Module music_df.humdrum_export.collate_spines

Functions

def collate_spines(files: List[str])

Expand source code

def collate_spines(files: t.List[str]):
    gcd = run_minrhy(files)
    tmp_paths = run_timebase(files, gcd)
    # rid strips null records (lines consisting only of ".")
    result = sh.rid("-d", _in=sh.assemble(*tmp_paths))  # type:ignore
    _clean_up(tmp_paths)
    return result

def run_minrhy(files: List[str]) ‑> int

Expand source code

def run_minrhy(files: t.List[str]) -> int:
    assert files
    out = sh.minrhy(*files)  # type:ignore
    # TODO: (Malcolm 2024-03-04) remove
    # print(out)
    # breakpoint()
    m = re.search(r"^all:\t(?P<result>\d+)", out, re.MULTILINE)
    if not m:
        # this occurs if there is only one file; I guess we should just
        #   inspect len(files)
        m = re.search(r"(?P<result>\d+)", out)
    assert m is not None
    out = int(m.group("result"))
    return out

def run_timebase(files: List[str], gcd: int) ‑> List[str]

Expand source code

def run_timebase(files: t.List[str], gcd: int) -> t.List[str]:
    tmp_paths = []
    for f in files:
        with open(f) as inf:
            data = inf.read()
        recips = re.findall(r"\d%(\d+)", data)
        if recips:
            # because `timebase`, being from the original humdrum toolkit, doesn't
            #   understand the "%" notation (e.g., 2%3), we first use rscale to scale up
            #   the file to the gcd of the recips (ensuring there will be no %), and
            #   then afterwards scale it back
            recip = math.gcd(*[int(r) for r in recips])

            _, tmp_path = tempfile.mkstemp(suffix=".krn")
            rscale1_result = sh.rscale("-f", f"1/{recip}", f)  # type:ignore
            with open(tmp_path, "w") as outf:
                outf.write(rscale1_result)
            timebase_result = sh.timebase(  # type:ignore
                "-t", str(gcd), tmp_path
            )
            with open(tmp_path, "w") as outf:
                outf.write(timebase_result)
            rscale2_result = sh.rscale("-f", f"{recip}", tmp_path)  # type:ignore
            with open(tmp_path, "w") as outf:
                outf.write(rscale2_result)
            tmp_paths.append(tmp_path)
        else:
            result = sh.timebase("-t", str(gcd), f)  # type:ignore
            _, tmp_path = tempfile.mkstemp(suffix=".krn")
            with open(tmp_path, "w") as outf:
                outf.write(result)
            tmp_paths.append(tmp_path)
    return tmp_paths