pybase16384 0.3.6__cp311-cp311-macosx_10_9_x86_64.whl → 0.3.8__cp311-cp311-macosx_10_9_x86_64.whl

pybase16384/__init__.py

@@ -19,54 +19,70 @@ if not _should_use_cffi():
     from pybase16384.backends.cython import (
         DECBUFSZ,
         ENCBUFSZ,
+        FLAG_DO_SUM_CHECK_FORCELY,
         FLAG_NOHEADER,
         FLAG_SUM_CHECK_ON_REMAIN,
-        SIMPLE_SUM_INIT_VALUE,
         _decode,
         _decode_into,
+        _decode_into_safe,
+        _decode_safe,
         _encode,
         _encode_into,
+        _encode_into_safe,
+        _encode_safe,
         decode_fd,
         decode_fd_detailed,
         decode_file,
+        decode_file_safe,
         decode_len,
         decode_local_file,
         decode_local_file_detailed,
+        decode_stream_detailed,
         encode_fd,
         encode_fd_detailed,
         encode_file,
+        encode_file_safe,
         encode_len,
         encode_local_file,
         encode_local_file_detailed,
+        encode_stream_detailed,
         is_64bits,
     )
 else:
     from pybase16384.backends.cffi import (
         DECBUFSZ,
         ENCBUFSZ,
+        FLAG_DO_SUM_CHECK_FORCELY,
         FLAG_NOHEADER,
         FLAG_SUM_CHECK_ON_REMAIN,
-        SIMPLE_SUM_INIT_VALUE,
         _decode,
         _decode_into,
+        _decode_into_safe,
+        _decode_safe,
         _encode,
         _encode_into,
+        _encode_into_safe,
+        _encode_safe,
         decode_fd,
         decode_fd_detailed,
         decode_file,
+        decode_file_safe,
         decode_len,
         decode_local_file,
         decode_local_file_detailed,
+        decode_stream_detailed,
         encode_fd,
         encode_fd_detailed,
         encode_file,
+        encode_file_safe,
         encode_len,
         encode_local_file,
         encode_local_file_detailed,
+        encode_stream_detailed,
         is_64bits,
     )
 
-__version__ = "0.3.6"
+__version__ = "0.3.8"
 
 
 def encode(data: bytes) -> bytes:
@@ -76,6 +92,13 @@ def encode(data: bytes) -> bytes:
     return out.getvalue()
 
 
+def encode_safe(data: bytes) -> bytes:
+    inp = BytesIO(data)
+    out = BytesIO()
+    encode_file_safe(inp, out, False, len(data) // 7)
+    return out.getvalue()
+
+
 def decode(data: bytes) -> bytes:
     inp = BytesIO(data)
     out = BytesIO()
@@ -83,6 +106,13 @@ def decode(data: bytes) -> bytes:
     return out.getvalue()
 
 
+def decode_safe(data: bytes) -> bytes:
+    inp = BytesIO(data)
+    out = BytesIO()
+    decode_file_safe(inp, out, len(data) // 8)
+    return out.getvalue()
+
+
 def encode_from_string(data: str, write_head: bool = False) -> bytes:
     bt = data.encode()
     inp = BytesIO(bt)
@@ -91,6 +121,14 @@ def encode_from_string(data: str, write_head: bool = False) -> bytes:
     return out.getvalue()
 
 
+def encode_from_string_safe(data: str, write_head: bool = False) -> bytes:
+    bt = data.encode()
+    inp = BytesIO(bt)
+    out = BytesIO()
+    encode_file_safe(inp, out, write_head, len(bt) // 7)
+    return out.getvalue()
+
+
 def encode_to_string(data: bytes) -> str:
     inp = BytesIO(data)
     out = BytesIO()
@@ -98,6 +136,13 @@ def encode_to_string(data: bytes) -> str:
     return out.getvalue().decode("utf-16-be")
 
 
+def encode_to_string_safe(data: bytes) -> str:
+    inp = BytesIO(data)
+    out = BytesIO()
+    encode_file_safe(inp, out, False, len(data) // 7)
+    return out.getvalue().decode("utf-16-be")
+
+
 def encode_string(data: str) -> str:
     data = data.encode()
     inp = BytesIO(data)
@@ -106,6 +151,14 @@ def encode_string(data: str) -> str:
     return out.getvalue().decode("utf-16-be")
 
 
+def encode_string_safe(data: str) -> str:
+    data = data.encode()
+    inp = BytesIO(data)
+    out = BytesIO()
+    encode_file_safe(inp, out, False, len(data) // 7)
+    return out.getvalue().decode("utf-16-be")
+
+
 def decode_from_bytes(data: bytes) -> str:
     inp = BytesIO(data)
     out = BytesIO()
@@ -113,6 +166,13 @@ def decode_from_bytes(data: bytes) -> str:
     return out.getvalue().decode()
 
 
+def decode_from_bytes_safe(data: bytes) -> str:
+    inp = BytesIO(data)
+    out = BytesIO()
+    decode_file_safe(inp, out, len(data) // 8)
+    return out.getvalue().decode()
+
+
 def decode_from_string(data: str) -> bytes:
     bt = data.encode("utf-16-be")
     inp = BytesIO(bt)
@@ -121,9 +181,25 @@ def decode_from_string(data: str) -> bytes:
     return out.getvalue()
 
 
+def decode_from_string_safe(data: str) -> bytes:
+    bt = data.encode("utf-16-be")
+    inp = BytesIO(bt)
+    out = BytesIO()
+    decode_file_safe(inp, out, len(bt) // 8)
+    return out.getvalue()
+
+
 def decode_string(data: str) -> str:
     bt = data.encode("utf-16-be")
     inp = BytesIO(bt)
     out = BytesIO()
     decode_file(inp, out, len(bt) // 8)
     return out.getvalue().decode()
+
+
+def decode_string_safe(data: str) -> str:
+    bt = data.encode("utf-16-be")
+    inp = BytesIO(bt)
+    out = BytesIO()
+    decode_file_safe(inp, out, len(bt) // 8)
+    return out.getvalue().decode()

pybase16384/backends/cffi/__init__.py

@@ -1,6 +1,7 @@
 """
 Copyright (c) 2008-2021 synodriver <synodriver@gmail.com>
 """
+
 from pathlib import Path
 from typing import IO
 
@@ -15,7 +16,7 @@ ENCBUFSZ = lib.get_encsize()
 DECBUFSZ = lib.get_decsize()
 FLAG_NOHEADER = lib.BASE16384_FLAG_NOHEADER_()
 FLAG_SUM_CHECK_ON_REMAIN = lib.BASE16384_FLAG_SUM_CHECK_ON_REMAIN_()
-SIMPLE_SUM_INIT_VALUE = lib.BASE16384_SIMPLE_SUM_INIT_VALUE_()
+FLAG_DO_SUM_CHECK_FORCELY = lib.BASE16384_FLAG_DO_SUM_CHECK_FORCELY_()
 
 
 # -----------------low level api------------------------------
@@ -29,10 +30,26 @@ def _encode(data: bytes) -> bytes:
     return ffi.unpack(output_buf, count)
 
 
+def _encode_safe(data: bytes) -> bytes:
+    length = len(data)
+    output_size = encode_len(length)
+    output_buf = ffi.new(f"char[{output_size}]")
+    if output_buf == ffi.NULL:
+        raise MemoryError
+    count = lib.base16384_encode_safe(ffi.from_buffer(data), length, output_buf)
+    return ffi.unpack(output_buf, count)
+
+
 def _encode_into(data: bytes, out: bytearray) -> int:
     return lib.base16384_encode(ffi.from_buffer(data), len(data), ffi.from_buffer(out))
 
 
+def _encode_into_safe(data: bytes, out: bytearray) -> int:
+    return lib.base16384_encode_safe(
+        ffi.from_buffer(data), len(data), ffi.from_buffer(out)
+    )
+
+
 def _decode(data: bytes) -> bytes:
     length = len(data)
     output_size = decode_len(length, 0) + 16
@@ -43,10 +60,26 @@ def _decode(data: bytes) -> bytes:
     return ffi.unpack(output_buf, count)
 
 
+def _decode_safe(data: bytes) -> bytes:
+    length = len(data)
+    output_size = decode_len(length, 0)
+    output_buf = ffi.new(f"char[{output_size}]")
+    if output_buf == ffi.NULL:
+        raise MemoryError
+    count = lib.base16384_decode_safe(ffi.from_buffer(data), length, output_buf)
+    return ffi.unpack(output_buf, count)
+
+
 def _decode_into(data: bytes, out: bytearray) -> int:
     return lib.base16384_decode(ffi.from_buffer(data), len(data), ffi.from_buffer(out))
 
 
+def _decode_into_safe(data: bytes, out: bytearray) -> int:
+    return lib.base16384_decode_safe(
+        ffi.from_buffer(data), len(data), ffi.from_buffer(out)
+    )
+
+
 def is_64bits() -> bool:
     return bool(lib.pybase16384_64bits())
 
@@ -73,11 +106,15 @@ def encode_file(input: IO, output: IO, write_head: bool = False, buf_rate: int =
         output.write(b"\xfe\xff")
 
     current_buf_len: int = buf_rate * 7  # 一次读取这么多字节
-    output_size: int = encode_len(current_buf_len) + 16  # 因为encode_len不是单调的 这16备用
+    output_size: int = (
+        encode_len(current_buf_len) + 16
+    )  # 因为encode_len不是单调的 这16备用
     output_buf = ffi.new(f"char[{output_size}]")
     if output_buf == ffi.NULL:
         raise MemoryError
-    first_check: int = 1  # 检查一次就行了 怎么可能出现第一次读出来是bytes 以后又变卦了的对象呢 不会吧不会吧
+    first_check: int = (
+        1  # 检查一次就行了 怎么可能出现第一次读出来是bytes 以后又变卦了的对象呢 不会吧不会吧
+    )
     while True:
         chunk = input.read(current_buf_len)
         if first_check:
@@ -88,7 +125,9 @@ def encode_file(input: IO, output: IO, write_head: bool = False, buf_rate: int =
                 )
         size = len(chunk)
         if size < current_buf_len:  # 数据不够了 要减小一次读取的量
-            if buf_rate > 1:  # 重新设置一次读取的大小 重新设置流的位置 当然要是已经是一次读取7字节了 那就不能再变小了 直接encode吧
+            if (
+                buf_rate > 1
+            ):  # 重新设置一次读取的大小 重新设置流的位置 当然要是已经是一次读取7字节了 那就不能再变小了 直接encode吧
                 buf_rate = buf_rate // 2
                 current_buf_len = buf_rate * 7
                 input.seek(-size, 1)
@@ -100,6 +139,56 @@ def encode_file(input: IO, output: IO, write_head: bool = False, buf_rate: int =
             break
 
 
+def encode_file_safe(
+    input: IO, output: IO, write_head: bool = False, buf_rate: int = 10
+):
+    if not _check_file(input):
+        raise TypeError(
+            "input except a file-like object, got %s" % type(input).__name__
+        )
+    if not _check_file(output):
+        raise TypeError(
+            "output except a file-like object, got %s" % type(input).__name__
+        )
+    if buf_rate <= 0:
+        buf_rate = 1
+    if write_head:
+        output.write(b"\xfe\xff")
+
+    current_buf_len: int = buf_rate * 7  # 一次读取这么多字节
+    output_size: int = encode_len(
+        current_buf_len
+    )  # 因为encode_len不是单调的 safe不用加16
+    output_buf = ffi.new(f"char[{output_size}]")
+    if output_buf == ffi.NULL:
+        raise MemoryError
+    first_check: int = (
+        1  # 检查一次就行了 怎么可能出现第一次读出来是bytes 以后又变卦了的对象呢 不会吧不会吧
+    )
+    while True:
+        chunk = input.read(current_buf_len)
+        if first_check:
+            first_check = 0
+            if not isinstance(chunk, bytes):
+                raise TypeError(
+                    f"input must be a file-like rb object, got {type(input).__name__}"
+                )
+        size = len(chunk)
+        if size < current_buf_len:  # 数据不够了 要减小一次读取的量
+            if (
+                buf_rate > 1
+            ):  # 重新设置一次读取的大小 重新设置流的位置 当然要是已经是一次读取7字节了 那就不能再变小了 直接encode吧
+                buf_rate = buf_rate // 2
+                current_buf_len = buf_rate * 7
+                input.seek(-size, 1)
+                continue
+
+        count = lib.base16384_encode_safe(ffi.from_buffer(chunk), size, output_buf)
+        output.write(ffi.unpack(output_buf, count))
+        if size < 7:
+            break
+
+
 def decode_file(input: IO, output: IO, buf_rate: int = 10):
     if not _check_file(input):
         raise TypeError(
@@ -150,6 +239,56 @@ def decode_file(input: IO, output: IO, buf_rate: int = 10):
         output.write(ffi.unpack(output_buf, count))
 
 
+def decode_file_safe(input: IO, output: IO, buf_rate: int = 10):
+    if not _check_file(input):
+        raise TypeError(
+            "input except a file-like object, got %s" % type(input).__name__
+        )
+    if not _check_file(output):
+        raise TypeError(
+            "output except a file-like object, got %s" % type(output).__name__
+        )
+    if buf_rate <= 0:
+        buf_rate = 1
+
+    chunk = input.read(1)  # type: bytes
+    if not isinstance(chunk, bytes):
+        raise TypeError(
+            f"input must be a file-like rb object, got {type(input).__name__}"
+        )
+    if chunk == b"\xfe":  # 去头
+        input.read(1)
+    else:
+        input.seek(0, 0)  # 没有头 回到开头
+
+    current_buf_len: int = buf_rate * 8
+    output_size: int = decode_len(current_buf_len, 0)
+    output_buf = ffi.new(f"char[{output_size}]")
+    if output_buf == ffi.NULL:
+        raise MemoryError
+    while True:
+        chunk = input.read(current_buf_len)  # 8的倍数
+        size = len(chunk)
+        if size == 0:
+            break
+        if size < current_buf_len:  # 长度不够了
+            if buf_rate > 1:  # 还能继续变小
+                buf_rate = buf_rate // 2  # 重新设置一次读取的大小
+                current_buf_len = buf_rate * 8
+                input.seek(-size, 1)
+                continue
+        tmp = input.read(2)  # type: bytes
+        if len(tmp) == 2:
+            if tmp[0] == 61:  # = stream完了   一次解码8n+2个字节
+                chunk += tmp
+                size += 2
+            else:
+                input.seek(-2, 1)
+
+        count = lib.base16384_decode_safe(ffi.from_buffer(chunk), size, output_buf)
+        output.write(ffi.unpack(output_buf, count))
+
+
 def ensure_bytes(inp) -> bytes:
     if isinstance(inp, str):
         return inp.encode()
@@ -265,3 +404,60 @@ def decode_fd_detailed(inp: int, out: int, flag: int) -> None:
     ret = lib.base16384_decode_fd_detailed(inp, out, encbuf, decbuf, flag)
     if ret != lib.base16384_err_ok:
         raise ValueError(err_to_str(ret))
+
+
+# stream
+@ffi.def_extern()
+def b14_readcallback(client_data, buffer, count: int):
+    file = ffi.from_handle(client_data)
+    data: bytes = file.read(count)
+    data_size: int = len(data)
+    ffi.memmove(buffer, data, data_size)
+    return data_size
+
+
+@ffi.def_extern()
+def b14_writecallback(client_data, buffer, count: int):
+    file = ffi.from_handle(client_data)
+    data: bytes = ffi.unpack(ffi.cast("char*", buffer), count)
+    return file.write(data)
+
+
+def encode_stream_detailed(inp, out, flag: int):
+    encbuf = ffi.new(f"char[{ENCBUFSZ}]")
+    decbuf = ffi.new(f"char[{DECBUFSZ}]")
+
+    inpstream = ffi.new(f"base16384_stream_t *")
+    inpstream.f.reader = lib.b14_readcallback
+    handle_inp = ffi.new_handle(inp)
+    inpstream.client_data = handle_inp
+
+    outstream = ffi.new(f"base16384_stream_t *")
+    outstream.f.writer = lib.b14_writecallback
+    handle_out = ffi.new_handle(out)
+    outstream.client_data = handle_out
+    ret = lib.base16384_encode_stream_detailed(
+        inpstream, outstream, encbuf, decbuf, flag
+    )
+    if ret != lib.base16384_err_ok:
+        raise ValueError(err_to_str(ret))
+
+
+def decode_stream_detailed(inp, out, flag: int):
+    encbuf = ffi.new(f"char[{ENCBUFSZ}]")
+    decbuf = ffi.new(f"char[{DECBUFSZ}]")
+
+    inpstream = ffi.new(f"base16384_stream_t *")
+    inpstream.f.reader = lib.b14_readcallback
+    handle_inp = ffi.new_handle(inp)
+    inpstream.client_data = handle_inp
+
+    outstream = ffi.new(f"base16384_stream_t *")
+    outstream.f.writer = lib.b14_writecallback
+    handle_out = ffi.new_handle(out)
+    outstream.client_data = handle_out
+    ret = lib.base16384_decode_stream_detailed(
+        inpstream, outstream, encbuf, decbuf, flag
+    )
+    if ret != lib.base16384_err_ok:
+        raise ValueError(err_to_str(ret))

pybase16384/backends/cffi/build.py

@@ -1,6 +1,7 @@
 """
 Copyright (c) 2008-2021 synodriver <synodriver@gmail.com>
 """
+
 import platform
 import sys
 
@@ -74,6 +75,8 @@ base16384_err_t base16384_decode_fd(int input, int output, char* encbuf, char* d
 
 int base16384_encode_unsafe(const char * data, int dlen, char * buf);
 int base16384_decode_unsafe(const char * data, int dlen, char * buf);
+int base16384_encode_safe(const char * data, int dlen, char * buf);
+int base16384_decode_safe(const char * data, int dlen, char * buf);
 
 base16384_err_t base16384_encode_file_detailed(const char* input, const char* output, char* encbuf, char* decbuf, int flag);
 base16384_err_t base16384_decode_file_detailed(const char* input, const char* output, char* encbuf, char* decbuf, int flag);
@@ -92,7 +95,40 @@ int BASE16384_FLAG_NOHEADER_();
 
 int BASE16384_FLAG_SUM_CHECK_ON_REMAIN_();
 
-int BASE16384_SIMPLE_SUM_INIT_VALUE_();
+int BASE16384_FLAG_DO_SUM_CHECK_FORCELY_();
+
+typedef ssize_t (*base16384_reader_t)(const void *client_data, void *buffer, size_t count);
+
+/**
+ * @brief custom writer function interface
+ * @param client_data the data pointer defined by the client
+ * @param buffer from where read data
+ * @param count write bytes count
+ * @return the size written
+*/
+typedef ssize_t (*base16384_writer_t)(const void *client_data, const void *buffer, size_t count);
+
+union base16384_io_function_t {
+	base16384_reader_t reader;
+	base16384_writer_t writer;
+};
+typedef union base16384_io_function_t base16384_io_function_t;
+
+struct base16384_stream_t {
+	base16384_io_function_t f;
+	void *client_data;
+};
+/**
+ * @brief for stream encode/decode
+*/
+typedef struct base16384_stream_t base16384_stream_t;
+
+base16384_err_t base16384_encode_stream_detailed(base16384_stream_t* input, base16384_stream_t*  output, char* encbuf, char* decbuf, int flag);
+base16384_err_t base16384_decode_stream_detailed(base16384_stream_t* input, base16384_stream_t*  output, char* encbuf, char* decbuf, int flag);
+
+extern "Python" ssize_t b14_readcallback(const void *client_data, void *buffer, size_t count);
+
+extern "Python" ssize_t b14_writecallback(const void *client_data, const void *buffer, size_t count);
     """
 )
 
@@ -100,9 +136,9 @@ source = """
 #include "base16384.h"
 
 #ifdef CPUBIT32
-#define pybase16384_64bits() 0
+#define pybase16384_64bits() (0)
 #else
-#define pybase16384_64bits() 1
+#define pybase16384_64bits() (1)
 #endif
 
 int get_encsize()
@@ -125,16 +161,20 @@ int BASE16384_FLAG_SUM_CHECK_ON_REMAIN_()
     return BASE16384_FLAG_SUM_CHECK_ON_REMAIN;
 }
 
-int BASE16384_SIMPLE_SUM_INIT_VALUE_()
+int BASE16384_FLAG_DO_SUM_CHECK_FORCELY_()
 {
-    return BASE16384_SIMPLE_SUM_INIT_VALUE;
+    return BASE16384_FLAG_DO_SUM_CHECK_FORCELY;
 }
 """
 
 ffibuilder.set_source(
     "pybase16384.backends.cffi._core",
     source,
-    sources=[f"./base16384/base14{CPUBIT}.c", "./base16384/file.c"],
+    sources=[
+        f"./base16384/base14{CPUBIT}.c",
+        "./base16384/file.c",
+        "./base16384/wrap.c",
+    ],
     include_dirs=["./base16384"],
     define_macros=macro_base,
 )

pybase16384/backends/cython/__init__.py

@@ -1,27 +1,36 @@
 """
 Copyright (c) 2008-2021 synodriver <synodriver@gmail.com>
 """
+
 from pybase16384.backends.cython._core import (
     DECBUFSZ,
     ENCBUFSZ,
+    FLAG_DO_SUM_CHECK_FORCELY,
     FLAG_NOHEADER,
     FLAG_SUM_CHECK_ON_REMAIN,
-    SIMPLE_SUM_INIT_VALUE,
     _decode,
     _decode_into,
+    _decode_into_safe,
+    _decode_safe,
     _encode,
     _encode_into,
+    _encode_into_safe,
+    _encode_safe,
     decode_fd,
     decode_fd_detailed,
     decode_file,
+    decode_file_safe,
     decode_len,
     decode_local_file,
     decode_local_file_detailed,
+    decode_stream_detailed,
     encode_fd,
     encode_fd_detailed,
     encode_file,
+    encode_file_safe,
     encode_len,
     encode_local_file,
     encode_local_file_detailed,
+    encode_stream_detailed,
     is_64bits,
 )