Skip to content

Commit

Permalink
Merge pull request adafruit#6754 from jepler/check-read-utf8
Browse files Browse the repository at this point in the history
When reading data from a file into a str, check if it's utf-8
  • Loading branch information
tannewt authored Aug 15, 2022
2 parents a223102 + 606c75a commit 7717ab8
Show file tree
Hide file tree
Showing 2 changed files with 13 additions and 6 deletions.
17 changes: 12 additions & 5 deletions py/stream.c
Original file line number Diff line number Diff line change
Expand Up @@ -31,6 +31,7 @@
#include "py/objstr.h"
#include "py/stream.h"
#include "py/runtime.h"
#include "py/unicode.h"
#include "supervisor/shared/translate/translate.h"

// This file defines generic Python stream read/write methods which
Expand All @@ -43,6 +44,13 @@ STATIC mp_obj_t stream_readall(mp_obj_t self_in);

#define STREAM_CONTENT_TYPE(stream) (((stream)->is_text) ? &mp_type_str : &mp_type_bytes)

static mp_obj_t mp_obj_new_str_from_vstr_check(const mp_obj_type_t *type, vstr_t *vstr) {
if (type == &mp_type_str && !utf8_check((void *)vstr->buf, vstr->len)) {
mp_raise_msg(&mp_type_UnicodeError, NULL);
}
return mp_obj_new_str_from_vstr(type, vstr);
}

// Returns error condition in *errcode, if non-zero, return value is number of bytes written
// before error condition occurred. If *errcode == 0, returns total bytes written (which will
// be equal to input size).
Expand Down Expand Up @@ -201,8 +209,7 @@ STATIC mp_obj_t stream_read_generic(size_t n_args, const mp_obj_t *args, byte fl
}
}
}

return mp_obj_new_str_from_vstr(&mp_type_str, &vstr);
return mp_obj_new_str_from_vstr_check(&mp_type_str, &vstr);
}
#endif

Expand All @@ -223,7 +230,7 @@ STATIC mp_obj_t stream_read_generic(size_t n_args, const mp_obj_t *args, byte fl
mp_raise_OSError(error);
} else {
vstr.len = out_sz;
return mp_obj_new_str_from_vstr(STREAM_CONTENT_TYPE(stream_p), &vstr);
return mp_obj_new_str_from_vstr_check(STREAM_CONTENT_TYPE(stream_p), &vstr);
}
}

Expand Down Expand Up @@ -364,7 +371,7 @@ STATIC mp_obj_t stream_readall(mp_obj_t self_in) {
}

vstr.len = total_size;
return mp_obj_new_str_from_vstr(STREAM_CONTENT_TYPE(stream_p), &vstr);
return mp_obj_new_str_from_vstr_check(STREAM_CONTENT_TYPE(stream_p), &vstr);
}

// Unbuffered, inefficient implementation of readline() for raw I/O files.
Expand Down Expand Up @@ -417,7 +424,7 @@ STATIC mp_obj_t stream_unbuffered_readline(size_t n_args, const mp_obj_t *args)
}
}

return mp_obj_new_str_from_vstr(STREAM_CONTENT_TYPE(stream_p), &vstr);
return mp_obj_new_str_from_vstr_check(STREAM_CONTENT_TYPE(stream_p), &vstr);
}
MP_DEFINE_CONST_FUN_OBJ_VAR_BETWEEN(mp_stream_unbuffered_readline_obj, 1, 2, stream_unbuffered_readline);

Expand Down
2 changes: 1 addition & 1 deletion tests/extmod/qrio.py
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@
raise SystemExit

loc = __file__.rsplit("/", 1)[0]
with open(f"{loc}/data/qr.pgm") as f:
with open(f"{loc}/data/qr.pgm", "rb") as f:
content = f.read()[-320 * 240 :]

decoder = qrio.QRDecoder(320, 240)
Expand Down

0 comments on commit 7717ab8

Please sign in to comment.