Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Add Cython implementation for VOTable binary converters #18454

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 5 commits into
base: main
Choose a base branch
from
Open

Add Cython implementation for VOTable binary converters #18454

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
1dfea18
Add Cython implementation for VOTable binary converters
stvoutsin Jul 12, 2025
11d9403
Fix buffer handling and null detection to match original converter be…
stvoutsin Jul 24, 2025
1babe2e
Reduce duplication of binparse methods for Numeric types
stvoutsin Jul 24, 2025
70674cc
Move null handling to Cython parser implementations to address perfor…
stvoutsin Jul 25, 2025
96b3904
Update docs/change entry with some notes on benchmark results
stvoutsin Jul 26, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
Fix buffer handling and null detection to match original converter be… 
…havior
  • Loading branch information
@stvoutsin
stvoutsin committed Jul 24, 2025
commit 11d94038c51f4debc5ac718a284d3229f592e1d8
13 changes: 9 additions & 4 deletions astropy/io/votable/converters.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -832,7 +832,9 @@ def binparse(self, read):
data = read(8)
if len(data) != 8:
raise EOFError("Unexpected end of file")
return fast_binparse_double(data)
value, _ = fast_binparse_double(data)
is_null = self.is_null(value)
return value, is_null


class Float(FloatingPoint):
Expand Down Expand Up @@ -971,7 +973,9 @@ def binparse(self, read):
data = read(4)
if len(data) != 4:
raise EOFError("Unexpected end of file")
return fast_binparse_int(data)
value, _ = fast_binparse_int(data)
is_null = self.is_null(value)
return value, is_null


class Long(Integer):
Expand All @@ -987,7 +991,9 @@ def binparse(self, read):
data = read(8)
if len(data) != 8:
raise EOFError("Unexpected end of file")
return fast_binparse_long(data)
value, _ = fast_binparse_long(data)
is_null = self.is_null(value)
return value, is_null


class ComplexArrayVarArray(VarArray):
Expand Down Expand Up @@ -1173,7 +1179,6 @@ def binparse(self, read):
def binoutput(self, value, mask):
if np.any(mask):
vo_warn(W39)

return bool_to_bitarray(value)


Expand Down
32 changes: 24 additions & 8 deletions astropy/io/votable/src/fast_converters.pyx
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,17 +29,26 @@ ctypedef cnp.uint8_t bool_t

cdef bint NEED_BYTESWAP = sys.byteorder == 'little'

def fast_bitarray_to_bool(const unsigned char[::1] data, int length):
def fast_bitarray_to_bool(data, int length):
"""Convert packed bits to bool array"""
cdef const unsigned char[::1] byte_data

if hasattr(data, 'tobytes'):
byte_data = memoryview(data.tobytes())
elif isinstance(data, (bytes, bytearray)):
byte_data = memoryview(data)
else:
byte_data = data

cdef cnp.ndarray[bool_t, ndim=1] results = np.empty(length, dtype=np.bool_)
cdef int i = 0
cdef int byte_idx = 0
cdef int bit_no
cdef unsigned char byte_val
cdef bool_t bit_val

while i < length and byte_idx < data.shape[0]:
byte_val = data[byte_idx]
while i < length and byte_idx < byte_data.shape[0]:
byte_val = byte_data[byte_idx]
for bit_no in range(7, -1, -1):
if i >= length:
break
Expand All @@ -50,21 +59,28 @@ def fast_bitarray_to_bool(const unsigned char[::1] data, int length):

return results

def fast_bool_to_bitarray(const bool_t[::1] value):
"""Pack bools into bit array - reverse of above."""
cdef int length = value.shape[0]
def fast_bool_to_bitarray(value):
"""
Cython implementation of bool_to_bitarray.
"""
if np.isscalar(value):
flat_array = np.array([bool(value)], dtype=np.bool_)
else:
value_array = np.asarray(value, dtype=np.bool_)
flat_array = value_array.ravel()

cdef int length = len(flat_array)
cdef int num_bytes = (length + 7) // 8
cdef cnp.ndarray[unsigned char, ndim=1] result = np.zeros(num_bytes, dtype=np.uint8)

cdef int i
cdef int byte_idx
cdef int bit_no
cdef unsigned char byte_val

for i in range(length):
byte_idx = i // 8
bit_no = 7 - (i % 8)
if value[i]:
if flat_array[i]:
result[byte_idx] |= (1 << bit_no)

return result.tobytes()
Expand Down
Loading