opensource/galaxy
3
0
Fork 0
mirror of https://github.com/galaxyproject/galaxy.git synced 2026-10-25 10:24:38 +08:00
Code Issues Packages Projects Releases Wiki Activity

use fingerprint for extxyz sniffer

Browse Source
This commit is contained in:
elichad
2022-04-01 09:35:04 +00:00
parent a87bcf0b8e
commit b6c0b0074d
3 changed files with 6 additions and 54 deletions
Download Patch File Download Diff File Expand all files Collapse all files
lib/galaxy/datatypes/molecules.py
+6 -31
View File
@@ -1401,42 +1401,17 @@ class ExtendedXYZ(XYZ):
>>> fname = get_test_fname('Si.extxyz')
>>> ExtendedXYZ().sniff(fname)
True
>>> from galaxy.datatypes.sniff import get_test_fname
>>> fname = get_test_fname('Si_multi.extxyz')
>>> ExtendedXYZ().sniff(fname)
True
>>> fname = get_test_fname('Si.xyz')
>>> ExtendedXYZ().sniff(fname)
False
"""
with open(filename) as f:
xyz_lines = f.readlines(1000)[:-1]
try:
blocks = self.read_blocks(xyz_lines)
except (TypeError, ValueError):
return False
except IndexError as e:
if "pop from empty list" in str(e):
# xyz_lines ran out mid-block – check full file
with open(filename) as f:
xyz_lines = f.readlines()
try:
blocks = self.read_blocks(xyz_lines)
except (TypeError, ValueError, IndexError):
# any error is now a failure
return False
else:
# some other IndexError - invalid input
return False
# check blocks are valid
for block in blocks:
if block["number_of_atoms"] is None:
return False
# all checks passed
return True
# fingerprint is 'Properties="<name>:<type>:<number>:..."' in the second line
# e.g. Properties="species:S:1:pos:R:3:vel:R:3:select:I:1"
f.readline()
comment = f.readline()
properties = re.search(r"Properties=\"?([a-zA-Z0-9:]+)\"?", comment) # returns None if no match
return True if properties else False
def read_blocks(self, lines):
"""
lib/galaxy/datatypes/sniff.py
-3
View File
@@ -390,9 +390,6 @@ def guess_ext(fname, sniff_order, is_binary=False):
>>> fname = get_test_fname('Si.extxyz')
>>> guess_ext(fname, sniff_order)
'extxyz'
>>> fname = get_test_fname('Si_multi.extxyz')
>>> guess_ext(fname, sniff_order)
'extxyz'
>>> fname = get_test_fname('mothur_datatypetest_true.mothur.otu')
>>> guess_ext(fname, sniff_order)
'mothur.otu'
lib/galaxy/datatypes/test/Si_multi.extxyz
-20
View File
@@ -1,20 +0,0 @@
8
Lattice="5.44 0.0 0.0 0.0 5.44 0.0 0.0 0.0 5.44" Properties="species:S:1:pos:R:3:vel:R:3:select:I:1" Time=0.0
Si 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 1
Si 1.36000000 1.36000000 1.36000000 0.00000000 0.00000000 0.00000000 1
Si 2.72000000 2.72000000 0.00000000 0.00000000 0.00000000 0.00000000 1
Si 4.08000000 4.08000000 1.36000000 0.00000000 0.00000000 0.00000000 1
Si 2.72000000 0.00000000 2.72000000 0.00000000 0.00000000 0.00000000 1
Si 4.08000000 1.36000000 4.08000000 0.00000000 0.00000000 0.00000000 1
Si 0.00000000 2.72000000 2.72000000 0.00000000 0.00000000 0.00000000 1
Si 1.36000000 4.08000000 4.08000000 0.00000000 0.00000000 0.00000000 1
8
Lattice="5.44 0.0 0.0 0.0 5.44 0.0 0.0 0.0 5.44" Properties="species:S:1:pos:R:3:vel:R:3:select:I:1" Time=0.0
Si 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 0.00000000 1
Si 1.36000000 1.36000000 1.36000000 0.00000000 0.00000000 0.00000000 1
Si 2.72000000 2.72000000 0.00000000 0.00000000 0.00000000 0.00000000 1
Si 4.08000000 4.08000000 1.36000000 0.00000000 0.00000000 0.00000000 1
Si 2.72000000 0.00000000 2.72000000 0.00000000 0.00000000 0.00000000 1
Si 4.08000000 1.36000000 4.08000000 0.00000000 0.00000000 0.00000000 1
Si 0.00000000 2.72000000 2.72000000 0.00000000 0.00000000 0.00000000 1
Si 1.36000000 4.08000000 4.08000000 0.00000000 0.00000000 0.00000000 1