symbtrdataextractor: symbtrdataextractor v2.1.0

symbtrdataextractor

Python tools for extracting relevant (meta)data from SymbTr-scores

Introduction

This repository contains to algorithms to extract metadata, related music knowledge and section information from SymbTr-scores and related information sources.

Currently you can:

• Obtain the makam, usul, form, name and composer of the given SymbTr score
• Extract section boundaries from both the implicit and explicit section information given in the SymbTr scores
• Extract annotated phrase boundaries in the SymbTr scores
• Add user provided segment boundaries in the SymbTr scores
• Analyse the melody and the lyrics of the sections, phrase annotations and user provided segments, and apply semiotic labels
• Query relevant metadata from MusicBrainz, if the MBID is supplied.
• Read the metadata stored in the header of the mu2 file.

If you are using this code for academic purposes please cite the software as:

Sertan Senturk. (2016). symbtrdataextractor: symbtrdataextractor v2.0.0. Zenodo. 10.5281/zenodo.50033

Usage Extracting (meta)data from the txt-score:

from symbtrdataextractor.dataextractor import DataExtractor extractor = DataExtractor( melody_sim_thres=0.7, lyrics_sim_thres=0.7, save_structure_sim=True, extract_all_labels=False, crop_consec_bounds=True, get_recording_rels=False, print_warnings=True) """ Inputs ---------- melody_sim_thres : (optional) the maximum similarity threshold for two melodic stuctures to be considered as variant of each other. The default is 0.75. lyrics_sim_thres : (optional) the maximum similarity threshold for two lyric stuctures to be considered as variant of each other. The default is 0.75. save_structure_sim : (optional) boolean to add the melodic and lyrics similarity between each section and segment to the output or not extract_all_labels: (optional) boolean to treat all (explicit) annotations in the lyrics as a section or not (e.g. INSTRUMENTATION labels). The default is False. crop_consec_bounds: (optional) remove the first of the two consecutive boundaries inside the user given segmentation boundaries get_recording_rels: (optional) boolean to extract the relevant recording relations from MusicBrainz. The default is False. print_warnings : (optional) boolean to print possible warnings during reading the scores. Note that errors will always be printed. The default is True """ txt_data, is_data_valid = extractor.extract(txt_filename, symbtr_name=scorename, segment_note_bound_idx=auto_seg_bounds, mbid=mbid) """ Inputs ---------- txt_filename : the filepath of the SymbTr-txt score symbtr_name : (optional) the SymbTr-name in the "makam--form--usul--name--composer" format. mbid : (optional) the work or recording mbid of the composition/performance related to the score segment_note_bound_idx : (optional) user provided segment boundaries. NOTE - the indexing should start from 1, as this is the convention in the SymbTr-txt scores. TIP - makam-symbolic-phrase-segmentation package (https://github.com/MTG/makam-symbolic-phrase-segmentation) can be used to segment the SymbTr-txt scores automatically. """

Important note: For the sake of consistency with the indexing in the SymbTr-txt scores, the output note indices start from 1, not 0! Likewise, segment_note_bound_idx input should also obey to the 1 indexing convention, otherwise the semiotic labels will be wrong.

Extracting metadata stored in the mu2 headers:

from symbtrdataextractor.reader.mu2 import Mu2Reader mu2_header, header_row, is_header_valid = Mu2Reader.read( mu2_filename, symbtr_name=scorename) """ Inputs ---------- mu2filename : the filepath of the mu2 score symbtr_name : (optional) the SymbTr-name in the "makam--form--usul--name--composer" format. """

For an interactive demo please refer to extractsymbtrdata.ipynb

Installation

If you want to install symbtrmetadataextractor, it is recommended to install symbtrmetadataextractor and dependencies into a virtualenv. In the terminal, do the following:

virtualenv env source env/bin/activate python setup.py install

If you want to be able to edit files and have the changes be reflected, then install the repository like this instead:

pip install -e .

Now you can install the rest of the dependencies:

pip install -r requirements Changelog

• Melodic organization is not replaced by the instrumental section names anymore
• Added melodic and lyrics similarity to output
• Default similarity threshold is lowered to 0.7 from 0.75
• Fixed typos (lyric -> lyrics) in variable names and output dictionary keys.
• Improved warnings

Authors

Sertan Senturk contact@sertansenturk.com

Reference

Thesis