Reputation: 19172
I have a handful of different type of C-structs that are all compressed into a binary file.
struct-id serialized-struct struct-id serialized-struct ...
If it were the same struct over and over, it would make sense to use the struct
package, but I want to switch between previously defined structs all the time.
STRUCT1_ID = '\xAA'
STRUCT2_ID = '\xBB'
STRUCT_IDS = frozenset([STRUCT1_ID, STRUCT2_ID])
struct1s = []
struct2s = []
def create_test_file(filepath):
with open(filepath, 'wb') as f:
# Write an example struct1 id followed by struct
f.write(STRUCT1_ID)
f.write(b'\x01\x02\x03\x04\x05\x06')
# Write an example struct2 id followed by struct
f.write(STRUCT2_ID)
f.write(b'\x07\x08\x09\x0A')
def parse_test_file(filepath):
with open(filepath, 'rb') as f:
msg_type = f.read(1)
while msg_type:
print(byte)
if byte in STRUCT_IDS:
# Parse the next however many bytes needed by struct
# logic breaks down here
struct1s.append(turnIntoStruct(f.read(?)))
msg_type = f.read(1)
else:
print('Corrupted file. Unrecognized id')
In C, the structs would be:
typedef struct struct1_s {
uint16_t a;
uint16_t b;
uint16_t c;
} struct1_t;
typedef struct struct2_s {
uint16_t d;
uint16_t e;
} struct2_t;
// Declare and initialize the structs
struct1_t s1 = {
.a = 0x0201,
.b = 0x0403,
.c = 0x0605
};
struct2_t s2 = {
.d = 0x0807,
.e = 0x0A09
};
I'm less python than I am C right now. I seem unable to bring construct to python 3.4.3?
Upvotes: 0
Views: 380
Reputation: 798676
Map the ID to the struct pattern, and use the appropriate one.
structmap = {
b'\xaa': ('3H', struct1s),
b'\xbb': ('2H', struct2s)
}
...
structmap[msg_type][1].append(struct.unpack(structmap[msg_type][0],
f.read(struct.calcsize(structmap[msg_type][0]))))
Upvotes: 1