gtfreader 0.1.1__tar.gz → 0.2.0__tar.gz

gtfreader-0.2.0/PKG-INFO

@@ -0,0 +1,35 @@
+Metadata-Version: 2.4
+Name: gtfreader
+Version: 0.2.0
+Summary: Fast Cython-backed parsing for GTF attribute columns.
+Author: Endre Bakken Stovner
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Cython
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Operating System :: OS Independent
+Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
+Requires-Python: >=3.12
+Description-Content-Type: text/markdown
+Requires-Dist: pandas>=2.0
+
+# gtfreader
+
+Fast GTF reading into pandas DataFrames.
+
+## Install
+
+```bash
+python -m pip install -e .
+```
+
+## Example
+
+```python
+from gtfreader import read_gtf
+
+df = read_gtf("annotation.gtf")
+print(df.columns)
+print(df.head())
+```

gtfreader-0.2.0/README.md

@@ -0,0 +1,19 @@
+# gtfreader
+
+Fast GTF reading into pandas DataFrames.
+
+## Install
+
+```bash
+python -m pip install -e .
+```
+
+## Example
+
+```python
+from gtfreader import read_gtf
+
+df = read_gtf("annotation.gtf")
+print(df.columns)
+print(df.head())
+```

{gtfreader-0.1.1 → gtfreader-0.2.0}/gtfreader/__init__.py

@@ -14,8 +14,6 @@ from .readers import (
     read_gtf_full,
     read_gtf_full_python,
     read_gtf_python,
-    read_gtf_restricted,
-    read_gtf_restricted_python,
     to_rows,
     to_rows_keep_duplicates,
 )
@@ -28,8 +26,6 @@ __all__ = [
     "read_gtf_full",
     "read_gtf_full_python",
     "read_gtf_python",
-    "read_gtf_restricted",
-    "read_gtf_restricted_python",
     "to_rows",
     "to_rows_keep_duplicates",
 ]

gtfreader-0.2.0/gtfreader/_parser.pyx

@@ -0,0 +1,197 @@
+# cython: language_level=3
+# cython: boundscheck=False
+# cython: wraparound=False
+# cython: initializedcheck=False
+# cython: infer_types=True
+# cython: nonecheck=False
+
+cdef inline int _key_code(str s, Py_ssize_t a, Py_ssize_t b):
+    cdef Py_ssize_t n
+    n = b - a
+
+    if n == 7:
+        if (
+            s[a] == 'g' and s[a + 1] == 'e' and s[a + 2] == 'n' and
+            s[a + 3] == 'e' and s[a + 4] == '_' and s[a + 5] == 'i' and
+            s[a + 6] == 'd'
+        ):
+            return 1
+        if (
+            s[a] == 'e' and s[a + 1] == 'x' and s[a + 2] == 'o' and
+            s[a + 3] == 'n' and s[a + 4] == '_' and s[a + 5] == 'i' and
+            s[a + 6] == 'd'
+        ):
+            return 6
+
+    elif n == 9:
+        if (
+            s[a] == 'g' and s[a + 1] == 'e' and s[a + 2] == 'n' and
+            s[a + 3] == 'e' and s[a + 4] == '_' and s[a + 5] == 'n' and
+            s[a + 6] == 'a' and s[a + 7] == 'm' and s[a + 8] == 'e'
+        ):
+            return 3
+
+    elif n == 11:
+        if (
+            s[a] == 'e' and s[a + 1] == 'x' and s[a + 2] == 'o' and
+            s[a + 3] == 'n' and s[a + 4] == '_' and s[a + 5] == 'n' and
+            s[a + 6] == 'u' and s[a + 7] == 'm' and s[a + 8] == 'b' and
+            s[a + 9] == 'e' and s[a + 10] == 'r'
+        ):
+            return 5
+
+    elif n == 13:
+        if (
+            s[a] == 't' and s[a + 1] == 'r' and s[a + 2] == 'a' and
+            s[a + 3] == 'n' and s[a + 4] == 's' and s[a + 5] == 'c' and
+            s[a + 6] == 'r' and s[a + 7] == 'i' and s[a + 8] == 'p' and
+            s[a + 9] == 't' and s[a + 10] == '_' and s[a + 11] == 'i' and
+            s[a + 12] == 'd'
+        ):
+            return 2
+
+    elif n == 15:
+        if (
+            s[a] == 't' and s[a + 1] == 'r' and s[a + 2] == 'a' and
+            s[a + 3] == 'n' and s[a + 4] == 's' and s[a + 5] == 'c' and
+            s[a + 6] == 'r' and s[a + 7] == 'i' and s[a + 8] == 'p' and
+            s[a + 9] == 't' and s[a + 10] == '_' and s[a + 11] == 'n' and
+            s[a + 12] == 'a' and s[a + 13] == 'm' and s[a + 14] == 'e'
+        ):
+            return 4
+
+    return 0
+
+
+cdef inline str _cached_str(dict cache, str value):
+    cdef object obj
+    obj = cache.get(value)
+    if obj is None:
+        cache[value] = value
+        return value
+    return <str> obj
+
+
+def parse_chunk_columns(lines):
+    cdef Py_ssize_t n, row_i
+    cdef Py_ssize_t pos, m, key_start, key_end, val_start, val_end
+    cdef dict columns
+    cdef dict gene_name_cache
+
+    cdef str s, key, value
+    cdef list col
+    cdef int code
+
+    cdef object obj_col
+
+    cdef list gene_id_col
+    cdef list transcript_id_col
+    cdef list gene_name_col
+    cdef list transcript_name_col
+    cdef list exon_number_col
+    cdef list exon_id_col
+
+    n = len(lines)
+    columns = {}
+    gene_name_cache = {}
+
+    gene_id_col = None
+    transcript_id_col = None
+    gene_name_col = None
+    transcript_name_col = None
+    exon_number_col = None
+    exon_id_col = None
+
+    for row_i in range(n):
+        s = <str> lines[row_i]
+        m = len(s)
+        pos = 0
+
+        while pos < m:
+            while pos < m and (s[pos] == ' ' or s[pos] == '\t' or s[pos] == ';'):
+                pos += 1
+            if pos >= m:
+                break
+
+            key_start = pos
+
+            while pos < m and s[pos] != ' ' and s[pos] != '\t':
+                pos += 1
+            key_end = pos
+
+            while pos < m and (s[pos] == ' ' or s[pos] == '\t'):
+                pos += 1
+            if pos >= m:
+                break
+
+            if s[pos] == '"':
+                val_start = pos + 1
+                pos += 1
+
+                while pos < m and s[pos] != '"':
+                    pos += 1
+                if pos >= m:
+                    break
+
+                val_end = pos
+            else:
+                val_start = pos
+                while pos < m and s[pos] != ';':
+                    pos += 1
+                val_end = pos
+                while val_end > val_start and (s[val_end - 1] == ' ' or s[val_end - 1] == '\t'):
+                    val_end -= 1
+
+            value = s[val_start:val_end]
+
+            code = _key_code(s, key_start, key_end)
+
+            if code == 1:
+                if gene_id_col is None:
+                    gene_id_col = [None] * n
+                    columns["gene_id"] = gene_id_col
+                gene_id_col[row_i] = value
+
+            elif code == 2:
+                if transcript_id_col is None:
+                    transcript_id_col = [None] * n
+                    columns["transcript_id"] = transcript_id_col
+                transcript_id_col[row_i] = value
+
+            elif code == 3:
+                if gene_name_col is None:
+                    gene_name_col = [None] * n
+                    columns["gene_name"] = gene_name_col
+                gene_name_col[row_i] = _cached_str(gene_name_cache, value)
+
+            elif code == 4:
+                if transcript_name_col is None:
+                    transcript_name_col = [None] * n
+                    columns["transcript_name"] = transcript_name_col
+                transcript_name_col[row_i] = value
+
+            elif code == 5:
+                if exon_number_col is None:
+                    exon_number_col = [None] * n
+                    columns["exon_number"] = exon_number_col
+                exon_number_col[row_i] = value
+
+            elif code == 6:
+                if exon_id_col is None:
+                    exon_id_col = [None] * n
+                    columns["exon_id"] = exon_id_col
+                exon_id_col[row_i] = value
+
+            else:
+                key = s[key_start:key_end]
+                obj_col = columns.get(key)
+                if obj_col is None:
+                    col = [None] * n
+                    columns[key] = col
+                else:
+                    col = <list> obj_col
+                col[row_i] = value
+
+            pos += 1
+
+    return columns

{gtfreader-0.1.1 → gtfreader-0.2.0}/gtfreader/readers.py

@@ -23,7 +23,6 @@ GTF_DTYPES = {
     "Frame": "category",
 }
 GTF_NAMES = ["Chromosome", "Source", "Feature", "Start", "End", "Score", "Strand", "Frame", "Attribute"]
-RESTRICTED_ATTRIBUTE_COLUMNS = ["gene_id", "transcript_id", "exon_number", "exon_id"]
 
 
 def find_first_data_line_index(file_path: str | Path) -> int:
@@ -162,6 +161,7 @@ def _open_gtf_reader(
         header=None,
         names=GTF_NAMES,
         dtype=GTF_DTYPES,
+        comment="#",
         chunksize=chunksize,
         skiprows=skiprows,
         nrows=nrows,
@@ -204,22 +204,6 @@ def _parse_attributes_python(
     return to_rows(attribute_column, ignore_bad=ignore_bad)
 
 
-def _parse_restricted_attributes_compiled(attribute_column: pd.Series) -> pd.DataFrame:
-    if _parse_chunk_columns_compiled is None:
-        return _parse_restricted_attributes_python(attribute_column)
-
-    attribute_column = _normalize_attribute_series(attribute_column)
-    columns = _parse_chunk_columns_compiled(attribute_column.to_numpy(copy=False))
-    return pd.DataFrame(
-        {name: columns[name] for name in RESTRICTED_ATTRIBUTE_COLUMNS},
-        index=attribute_column.index,
-    )
-
-
-def _parse_restricted_attributes_python(attribute_column: pd.Series) -> pd.DataFrame:
-    return to_rows(_normalize_attribute_series(attribute_column)).reindex(columns=RESTRICTED_ATTRIBUTE_COLUMNS)
-
-
 def _read_gtf_full(
     path: Path,
     *,
@@ -242,55 +226,24 @@ def _read_gtf_full(
     return _finalize_gtf_frame(pd.concat(dfs, sort=False))
 
 
-def _read_gtf_restricted(
-    path: Path,
-    *,
-    skiprows: int,
-    nrows: int | None,
-    chunksize: int,
-    parse_attributes,
-) -> pd.DataFrame:
-    dfs: list[pd.DataFrame] = []
-    with _open_gtf_reader(path, chunksize=chunksize, skiprows=skiprows, nrows=nrows) as df_iter:
-        for df in df_iter:
-            subset = parse_attributes(df["Attribute"])
-            dfs.append(
-                pd.concat(
-                    [df[["Chromosome", "Source", "Feature", "Start", "End", "Score", "Strand", "Frame"]], subset],
-                    axis=1,
-                    sort=False,
-                )
-            )
-
-    if not dfs:
-        return pd.DataFrame(columns=GTF_NAMES[:-1] + RESTRICTED_ATTRIBUTE_COLUMNS)
-
-    return _finalize_gtf_frame(pd.concat(dfs, sort=False))
-
-
 def read_gtf(
     f: str | Path,
     /,
     *,
     nrows: int | None = None,
-    full: bool = True,
     duplicate_attr: bool = False,
     ignore_bad: bool = False,
 ) -> pd.DataFrame:
     """Read a GTF file using the compiled parser path when available."""
     path = Path(f)
     skiprows = find_first_data_line_index(path)
-
-    if full:
-        return read_gtf_full(
-            path,
-            nrows=nrows,
-            skiprows=skiprows,
-            duplicate_attr=duplicate_attr,
-            ignore_bad=ignore_bad,
-        )
-
-    return read_gtf_restricted(path, skiprows=skiprows, nrows=nrows)
+    return read_gtf_full(
+        path,
+        nrows=nrows,
+        skiprows=skiprows,
+        duplicate_attr=duplicate_attr,
+        ignore_bad=ignore_bad,
+    )
 
 
 def read_gtf_python(
@@ -298,24 +251,19 @@ def read_gtf_python(
     /,
     *,
     nrows: int | None = None,
-    full: bool = True,
     duplicate_attr: bool = False,
     ignore_bad: bool = False,
 ) -> pd.DataFrame:
     """Read a GTF file using the pure Python attribute parser."""
     path = Path(f)
     skiprows = find_first_data_line_index(path)
-
-    if full:
-        return read_gtf_full_python(
-            path,
-            nrows=nrows,
-            skiprows=skiprows,
-            duplicate_attr=duplicate_attr,
-            ignore_bad=ignore_bad,
-        )
-
-    return read_gtf_restricted_python(path, skiprows=skiprows, nrows=nrows)
+    return read_gtf_full_python(
+        path,
+        nrows=nrows,
+        skiprows=skiprows,
+        duplicate_attr=duplicate_attr,
+        ignore_bad=ignore_bad,
+    )
 
 
 def read_gtf_full(
@@ -368,48 +316,6 @@ def read_gtf_full_python(
     )
 
 
-def read_gtf_restricted(
-    f: str | Path,
-    /,
-    skiprows: int = 0,
-    nrows: int | None = None,
-    chunksize: int = int(1e5),
-    *,
-    chunk_size: int | None = None,
-) -> pd.DataFrame:
-    """Read core GTF columns plus a small compiled-parser attribute subset."""
-    path = Path(f)
-    chunksize = _resolve_chunksize(chunksize, chunk_size)
-    return _read_gtf_restricted(
-        path,
-        skiprows=skiprows,
-        nrows=nrows,
-        chunksize=chunksize,
-        parse_attributes=_parse_restricted_attributes_compiled,
-    )
-
-
-def read_gtf_restricted_python(
-    f: str | Path,
-    /,
-    skiprows: int = 0,
-    nrows: int | None = None,
-    chunksize: int = int(1e5),
-    *,
-    chunk_size: int | None = None,
-) -> pd.DataFrame:
-    """Read core GTF columns plus a small pure-Python attribute subset."""
-    path = Path(f)
-    chunksize = _resolve_chunksize(chunksize, chunk_size)
-    return _read_gtf_restricted(
-        path,
-        skiprows=skiprows,
-        nrows=nrows,
-        chunksize=chunksize,
-        parse_attributes=_parse_restricted_attributes_python,
-    )
-
-
 __all__ = [
     "find_first_data_line_index",
     "parse_kv_fields",
@@ -417,8 +323,6 @@ __all__ = [
     "read_gtf_full",
     "read_gtf_full_python",
     "read_gtf_python",
-    "read_gtf_restricted",
-    "read_gtf_restricted_python",
     "to_rows",
     "to_rows_keep_duplicates",
 ]

gtfreader-0.2.0/gtfreader.egg-info/PKG-INFO

@@ -0,0 +1,35 @@
+Metadata-Version: 2.4
+Name: gtfreader
+Version: 0.2.0
+Summary: Fast Cython-backed parsing for GTF attribute columns.
+Author: Endre Bakken Stovner
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Cython
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Operating System :: OS Independent
+Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
+Requires-Python: >=3.12
+Description-Content-Type: text/markdown
+Requires-Dist: pandas>=2.0
+
+# gtfreader
+
+Fast GTF reading into pandas DataFrames.
+
+## Install
+
+```bash
+python -m pip install -e .
+```
+
+## Example
+
+```python
+from gtfreader import read_gtf
+
+df = read_gtf("annotation.gtf")
+print(df.columns)
+print(df.head())
+```

{gtfreader-0.1.1 → gtfreader-0.2.0}/pyproject.toml

@@ -8,10 +8,10 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "gtfreader"
-version = "0.1.1"
+version = "0.2.0"
 description = "Fast Cython-backed parsing for GTF attribute columns."
 readme = "README.md"
-requires-python = ">=3.9"
+requires-python = ">=3.12"
 authors = [
     { name = "Endre Bakken Stovner" },
 ]
@@ -22,6 +22,8 @@ classifiers = [
     "Programming Language :: Python :: 3",
     "Programming Language :: Cython",
     "Programming Language :: Python :: 3 :: Only",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
     "Operating System :: OS Independent",
     "Topic :: Scientific/Engineering :: Bio-Informatics",
 ]
@@ -39,7 +41,7 @@ include = ["gtfreader*"]
 testpaths = ["tests"]
 
 [tool.cibuildwheel]
-build = "cp39-* cp310-* cp311-* cp312-* cp313-*"
+build = "cp312-* cp313-*"
 skip = "pp* *-musllinux_* *-manylinux_i686"
 test-command = "pytest -q {project}/tests"
 test-requires = ["pytest"]

{gtfreader-0.1.1 → gtfreader-0.2.0}/tests/test_parser.py

@@ -25,8 +25,14 @@ def test_known_and_dynamic_columns_are_parsed():
     assert columns["exon_number"] == ["1", None]
     assert columns["custom_key"] == ["custom", None]
     assert columns["gene_name"][0] is columns["gene_name"][1]
-    assert columns["gene_type"][0] is columns["gene_type"][1]
-    assert columns["level"][0] is columns["level"][1]
+    assert columns["gene_type"] == ["protein_coding", "protein_coding"]
+    assert columns["level"] == ["2", "2"]
+
+
+def test_compiled_parser_omits_unseen_fast_path_columns():
+    columns = parse_chunk_columns(['gene_id "GENE1"; transcript_id "TX1";'])
+
+    assert set(columns) == {"gene_id", "transcript_id"}
 
 
 def test_compiled_parser_supports_semicolons_in_quotes_and_unquoted_values():

{gtfreader-0.1.1 → gtfreader-0.2.0}/tests/test_readers.py

@@ -11,8 +11,6 @@ from gtfreader import (
     read_gtf_full,
     read_gtf_full_python,
     read_gtf_python,
-    read_gtf_restricted,
-    read_gtf_restricted_python,
 )
 
 
@@ -120,6 +118,23 @@ def test_full_readers_support_mixed_attribute_formats(tmp_path: Path, reader):
     assert result.iloc[1]["exon_id"] == "EX2"
 
 
+@pytest.mark.parametrize("reader", [read_gtf, read_gtf_python], ids=["default", "python"])
+def test_readers_ignore_comment_lines_inside_body(tmp_path: Path, reader):
+    path = _write_temp_gtf(
+        tmp_path,
+        "# header\n"
+        'chr1\thavana\tgene\t11869\t14409\t.\t+\t.\tgene_id "G1";\n'
+        "# another comment\n"
+        'chr1\thavana\ttranscript\t11869\t14409\t.\t+\t.\tgene_id "G1"; transcript_id "T1";\n',
+    )
+
+    result = _normalize_frame_for_compare(reader(path))
+
+    assert list(result["Feature"]) == ["gene", "transcript"]
+    assert list(result["gene_id"]) == ["G1", "G1"]
+    assert list(result["transcript_id"]) == [None, "T1"]
+
+
 def test_read_gtf_duplicate_attr_keeps_all_values(tmp_path: Path):
     path = _write_temp_gtf(
         tmp_path,
@@ -153,15 +168,14 @@ def test_read_gtf_full_multiple_chunks_match(tmp_path: Path):
     assert_frame_equal(chunked, unchunked, check_dtype=False)
 
 
-def test_read_gtf_restricted_returns_core_columns(tmp_path: Path):
+def test_read_gtf_omits_unseen_fast_path_columns(tmp_path: Path):
     path = _write_temp_gtf(
         tmp_path,
         "# header\n"
-        'chr1\thavana\texon\t12010\t12057\t.\t+\t.\tgene_id "ENSG1"; transcript_id "ENST1"; exon_number "1"; exon_id "ENSE1";\n',
+        'chr1\thavana\tgene\t12010\t12057\t.\t+\t.\tgene_id "ENSG1"; gene_name "DDX11L1";\n',
     )
 
-    skiprows = find_first_data_line_index(path)
-    result = read_gtf_restricted(path, skiprows=skiprows)
+    result = read_gtf(path)
 
     assert list(result.columns) == [
         "Chromosome",
@@ -173,33 +187,10 @@ def test_read_gtf_restricted_returns_core_columns(tmp_path: Path):
         "Strand",
         "Frame",
         "gene_id",
-        "transcript_id",
-        "exon_number",
-        "exon_id",
+        "gene_name",
     ]
     assert result.iloc[0]["Start"] == 12009
-    assert result.iloc[0]["transcript_id"] == "ENST1"
-
-
-@pytest.mark.parametrize(
-    "reader",
-    [read_gtf_restricted, read_gtf_restricted_python],
-    ids=["default", "python"],
-)
-def test_restricted_readers_support_unquoted_values(tmp_path: Path, reader):
-    path = _write_temp_gtf(
-        tmp_path,
-        "# header\n"
-        'chr1\tensembl_havana\texon\t3069203\t3069296\t.\t+\t.\tgene_id "G1"; transcript_id "T1"; exon_number 4; exon_id "EX4";\n',
-    )
-
-    skiprows = find_first_data_line_index(path)
-    result = _normalize_frame_for_compare(reader(path, skiprows=skiprows))
-
-    assert result.iloc[0]["gene_id"] == "G1"
-    assert result.iloc[0]["transcript_id"] == "T1"
-    assert result.iloc[0]["exon_number"] == "4"
-    assert result.iloc[0]["exon_id"] == "EX4"
+    assert result.iloc[0]["gene_name"] == "DDX11L1"
 
 
 def test_read_gtf_python_matches_compiled_reader(tmp_path: Path):
@@ -217,12 +208,7 @@ def test_read_gtf_python_matches_compiled_reader(tmp_path: Path):
     compiled = _normalize_frame_for_compare(compiled)
     python = _normalize_frame_for_compare(python)
 
-    assert list(python["gene_id"]) == list(compiled["gene_id"])
-    assert list(python["gene_name"]) == ["DDX11L1", None]
-    assert list(python["transcript_id"]) == [None, "ENST1"]
-    assert list(python["transcript_name"]) == [None, "DDX11L1-201"]
-    assert list(python["tag"]) == [None, "basic"]
-    assert list(python["Start"]) == list(compiled["Start"])
+    assert_frame_equal(compiled, python, check_dtype=False)
 
 
 def test_read_gtf_python_duplicate_attr_keeps_all_values(tmp_path: Path):
@@ -237,20 +223,3 @@ def test_read_gtf_python_duplicate_attr_keeps_all_values(tmp_path: Path):
 
     assert compiled.iloc[0]["tag"] == "CCDS,basic"
     assert python.iloc[0]["tag"] == "CCDS,basic"
-
-
-def test_read_gtf_restricted_python_matches_compiled(tmp_path: Path):
-    path = _write_temp_gtf(
-        tmp_path,
-        "# header\n"
-        'chr1\thavana\texon\t12010\t12057\t.\t+\t.\tgene_id "ENSG1"; transcript_id "ENST1"; exon_number "1"; exon_id "ENSE1";\n',
-    )
-
-    skiprows = find_first_data_line_index(path)
-    compiled = read_gtf_restricted(path, skiprows=skiprows)
-    python = read_gtf_restricted_python(path, skiprows=skiprows)
-
-    compiled = _normalize_frame_for_compare(compiled)
-    python = _normalize_frame_for_compare(python)
-
-    assert_frame_equal(compiled, python, check_dtype=False)

gtfreader-0.1.1/PKG-INFO

@@ -1,66 +0,0 @@
-Metadata-Version: 2.4
-Name: gtfreader
-Version: 0.1.1
-Summary: Fast Cython-backed parsing for GTF attribute columns.
-Author: Endre Bakken Stovner
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Cython
-Classifier: Programming Language :: Python :: 3 :: Only
-Classifier: Operating System :: OS Independent
-Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
-Requires-Python: >=3.9
-Description-Content-Type: text/markdown
-Requires-Dist: pandas>=2.0
-
-# gtfreader
-
-`gtfreader` is a small package for parsing and reading GTF files into pandas dataframes.
-
-## Install
-
-```bash
-python -m pip install -e .
-```
-
-## Usage
-
-```python
-from gtfreader import read_gtf, read_gtf_python
-
-df = read_gtf("annotation.gtf")
-df_python = read_gtf_python("annotation.gtf")
-```
-
-`read_gtf(...)` uses the compiled parser path when available. `read_gtf_python(...)` uses the high-level pure Python parser path used by the current `pyrunges` reader style.
-
-If you want to use the compiled low-level parser directly, pass it raw attribute strings from column 9 of the GTF before they have been expanded:
-
-```python
-import pandas as pd
-
-from gtfreader import find_first_data_line_index, parse_chunk_columns
-
-skiprows = find_first_data_line_index("annotation.gtf")
-attribute_lines = pd.read_csv(
-    "annotation.gtf",
-    sep="\t",
-    header=None,
-    usecols=[8],
-    names=["Attribute"],
-    skiprows=skiprows,
-)["Attribute"].tolist()
-
-compiled_columns = parse_chunk_columns(attribute_lines)
-```
-
-## Build
-
-```bash
-python -m build
-```
-
-## Test
-
-```bash
-python -m pytest -q
-```

gtfreader-0.1.1/README.md

@@ -1,52 +0,0 @@
-# gtfreader
-
-`gtfreader` is a small package for parsing and reading GTF files into pandas dataframes.
-
-## Install
-
-```bash
-python -m pip install -e .
-```
-
-## Usage
-
-```python
-from gtfreader import read_gtf, read_gtf_python
-
-df = read_gtf("annotation.gtf")
-df_python = read_gtf_python("annotation.gtf")
-```
-
-`read_gtf(...)` uses the compiled parser path when available. `read_gtf_python(...)` uses the high-level pure Python parser path used by the current `pyrunges` reader style.
-
-If you want to use the compiled low-level parser directly, pass it raw attribute strings from column 9 of the GTF before they have been expanded:
-
-```python
-import pandas as pd
-
-from gtfreader import find_first_data_line_index, parse_chunk_columns
-
-skiprows = find_first_data_line_index("annotation.gtf")
-attribute_lines = pd.read_csv(
-    "annotation.gtf",
-    sep="\t",
-    header=None,
-    usecols=[8],
-    names=["Attribute"],
-    skiprows=skiprows,
-)["Attribute"].tolist()
-
-compiled_columns = parse_chunk_columns(attribute_lines)
-```
-
-## Build
-
-```bash
-python -m build
-```
-
-## Test
-
-```bash
-python -m pytest -q
-```

gtfreader-0.1.1/gtfreader/_parser.pyx

@@ -1,352 +0,0 @@
-# cython: language_level=3
-# cython: boundscheck=False
-# cython: wraparound=False
-# cython: initializedcheck=False
-# cython: infer_types=True
-# cython: nonecheck=False
-
-cdef inline int _key_code(str s, Py_ssize_t a, Py_ssize_t b):
-    cdef Py_ssize_t n
-    n = b - a
-
-    if n == 3:
-        if s[a] == 't' and s[a + 1] == 'a' and s[a + 2] == 'g':
-            return 9
-        if s[a] == 'o' and s[a + 1] == 'n' and s[a + 2] == 't':
-            return 17
-
-    elif n == 5:
-        if (
-            s[a] == 'l' and s[a + 1] == 'e' and s[a + 2] == 'v' and
-            s[a + 3] == 'e' and s[a + 4] == 'l'
-        ):
-            return 16
-
-    elif n == 6:
-        if (
-            s[a] == 'c' and s[a + 1] == 'c' and s[a + 2] == 'd' and
-            s[a + 3] == 's' and s[a + 4] == 'i' and s[a + 5] == 'd'
-        ):
-            return 14
-
-    elif n == 7:
-        if (
-            s[a] == 'g' and s[a + 1] == 'e' and s[a + 2] == 'n' and
-            s[a + 3] == 'e' and s[a + 4] == '_' and s[a + 5] == 'i' and
-            s[a + 6] == 'd'
-        ):
-            return 1
-        if (
-            s[a] == 'e' and s[a + 1] == 'x' and s[a + 2] == 'o' and
-            s[a + 3] == 'n' and s[a + 4] == '_' and s[a + 5] == 'i' and
-            s[a + 6] == 'd'
-        ):
-            return 8
-        if (
-            s[a] == 'h' and s[a + 1] == 'g' and s[a + 2] == 'n' and
-            s[a + 3] == 'c' and s[a + 4] == '_' and s[a + 5] == 'i' and
-            s[a + 6] == 'd'
-        ):
-            return 13
-
-    elif n == 9:
-        if (
-            s[a] == 'g' and s[a + 1] == 'e' and s[a + 2] == 'n' and
-            s[a + 3] == 'e' and s[a + 4] == '_'
-        ):
-            if (
-                s[a + 5] == 'n' and s[a + 6] == 'a' and
-                s[a + 7] == 'm' and s[a + 8] == 'e'
-            ):
-                return 3
-            if (
-                s[a + 5] == 't' and s[a + 6] == 'y' and
-                s[a + 7] == 'p' and s[a + 8] == 'e'
-            ):
-                return 4
-
-    elif n == 10:
-        if (
-            s[a] == 'a' and s[a + 1] == 'r' and s[a + 2] == 't' and
-            s[a + 3] == 'i' and s[a + 4] == 'f' and s[a + 5] == '_' and
-            s[a + 6] == 'd' and s[a + 7] == 'u' and s[a + 8] == 'p' and
-            s[a + 9] == 'l'
-        ):
-            return 15
-
-    elif n == 11:
-        if (
-            s[a] == 'e' and s[a + 1] == 'x' and s[a + 2] == 'o' and
-            s[a + 3] == 'n' and s[a + 4] == '_' and s[a + 5] == 'n' and
-            s[a + 6] == 'u' and s[a + 7] == 'm' and s[a + 8] == 'b' and
-            s[a + 9] == 'e' and s[a + 10] == 'r'
-        ):
-            return 7
-        if (
-            s[a] == 'h' and s[a + 1] == 'a' and s[a + 2] == 'v' and
-            s[a + 3] == 'a' and s[a + 4] == 'n' and s[a + 5] == 'a' and
-            s[a + 6] == '_' and s[a + 7] == 'g' and s[a + 8] == 'e' and
-            s[a + 9] == 'n' and s[a + 10] == 'e'
-        ):
-            return 11
-
-    elif n == 13:
-        if (
-            s[a] == 't' and s[a + 1] == 'r' and s[a + 2] == 'a' and
-            s[a + 3] == 'n' and s[a + 4] == 's' and s[a + 5] == 'c' and
-            s[a + 6] == 'r' and s[a + 7] == 'i' and s[a + 8] == 'p' and
-            s[a + 9] == 't' and s[a + 10] == '_' and s[a + 11] == 'i' and
-            s[a + 12] == 'd'
-        ):
-            return 2
-
-    elif n == 15:
-        if (
-            s[a] == 't' and s[a + 1] == 'r' and s[a + 2] == 'a' and
-            s[a + 3] == 'n' and s[a + 4] == 's' and s[a + 5] == 'c' and
-            s[a + 6] == 'r' and s[a + 7] == 'i' and s[a + 8] == 'p' and
-            s[a + 9] == 't' and s[a + 10] == '_'
-        ):
-            if (
-                s[a + 11] == 'n' and s[a + 12] == 'a' and
-                s[a + 13] == 'm' and s[a + 14] == 'e'
-            ):
-                return 5
-            if (
-                s[a + 11] == 't' and s[a + 12] == 'y' and
-                s[a + 13] == 'p' and s[a + 14] == 'e'
-            ):
-                return 6
-
-    elif n == 17:
-        if (
-            s[a] == 'h' and s[a + 1] == 'a' and s[a + 2] == 'v' and
-            s[a + 3] == 'a' and s[a + 4] == 'n' and s[a + 5] == 'a' and
-            s[a + 6] == '_' and s[a + 7] == 't' and s[a + 8] == 'r' and
-            s[a + 9] == 'a' and s[a + 10] == 'n' and s[a + 11] == 's' and
-            s[a + 12] == 'c' and s[a + 13] == 'r' and s[a + 14] == 'i' and
-            s[a + 15] == 'p' and s[a + 16] == 't'
-        ):
-            return 10
-
-    elif n == 24:
-        if (
-            s[a] == 't' and s[a + 1] == 'r' and s[a + 2] == 'a' and
-            s[a + 3] == 'n' and s[a + 4] == 's' and s[a + 5] == 'c' and
-            s[a + 6] == 'r' and s[a + 7] == 'i' and s[a + 8] == 'p' and
-            s[a + 9] == 't' and s[a + 10] == '_' and s[a + 11] == 's' and
-            s[a + 12] == 'u' and s[a + 13] == 'p' and s[a + 14] == 'p' and
-            s[a + 15] == 'o' and s[a + 16] == 'r' and s[a + 17] == 't' and
-            s[a + 18] == '_' and s[a + 19] == 'l' and s[a + 20] == 'e' and
-            s[a + 21] == 'v' and s[a + 22] == 'e' and s[a + 23] == 'l'
-        ):
-            return 12
-
-    return 0
-
-
-cdef inline str _cached_str(dict cache, str value):
-    cdef object obj
-    obj = cache.get(value)
-    if obj is None:
-        cache[value] = value
-        return value
-    return <str> obj
-
-
-def parse_chunk_columns(lines):
-    cdef Py_ssize_t n, row_i
-    cdef Py_ssize_t pos, m, key_start, key_end, val_start, val_end
-    cdef dict columns
-
-    cdef dict gene_name_cache
-    cdef dict gene_type_cache
-    cdef dict transcript_type_cache
-    cdef dict tag_cache
-    cdef dict level_cache
-    cdef dict ont_cache
-    cdef dict transcript_support_level_cache
-    cdef dict artif_dupl_cache
-
-    cdef str s, key, value
-    cdef list col
-    cdef int code
-
-    cdef object obj_col
-
-    cdef list gene_id_col
-    cdef list transcript_id_col
-    cdef list gene_name_col
-    cdef list gene_type_col
-    cdef list transcript_name_col
-    cdef list transcript_type_col
-    cdef list exon_number_col
-    cdef list exon_id_col
-    cdef list tag_col
-    cdef list havana_transcript_col
-    cdef list havana_gene_col
-    cdef list transcript_support_level_col
-    cdef list hgnc_id_col
-    cdef list ccdsid_col
-    cdef list artif_dupl_col
-    cdef list level_col
-    cdef list ont_col
-
-    n = len(lines)
-    columns = {}
-
-    gene_name_cache = {}
-    gene_type_cache = {}
-    transcript_type_cache = {}
-    tag_cache = {}
-    level_cache = {}
-    ont_cache = {}
-    transcript_support_level_cache = {}
-    artif_dupl_cache = {}
-
-    gene_id_col = [None] * n
-    transcript_id_col = [None] * n
-    gene_name_col = [None] * n
-    gene_type_col = [None] * n
-    transcript_name_col = [None] * n
-    transcript_type_col = [None] * n
-    exon_number_col = [None] * n
-    exon_id_col = [None] * n
-    tag_col = [None] * n
-    havana_transcript_col = [None] * n
-    havana_gene_col = [None] * n
-    transcript_support_level_col = [None] * n
-    hgnc_id_col = [None] * n
-    ccdsid_col = [None] * n
-    artif_dupl_col = [None] * n
-    level_col = [None] * n
-    ont_col = [None] * n
-
-    columns["gene_id"] = gene_id_col
-    columns["transcript_id"] = transcript_id_col
-    columns["gene_name"] = gene_name_col
-    columns["gene_type"] = gene_type_col
-    columns["transcript_name"] = transcript_name_col
-    columns["transcript_type"] = transcript_type_col
-    columns["exon_number"] = exon_number_col
-    columns["exon_id"] = exon_id_col
-    columns["tag"] = tag_col
-    columns["havana_transcript"] = havana_transcript_col
-    columns["havana_gene"] = havana_gene_col
-    columns["transcript_support_level"] = transcript_support_level_col
-    columns["hgnc_id"] = hgnc_id_col
-    columns["ccdsid"] = ccdsid_col
-    columns["artif_dupl"] = artif_dupl_col
-    columns["level"] = level_col
-    columns["ont"] = ont_col
-
-    for row_i in range(n):
-        s = <str> lines[row_i]
-        m = len(s)
-        pos = 0
-
-        while pos < m:
-            while pos < m and (s[pos] == ' ' or s[pos] == '\t' or s[pos] == ';'):
-                pos += 1
-            if pos >= m:
-                break
-
-            key_start = pos
-
-            while pos < m and s[pos] != ' ' and s[pos] != '\t':
-                pos += 1
-            key_end = pos
-
-            while pos < m and (s[pos] == ' ' or s[pos] == '\t'):
-                pos += 1
-            if pos >= m:
-                break
-
-            if s[pos] == '"':
-                val_start = pos + 1
-                pos += 1
-
-                while pos < m and s[pos] != '"':
-                    pos += 1
-                if pos >= m:
-                    break
-
-                val_end = pos
-            else:
-                val_start = pos
-                while pos < m and s[pos] != ';':
-                    pos += 1
-                val_end = pos
-                while val_end > val_start and (s[val_end - 1] == ' ' or s[val_end - 1] == '\t'):
-                    val_end -= 1
-
-            value = s[val_start:val_end]
-
-            code = _key_code(s, key_start, key_end)
-
-            if code == 1:
-                gene_id_col[row_i] = value
-
-            elif code == 2:
-                transcript_id_col[row_i] = value
-
-            elif code == 3:
-                gene_name_col[row_i] = _cached_str(gene_name_cache, value)
-
-            elif code == 4:
-                gene_type_col[row_i] = _cached_str(gene_type_cache, value)
-
-            elif code == 5:
-                transcript_name_col[row_i] = value
-
-            elif code == 6:
-                transcript_type_col[row_i] = _cached_str(transcript_type_cache, value)
-
-            elif code == 7:
-                exon_number_col[row_i] = value
-
-            elif code == 8:
-                exon_id_col[row_i] = value
-
-            elif code == 9:
-                tag_col[row_i] = _cached_str(tag_cache, value)
-
-            elif code == 10:
-                havana_transcript_col[row_i] = value
-
-            elif code == 11:
-                havana_gene_col[row_i] = value
-
-            elif code == 12:
-                transcript_support_level_col[row_i] = _cached_str(
-                    transcript_support_level_cache, value
-                )
-
-            elif code == 13:
-                hgnc_id_col[row_i] = value
-
-            elif code == 14:
-                ccdsid_col[row_i] = value
-
-            elif code == 15:
-                artif_dupl_col[row_i] = _cached_str(artif_dupl_cache, value)
-
-            elif code == 16:
-                level_col[row_i] = _cached_str(level_cache, value)
-
-            elif code == 17:
-                ont_col[row_i] = _cached_str(ont_cache, value)
-
-            else:
-                key = s[key_start:key_end]
-                obj_col = columns.get(key)
-                if obj_col is None:
-                    col = [None] * n
-                    columns[key] = col
-                else:
-                    col = <list> obj_col
-                col[row_i] = value
-
-            pos += 1
-
-    return columns

gtfreader-0.1.1/gtfreader.egg-info/PKG-INFO

@@ -1,66 +0,0 @@
-Metadata-Version: 2.4
-Name: gtfreader
-Version: 0.1.1
-Summary: Fast Cython-backed parsing for GTF attribute columns.
-Author: Endre Bakken Stovner
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Cython
-Classifier: Programming Language :: Python :: 3 :: Only
-Classifier: Operating System :: OS Independent
-Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
-Requires-Python: >=3.9
-Description-Content-Type: text/markdown
-Requires-Dist: pandas>=2.0
-
-# gtfreader
-
-`gtfreader` is a small package for parsing and reading GTF files into pandas dataframes.
-
-## Install
-
-```bash
-python -m pip install -e .
-```
-
-## Usage
-
-```python
-from gtfreader import read_gtf, read_gtf_python
-
-df = read_gtf("annotation.gtf")
-df_python = read_gtf_python("annotation.gtf")
-```
-
-`read_gtf(...)` uses the compiled parser path when available. `read_gtf_python(...)` uses the high-level pure Python parser path used by the current `pyrunges` reader style.
-
-If you want to use the compiled low-level parser directly, pass it raw attribute strings from column 9 of the GTF before they have been expanded:
-
-```python
-import pandas as pd
-
-from gtfreader import find_first_data_line_index, parse_chunk_columns
-
-skiprows = find_first_data_line_index("annotation.gtf")
-attribute_lines = pd.read_csv(
-    "annotation.gtf",
-    sep="\t",
-    header=None,
-    usecols=[8],
-    names=["Attribute"],
-    skiprows=skiprows,
-)["Attribute"].tolist()
-
-compiled_columns = parse_chunk_columns(attribute_lines)
-```
-
-## Build
-
-```bash
-python -m build
-```
-
-## Test
-
-```bash
-python -m pytest -q
-```