rsrtools 0.1.2

Tools for creating Rocksmith 2014 songlists and managing Rocksmith save files

rsrtools is a package for creating and saving Rocksmith 2014 songlists to Rocksmith save files (profiles). Incidentally, it also provides tools for managing Rocksmith profiles.

Contents

• Acknowledgements

• Breaking Changes

• Warnings

• TL:DNR

• Motivation

• Introduction

• Alternatives

• Documentation and Tutorial

  • Pre-requisites

  • Set up a Testing Profile!

• Command line work flow summary

  • Preliminaries

  • Clone Profile

  • Song List Testing

• The Configuration File

  • Settings

  • Song List Sets

  • Filters

    • Basic Filter Definition

    • List Type Sub-filter

    • Range Type Sub-filter

  • Filter Examples

  • Nested vs. Flat Filters

• Something Went Wrong!

• Database Structure

• Package Caveats

• Sidebar: Rocksmith Save File Editing

• TODO

• Changelog

• Development notes

Acknowledgements

@0x0L for rs-utils and rocksmith Rocksmith tools. All of the save file and data file handling routines are based on this code.

@sandiz for rs-manager, which is an awesome set listing tool. This package also gave me a deeper understanding of the Rocksmith PSARC structure.

Breaking Changes

0.1.0 to 0.1.1+ Terminology correction from Steam user id to Steam account id. Probably the only effect for most people is to edit ‘config.toml’ and replace steam_user_id with steam_account_id. All Steam functions moved to steam.py, some windows registry functions for Steam removed and replaced with functions based on Steam vdf files.

Warnings

As this package is all about editing game saves, here are a couple of warnings.

0. This package is late Alpha/early Beta. I’ve been using it for about a year, and it has been robust for my application. However, until this warning disappears, please assume that you are the second ever user and that you will find bugs. Please report these to me via github issues so I can implement fixes.

1. This package edits Rocksmith profiles. Use at your own risk and with the understanding that this package carries the risk of corrupting your save files (to date it has worked fine for me - YMMV, and it will definitely stop working if Ubisoft make any changes to the Rocksmith save file format). However, the package includes many self checks and tries to make backups of profiles before making changes, so (hopefully) the risk of profile loss or corruption should be low.

2. This package is (obviously) not endorsed by Ubisoft - if you use this package and run into problems with your save files, Ubisoft will not be very interested in helping you. If this happens, I will try to help, but will be limited by my available time and the complexity of your problem. So, in effect repeating the first warning: use this package at your own risk.

3. Don’t run this package at the same time as Rocksmith is running. You’ll end up crossing the save files and nobody will be happy (mostly you though).

4. This package will only work on Windows and Mac OS X at the moment. I have no idea what would be needed for Linux/PS4/XBox song lists.

TL:DNR

If you know what you are doing with python and know your way around the Customs Forge Song Manager, here are the recommended quick start steps.

0. The package only works on Windows and Mac OS X for now.

1. READ the section on setting up a test profile (Set up a Testing Profile!). Until you are familiar with the package, this will be your best safeguard against damaging your precious save game(s).

2. SET UP a Rocksmith test profile.

3. Install python 3.7.x (I’m on 3.7.3, and you will need some 3.7 features).

4. Create a virtual environment.

5. Install rsrtools into your virtual environment with:

   pip install rsrtools

6. Create a working folder/directory.

7. Create an ArrangementsGrid.xml file from Customs Forge Song Manger and copy or move it into the working directory (hopefully this step will become optional with the implementation of a PSARC scanner).

8. Start your virtual environment and run the package (with appropriate substitution for <path_to_your_working_directory>):

   python -m rsrtools.songlists.songlists <path_to_your_working_directory>

   If you start python in the working directory, you could use:

   python -m rsrtools.songlists.songlists .

9. Clone your save game into the test profile and do all of your testing on this test profile until you are comfortable that the package is working and doing what you want it to do.

Motivation

Hopefully this section doesn’t read too much like a food blog.

I’ve implemented this package because, while I really enjoy Rocksmith 2014 Remastered as a learning tool, I’ve had ongoing frustration with creating custom play lists. I thought there had to be a better way (and I also wanted a project I could use to learn python). My initial goal for this package was to be able to easily create song lists for a specific tuning and play counts - I break my practice sessions up into new stuff, moderately new and old - and it’s a real pain in the backside scrolling through 500 odd tracks. And it’s also a pain in the backside setting up custom song lists in Rocksmith. So that’s the motivation for this project. During implementation, I realised it would be possible to create much more varied song lists (not so useful for me, but maybe so for others).

Introduction

The purpose of this package is to provide an improved song list creator for Rocksmith. This package allows creation of song lists based on a variety of criteria, and allows the criteria to be built up hierarchically. Here is an incomplete list of the type of song lists you can create with this package.

• All arrangements with E Standard tunings (not very exciting).

• All songs with E Standard tunings at 440 pitch (still not exciting).

• All D standard 440 songs with a played count between 12 and 18 (getting somewhere now).

• All Eb standard 440 songs with a mastery between 40 and 65%.

• All E standard songs that I have played at least once on score attack, but haven’t got a platinum badge (yet).

• All easy E Standard songs that I haven’t yet got a platinum badge for (OK. So it’s a long list for me, but something to work on).

I’m simplifying a bit here, but it gives an idea of the type of thing that this package is intended to do. Extending the above examples, the song lists could be created for songs or a specific arrangement type (Bass, Lead, Rhythm).

Criteria that can be used for song list creation include:

• List criteria:

  • Tuning

  • ArrangementName (Bass, Lead, Lead1, Lead2, Lead3, Rhythm, Rhythm1, Rhythm2, Combo, Combo1, Combo2, Combo3)

  • Song key (typically the unique part of DLC/song file names)

  • ArrangementId (expert functionality)

  • Artist Name

  • Track Title

  • Album Name

• Range criteria:

  • Album Year

  • Pitch (A440 or otherwise)

  • Tempo

  • Note Count

  • Played Count

  • Mastery Peak

  • SA Easy Badges

  • SA Medium Badges

  • SA Hard Badges

  • SA Master Badges

  • and a few more.

Filtering can be by inclusion or exclusion. A more complicated example would be: all E Standard, D Standard and C Standard lead tracks, but nothing by the Foo Fighters or Green Day and nothing in the decade 2000-2010, only tracks I haven’t completed a hard platinum score attack, and only tracks I’ve played at least 4 times. (I can’t imagine using this filter myself, but somebody with a grudge against Dave Grohl might care).

If you want a particular type of song list and can’t see how to build it from the help, ask me and I’ll see if I can either come up with a solution or add the needed functionality.

Alternatives

1. The Customs Forge Song Manager (CFSM) provides a different and better supported mechanism for creating song lists based on moving files in and out of directories. My approach provides some of the same functionality, with the following variations:

   • I don’t move song files, but rather edit the song lists directly in the Rocksmith profiles/save files.

   • I support building song lists based on data in save files (played counts, score attack performance, mastery, etc.).

   The CFSM approach is very actively supported, so if you aren’t interested in the specific functionality my approach provides, I’d go with their tool, which is available from: http://customsforge.com/.

2. rs-manager (https://github.com/sandiz/rs-manager) is a GUI application that can create set lists manually or from procedural filtering similar to rsrtools. It is a much friendlier way to generate song/set lists than rsrtools. The rs-manager set lists are for reference only (i.e. rs-manager does not support loading set lists into Rocksmith save files).

   However, @sandiz, the rs-manager developer, has implemented functionality to export rs-manager set lists in a format that can be used by rsrtools. A near term update of rsrtools will allow loading of these set lists into Rocksmith save files. Once this is done, we will have a work flow where set lists can be generated using the rs-manager GUI and then exported for loading into Rocksmith by a simple rsrtools command (bypassing the joys of setting up text filters for rsrtools).

That’s the Long Intro over.

Documentation and Tutorial

The documentation provided here is fairly detailed. I’ve done this on the basis that a significant portion of users will be interested in using the system, but not interested in the details of the python. Consequently, there is a lot of step by step detail included. If you know your way around python, programming and CFSM, you should be able to skim through a lot of the content very quickly (and you can modify the set up to match your own environment).

This package provides:

• A command line tool for creating Rocksmith song lists from a series of filters, and writing the resulting song lists into a Rocksmith profile. The command line work flow is described below.

• A set of routines that can be used to implement a GUI version of the command line tools (I have not implemented a GUI, as the command line is sufficient for my requirements - see the section on Alternatives for more GUI oriented solutions).

Repeating warning #4, this package is currently only supported on Windows (and only tested on Windows 10).

Pre-requisites

• Download and install Python 3.7+ from www.python.org. (I’d recommend 3.7.3, which is what I’m using).

• Create a folder/directory for running rsrtools. For this tutorial, I’m assuming this is: D:\RS Stuff, and create an environment sub-directory Env and a working sub-directory Working in the rsrtools directory. At the end of this step, my folders are:

  D:\RS Stuff
  D:\RS Stuff\Env
  D:\RS Stuff\Working

• Set up a python virtual environment for rsrtools and install the package via pip. If you are unfamiliar with python, follow these steps:

  1. Open a command window (cmd.exe).

  2. Type the following commands. The hashed lines are comments that explain what each command does and can be ignored:

     # Change paths as required to match your rsrtools directory
     # Create the environment in D:\RS Stuff\Env
     python -m venv "d:\RS Stuff\Env"
     
     # Activate the python environment
     "d:\RS Stuff\Env\Scripts\activate.bat"
     
     # install rsrtools and supporting libraries
     pip install rsrtools

  3. Exit the command window.

Set up a Testing Profile!

Until you are confident that this package is working properly, I strongly suggest you do some testing on a temporary Rocksmith profile. I’d also suggest testing all new song list filters on the temporary profile before applying them to your main profile.

The process I follow for testing changes before applying them to my main profile is:

• Create the Testing profile (described in this section).

• Clone my profile into the Testing profile. This is very useful if you want to test song lists based on played counts, score attack, mastery, etc. The command line work flow in the following section explains how to clone your profile.

• Try out the song filters on the Testing profile.

The process for setting up a temporary profile is as easy as it sounds:

1. Start Rocksmith.

2. At the Select Profile Menu, click New Profile, name the profile and go through set up (the set up step can’t be avoided unfortunately).

Command line work flow summary

Repeating an important warning: Don’t run this package at the same time as Rocksmith is running. You’ll end up crossing the save files and nobody will be happy (mostly you though).

Preliminaries

1. Create a working directory that will contain working copies of Rocksmith files, the arrangement database, and the song list configuration file. For this tutorial I will use the folder/directory set up in the previous section:

   D:\RS Stuff\Working

2. Download and install the Customs Forge Song Manager from: http://customsforge.com/

   The rsrtools song list creator needs information about song arrangements. I plan to add a song library scanner in the future, but in the interim, the easiest way to get this information is from a CFSM report. The steps required are:

   • Run CFSM.

   • Go to Arrangement Analyzer.

   • Go to Settings (check that the settings are for Arrangement Analyzer).

     • Tick ‘Include RS2014 Base Songs’

     • Tick ‘Include RS1 Compatibility Packs’

   • Go Back to Arrangement Analyzer.

   • Click the ‘Rescan’ button (this will take a while).

   • Export to XML. This should create the ArrangementsGrid.xml file needed by rsrtools.

   • Put the xml file somewhere easy to find.

   I normally drop the xml file into my working directory - this allows automatic loading of the arrangement data into the database.

3. Optional, but strongly recommended: Create a temporary/testing profile so that you can get comfortable with how this package works on Rocksmith save files (I use this approach any time I’m experimenting with major changes). See the previous section for a description of this process.

   For this tutorial, I’ll assume the test profile is called ‘Testing’. I’ll go through the steps to clone data from your normal profile into the test profile later on.

4. Because I’m lazy, at this point I put together a batch file in the working directory. Let’s call it ‘song_lists.bat’ and put the following lines in it:

   echo on
   Call "D:\RS Stuff\Env\Scripts\Activate.bat"
   python -m rsrtools.songlists.songlists "D:\RS Stuff\Working"
   Deactivate.bat

   You will need to edit your paths to match where you have put your python environment and your working directory.

   When I say run the batch file below, I suggest that you do this initially from a command shell (cmd.exe). This will allow you to see any errors (otherwise if you double click on the batch file, the screen will flash up and close before you have a chance to read anything). Once you are confident everything is working, you can run it with a double click.

5. Skip this step if you have put the ArrangementsGrid.xml file in the working directory.

   Otherwise you need to set up the core arrangement table in the database with the following commands in command shell, substituting in the path to your working folder and the path to the arrangements file (<path_to_xml_file>):

   Call "D:\RS Stuff\Env\Scripts\Activate.bat"
   python -m rsrtools.songlists.songlists "D:\RS Stuff\Working" --CFSMxml <path_to_xml_file>

   When the menu comes up, choose 0 to exit the package, and then choose y to save the configuration. (hit enter after making a choice). Then exit the command shell.

6. Run the batch file to set up the default configuration. This should result in text menu something like the following:

   Rocksmith song list generator main menu.
   
       Steam account id:    'not set'
       Rocksmith profile:   'not set'
       Reporting to:        Standard output/console
       Working directory:   D:\RS Stuff\Working
   
   Please choose from the following options:
   
     1) Change/select Steam account id. This also clears the profile selection.
     2) Change/select Rocksmith player profile.
     3) Toggle the report destination.
     4) Choose a single filter and create a song list report.
     5) Choose a song list set and create a song list report.
     6) Choose a song list set and write the list(s) to Song Lists in the Rocksmith profile.
     7) Choose a filter and write the resulting song list to Favorites in the Rocksmith profile.
     8) Utilities (database reports, profile management.)
     0) Exit program.
     h) Help.
   
   Choose>

   All of the text menus and text prompts will ask you to either select a number or select y/n (followed by enter to action).

7. At this menu, you first need to select a Steam account id, so choose 1 to start a text menu for selecting from the available Steam account ids. For this tutorial, our selection options look like this:

   Please select a Steam account id/Rocksmith file set from the following options.
   
   1) Steam user '12345678'. This is the user logged into Steam now. (Sat Sep  1 16:47:25 2018).
   0) Do nothing and raise error.

   We get a bit of help here - only one Steam id is available, and it is the user logged into Steam now. So we choose 1 to select user 12345678.

   Most people will only have one account id available - if you have more than one, you may need a bit of trial and error to work out which one in is yours. The easiest way to do this is select an id and then check if the Testing profile can be selected (next step). If not, you have the wrong Steam id and need to try another one.

8. After selecting a Steam id, you need to select a user profile for song list creation. Choose 2 to start this process, and then choose a profile (‘Testing’ for this tutorial). After completing this process, the first two information lines of the song list menu should be similar to:

   Steam account id:    '12345678'
   Rocksmith profile:   'Testing'

9. At this point, it’s worth saving the changes you have made.

   Select 0 to exit the program.

   You will then be offered the option to save changes to the configuration file. Choose y.

   After this, your working directory should contain the following files and sub-directories:

   ArrangementsGrid.xml    - If you put this file in the working directory.
   RS_Arrangements.sqlite  - The song list arrangements database.
   config.toml             - The default configuration file. Heart and brains of the
                             system. More on this below.
   song_lists.bat          - If you created it.
   .\RS_backup             - Backups of Rocksmith save files will be stored here.
   .\RS_update             - Changed save files will be stored here before copying
                             back to Steam.
   .\RS_working            - Save files will be copied from Steam to this folder
                             before working on them.

   If your working directory doesn’t match this, try this step again.

Clone Profile

Optional, but recommended. Clone data into the Testing profile. If you clone data from your main profile, you can test out the song list filters before overwriting the song list in your main profile.

I’ll assume we are cloning data from the profile ‘Eric the Half a Bee’ into ‘Testing’. This will replace all data in the Testing profile.

Run the batch file.

Select the utilities sub-menu, and then select Clone profile.

Make sure you get the next two right. Cloning destroys data in the profile you are copying to (the target).

Select the source profile for cloning. For the tutorial, I’m copying FROM ‘Eric the Half a Bee’.

Select the target profile for cloning. For the tutorial, I’m copying TO ‘Testing’.

A yes/non confirmation message will pop up. Check that the cloning operation is doing what you expect, and if so choose y.

Return to the main menu and exit the program. No need to save config changes this time.

Now is a good time to start up Rocksmith and check the Testing profile:

• To see that it still works after cloning.

• To check that the data from your main profile has been copied in correctly.

Song List Testing

The package is now set up with a default configuration, which you can use for some basic testing before creating your own song list filters - or you can skip this step and go straight to making your own.

Run the batch file and check that the Steam account id and profile are as expected:

Steam account id:     '12345678'
Rocksmith profile:   'Testing'

Experiment with the reporting options:

• Toggle between reporting to file and console (File reports are saved in the working directory).

• Test out reports on a single filter and on a filter set.

If you are reporting to the console, you will almost certainly need to scroll up to see the report output, as the song list menu takes up most of the normal console window.

Also experiment with the reporting options in the utility sub-menu. These reports may be useful when developing your own filters.

If you are happy with the reporting, you can try writing one of the default song list sets to Rocksmith - either "E Standard" for lead players or "Bass or Rhythm" for bass and rhythm players. Before you do this, I would recommend doing a text report for the song list set and checking it looks sensible. And finally, before writing to Rocksmith, please remember that this is going to replace existing song lists in the profile (use a test profile for testing!).

The default E Standard song list for lead players will create the following song lists:

1. E Standard 440 leads that have been played 1-12 times in Learn a song.

2. E Standard 440 leads that have been played 13-27 times in Learn a song.

3. E Standard 440 leads that have been played 27 or more times in Learn a song.

4. E Standard songs with an off concert pitch (i.e. not A440) that have been played once.

5. Will not be changed.

6. All E Standard songs that you have played in easy score attack, but haven’t yet got a platinum pick.

The bass or rhythm song list set generates a similar set of song lists.

Once you have written a song list set to Rocksmith, exit the package, open up Rocksmith, load the test profile and check the song lists to see if they match expectation (song lists 1, 2 or 3 may be empty you if haven’t played any songs that match the filter criteria.

If you are happy with all of this, the next step is to edit config.toml to create your own song list filters.

The Configuration File

All song lists are driven by the config.toml file in the working directory. This section describes the structure of this file. If you end up with major problems with this file, I suggest renaming the problem file and creating a new config file by following the set up steps in the tutorial (you can also try contacting me for help).

TOML is somewhat similar to windows .ini files. I’ve used it because it is a human readable/editable text form that “just works” and because python appears to be leaning towards it as a standard for configuration files. It’s a bit fiddly to edit for the data structures used in rsrtools, but it’s nowhere near as bad as JSON (which was the likely alternative).

Unfortunately, if any of the the TOML is malformed, the song list creator will throw an error and exit. However, when this happens, you will (hopefully) get an informative error message that will help you track the problem down. And a gotcha - the input is validated in two stages - some checking when loading, and some checking values when creating the song lists. So your debugging may need to be two stage as well. I’d also suggesting setting up one song list at a time to minimise your pain.

TODO I’m planning to put together some form of primitive filter builder as part of the next round of updates

I suggest that you open and look at `config.toml` while reading the rest of this section.

The configuration file is broken into three sections:

[settings]
...

[filters]
...

[song_list_sets]
...

Note that correct parenthesis type and double quoting is vital, and ... shows something I will fill in more detail on later. For this section, text should be typed as shown with the exception of text in angle brackets <>, which represents user defined names and input. You should replace both the angle brackets and the guide text with your own text. For example:

• "<filter name>" would become "E Standard".

• "<value 1>" would become "David Bowie".

• <list field name> would become Tuning.

Note that double quoting is typically required where shown – this provides protection for fields with spaces and non-standard characters. The only fields that do not need double quoting are <list field name> and <range field name> as these have a limited set of valid values, and none of them contain spaces or special characters. The values for include (true or false) and ranges (numbers) must not be quoted, and the values for mode ("AND", "OR") should be quoted.

Settings

The settings section is the simplest of the three, describing the location of the CFSM xml file (optional), the default Steam account id, and the default profile name:

[settings]
CFSM_file_path: "D:\\RS Stuff\\Working\\ArrangementsGrid.xml"
steam_account_id": "12345678"
player_profile": "Testing"
version = "x.x.x"

Version is for future functionality.

Song List Sets

The song list sets section is just about this simple as the settings - each song list set is a named list containing up to six filter names that will be used to create the song lists in the Rocksmith profile (the next part of this section describes filter definitions). The following example shows the structure:

[song_list_sets]
"E Standard" = [ "E Std Low Plays", "E Std Mid Plays", "E Std High Plays",
    "E Std Non Concert", "", "Easy E Std Plat Badge in progress",]
"Non E Std Tunings" = [ "Drop D", "Eb Standard", "Eb Drop Db", "D Standard",
    "D Drop C", "Other Tunings",]
Testing = [ "Artist test", "Played Count of 1 to 15",]

The song list set names are “E Standard”, “Non E Std Tunings”, and “Testing”. You can choose your own unique names for filter sets when you add them. The “E Standard” song list set consists of five unique filters - three filters for E 440 with differing play counts, an E standard non 440, and an easy platinum score attack in progress filter. It also includes “” for the fifth filter - this tells the song list creator to leave the fifth song list in the profile unchanged.

In summary, the format of a song list set is:

"<set name>" = [ "<filter 1>", "<filter 2>", "<filter 3>", ... "<filter 6>"]

where the values in <> are the song list set names, the filter names or empty to skip a song list ("").

The song list creator will only modify as many song lists as there are filters defined (up to six), and will not change any list with “” specified for the filter. So the “Testing” filter set will only modify song list one and two and will leave lists 3-6 unchanged.

Filters

The filters section consists of a list of named filters, where each named filter is made up of the following elements:

• The basic filter definition (one only per filter).

• One or more sub-filters, which in turn may be either list type or range type.

The following sections detail these elements.

Basic Filter Definition

A basic filter definition has the form:

[filters."<filter name>"]
base = "<base filter name>"
mode = "<mode value>"

The filter can either have a base filter, in which case the filter criteria will be applied to records generated from the base filter, or if base filter is the empty string ("") the filter will be applied to all records in the arrangements database. That is, the base filter is an optional field that allows building of nested or hierarchical filters.

Mode must be either "AND" or "OR", and specifies the way that sub-filters will be combined. For "AND", the filter will only return the records that match all of the sub-filters, while for "OR", the filter will return all records that match at least one of the sub-filters (i.e. AND narrows, while OR is inclusive).

List Type Sub-filter

The list type sub-filter is of the form:

[filters."<filter name>".sub_filters.<list field name>]
include = <true or false>
values = [ "<value 1>", "<value 2>", ... , "<value N>",]

<list field name> must be one of the list type field names:

SongKey
Tuning
ArrangementName
ArrangementId
Artist
Title
Album

ArrangementId may be useful for building song lists of alternative/bonus arrangements.

In future, the utilities menu will provide a list of these field names.

include must be true or false. If true, the filter will return the records for song arrangements whose field value matches any of the values in the list. If false, the filter will return the records for song arrangements whose field value does not match any of the values in the list. E.g. if the field name is Artist and the values are “Queen” and “Roxette”, then an include value of true will return only song arrangements by Queen and Roxette. If include is false, then all arrangements except songs by Queen and Roxette will be returned.

The list values must match values in the arrangements data and must be double quoted - the easiest way to check on validity is to run the relevant reports in the utilities menu of the song list creator (e.g. Tunings, Arrangement Types, Artists, Album Names and Track Titles).

GOTCHA: Values must be exact matches on content and case. So “E Standard” works, but “e standard” doesn’t, likewise it must be “Foo Fighters”, and not “Foo f” or “foo fighters”. I may add wild card support at some point in the future if there is strong support for it.

Range Type Sub-filter

The range type sub-filter is of the form:

[filters."<filter name>".sub_filters.<range field name>]
include = <true or false>
ranges  = [ [<low1>, <high1>], [<low2>, <high2>] ]

<range field name> must be one of the range type field names:

Pitch
Tempo
NoteCount
Year
PlayedCount
MasteryPeak
SAEasyCount
SAMediumCount
SAHardCount
SAMasterCount
SAPlayedCount
SAEasyBadges
SAMediumBadges
SAHardBadges
SAMasterBadges

In future, the utilities menu will provide a list of these field names.

A note of caution: I’m pretty sure the MasteryPeak values are not quite right. At the moment, I’m calculating these by multiplying the raw mastery peak value from the player profile by 100. However, this value doesn’t quite match the reported value in Rocksmith. I can fix this quickly if anybody knows the correct calculation.

SA stands for score attack, SA*Count is the score attack play account at the level, and SAPlayedCount is the total score attack play count.

The SA*Badges values have the following meanings:

• 0 No badge/not played yet.

• 1 Strike out/three red crosses.

• 2 Bronze/two red crosses

• 3 Silver/one red cross

• 4 Gold

• 5 Platinum

When I set up a badge filter, I’m normally only interested in songs I have played and haven’t yet got a a platinum badge for, so I use a range value of [[1, 4]]. I generally filter zero out, as otherwise the filter returns all un-played arrangements.

include must be true or false. If true, the filter will return only those song arrangement records that have field values in the ranges specified in the ranges list. If false, the filter will return those song arrangement records that have field values that do not appear in any of ranges in the ranges list.

ranges is a list of numeric low/high value pairs. The only constraint on the values is that they must be greater than or equal to zero. Note that the number values are not double quoted. If you enter a low value that is greater than the high value, the package will assume you have your numbers backward and will swap them silently.

For example, for a field name of PlayedCount and ranges = [[1,10],[18,19]] and include = true, the filter will return all arrangements with Learn a Song play count in the range 1 to 10 or 18 to 19. If include is false, the filter will return all arrangements that have a play count that is either: 0, in the range 11 to 17, or greater than or equal to 20.

Filter Examples

The following examples taken from the default set of filters illustrate most of the filter features.

First up, a filter for songs with (mostly) lead arrangements:

[filters."Not Bass, Rhythm"]
base = ""
mode = "OR"

[filters."Not Bass, Rhythm".sub_filters.ArrangementName]
include = false
values = [ "Bass", "Bass2", "Rhythm", "Rhythm1", "Rhythm2",]

[filters."Not Bass, Rhythm".sub_filters.Title]
include = true
values = [ "Cissy Strut",]

This filter is interpreted as follows:

• The filter is named “Not Bass, Rhythm”.

• It does not have a base filter, so it will apply the filter to the entire record set in the arrangement database.

• There are two sub-filters. The first filter excludes all arrangements with bass or rhythm types (this could have as easily been an include on all lead and combo types). The second filter includes the arrangements for one song: Cissy Strut by The Meters.

• The "OR" mode combines the results of the sub-filters.

In effect, this filter results in the records for all arrangements that are lead type along with the arrangements for Cissy Strut. This filter ensures that I can see all lead tracks and Cissy Strut, which only has bass and rhythm arrangements, but I still want it to appear in my song lists.

The following filter narrows the lead-ish filter to E Standard tunings:

[filters."E Standard"]
base = "Not Bass, Rhythm"
mode = "AND"

[filters."E Standard".sub_filters.Tuning]
include = true
values = [ "E Standard",]

This nested filter is interpreted as taking the records generated by the “Not Bass, Rhythm” filter and keeping only those arrangements with an E Standard tuning.

The final filter generates a list of E Standard tunings which are off concert pitch (i.e. not A440 tunings):

[filters."E Std Non Concert"]
base = "E Standard"
mode = "AND"

[filters."E Std Non Concert".sub_filters.Pitch]
include = false
ranges = [ [ 439.5, 440.5,],]

[filters."E Std Non Concert".sub_filters.PlayedCount]
include = true
ranges = [ [ 1.0, 5000.0,],]

This filter builds on the results of the “E Standard filter” by keeping only records which:

• Have a pitch outside the range 439.5 to 440.5 Hz (include = false). That is, this removes all A440 tunings,

• AND (mode = "AND") have a play count between 1 and 5000 (i.e. this removes tracks with a play count of zero - at least if like me, none of your play counts are within any sort of distance of 5000).

Nested vs. Flat Filters

The examples in the previous section demonstrate how to build up filters using a nested or hierarchical approach.

This nesting capability improves re-usability of filter logic and makes assembling complex filters quite a lot simpler. (This mechanism could definitely be improved further still, but hey, it’s only a simple play list creator.)

You can build also build up a complex filters by using multiple sub-filters in a single filter. For example, something close to the nested filters for the off concert pitch E Standard arrangements could have been built in with a single filter applying the following sub-filters:

[filters."One Step E Std Non Concert"]
base = ""
mode = "AND"

[filters."One Step E Std Non Concert".sub_filters.ArrangementName]
include = false
values = [ "Bass", "Bass2", "Rhythm", "Rhythm1", "Rhythm2",]

[filters."One Step E Std Non Concert".sub_filters.Tuning]
include = true
values = [ "E Standard",]

[filters."One Step E Std Non Concert".sub_filters.Pitch]
include = false
ranges = [ [ 439.5, 440.5,],]

[filters."One Step E Std Non Concert".sub_filters.PlayedCount]
include = true
ranges = [ [ 1.0, 5000.0,],]

(This is something close, because it’s not possible to build a one shot filter like this that also capture the Cissy Strut arrangements).

To date I have always found the most effective way to build the filters is to use simpler filters based on one or two sub-filters, and then build complexity by nesting. (Either way is fine of course, so go with whatever works best for you.)

Something Went Wrong!

Something unexpected has happened with loading a profile in Rocksmith? All is (probably) not lost. Before rsrtools writes files to the Rocksmith Steam folders, it creates a zip archive of all of the key files associated with the Steam account id. These backups are kept in the working directory under RS_backup.

To restore a backup, extract the contents of the zip file and copy the contents into your Steam Rocksmith save folder. For most people, this should be in your Steam install directory under:

<Steam directory>\userdata\<steam_account_id>\221680

<steam_account_id> is the same Steam account id used in the rsrtools songlists menu.

As a check, this folder should contain a remotecache.vdf file and a remote sub-directory. The remote subdirectory should contain a file named LocalProfiles.json and one or or more files with names ending in _PRFLDB.

Database Structure

For those who are interested, the database is structured as two tables, which contain song arrangement data and player performance data. The filters are executed on a join of these two tables.

The string fields are the same fields defined in the List Type Sub-filter section, and the numeric fields are those defined in the Range Type Sub-filter section.

Package Caveats

Be aware that the package currently has a couple of irritating quirks:

• It can’t distinguish between the representative (default) arrangement on a path and the alternative/bonus arrangements on that path (i.e. it can’t tell which of the leads is the default).

• A related issue. It can’t tell which path Rocksmith (OG) combo tracks should be allocated to.

I know how to resolve the issue, but it is waiting on the song scanner implementation. The way I work around this is to play all of the tracks that I want to show up in a filter at least once, and then apply a minimum play count criteria. For my use case, this is mainly an issue for E standard arrangements - I don’t tend to worry about this for the alternate tunings.

Sidebar: Rocksmith Save File Editing

The primary purpose of this package is to provide facilities for customising Rocksmith song lists. However, along the way I needed to develop classes for opening, editing and saving Rocksmith save files (profiles).

If you are interested in using this functionality, you should start with RSProfileManager in profilemanager.py, which is the primary class for managing Rocksmith profiles and their associated steam and Rocksmith metadata. The class methods are currently only documented in their docstrings, although I plan to provide more detail in this document in the future (and I’m happy to answer questions via github issues).

The RSProfileManager class provides two simple examples of profile editing:

• RSProfileManager.cl_set_play_counts(), which is a command line mechanism for setting the ‘Learn a Song’ play counts for one or more song arrangements.

• RSProfileManager.cl_clone_profile(), which is a command line mechanism for cloning a player data from one profile into another (a destructive copy).

Both of these routines can be run from the command line. For further details see the profile manager help, which can be obtained from the command line:

python -m rsrtools.files.profilemanager -h

The song list creator also uses the profile manager to obtain player data and to write song lists into player profiles.

These methods either a) implement very small changes to save files with a lot of care to maintain Rocksmith formats or b) replace Rocksmith data with Rocksmith data. Consequently their implementations are buried within classes used by the profile manager.

If you want to make more general changes to Rocksmith profiles, you can use the methods:

RSProfileManager.get_json_subtree()
RSProfileManager.set_json_subtree()
RSProfileManager.mark_as_dirty()

Before explaining these methods, please note that I haven’t yet done any testing of these three methods - they are so simple that they should work out of the box, but bugs are possible, so please be careful with your save files. If you want a somewhat safer path for changes, please make a feature request on github and we’ll see what we can work up. I will do some testing when I add a demonstration utility in a future release (specifically, the feature request to delete progress for specified CDLCs).

With that warning out of the way, onto the approach. The general steps are:

1. Create a profile manager instance (pm), which will need a working directory.

2. Read json data from a profile using:

   pm.get_json_subtree(profile_name, json_path).

   Keep in mind this may return a mutable (list, dict), in which case, editing the json data is effectively editing the profile data. (My preferred approach is to edit a copy and write the copy back using set_json_subtree). If you do choose to edit a mutable json object, you need to let the profile manager know that you have done this by calling:

   pm.mark_as_dirty(profile_name)

3. If you are working on new data, a copy of data obtained from get_json_subtree, or a non-mutable value, replace the instance data in the profile manager with the new data by:

   pm.set_json_subtree(profile_name, json_path, new_values)

   This approach will automatically mark the instance data for profile_name as dirty.

4. Write the files to the update folder (and generate backups along the way):

   pm.write_files()

5. Move the updated files to the Steam folder:

   pm.move_update_to_steam(steam_account_id)

   Note that it’s up to you to ensure that the save files match up with the steam account id (the method doesn’t check this).

And finally, a brief explanation of json_path: the get/set_subtree methods use a JSON path to navigate save data elements in the Rocksmith profile JSON dictionary. A JSON path is a list or tuple of the elements used to locate a specific value or subtree in the save data. E.g. the json_path to song list 2 is:

('SongListsRoot', 'SongLists', 1)

and the Learn a song play count for Take Me Out is:

("Stats", "Songs", "AB6880DBE00E6E059A5B8449873BE187", "PlayedCount")

(I grabbed the Take Me Out Arrangement Id of AB6880DBE00E6E059A5B8449873BE187 from an rsrtools report.)

TODO

• Convert major TODO items to issues.

• Add whatever functionality is needed for rs-manager to use rsrtools as an integration option.

• Complete PSARC scanner (welder.py).

• Add more substantial documentation on profile manager (for Rocksmith file editing), database, and song lists (hooks for GUI implementations).

• Add command line option for profilemanager to dump profile json to file.

Changelog

0.1.2beta 2019-04-26 Mac OS X support added.

0.1.1beta 2019-04-26 Minor updates to refer to Steam account id and Steam user id correctly. All Steam support functions moved to steam.py. Some windows specific Steam functions removed and replaced with methods based on Steam vdf files.

0.1.0beta 2019-04-22 First functional beta release for rsrtools. Windows only.

0.0.1 2019-03-12 Place holder release to lock package name down in pypi.

Development notes

20190421 Song list creator and database modules functional, first draft of documentation complete. 0.1 release imminent.