pianosnake / ireal-reader Goto Github PK

#4 fixes a bug with filenames and case sensitivity. However, the version of the code I get from npm doesn't have this fix. Is it possible to push this to npm again?

Hi,
thanks for this nice library! I use it to extract chord progressions. When doing some tests, I saw that chords containing "add" and "sus" are not correctly parsed. I guess this is due to the fact that they are not captured by this regexp expression.

Are you interested in me working on this? I'd like to understand why it's necessary to do the regexp matching so strictly. Can we just add 'a', 'd', 's', and 'u' to the character classes - which would allow "add" and "sus" but also allow rubbish like "Caus"? Or do you think it should be a proper regexp that refuses wrong chords?

Cheers, Boris

Hey there, nice work!
Would you accept a pull request that extends the lib with a method to return all unprocessed ireal tokens? I am currently tinkering with leadsheets, trying to get a nice format for a playalong lib. The unscrambling and parsing of the song data works well will ireal-reader, but the returned measures lack the meta info (repeat signs, sections, comments etc).
My current solution is a combination of your lib to parse the links + modified parts of ireal-renderer to get the raw ireal tokens. The parsed tokens look like this:

[
  {
    "annots": [
      "*A", // section A
      "T44", // 44 time
      "{" // repeat start
    ],
    "comments": [],
    "bars": "{",
    "spacer": 0,
    "chord": {
      "note": "E",
      "modifiers": "h7",
      "over": null,
      "alternate": null
    },
    "token": " "
  },
  { // this is just empty space
    "annots": [],
    "comments": [],
    "bars": "",
    "spacer": 0,
    "chord": null,
    "token": " "
  },
  /* ... */
  {
    "annots": [],
    "comments": [],
    "bars": "Z", // last bar
    "spacer": 0,
    "chord": null
  }
]

At this point, no information of the sheet is lost. To simplify the format, I can then parse the tokens to a format that omits everything that is just layout related (empty spacers, unused props):

[
    {
        "chords": ["Eh7"],
        "section": "A",
        "time": "44",
        "signs": ["{"]
    },
    {
        "chords": ["A7b9"]
    },
    /* ... */
    {
        "chords": ["Eh7", "A7b9"]
    },
    {
        "chords": ["D-"],
        "section": "B",
        "house": 1
    },
    {
        "chords": ["G-7"]
    },
   /* .. */
    {
        "chords": ["A7b9"],
        "signs": ["}"]
    },
    {
        "chords": ["D-"],
        "section": "C",
        "house": 2
    },
    {
        "chords": ["G-7"]
    },
    /* .. */
    {
        "chords": ["D-"]
    }
]

This format is suitable for rendering a sheet visually. To render it aurally, all repeat signs/houses must be resolved.
I am also using another snippet format that allows easy editing with a textarea:

|: Eh7    |  A7b9         |  D-   |  D-        |
|  G-7    |  C7           |  F^7  |  Eh7 A7b9  |
|1 D-     |  G-7          |  Bb7  |  A7b9      |
|  D-     |  G7#11        |  Eh7  |  A7b9     :|
|2 D-     |  G-7          |  Bb7  |  A7b9      |
|  D- B7  |  Bb7#11 A7b9  |  D-   |  D-        |

If you want, you can play around with it here.
To spare you from more details, I would be happy if this lib could handle the token magic, maybe even parsing the tokens to the format of the second snippet I wrote.
Greetings,
Felix

I found the error below after running the code.
May I get some help please?
Error:

 const percentDecoded = decodeURIComponent(percentEncoded[1]);
                                                          ^

TypeError: Cannot read properties of null (reading '1')

Code:

const fs = require("fs");
const iRealReader = require("ireal-reader");

fs.readFile("./On The Sunny Side Of The Street.xml", "utf8", function (err, data) {
  if (err) throw err;
  const playlist = iRealReader(data);
  console.log(playlist);
});

On The Sunny Side Of The Street.txt