16442527952

Committed 22 Jul 2025 11:08AM UTC coverage: 90.638% (+0.05%) from 90.588%

Build # 16442527952

Build Type

push

github

Committed by

web-flow

Commit Message

feat: Add DocumentTypeRouter (#9634)

* Add DocumentTypeRouter

* PR comments

* Turn off isort for one line so pylint will pass

Run Details

12325 of 13598 relevant lines covered (90.64%)

0.91 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

75.76

haystack/utils/misc.py

# SPDX-FileCopyrightText: 2022-present deepset GmbH <info@deepset.ai>
#
# SPDX-License-Identifier: Apache-2.0

import mimetypes
from pathlib import Path
from typing import Any, List, Optional, Union, overload

from numpy import exp, ndarray

CUSTOM_MIMETYPES = {
    # we add markdown because it is not added by the mimetypes module
    # see https://github.com/python/cpython/pull/17995
    ".md": "text/markdown",
    ".markdown": "text/markdown",
    # we add msg because it is not added by the mimetypes module
    ".msg": "application/vnd.ms-outlook",
}


def expand_page_range(page_range: List[Union[str, int]]) -> List[int]:
    """
    Takes a list of page numbers and ranges and expands them into a list of page numbers.

    For example, given a page_range=['1-3', '5', '8', '10-12'] the function will return [1, 2, 3, 5, 8, 10, 11, 12]

    :param page_range: List of page numbers and ranges
    :returns:
        An expanded list of page integers

    """
    expanded_page_range = []

    for page in page_range:
        if isinstance(page, int):
            # check if it's a range wrongly passed as an integer expression
            if "-" in str(page):
                msg = "range must be a string in the format 'start-end'"
                raise ValueError(f"Invalid page range: {page} - {msg}")
            expanded_page_range.append(page)

        elif isinstance(page, str) and page.isdigit():
            expanded_page_range.append(int(page))

        elif isinstance(page, str) and "-" in page:
            start, end = page.split("-")
            expanded_page_range.extend(range(int(start), int(end) + 1))

        else:
            msg = "range must be a string in the format 'start-end' or an integer"
            raise ValueError(f"Invalid page range: {page} - {msg}")

    if not expanded_page_range:
        raise ValueError("No valid page numbers or ranges found in the input list")

    return expanded_page_range


@overload
def expit(x: float) -> float: ...
@overload
def expit(x: ndarray[Any, Any]) -> ndarray[Any, Any]: ...
def expit(x: Union[float, ndarray[Any, Any]]) -> Union[float, ndarray[Any, Any]]:
    """
    Compute logistic sigmoid function. Maps input values to a range between 0 and 1

    :param x: input value. Can be a scalar or a numpy array.
    """
    return 1 / (1 + exp(-x))


def _guess_mime_type(path: Path) -> Optional[str]:
    """
    Guess the MIME type of the provided file path.

    :param path: The file path to get the MIME type for.

    :returns: The MIME type of the provided file path, or `None` if the MIME type cannot be determined.
    """
    extension = path.suffix.lower()
    mime_type = mimetypes.guess_type(path.as_posix())[0]
    # lookup custom mappings if the mime type is not found
    return CUSTOM_MIMETYPES.get(extension, mime_type)

1	# SPDX-FileCopyrightText: 2022-present deepset GmbH <info@deepset.ai>
2	#
3	# SPDX-License-Identifier: Apache-2.0
4
5	import mimetypes	1✔
6	from pathlib import Path	1✔
7	from typing import Any, List, Optional, Union, overload	1✔
8
9	from numpy import exp, ndarray	1✔
10
11	CUSTOM_MIMETYPES = {	1✔
12	# we add markdown because it is not added by the mimetypes module
13	# see https://github.com/python/cpython/pull/17995
14	".md": "text/markdown",
15	".markdown": "text/markdown",
16	# we add msg because it is not added by the mimetypes module
17	".msg": "application/vnd.ms-outlook",
18	}
19
20
21	def expand_page_range(page_range: List[Union[str, int]]) -> List[int]:	1✔
22	"""
23	Takes a list of page numbers and ranges and expands them into a list of page numbers.
24
25	For example, given a page_range=['1-3', '5', '8', '10-12'] the function will return [1, 2, 3, 5, 8, 10, 11, 12]
26
27	:param page_range: List of page numbers and ranges
28	:returns:
29	An expanded list of page integers
30
31	"""
32	expanded_page_range = []	1✔
33
34	for page in page_range:	1✔
35	if isinstance(page, int):	1✔
36	# check if it's a range wrongly passed as an integer expression
37	if "-" in str(page):	×
38	msg = "range must be a string in the format 'start-end'"	×
39	raise ValueError(f"Invalid page range: {page} - {msg}")	×
40	expanded_page_range.append(page)	×
41
42	elif isinstance(page, str) and page.isdigit():	1✔
43	expanded_page_range.append(int(page))	×
44
45	elif isinstance(page, str) and "-" in page:	1✔
46	start, end = page.split("-")	1✔
47	expanded_page_range.extend(range(int(start), int(end) + 1))	1✔
48
49	else:
50	msg = "range must be a string in the format 'start-end' or an integer"	×
51	raise ValueError(f"Invalid page range: {page} - {msg}")	×
52
53	if not expanded_page_range:	1✔
54	raise ValueError("No valid page numbers or ranges found in the input list")	×
55
56	return expanded_page_range	1✔
57
58
59	@overload	1✔
60	def expit(x: float) -> float: ...	1✔
61	@overload	1✔
62	def expit(x: ndarray[Any, Any]) -> ndarray[Any, Any]: ...	1✔
63	def expit(x: Union[float, ndarray[Any, Any]]) -> Union[float, ndarray[Any, Any]]:	1✔
64	"""
65	Compute logistic sigmoid function. Maps input values to a range between 0 and 1
66
67	:param x: input value. Can be a scalar or a numpy array.
68	"""
69	return 1 / (1 + exp(-x))	1✔
70
71
72	def _guess_mime_type(path: Path) -> Optional[str]:	1✔
73	"""
74	Guess the MIME type of the provided file path.
75
76	:param path: The file path to get the MIME type for.
77
78	:returns: The MIME type of the provided file path, or `None` if the MIME type cannot be determined.
79	"""
80	extension = path.suffix.lower()	1✔
81	mime_type = mimetypes.guess_type(path.as_posix())[0]	1✔
82	# lookup custom mappings if the mime type is not found
83	return CUSTOM_MIMETYPES.get(extension, mime_type)	1✔

deepset-ai / haystack / 16442527952

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous