File "midiToText.py", line 10, in convertMidiFile csv_string = pm.midi_to_csv(midiName) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midicsv.py", line 22, in parse pattern = read_midifile(file) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 221, in read_midifile return read_midifile(inp) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 223, in read_midifile return reader.read(midifile) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 45, in read self.parse_track(midifile, track) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 87, in parse_track event = self.parse_midi_event(trackdata) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 133, in parse_midi_event data = [trackdata.get_data_byte() for _ in range(cls.length)] File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 133, in <listcomp> data = [trackdata.get_data_byte() for _ in range(cls.length)] File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 36, in get_data_byte self.assert_data_byte(byte) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 29, in assert_data_byte assert data & 0x80 == 0, self.errmsg("Unexpected status byte", data) AssertionError: Unexpected status byte 0x80 at position 35078

https://easyupload.io/zz2wf6 midi file that causes the issue

when i comment out the assert the conversion passes, though i get a different error later with a different midi

Traceback (most recent call last): File "midiToText.py", line 35, in <module> convertMidiFolderOfFolders(midiName) File "midiToText.py", line 20, in convertMidiFolderOfFolders convertMidiFolder(path + "/" + midiName) File "midiToText.py", line 16, in convertMidiFolder convertMidiFile(path + "/" + midiName) File "midiToText.py", line 10, in convertMidiFile csv_string = pm.midi_to_csv(midiName) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midicsv.py", line 22, in parse pattern = read_midifile(file) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 221, in read_midifile return read_midifile(inp) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 223, in read_midifile return reader.read(midifile) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 45, in read self.parse_track(midifile, track) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 87, in parse_track event = self.parse_midi_event(trackdata) File "/home/user/anaconda3/lib/python3.8/site-packages/py_midicsv/midi/fileio.py", line 103, in parse_midi_event raise Warning("Unknown Meta MIDI Event: " + repr(cmd)) Warning: Unknown Meta MIDI Event: 10 https://easyupload.io/8y3mz2

file that causes the issue

you should rename the command line midicsv and csvmidi scripts that you included. When I did the pip install, it overwrote my existing binaries from John Walker that were already located in /user/local/bin...

Secondly, if you're going to overwrite, then the pip installer should prompt me to make sure that is what I really want to do

thirdly, you should try to make your version of midicsv and csvmidi use the same command line options as the original binary so that if for some reason it does get overwritten and maybe I would even prefer to use your version for the Latin decoding...but if I have other scripts elsewhere based on using the binary midicsv, I don't want to break all those scripts.

Hello and thank you for this great library! I am using it to do some processing on midi files, but I am having an issue with the TimeSignature of certain midi files.

Making a csv csv_string = py_midicsv.midi_to_csv(source) gives me the line 1, 0, Time_signature, 4, 2. Then when I try to parse it back to midi (without any processing) midi = py_midicsv.csv_to_midi(csv_string) I get the following error:

File ".../py_midicsv/csvmidi.py", line 46, in parse
    event = csv_to_midi_map[identifier](tr, time, identifier, line[3:])
File ".../py_midicsv/csv_converters.py", line 142, in to_TimeSignatureEvent
    num, denom, click, notesq = map(int, line)
ValueError: not enough values to unpack (expected 4, got 2)

Is this a known issue and how can I prevent it?

in https://pypi.org/project/py-midicsv/#description

we have

import py_midicsv

# Load the MIDI file and parse it into CSV format
csv_string = py_midicsv.midi_to_csv("example.mid")

# Parse the CSV output of the previous command back into a MIDI file
midi_object = py_midicsv.csv_to_midi(csv_string)

# Save the parsed MIDI file to disk
with open("example_converted.mid", "wb") as output_file:
    midi_writer = py_midicsv.FileWriter(output_file)
    midi_writer.write(midi_object)

in https://github.com/timwedde/py_midicsv

we have

import py_midicsv as pm

# Load the MIDI file and parse it into CSV format
csv_string = pm.midi_to_csv("example.mid")

# Parse the CSV output of the previous command back into a MIDI file
midi_object = pm.csv_to_midi(csv_string)

# Save the parsed MIDI file to disk
with open("example_converted.mid", "wb") as output_file:
    midi_writer = pm.FileWriter(output_file)
    midi_writer.write(midi_object)

but none of them work or i did something wrong

Traceback (most recent call last):
  File "C:/Users/felipe/PycharmProjects/2020/PORRA.py", line 9, in <module>
    midi_writer.write(midi_object)
NameError: name 'midi_object' is not defined

First of all, congrats for this Python library, which makes easier and integrable the parsing from MIDI files to CSV in Python. I'm encountering a problem while parsing a certain file, which gives the following error:

Traceback (most recent call last): File "/Users/josepdecidrodriguez/Google Drive/IU/AML/Project/BachPropagation/src/dataset/parser.py", line 27, in data = py_midicsv.midi_to_csv(file) File "/Users/josepdecidrodriguez/Envs/BachPropagation-L4F7V3ER/lib/python3.7/site-packages/py_midicsv/midicsv.py", line 31, in parse csv_file.append(midi_to_csv_map[type(event)](index + 1, abstime, event)) File "/Users/josepdecidrodriguez/Envs/BachPropagation-L4F7V3ER/lib/python3.7/site-packages/py_midicsv/midi_converters.py", line 117, in from_KeySignatureEvent return write_event(track, time, "Key_signature", [event.get_alternatives(), '"major"' if event.data[1] == 0 else '"minor"']) IndexError: list index out of range

This is given by the following file BWV_809____piano.txt (I've changed the extension to .txt to be able to upload it here).

The commits in this pull request together enable the package to read in any(?) valid midi file, convert it to CSV, and convert the CSV back to midi, without losing any data (or introducing any other changes). This was "almost" true already; most of the changes are small adjustments for particular event types.

My test set of almost 5000 midi files inadvertently contained an invalid file, which (at first) the package happily translated to CSV and (almost) back to the original, without indicating a problem. The "Validate byte types in midi input" commit makes the package catch the bad file and report the place in the input where the problem was found. (Big points to Python's design for enabling that to be implemented cleanly!)

See the commit comments for more details on the individual changes.

Hi timwedde. I'm trying to build a dataset of midi-files for use in a machine learning project. I've been using py_midicsv to try and remove the drum track from my files. I've tried looking for "Program_c, 9" in the file after conversion to CSV, deleting that whole section and converting back to MIDI, but something is breaking. Could you maybe suggest a better way to do this? Thanks for the cool module!

Tree project

├── csv
│   └── AllBlues.csv
├── mid
│   └── AllBlues.mid
├── midi
│   └── AllBlues.mid

Implementation

import os
import py_midicsv as pm

#MIDI folder
file_list=os.listdir(r"midi")

#Custom Path
path = ''

#Create folders where to save files
try:
    os.mkdir(path +"csv")
    os.mkdir(path +"mid") 

except OSError:
    print ("Creation of the directory %s failed" % path)
else:
    print ("Successfully created the directory %s" % path)
    
for file_name_song in file_list:
    
    name_file_without_ext = file_name_song.rsplit('.', 1)[0]
    
    # Load the MIDI file and parse it into CSV format
    csv_string = pm.midi_to_csv("midi/"+ name_file_without_ext + ".mid")

    with open("csv/" + name_file_without_ext + ".csv", "w") as f:
        f.writelines(csv_string)

    # Parse the CSV output of the previous command back into a MIDI file
    midi_object = pm.csv_to_midi(csv_string)

    # Save the parsed MIDI file to disk
    with open("mid/" + name_file_without_ext + ".mid", "wb") as output_file:
        midi_writer = pm.FileWriter(output_file)
        midi_writer.write(midi_object)
    print("processed file name:",file_name_song)

As you mention here, the midi_to_csv function returns a list of strings. write is expecting a string rather than a list of strings, so running the example results in an error. Your provided fix was to use the writelines function instead, but you forgot to update the README.

Ran this on Google Colab, I uploaded a midi file and tested multiple and kept getting this error. What's going on? (Also, IDK why the indentations arent working on the code) ` #MIDI TO CSV

import py_midicsv as pm def midi2csv(): save=input('filename\n') print("Step 1") owo=open(save,mode='w') print("Step 2") owo.write(str(pm.midi_to_csv(str(save)+'.mid'))) print("Step 3") owo.close print("Step 4") messagebox.showinfo('Complete','file saved as {}'.format(save)) print("Step 5") while 1==1: x=input('midi to CSV: 0\nend: 2\n') if x == '0': midi2csv() elif x == '2': break `

When converting CSV to MIDI, csv.reader requires a file object. I added the option to pass a file path that can be activated with an extra boolean parameter in order to not break current version behaviour.