBUG"): void =
+  if not logService.isSome: discard enableLogging()
+  let val = getEnv(envVar, "false").toLower()
+
+  configureLoggingThresholds(
+   "true".startsWith(val) or
+   "yes".startsWith(val) or
+   "on".startsWith(val) or
+    val == "1")
+
+proc getLogger*(name: string): Option[Logger] =
+  if logService.isSome: return some(logService.get.getLogger(name))
+  else: return none[Logger]()
--- a/src/main/nim/wdiwtlt/medialibrary.nim
+++ b/src/main/nim/wdiwtlt/medialibrary.nim
@ -0,0 +1,247 @@
+import std/[dirs, files, json, options, paths, sequtils, strutils, sugar, times]
+import namespaced_logging, timeutils, uuids
+
+import ./[db, incremental_md5, logging, models, taglib]
+
+type
+  MediaLibrary* = ref object
+    db: WdiwtltDb
+    root: Path
+
+proc initMediaLibrary*(rootDir: Path, dbPath: Path): MediaLibrary =
+  ## Initialize the media library.
+
+  let db = initDB(dbPath)
+
+  return MediaLibrary(db: db, root: rootDir)
+
+proc clean*(lib: MediaLibrary) =
+  removeEmptyAlbums(lib.db)
+  removeEmptyArtists(lib.db)
+  removeEmptyPlaylists(lib.db)
+
+  let expirationDate = now() - weeks(1)
+
+  let expiredPlaylists = lib.db.findPlaylistsWhere(
+    "user_created = false AND last_used_at < ?",
+    [expirationDate.formatIso8601])
+
+  let expiredBookmarks = lib.db.findBookmarksWhere(
+    "user_created = false AND last_used_at < ?",
+    [expirationDate.formatIso8601])
+
+  for p in expiredPlaylists.records: discard lib.db.deletePlaylist(p.id)
+  for b in expiredBookmarks.records: discard lib.db.deleteBookmark(b.id)
+
+proc findOrCreateArtist*(lib: MediaLibrary, name: string): Artist =
+  ## Find or create an artist record.
+
+  let existing = lib.db.findArtistsByName(name)
+  if existing.records.len > 0:
+    return existing.records[0]
+
+  getLogger("wdiwtlt/medialibrary").debug(%*{
+    "msg": "Creating missing Artist record.",
+    "method": "findOrCreateArtist",
+    "artistName": name })
+  return lib.db.createArtist(Artist(id: genUuid(), name: name))
+
+
+proc findOrCreateAlbum*(
+    lib: MediaLibrary,
+    name: string,
+    artists: seq[Artist],
+    year: Option[int]): Album =
+
+  var foundAlbum = none[Album]()
+
+  # If we know the year the album was released, we can use that to narrow
+  # down the list of matching albums
+  if year.isSome:
+    # First look only at albums already associated with this artist
+    for artist in artists:
+      let existing = lib.db.findAlbumsByArtistNameAndYear(artist, name, year.get)
+      if existing.records.len > 0:
+        foundAlbum = some(existing.records[0])
+        break
+
+    # Then look at all albums with that name and year
+    if foundAlbum.isNone:
+      let existing = lib.db.findAlbumsByNameAndYear(name, $year.get)
+      if existing.records.len > 0: foundAlbum = some(existing.records[0])
+
+  if foundAlbum.isNone:
+    # If we don't know the year, or if there are no albums with that year,
+    # look at all albums by this artist
+    for artist in artists:
+      let existing = lib.db.findAlbumsByArtistAndName(artist, name)
+      if existing.records.len > 0:
+        foundAlbum = some(existing.records[0])
+        break
+
+  # If we still don't have a match, look at all albums only by name
+  if foundAlbum.isNone:
+    let existing = lib.db.findAlbumsByName(name)
+    if existing.records.len > 0: foundAlbum = some(existing.records[0])
+
+  # If we still don't have a match, create a new album
+  if foundAlbum.isNone:
+    getLogger("wdiwtlt/medialibrary").debug(%*{
+      "msg": "Creating missing Album record.",
+      "method": "findOrCreateAlbum",
+      "albumName": name,
+      "year": year.map((y) => $y).get("") })
+
+    foundAlbum = some(lib.db.createAlbum(Album(
+      id: genUuid(),
+      name: name,
+      year: year,
+      trackTotal: 0)))
+
+  return foundAlbum.get
+
+proc associateWithAristsAndAlbums*(
+    lib: MediaLibrary,
+    mf: MediaFile,
+    artistNames: seq[string],
+    albumNames: seq[string],
+    year: Option[int]) =
+
+  # Find or create artist and album records.
+  let artists = artistNames.mapIt(lib.findOrCreateArtist(it))
+
+  # Find or create album records.
+  let albums = albumNames.mapIt(lib.findOrCreateAlbum(it, artists, year))
+
+  # Associate this file with the artists and albums.
+  for artist in artists: lib.db.associate(artist, mf)
+  for album in albums: lib.db.associate(album, mf)
+
+  # Make sure we have associations between all artists and albums.
+  for artist in artists:
+    let albumsForArtist = lib.db.findAlbumsByArtistId($artist.id)
+    for album in albums:
+      if not albumsForArtist.records.anyIt(album.id == it.id):
+        lib.db.associate(artist, album)
+
+proc addFile*(lib: MediaLibrary, relativeFilePath: Path): MediaFile =
+
+  let fullPath = lib.root / relativeFilePath
+  let pathParts = splitFile(fullPath)
+
+  if not fullPath.fileExists:
+    raise newException(ValueError, "File does not exist: " & $fullPath)
+
+  let existing = lib.db.findMediaFilesByFilePath($relativeFilePath)
+  if existing.records.len > 0:
+    getLogger("wdiwtlt/medialibrary").debug(%*{
+      "msg": "File already exists in library, using existing record.",
+      "method": "addFile",
+      "relativeFilePath": $relativeFilePath,
+      "existingId": existing.records[0].id})
+    return existing.records[0]
+
+  var newMf = MediaFile(
+    id: genUuid(),
+    playCount: 0,
+    filePath: $relativeFilePath,
+    dateAdded: now(),
+    lastPlayed: none[DateTime](),
+    presentLocally: true,
+    comment: "")
+
+  newMf.fileHash = fileToMD5(fullPath)
+
+  let existingForHash = lib.db.findMediaFilesByFileHash(newMf.fileHash)
+  if existingForHash.records.len > 0:
+    getLogger("wdiwtlt/medialibrary").debug(%*{
+      "msg": "File with the same hash already exists in library, using existing record.",
+      "method": "addFile",
+      "relativeFilePath": $relativeFilePath,
+      "existingId": existingForHash.records[0].id})
+    return existingForHash.records[0]
+
+  var tagFile = openTags(fullPath)
+  defer: close(tagFile)
+
+  newMf.name =
+    if tagFile.title.strip().len == 0: $pathParts.name
+    else: tagFile.title.strip()
+
+  newMf.comment = tagFile.comment.strip()
+  newMf.discNumber = tagFile.discNumber
+  newMf.trackNumber =
+    if tagFile.track == 0: none[int]()
+    else: some(tagFile.track)
+
+  let folderParts = ($pathParts.dir).split("/")
+
+  var artistNames = newSeq[string]()
+  if tagFile.artist.strip().len > 0:
+    newMf.metaInfoSource = "tag info"
+    artistNames = tagFile.artist.split({'/', ';'}).mapIt(strip(it))
+  elif tagFile.albumArtist.strip().len > 0:
+    newMf.metaInfoSource = "tag info"
+    artistNames = tagFile.albumArtist.split({'/', ';'}).mapIt(strip(it))
+  elif folderParts.len > 1:
+    newMf.metaInfoSource = "folder"
+    artistNames = @[folderParts[folderParts.len - 2].strip()]
+  else:
+    newMf.metaInfoSource = "unknown"
+    artistNames = @[]
+
+  var albumNames = newSeq[string]()
+  if tagFile.album.strip().len > 0:
+    newMf.metaInfoSource = "tag info"
+    albumNames = tagFile.album.split({'/', ';'}).mapIt(strip(it))
+  elif folderParts.len > 0:
+    newMf.metaInfoSource = "folder"
+    albumNames = @[folderParts[folderParts.len - 1].strip()]
+  else:
+    newMf.metaInfoSource = "unknown"
+    albumNames = @[]
+
+  result = lib.db.createMediaFile(newMf)
+
+  let albumYear =
+    if tagFile.year == 0: none[int]()
+    else: some(tagFile.year)
+
+  lib.associateWithAristsAndAlbums(result, artistNames, albumNames, albumYear)
+
+proc rescanLibrary*(lib: MediaLibrary):
+    tuple[ total, ignored, new, present, absent: int] =
+
+  var missingFiles = lib.db.getAllMediaFiles().records
+  var foundFiles = newSeq[MediaFile]()
+
+  let startDate = now()
+  for p in lib.root.walkDir(
+      relative = true,
+      checkDir = true,
+      skipSpecial = true):
+
+    if not p.path.fileExists: continue
+    let mf = lib.addFile(p.path)
+    result.total += 1
+
+    foundFiles.add(mf)
+    missingFiles = missingFiles.filterIt(it.id != mf.id)
+
+    if mf.dateAdded > startDate:
+      result.new += 1
+
+  for mf in foundFiles:
+    if not mf.presentLocally:
+      var updated = mf
+      updated.presentLocally = true
+      discard lib.db.updateMediaFile(updated)
+
+  for mf in missingFiles:
+    if mf.presentLocally:
+      var updated = mf
+      updated.presentLocally = false
+      discard lib.db.updateMediaFile(updated)
+
+  result.present = foundFiles.len
+  result.absent = missingFiles.len
--- a/src/main/nim/wdiwtlt/messagebuffer.nim
+++ b/src/main/nim/wdiwtlt/messagebuffer.nim
@ -0,0 +1,167 @@
+import std/[options, sequtils, strutils, times, wordwrap]
+
+import ./[scrolltext, terminal, util]
+
+type
+  BufferDimensions* = object
+    width*, height*, x*, y*: int
+
+  MessageBuffer*[T: ScrollText or string] = ref object
+    content: T
+    dims*: BufferDimensions
+    dismissAfter: Option[DateTime]
+    firstColumnIdx: int
+    firstLineIdx: int
+
+
+proc getLinePortion(line: string, start: int, width: int): string =
+  let cleanLine = stripFormatting(line)
+
+  if start == 0 and len(cleanLine) <= width:
+      result = line & ' '.repeat(width - len(cleanLine))
+  else:
+    result = ansiAwareSubstring(line, start, width)
+
+
+proc render[T: ScrollText or string](mb: var MessageBuffer[T]) =
+  hideCursor()
+  when T is ScrollText: write(mb.dims.x, mb.dims.y, $mb.content)
+  when T is string:
+    let renderedContent = splitLines(mb.content)
+      .mapIt(wrapWords(
+        s = it,
+        maxLineWidth = mb.dims.width,
+        newLine = "\p").alignLeft(mb.dims.width))
+      .mapIt(getLinePortion(it, mb.firstColumnIdx, mb.dims.width))
+      .join("\p")
+    let lastLineIdx = min(mb.firstLineIdx + mb.dims.height, len(renderedContent))
+    write(mb.dims.x, mb.dims.y, renderedContent[mb.firstLineIdx..<lastLineIdx])
+  showCursor()
+
+
+func variableMsgDuration[T: ScrollText or string](
+    mb: MessageBuffer[T],
+    initDur: Duration): Duration =
+  when T is ScrollText:
+    # text moves at 5 characters per second (192ms per character)
+    result = initDur + initDuration(milliseconds = len($mb.content) * 192)
+  when T is string:
+    result = initDur + initDuration(milliseconds = len(splitLines(mb.content)) * 250)
+
+
+proc initMessageBuffer*[T: ScrollText or string](
+    dims: BufferDimensions,
+    initialMsg = ""): MessageBuffer[T] =
+
+  when T is ScrollText:
+    result = MessageBuffer[T](
+      content: initScrollText(initialMsg, max(dims.width, 1)),
+      dims: BufferDimensions(
+        width: max(dims.width, 1),
+        height: 1,
+        x: dims.x,
+        y: dims.y),
+      dismissAfter: none[DateTime](),
+      firstColumnIdx: 0,
+      firstLineIdx: 0)
+
+  when T is string:
+    result = MessageBuffer[T](
+      content: initialMsg,
+      dims: BufferDimensions(
+        width: max(dims.width, 1),
+        height: dims.height,
+        x: dims.x,
+        y: dims.y),
+      dismissAfter: none[DateTime](),
+      firstColumnIdx: 0,
+      firstLineIdx: 0)
+
+  render(result)
+
+
+proc showMsg*[T: ScrollText or string](
+    mb: var MessageBuffer[T],
+    msg: string,
+    dur: Option[Duration] = none[Duration]()) =
+
+  when T is ScrollText: mb.content.text = msg
+  when T is string: mb.content = msg
+
+  if dur.isSome:
+    mb.dismissAfter = some(now() + mb.variableMsgDuration(dur.get))
+  else:
+    mb.dismissAfter = none[DateTime]()
+
+  render(mb)
+
+
+proc clear*[T: ScrollText or string](mb: var MessageBuffer[T]) =
+  when T is ScrollText:
+    let clearText = ' '.repeat(mb.dims.width)
+    mb.content.text = ""
+
+  when T is string:
+    let clearText = splitLines(mb.content)
+      .mapIt(' '.repeat(mb.dims.width))
+      .join("\p")
+
+    mb.content = ""
+
+  hideCursor()
+  write(mb.dims.x, mb.dims.y, clearText)
+  showCursor()
+  mb.dismissAfter = none[DateTime]()
+
+
+#[ TODO: re-implement taking BufferDimensions if needed
+proc `maxWidth=`*[T: ScrollText or string](
+    mb: var MessageBuffer[T],
+    maxWidth: int) =
+  mb.maxWidth = max(maxWidth, 1)
+
+  when T is ScrollText: mb.content.maxWidth = mb.maxWidth
+  when T is string: mb.renderStringContent()
+]#
+
+
+proc nextTick*[T: ScrollText or string](mb: var MessageBuffer[T]) =
+
+  if mb.dismissAfter.isSome and now() > mb.dismissAfter.get:
+    mb.clear()
+
+  when T is ScrollText:
+    if mb.content.isScrolling:
+      discard mb.content.nextTick()
+      render(mb)
+
+
+func hasContent*[T: ScrollText or string](
+    mb: MessageBuffer[T]): bool =
+  when T is ScrollText: return len($mb.content) > 0
+  when T is string: return len(mb.content) > 0
+
+
+proc scrollLines*[T](mb: var MessageBuffer[T], lines: int) =
+  mb.firstLineIdx = clamp(
+    mb.firstLineIdx + lines,
+    0,
+    max(0, len(mb.content) - mb.dims.height))
+  render(mb)
+
+proc scrollPages*[T](mb: var MessageBuffer[T], pages: int) =
+  scrollLines(mb, pages * mb.dims.height)
+
+
+proc scrollColumns*(mb: var MessageBuffer[string], chars: int) =
+  mb.firstColumnIdx = clamp(
+    mb.firstColumnIdx + chars,
+    0,
+    max(0, len(mb.content) - mb.dims.width))
+  render(mb)
+
+proc resetScroll*(
+    mb: var MessageBuffer[ScrollText]) =
+  mb.firstColumnIdx = 0
+  mb.firstLineIdx = 0
+  render(mb)
--- a/src/main/nim/wdiwtlt/models.nim
+++ b/src/main/nim/wdiwtlt/models.nim
@ -0,0 +1,56 @@
+import std/[options, times]
+import uuids
+
+type
+  Artist* = object
+    id*: UUID
+    name*: string
+
+  Album* = object
+    id*: UUID
+    name*: string
+    year*: Option[int]
+    trackTotal*: int
+
+  MediaFile* = object
+    id*: UUID
+    name*: string
+    discNumber*: int
+    trackNumber*: Option[int]
+    playCount*: int
+    filePath*: string
+    fileHash*: string
+    metaInfoSource*: string
+    dateAdded*: DateTime
+    lastPlayed*: Option[DateTime]
+    presentLocally*: bool
+    comment*: string
+
+  Tag* = object
+    id*: UUID
+    name*: string
+    description*: string
+
+  Playlist* = object
+    id*: UUID
+    userCreated*: bool
+    name*: string
+    mediaFileCount*: int
+    copiedFromId*: Option[UUID]
+    createdAt*: DateTime
+    lastUsedAt*: DateTime
+
+  Bookmark* = object
+    id*: UUID
+    name*: Option[string]
+    userCreated*: bool
+    playlistId*: UUID
+    mediaFileId*: UUID
+    playIndex*: int
+    playTimeMs*: int
+    createdAt*: DateTime
+    lastUsedAt*: DateTime
+
+  Image* = object
+    id*: UUID
+    url*: string
--- a/src/main/nim/wdiwtlt/nonblockingio.nim
+++ b/src/main/nim/wdiwtlt/nonblockingio.nim
@ -0,0 +1,263 @@
+# Adapted from [illwill](https://github.com/johnnovak/illwill) by John Novak
+import std/[posix, termios]
+
+type
+  Key* {.pure.} = enum      ## Supported single key presses and key combinations
+    None = (-1, "None"),
+
+    # Special ASCII characters
+    CtrlA  = (1, "CtrlA"),
+    CtrlB  = (2, "CtrlB"),
+    CtrlC  = (3, "CtrlC"),
+    CtrlD  = (4, "CtrlD"),
+    CtrlE  = (5, "CtrlE"),
+    CtrlF  = (6, "CtrlF"),
+    CtrlG  = (7, "CtrlG"),
+    CtrlH  = (8, "CtrlH"),
+    Tab    = (9, "Tab"),     # Ctrl-I
+    CtrlJ  = (10, "CtrlJ"),
+    CtrlK  = (11, "CtrlK"),
+    CtrlL  = (12, "CtrlL"),
+    Enter  = (13, "Enter"),  # Ctrl-M
+    CtrlN  = (14, "CtrlN"),
+    CtrlO  = (15, "CtrlO"),
+    CtrlP  = (16, "CtrlP"),
+    CtrlQ  = (17, "CtrlQ"),
+    CtrlR  = (18, "CtrlR"),
+    CtrlS  = (19, "CtrlS"),
+    CtrlT  = (20, "CtrlT"),
+    CtrlU  = (21, "CtrlU"),
+    CtrlV  = (22, "CtrlV"),
+    CtrlW  = (23, "CtrlW"),
+    CtrlX  = (24, "CtrlX"),
+    CtrlY  = (25, "CtrlY"),
+    CtrlZ  = (26, "CtrlZ"),
+    Escape = (27, "Escape"),
+
+    CtrlBackslash    = (28, "CtrlBackslash"),
+    CtrlRightBracket = (29, "CtrlRightBracket"),
+
+    # Printable ASCII characters
+    Space           = (32, "Space"),
+    ExclamationMark = (33, "ExclamationMark"),
+    DoubleQuote     = (34, "DoubleQuote"),
+    Hash            = (35, "Hash"),
+    Dollar          = (36, "Dollar"),
+    Percent         = (37, "Percent"),
+    Ampersand       = (38, "Ampersand"),
+    SingleQuote     = (39, "SingleQuote"),
+    LeftParen       = (40, "LeftParen"),
+    RightParen      = (41, "RightParen"),
+    Asterisk        = (42, "Asterisk"),
+    Plus            = (43, "Plus"),
+    Comma           = (44, "Comma"),
+    Minus           = (45, "Minus"),
+    Dot             = (46, "Dot"),
+    Slash           = (47, "Slash"),
+
+    Zero  = (48, "Zero"),
+    One   = (49, "One"),
+    Two   = (50, "Two"),
+    Three = (51, "Three"),
+    Four  = (52, "Four"),
+    Five  = (53, "Five"),
+    Six   = (54, "Six"),
+    Seven = (55, "Seven"),
+    Eight = (56, "Eight"),
+    Nine  = (57, "Nine"),
+
+    Colon        = (58, "Colon"),
+    Semicolon    = (59, "Semicolon"),
+    LessThan     = (60, "LessThan"),
+    Equals       = (61, "Equals"),
+    GreaterThan  = (62, "GreaterThan"),
+    QuestionMark = (63, "QuestionMark"),
+    At           = (64, "At"),
+
+    ShiftA  = (65, "ShiftA"),
+    ShiftB  = (66, "ShiftB"),
+    ShiftC  = (67, "ShiftC"),
+    ShiftD  = (68, "ShiftD"),
+    ShiftE  = (69, "ShiftE"),
+    ShiftF  = (70, "ShiftF"),
+    ShiftG  = (71, "ShiftG"),
+    ShiftH  = (72, "ShiftH"),
+    ShiftI  = (73, "ShiftI"),
+    ShiftJ  = (74, "ShiftJ"),
+    ShiftK  = (75, "ShiftK"),
+    ShiftL  = (76, "ShiftL"),
+    ShiftM  = (77, "ShiftM"),
+    ShiftN  = (78, "ShiftN"),
+    ShiftO  = (79, "ShiftO"),
+    ShiftP  = (80, "ShiftP"),
+    ShiftQ  = (81, "ShiftQ"),
+    ShiftR  = (82, "ShiftR"),
+    ShiftS  = (83, "ShiftS"),
+    ShiftT  = (84, "ShiftT"),
+    ShiftU  = (85, "ShiftU"),
+    ShiftV  = (86, "ShiftV"),
+    ShiftW  = (87, "ShiftW"),
+    ShiftX  = (88, "ShiftX"),
+    ShiftY  = (89, "ShiftY"),
+    ShiftZ  = (90, "ShiftZ"),
+
+    LeftBracket  = (91, "LeftBracket"),
+    Backslash    = (92, "Backslash"),
+    RightBracket = (93, "RightBracket"),
+    Caret        = (94, "Caret"),
+    Underscore   = (95, "Underscore"),
+    GraveAccent  = (96, "GraveAccent"),
+
+    A = (97, "A"),
+    B = (98, "B"),
+    C = (99, "C"),
+    D = (100, "D"),
+    E = (101, "E"),
+    F = (102, "F"),
+    G = (103, "G"),
+    H = (104, "H"),
+    I = (105, "I"),
+    J = (106, "J"),
+    K = (107, "K"),
+    L = (108, "L"),
+    M = (109, "M"),
+    N = (110, "N"),
+    O = (111, "O"),
+    P = (112, "P"),
+    Q = (113, "Q"),
+    R = (114, "R"),
+    S = (115, "S"),
+    T = (116, "T"),
+    U = (117, "U"),
+    V = (118, "V"),
+    W = (119, "W"),
+    X = (120, "X"),
+    Y = (121, "Y"),
+    Z = (122, "Z"),
+
+    LeftBrace  = (123, "LeftBrace"),
+    Pipe       = (124, "Pipe"),
+    RightBrace = (125, "RightBrace"),
+    Tilde      = (126, "Tilde"),
+    Backspace  = (127, "Backspace"),
+
+    # Special characters with virtual keycodes
+    Up       = (1001, "Up"),
+    Down     = (1002, "Down"),
+    Right    = (1003, "Right"),
+    Left     = (1004, "Left"),
+    Home     = (1005, "Home"),
+    Insert   = (1006, "Insert"),
+    Delete   = (1007, "Delete"),
+    End      = (1008, "End"),
+    PageUp   = (1009, "PageUp"),
+    PageDown = (1010, "PageDown"),
+
+    F1  = (1011, "F1"),
+    F2  = (1012, "F2"),
+    F3  = (1013, "F3"),
+    F4  = (1014, "F4"),
+    F5  = (1015, "F5"),
+    F6  = (1016, "F6"),
+    F7  = (1017, "F7"),
+    F8  = (1018, "F8"),
+    F9  = (1019, "F9"),
+    F10 = (1020, "F10"),
+    F11 = (1021, "F11"),
+    F12 = (1022, "F12"),
+
+    Mouse = (5000, "Mouse")
+
+const
+  KEYS_D = [Key.Up, Key.Down, Key.Right, Key.Left, Key.None, Key.End, Key.None, Key.Home]
+  KEYS_E = [Key.Delete, Key.End, Key.PageUp, Key.PageDown, Key.Home, Key.End]
+  KEYS_F = [Key.F1, Key.F2, Key.F3, Key.F4, Key.F5, Key.None, Key.F6, Key.F7, Key.F8]
+  KEYS_G = [Key.F9, Key.F10, Key.None, Key.F11, Key.F12]
+
+{.push warning[HoleEnumConv]:off.}
+
+func toKey(c: int): Key =
+  try:
+    result = Key(c)
+  except RangeDefect:  # ignore unknown keycodes
+    result = Key.None
+
+{.pop}
+
+proc setNonBlockingTty*(enabled: bool) =
+  var ttyState: Termios
+
+  # get the terminal state
+  discard tcGetAttr(STDIN_FILENO, ttyState.addr)
+
+  if enabled:
+    # turn off canonical mode & echo
+    ttyState.c_lflag = ttyState.c_lflag and not Cflag(ICANON or ECHO)
+
+    # minimum of number input read
+    ttyState.c_cc[VMIN] = 0.char
+
+  else:
+    # turn on canonical mode & echo
+    ttyState.c_lflag = ttyState.c_lflag or ICANON or ECHO
+
+  # set the terminal attributes.
+  discard tcSetAttr(STDIN_FILENO, TCSANOW, ttyState.addr)
+
+proc parseStdin[T](input: T): Key =
+  var ch1, ch2, ch3, ch4, ch5: char
+  result = Key.None
+  if read(input, ch1.addr, 1) > 0:
+    case ch1
+    of '\e':
+      if read(input, ch2.addr, 1) > 0:
+        if ch2 == 'O' and read(input, ch3.addr, 1) > 0:
+          if ch3 in "ABCDFH":
+            result = KEYS_D[int(ch3) - int('A')]
+          elif ch3 in "PQRS":
+            result = KEYS_F[int(ch3) - int('P')]
+        elif ch2 == '[' and read(input, ch3.addr, 1) > 0:
+          if ch3 in "ABCDFH":
+            result = KEYS_D[int(ch3) - int('A')]
+          elif ch3 in "PQRS":
+            result = KEYS_F[int(ch3) - int('P')]
+          elif ch3 == '1' and read(input, ch4.addr, 1) > 0:
+            if ch4 == '~':
+              result = Key.Home
+            elif ch4 in "12345789" and read(input, ch5.addr, 1) > 0 and ch5 == '~':
+              result = KEYS_F[int(ch4) - int('1')]
+          elif ch3 == '2' and read(input, ch4.addr, 1) > 0:
+            if ch4 == '~':
+              result = Key.Insert
+            elif ch4 in "0134" and read(input, ch5.addr, 1) > 0 and ch5 == '~':
+              result = KEYS_G[int(ch4) - int('0')]
+          elif ch3 in "345678" and read(input, ch4.addr, 1) > 0 and ch4 == '~':
+            result = KEYS_E[int(ch3) - int('3')]
+          else:
+            discard   # if cannot parse full seq it is discarded
+        else:
+          discard     # if cannot parse full seq it is discarded
+      else:
+        result = Key.Escape
+    of '\n':
+      result = Key.Enter
+    of '\b':
+      result = Key.Backspace
+    else:
+      result = toKey(int(ch1))
+
+proc kbhit(ms: int): cint =
+  var tv: Timeval
+  tv.tv_sec = Time(ms div 1000)
+  tv.tv_usec = 1000 * (int32(ms) mod 1000) # int32 because of macos
+
+  var fds: TFdSet
+  FD_ZERO(fds)
+  FD_SET(STDIN_FILENO, fds)
+  discard select(STDIN_FILENO+1, fds.addr, nil, nil, tv.addr)
+  return FD_ISSET(STDIN_FILENO, fds)
+
+proc getKeyAsync*(ms: int = 0): Key =
+  result = Key.None
+  if kbhit(ms) > 0:
+    result = parseStdin(cint(STDIN_FILENO))
--- a/src/main/nim/wdiwtlt/scrolltext.nim
+++ b/src/main/nim/wdiwtlt/scrolltext.nim
@ -0,0 +1,64 @@
+import std/strutils
+
+type ScrollText* = ref object
+  text: string
+  maxWidth: int
+  scrollIdx: int        # Current index of the first character to show.
+  lastRender: string
+
+const endBufLen = 4
+
+
+proc render(st: ScrollText): string =
+  if st.text.len <= st.maxWidth:
+    return st.text
+
+  if st.scrollIdx == 0: return st.text[0..<st.maxWidth]
+  elif st.scrollIdx >= st.text.len:
+    return ' '.repeat(endBufLen - (st.scrollIdx - st.text.len)) &
+      st.text[0..<st.maxWidth - 1]
+  elif st.scrollIdx + st.maxWidth < st.text.len:
+    return st.text[st.scrollIdx..<(st.scrollIdx + st.maxWidth)]
+  else:
+    return st.text[st.scrollIdx..<st.text.len] & ' '.repeat(endBufLen) &
+      st.text[0..<max(0, (st.scrollIdx + st.maxWidth - st.text.len - 1))]
+
+
+proc `text=`*(st: var ScrollText, text: string) =
+  st.text = text
+  st.scrollIdx = text.len
+  st.lastRender = render(st)
+
+
+proc `maxWidth=`*(st: var ScrollText, maxWidth: int) =
+  st.maxWidth = max(maxWidth, 1)
+  st.lastRender = render(st)
+
+
+proc initScrollText*(text: string, maxWidth: int): ScrollText =
+  result = ScrollText(
+    text: text,
+    maxWidth: max(1, maxWidth),
+    scrollIdx: 0)
+
+  result.lastRender = render(result)
+
+
+proc nextTick*(st: var ScrollText): string =
+
+  st.lastRender = render(st)
+
+  # Advance the scroll index by one.
+  st.scrollIdx = (st.scrollIdx + 1) mod (st.text.len + endBufLen)
+
+  return st.lastRender
+
+
+func isScrolling*(st: ScrollText): bool =
+  return st.text.len > st.maxWidth
+
+
+func `$`*(st: ScrollText): string = st.lastRender
+
+
+converter toString*(st: ScrollText): string = st.lastRender
--- a/src/main/nim/wdiwtlt/taglib.nim
+++ b/src/main/nim/wdiwtlt/taglib.nim
@ -0,0 +1,167 @@
+import std/[paths, strutils]
+
+{.passl: "-ltag_c".}
+{.passc: "-ltag_c".}
+
+type
+  TagLibFileType* {.size: sizeof(cint).} = enum
+    MPEG, OggVorbis, FLAC, MPC, OggFlac, WavPack, Speex, TrueAudio, MP4, ASF,
+    AIFF, WAV, APE, IT, Mod, S3M, XM, Opus, DSF, DSDIFF, SHORTEN
+
+  TagLibID3v2Encoding* {.size: sizeof(cint).} = enum
+    Latin1, UTF16, UTF16BE, UTF8, Encoding
+
+  CFile = pointer
+  CTag* = pointer
+  CAudioProperties = pointer
+
+  TagLibFIle* = object
+    path*: Path
+    cfile*: CFile
+    tag*: CTag
+    ap*: CAudioProperties
+
+{.push importc.}
+{.push cdecl.}
+proc taglib_set_strings_unicode*(unicode: cint)
+proc taglib_set_string_management_enabled*(enabled: cint)
+proc taglib_free*(p: pointer)
+
+proc taglib_file_new(filename: cstring): CFile
+proc taglib_file_new_type(filename: cstring, `type`: TagLibFileType): CFile
+proc taglib_file_free(file: CFile)
+proc taglib_file_is_valid(file: CFile): cint
+proc taglib_file_tag(file: CFile): CTag
+proc taglib_file_audioproperties(file: CFile): CAudioProperties
+proc taglib_file_save(file: CFile): cint
+
+proc taglib_tag_title(tag: CTag): cstring
+proc taglib_tag_artist(tag: CTag): cstring
+proc taglib_tag_album(tag: CTag): cstring
+proc taglib_tag_comment(tag: CTag): cstring
+proc taglib_tag_genre(tag: CTag): cstring
+proc taglib_tag_year(tag: CTag): cuint
+proc taglib_tag_track(tag: CTag): cuint
+proc taglib_tag_set_title(tag: CTag, title: cstring)
+proc taglib_tag_set_artist(tag: CTag, artist: cstring)
+proc taglib_tag_set_album(tag: CTag, album: cstring)
+proc taglib_tag_set_comment(tag: CTag, comment: cstring)
+proc taglib_tag_set_genre(tag: CTag, genre: cstring)
+proc taglib_tag_set_year(tag: CTag, year: cuint)
+proc taglib_tag_set_track(tag: CTag, year: cuint)
+proc taglib_tag_free_strings()
+
+proc taglib_audioproperties_length(ap: CAudioProperties): cint
+proc taglib_audioproperties_bitrate(ap: CAudioProperties): cint
+proc taglib_audioproperties_samplerate(ap: CAudioProperties): cint
+proc taglib_audioproperties_channels(ap: CAudioProperties): cint
+
+proc taglib_id3v2_set_default_text_encoding(encoding: TagLibID3v2Encoding)
+
+proc taglib_property_set(file: CFile, prop: cstring, value: cstring)
+proc taglib_property_set_append(file: CFile, prop: cstring, value: cstring)
+#proc taglib_property_keys(file: CFile): cstring[]
+proc taglib_property_get(file: CFile, prop: cstring): cstring
+proc taglib_property_free(props: pointer)
+
+{.pop.} # cdecl
+{.pop.} # importc
+
+taglib_set_strings_unicode(1)
+
+proc initTagLibFile(path: Path, cfile: CFile): TagLibFile =
+  if cfile.isNil:
+    raise newException(IOError, "Failed to open file: " & $path)
+
+  if taglib_file_is_valid(cfile) < 0:
+    taglib_file_free(cfile)
+    raise newException(IOError, "Invalid TagLib file: " & $path)
+
+  result.path = path
+  result.cfile = cfile
+  result.tag = taglib_file_tag(cfile)
+  result.ap = taglib_file_audioproperties(cfile)
+
+proc openTags*(path: Path): TagLibFile =
+  ## Open a file and return a TagLibFile object.
+  let cfile = taglib_file_new(path.cstring)
+  initTagLibFile(path, cfile)
+
+proc openTags*(path: Path, fileType: TagLibFileType): TagLibFile =
+  ## Open a file of a specific type and return a TagLibFile object.
+  let cfile = taglib_file_new_type(path.cstring, fileType)
+  initTagLibFile(path, cfile)
+
+proc writeTags*(file: TagLibFile) =
+  ## Write tags to the file.
+  discard taglib_file_save(file.cfile)
+
+proc close*(file: var TagLibFile) =
+  ## Close the file and free resources.
+  if not file.cfile.isNil:
+    taglib_tag_free_strings()
+    taglib_file_free(file.cfile)
+    file.cfile = nil
+    file.tag = nil
+    file.ap = nil
+
+#proc `=destroy`*(file: TagLibFile) =
+#  ## Destroy the file and free resources.
+#  close(file)
+
+{.push inline.}
+proc length*(file: TagLibFile): int = taglib_audioproperties_length(file.ap)
+proc bitrate*(file: TagLibFile): int = taglib_audioproperties_bitrate(file.ap)
+proc samplerate*(file: TagLibFile): int = taglib_audioproperties_samplerate(file.ap)
+proc channels*(file: TagLibFile): int = taglib_audioproperties_channels(file.ap)
+
+proc title*(file: TagLibFile): string = $taglib_tag_title(file.tag)
+proc artist*(file: TagLibFile): string = $taglib_tag_artist(file.tag)
+proc album*(file: TagLibFile): string = $taglib_tag_album(file.tag)
+proc comment*(file: TagLibFile): string = $taglib_tag_comment(file.tag)
+proc genre*(file: TagLibFile): string = $taglib_tag_genre(file.tag)
+proc year*(file: TagLibFile): int = taglib_tag_year(file.tag).int
+proc track*(file: TagLibFile): int = taglib_tag_track(file.tag).int
+
+proc `title=`*(file: var TagLibFile, title: string) = taglib_tag_set_title(file.tag, title.cstring)
+proc `artist=`*(file: var TagLibFile, artist: string) = taglib_tag_set_artist(file.tag, artist.cstring)
+proc `album=`*(file: var TagLibFile, album: string) = taglib_tag_set_album(file.tag, album.cstring)
+proc `comment=`*(file: var TagLibFile, comment: string) = taglib_tag_set_comment(file.tag, comment.cstring)
+proc `genre=`*(file: var TagLibFile, genre: string) = taglib_tag_set_genre(file.tag, genre.cstring)
+proc `year=`*(file: var TagLibFile, year: int) = taglib_tag_set_year(file.tag, year.cuint)
+proc `track=`*(file: var TagLibFile, track: int) = taglib_tag_set_track(file.tag, track.cuint)
+
+proc albumArtist*(file: TagLibFile): string =
+  ## Get the album artist of the file.
+  let albumArtist = taglib_property_get(file.cfile, "ALBUMARTIST")
+  if albumArtist.isNil:
+    return ""
+  else:
+    return $albumArtist
+
+proc `albumArtist=`*(file: var TagLibFile, albumArtist: string) =
+  ## Set the album artist of the file.
+  taglib_property_set(file.cfile, "ALBUMARTIST", albumArtist.cstring)
+
+proc discNumber*(file: TagLibFile): int =
+  ## Get the disc number of the file.
+  let discNumber = taglib_property_get(file.cfile, "DISCNUMBER")
+  if discNumber.isNil:
+    return 0
+  else:
+    return parseInt($discNumber)
+
+proc `discNumber=`*(file: var TagLibFile, discNumber: int) =
+  ## Set the disc number of the file.
+  taglib_property_set(file.cfile, "DISCNUMBER".cstring, ($discNumber).cstring)
+
+# Other properties we could use:
+# - SUBTITLE
+# - DATE
+# - COMPOSER
+# - TITLESORT
+# - ARTISTSORT
+# - ALBUMSORT
+# - ALBUMARTISTSORT
+
+{.pop.} # inline
--- a/src/main/nim/wdiwtlt/terminal.nim
+++ b/src/main/nim/wdiwtlt/terminal.nim
@ -0,0 +1,111 @@
+import std/[nre, sequtils]
+
+const CSI = "\x1b["
+const RESET_FORMATTING* = "\x1b[0m"
+const ANSI_ESCAPE_CODE_ENDINGS*: seq[char] = toSeq('A'..'Z') & toSeq('a'..'z')
+let FORMATTING_REGEX* = re("\x1b\\[([0-9;]*)([a-zA-Z])")
+
+type
+  CursorType* = enum
+    ctBlockBlink = 1, ctBlock, ctUnderlineBlink, ctUnderline, ctBarBlink, ctBar
+
+  EraseMode* = enum
+    emToEnd, emToStart, emAll
+
+  TerminalColors* = enum
+    cBlack, cRed, cGreen, cYellow, cBlue, cMagenta, cCyan, cWhite
+
+
+proc stripFormatting*(text: string): string =
+  text.replace(FORMATTING_REGEX, "")
+
+
+func ansiAwareSubstring*(s: string, start, length: int): string =
+  result = ""
+  var curAnsiEscCode = ""
+  var i = 0
+  var visibleLen = 0
+
+  while i < len(s) and visibleLen < length:
+
+    if len(s) > i + len(CSI) and
+      # We need to notice ANSI escape codes...
+       s[i..<i + len(CSI)] == CSI:
+      var j = i + len(CSI)
+      while j < s.len and s[j] notin ANSI_ESCAPE_CODE_ENDINGS: j += 1
+
+      # and remember it if we're before the start of the substring
+      if i < start: curAnsiEscCode = s[i..j]
+
+      # or add it without increasing our substring length
+      else: result.add(s[i..j])
+
+      # either way we want to pick up after it
+      i = j
+
+    else:
+      result.add(s[i])
+      visibleLen += 1
+
+    i += 1
+
+  result = curAnsiEscCode & result
+
+func color*(text: string, fg = cWhite, bg = cBlack): string =
+  return CSI & $int(fg) & ";" & $(int(bg) + 40) & "m" & text & RESET_FORMATTING
+
+
+func color*(text: string, fg: TerminalColors): string =
+  return CSI & $int(fg) & "m" & text & RESET_FORMATTING
+
+
+func color*(text: string, bg: TerminalColors): string =
+  return CSI & $(int(bg) + 40) & "m" & text & RESET_FORMATTING
+
+
+func invert*(text: string): string = return CSI & "7m" & text & RESET_FORMATTING
+func striketrough*(text: string): string = return CSI & "9m" & text & RESET_FORMATTING
+
+
+func eraseDisplay*(mode = emToEnd): string = return CSI & $int(mode) & "J"
+func eraseLine*(mode = emToEnd): string = return CSI & $int(mode) & "K"
+
+
+func blinkSlow*(text: string): string = return CSI & "5m" & text & RESET_FORMATTING
+func blinkFast*(text: string): string = return CSI & "6m" & text & RESET_FORMATTING
+
+
+func cursorUp*(n: int = 1): string = return CSI & $int(n) & "A"
+func cursorDown*(n: int = 1): string = return CSI & $int(n) & "B"
+func cursorForward*(n: int = 1): string = return CSI & $int(n) & "C"
+func cursorBackward*(n: int = 1): string = return CSI & $int(n) & "D"
+func cursorNextLine*(n: int = 1): string = return CSI & $int(n) & "E"
+func cursorPrevLine*(n: int = 1): string = return CSI & $int(n) & "F"
+func cursorHorizontalAbsolute*(col: int): string = return CSI & $int(col) & "G"
+
+
+func cursorPosition*(row, col: int): string =
+  return CSI & $int(row) & ";" & $int(col) & "H"
+
+
+proc cursorType*(ct: CursorType): string = return CSI & $int(ct) & " q"
+func saveCursorPosition*(): string = return CSI & "s"
+func restoreCursorPosition*(): string = return CSI & "u"
+
+
+func scrollUp*(n: int = 1): string = return CSI & $int(n) & "S"
+func scrollDown*(n: int = 1): string = return CSI & $int(n) & "T"
+
+
+proc write*(f: File, x, y: int, text: string) =
+  f.write(cursorPosition(y, x) & text)
+
+
+proc write*(x, y: int, text: string) =
+  stdout.write(cursorPosition(y, x) & text)
+
+
+proc setCursorType*(ct: CursorType) = stdout.write(cursorType(ct))
+proc setCursorPosition*(x, y: int) = stdout.write(cursorPosition(y, x))
+proc hideCursor*() = stdout.write(CSI & "?25l")
+proc showCursor*() = stdout.write(CSI & "?25l")
--- a/src/main/nim/wdiwtlt/usage.nim
+++ b/src/main/nim/wdiwtlt/usage.nim
@ -0,0 +1,5 @@
+func usageOf*(command: seq[string]): string =
+  case command[0]
+  of "log":
+    # TODO
+    discard
--- a/src/main/nim/wdiwtlt/util.nim
+++ b/src/main/nim/wdiwtlt/util.nim
@ -0,0 +1,7 @@
+func clamp*(value, min, max: int): int =
+  if value < min:
+    return min
+  elif value > max:
+    return max
+  else:
+    return value
--- a/core/src/main/sql/migrations/20151209054632-initial-schema-down.sql
+++ b/core/src/main/sql/migrations/20151209054632-initial-schema-down.sql
--- a/src/main/sql/media-library-schema.sql
+++ b/src/main/sql/media-library-schema.sql
@ -0,0 +1,112 @@
+CREATE TABLE artists (
+  id TEXT PRIMARY KEY,
+  name TEXT UNIQUE NOT NULL
+);
+
+CREATE INDEX artists_name_idx ON artists(name);
+
+CREATE TABLE albums (
+  id TEXT PRIMARY KEY,
+  name TEXT NOT NULL,
+  year INTEGER,
+  track_total INTEGER
+);
+
+CREATE INDEX albums_name_idx ON albums(name);
+
+CREATE TABLE media_files (
+  id TEXT PRIMARY KEY,
+  name TEXT NOT NULL,
+  disc_number INTEGER NOT NULL DEFAULT 1,
+  track_number INTEGER,
+  play_count INTEGER NOT NULL DEFAULT 0,
+  file_path TEXT NOT NULL,
+  file_hash TEXT NOT NULL,
+  meta_info_source TEXT NOT NULL, -- 'tag' or 'filesystem'
+  date_added TEXT NOT NULL DEFAULT (strftime('%F %TZ', datetime())),
+  last_played TEXT,
+  present_locally INTEGER NOT NULL DEFAULT TRUE,
+  comment TEXT DEFAULT ''
+);
+
+CREATE INDEX media_files_name_idx ON media_files(name);
+
+CREATE TABLE artists_media_files (
+  artist_id TEXT NOT NULL REFERENCES artists(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  media_file_id TEXT NOT NULL REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  PRIMARY KEY (artist_id, media_file_id)
+);
+
+CREATE TABLE albums_media_files (
+  album_id TEXT NOT NULL REFERENCES albums(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  media_file_id TEXT NOT NULL REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  PRIMARY KEY (album_id, media_file_id)
+);
+
+CREATE TABLE tags (
+  id TEXT PRIMARY KEY,
+  name TEXT UNIQUE NOT NULL,
+  description TEXT NOT NULL DEFAULT ''
+);
+
+CREATE TABLE playlists (
+  id TEXT PRIMARY KEY,
+  user_created INTEGER NOT NULL DEFAULT FALSE,
+  name TEXT NOT NULL,
+  media_file_count INTEGER NOT NULL DEFAULT 0,
+  copied_from_id TEXT DEFAULT NULL,
+  created_at TEXT NOT NULL DEFAULT (strftime('%F %TZ', datetime())),
+  last_used TEXT NOT NULL DEFAULT (strftime('%F %TZ', datetime()))
+);
+
+CREATE TABLE playlists_media_files (
+  playlist_id TEXT NOT NULL REFERENCES playlists(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  media_file_id TEXT NOT NULL REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  position INTEGER NOT NULL DEFAULT 0,
+  PRIMARY KEY (playlist_id, media_file_id, position),
+  UNIQUE (playlist_id, position)
+);
+
+CREATE TABLE bookmarks (
+  id TEXT PRIMARY KEY,
+  name TEXT,
+  user_created INTEGER NOT NULL DEFAULT FALSE,
+  playlist_id TEXT NOT NULL REFERENCES playlists(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  media_file_id TEXT NOT NULL REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  play_index INTEGER NOT NULL,
+  play_time_ms INTEGER NOT NULL DEFAULT 0,
+  created_at TEXT NOT NULL DEFAULT (strftime('%F %TZ', datetime())),
+  last_used TEXT NOT NULL DEFAULT (strftime('%F %TZ', datetime()))
+);
+
+CREATE INDEX bookmarks_playlist_id_idx ON bookmarks (playlist_id);
+CREATE INDEX bookmarks_media_file_id_idx ON bookmarks (media_file_id);
+
+CREATE TABLE media_files_tags (
+  media_file_id TEXT REFERENCES media_files(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  tag_id TEXT REFERENCES tags(id) ON DELETE CASCADE ON UPDATE CASCADE,
+  PRIMARY KEY (media_file_id, tag_id)
+);
+
+CREATE TABLE images (
+  id TEXT PRIMARY KEY,
+  url TEXT
+);
+
+CREATE TABLE artists_images (
+  artist_id TEXT REFERENCES artists (id) ON DELETE CASCADE ON UPDATE CASCADE,
+  image_id TEXT REFERENCES images (id) ON DELETE CASCADE ON UPDATE CASCADE,
+  PRIMARY KEY (artist_id, image_id)
+);
+
+CREATE TABLE albums_images (
+  album_id TEXT REFERENCES albums (id) ON DELETE CASCADE ON UPDATE CASCADE,
+  image_id TEXT REFERENCES images (id) ON DELETE CASCADE ON UPDATE CASCADE,
+  PRIMARY KEY (album_id, image_id)
+);
+
+CREATE TABLE artists_albums (
+  artist_id TEXT NOT NULL REFERENCES artists (id) ON DELETE CASCADE ON UPDATE CASCADE,
+  album_id TEXT NOT NULL REFERENCES albums (id) ON DELETE CASCADE ON UPDATE CASCADE,
+  PRIMARY KEY (artist_id, album_id)
+);
--- a/wdiwtlt.nimble
+++ b/wdiwtlt.nimble
@ -0,0 +1,18 @@
+# Package
+
+version       = "0.2.0"
+author        = "Jonathan Bernard"
+description   = "What Do I Want To Listen To (WDIWTLT) - CLI media player."
+license       = "GPL-3.0-or-later"
+srcDir        = "src/main/nim"
+bin           = @["wdiwtlt"]
+
+
+# Dependencies
+
+requires "nim >= 2.2.0"
+requires @["checksums", "docopt", "illwill", "mpv", "nimterop", "taglib",
+  "uuids", "waterpark"]
+
+# Dependencies from https://git.jdb-software.com/jdb/nim-packages
+requires @["cliutils", "db_migrate", "fiber_orm", "namespaced_logging"]
--- a/worklog.md
+++ b/worklog.md
@ -1,21 +0,0 @@
-WDIWTLT
-========================================
-
-* Local library
-* CLI interface
-* REST API
-
-
-Current track
----------------------------------------
-
-Building CLI:
-* More clarity around selection
-  management.
-* Tagging
-* Bookmark management
-* Playlist managment
-* Online command help.
-* Ability to actually play music.
-* Library management (split artists,
-  change other metadata)?
Author	SHA1	Message	Date
Jonathan Bernard	22d3966755	WIP: Rendering/UI refactor.	2025-04-02 06:13:27 -05:00
Jonathan Bernard	a7a9cf8620	Input line rendering logic in width-constrained situations.	2025-03-15 20:42:29 -05:00
Jonathan Bernard	c75438d409	WIP: vi-like input system.	2025-03-15 07:19:30 -05:00
Jonathan Bernard	eadf3946e7	MD5 streaming implementation.	2025-03-04 15:51:02 -06:00
Jonathan Bernard	2d44d2c328	TagLib Nim wrapper.	2025-03-04 15:49:54 -06:00
Jonathan Bernard	3f3a6b286b	Add DB implementation, models, logging.	2025-02-23 08:23:24 -06:00
Jonathan Bernard	bad288a24b	Proof-of-concept MPV-based player.	2025-02-20 22:25:30 -06:00
Jonathan Bernard	bbd7952232	Initial WIP data models, fiber_orm mappings.	2025-02-20 22:25:10 -06:00
Jonathan Bernard	85f37373b6	Updated schema for Sqlite.	2025-02-20 22:24:49 -06:00
Jonathan Bernard	c3a5b5f87f	Initial NIM stub (nimble file, gitignore.	2025-02-20 22:24:37 -06:00
Jonathan Bernard	bdd6c03129	Clean out old Groovy code.	2025-02-20 22:23:36 -06:00