| 1 |
#cython: language_level=3 |
| 2 |
|
| 3 |
''' |
| 4 |
Created on feb 20th 2018 |
| 5 |
|
| 6 |
@author: cmercier |
| 7 |
''' |
| 8 |
|
| 9 |
import types |
| 10 |
from obitools3.utils cimport __etag__ |
| 11 |
from obitools3.utils cimport str2bytes |
| 12 |
|
| 13 |
def tabIterator(lineiterator, |
| 14 |
bint header = False, |
| 15 |
bytes sep = None, |
| 16 |
bytes dec = b".", # TODO don't know how to use this to parse |
| 17 |
bint stripwhite=True, |
| 18 |
bint blanklineskip=True, |
| 19 |
bytes commentchar=b"#", |
| 20 |
bytes nastring=b"NA", |
| 21 |
int skip=0, |
| 22 |
only=None, |
| 23 |
firstline=None, |
| 24 |
int buffersize=100000000 |
| 25 |
): |
| 26 |
|
| 27 |
cdef LineBuffer lb |
| 28 |
cdef int lines_to_skip, ionly, read |
| 29 |
cdef list data |
| 30 |
cdef dict view_line |
| 31 |
cdef list keys |
| 32 |
cdef list key_types |
| 33 |
|
| 34 |
keys = [] |
| 35 |
key_types = [] |
| 36 |
skipped = 0 |
| 37 |
read = 0 |
| 38 |
|
| 39 |
if only is None: |
| 40 |
ionly = -1 |
| 41 |
else: |
| 42 |
ionly = int(only) |
| 43 |
|
| 44 |
if isinstance(lineiterator, (str, bytes)): |
| 45 |
lineiterator=uopen(lineiterator) |
| 46 |
if isinstance(lineiterator, LineBuffer): |
| 47 |
iterator = iter(lineiterator) |
| 48 |
else: |
| 49 |
if hasattr(lineiterator, "readlines"): |
| 50 |
iterator = iter(LineBuffer(lineiterator, buffersize)) |
| 51 |
elif hasattr(lineiterator, '__next__'): |
| 52 |
iterator = lineiterator |
| 53 |
else: |
| 54 |
raise Exception("Invalid line iterator") |
| 55 |
|
| 56 |
if firstline is None: |
| 57 |
line = next(iterator) |
| 58 |
else: |
| 59 |
line = firstline |
| 60 |
|
| 61 |
while True: |
| 62 |
|
| 63 |
if (not line.strip() and blanklineskip) or line[:1] == commentchar: |
| 64 |
line = next(iterator) |
| 65 |
|
| 66 |
if ionly >= 0 and read >= ionly: |
| 67 |
break |
| 68 |
|
| 69 |
if not keys: |
| 70 |
if header: |
| 71 |
# TODO read types eventually |
| 72 |
keys = line.split(sep) |
| 73 |
keys = [x.strip() for x in keys] |
| 74 |
line = next(iterator) |
| 75 |
continue |
| 76 |
else: |
| 77 |
# TODO ??? default column names? like R? |
| 78 |
keys = [str2bytes(str(i)) for i in range(len(line.split(sep)))] |
| 79 |
|
| 80 |
while skipped < skip : |
| 81 |
line = next(iterator) |
| 82 |
skipped += 1 |
| 83 |
|
| 84 |
view_line = {} |
| 85 |
|
| 86 |
# Parse |
| 87 |
data = line.split(sep) |
| 88 |
|
| 89 |
if stripwhite or key_types: |
| 90 |
data = [x.strip() for x in data] |
| 91 |
|
| 92 |
for i in range(len(data)): |
| 93 |
if key_types: # TODO handle None when key types are actually read |
| 94 |
view_line[keys[i]] = key_types[i](data[i]) |
| 95 |
else: |
| 96 |
view_line[keys[i]] = __etag__(data[i], nastring=nastring) |
| 97 |
|
| 98 |
yield view_line |
| 99 |
|
| 100 |
read+=1 |
| 101 |
|
| 102 |
try: |
| 103 |
line = next(iterator) |
| 104 |
except StopIteration: |
| 105 |
return |
| 106 |
|
| 107 |
|
| 108 |
|