Merge branch 'pack_writing'

Author: Byron

gitdb.pro

@@ -0,0 +1,48 @@
+
+OTHER_FILES += \
+    setup.py \
+    README.rst \
+    MANIFEST \
+    Makefile \
+    LICENSE \
+    AUTHORS \
+    gitdb/util.py \
+    gitdb/typ.py \
+    gitdb/stream.py \
+    gitdb/pack.py \
+    gitdb/__init__.py \
+    gitdb/fun.py \
+    gitdb/exc.py \
+    gitdb/base.py \
+    doc/source/tutorial.rst \
+    doc/source/intro.rst \
+    doc/source/index.rst \
+    doc/source/conf.py \
+    doc/source/changes.rst \
+    doc/source/api.rst \
+    doc/source/algorithm.rst \
+    gitdb/db/ref.py \
+    gitdb/db/pack.py \
+    gitdb/db/mem.py \
+    gitdb/db/loose.py \
+    gitdb/db/__init__.py \
+    gitdb/db/git.py \
+    gitdb/db/base.py \
+    gitdb/test/test_util.py \
+    gitdb/test/test_stream.py \
+    gitdb/test/test_pack.py \
+    gitdb/test/test_example.py \
+    gitdb/test/test_base.py \
+    gitdb/test/lib.py \
+    gitdb/test/__init__.py \
+    gitdb/test/performance/test_stream.py \
+    gitdb/test/performance/test_pack_streaming.py \
+    gitdb/test/performance/test_pack.py \
+    gitdb/test/performance/lib.py
+
+HEADERS += \
+    gitdb/_delta_apply.h
+
+SOURCES += \
+    gitdb/_fun.c \
+    gitdb/_delta_apply.c

gitdb.pro.user

@@ -1,4 +1,4 @@
-#include "_delta_apply.h"
+#include <_delta_apply.h>
 #include <stdint.h>
 #include <assert.h>
 #include <stdio.h>
@@ -463,7 +463,7 @@ void DIV_reset(DeltaInfoVector* vec)
 
 // Append one chunk to the end of the list, and return a pointer to it
 // It will not have been initialized !
-static inline
+inline
 DeltaInfo* DIV_append(DeltaInfoVector* vec)
 {
 	if (vec->size + 1 > vec->reserved_size){
@@ -703,7 +703,7 @@ typedef struct {
 } DeltaChunkList;
 
 
-static 
+ 
 int DCL_init(DeltaChunkList*self, PyObject *args, PyObject *kwds)
 {
 	if(args && PySequence_Size(args) > 0){
@@ -715,20 +715,20 @@ int DCL_init(DeltaChunkList*self, PyObject *args, PyObject *kwds)
 	return 0;
 }
 
-static
+
 void DCL_dealloc(DeltaChunkList* self)
 {
 	TSI_destroy(&(self->istream));
 }
 
-static
+
 PyObject* DCL_py_rbound(DeltaChunkList* self)
 {
 	return PyLong_FromUnsignedLongLong(self->istream.target_size);
 }
 
 // Write using a write function, taking remaining bytes from a base buffer
-static
+
 PyObject* DCL_apply(DeltaChunkList* self, PyObject* args)
 {
 	PyObject* pybuf = 0;
@@ -769,13 +769,13 @@ PyObject* DCL_apply(DeltaChunkList* self, PyObject* args)
 	Py_RETURN_NONE;
 }
 
-static PyMethodDef DCL_methods[] = {
+PyMethodDef DCL_methods[] = {
     {"apply", (PyCFunction)DCL_apply, METH_VARARGS, "Apply the given iterable of delta streams" },
     {"rbound", (PyCFunction)DCL_py_rbound, METH_NOARGS, NULL},
     {NULL}  /* Sentinel */
 };
 
-static PyTypeObject DeltaChunkListType = {
+PyTypeObject DeltaChunkListType = {
 	PyObject_HEAD_INIT(NULL)
 	0,						   /*ob_size*/
 	"DeltaChunkList",			/*tp_name*/
@@ -897,7 +897,7 @@ uint compute_chunk_count(const uchar* data, const uchar* dend, bool read_header)
 	return num_chunks;
 }
 
-static PyObject* connect_deltas(PyObject *self, PyObject *dstreams)
+PyObject* connect_deltas(PyObject *self, PyObject *dstreams)
 {
 	// obtain iterator
 	PyObject* stream_iter = 0;
@@ -1088,7 +1088,6 @@ static PyObject* connect_deltas(PyObject *self, PyObject *dstreams)
 
 // Write using a write function, taking remaining bytes from a base buffer
 // replaces the corresponding method in python
-static
 PyObject* apply_delta(PyObject* self, PyObject* args)
 {
 	PyObject* pybbuf = 0;

gitdb/_delta_apply.c

@@ -1,6 +1,6 @@
 #include <Python.h>
 
-static PyObject* connect_deltas(PyObject *self, PyObject *dstreams);
-static PyObject* apply_delta(PyObject* self, PyObject* args);
+extern PyObject* connect_deltas(PyObject *self, PyObject *dstreams);
+extern PyObject* apply_delta(PyObject* self, PyObject* args);
 
-static PyTypeObject DeltaChunkListType;
+extern PyTypeObject DeltaChunkListType;

gitdb/_delta_apply.h

@@ -17,6 +17,9 @@ class BadObject(ODBError):
 
 	def __str__(self):
 		return "BadObject: %s" % to_hex_sha(self.args[0])
+		
+class ParseError(ODBError):
+	"""Thrown if the parsing of a file failed due to an invalid format"""
 
 class AmbiguousObjectName(ODBError):
 	"""Thrown if a possibly shortened name does not uniquely represent a single object

gitdb/exc.py

@@ -48,7 +48,7 @@
 
 __all__ = ('is_loose_object', 'loose_object_header_info', 'msb_size', 'pack_object_header_info', 
 			'write_object', 'loose_object_header', 'stream_copy', 'apply_delta_data', 
-			'is_equal_canonical_sha', 'connect_deltas', 'DeltaChunkList')
+			'is_equal_canonical_sha', 'connect_deltas', 'DeltaChunkList', 'create_pack_object_header')
 
 
 #{ Structures
@@ -411,15 +411,25 @@ def pack_object_header_info(data):
 		size += (c & 0x7f) << s
 		s += 7
 	# END character loop
-	
-	try:
-		return (type_id, size, i)
-	except KeyError:
-		# invalid object type - we could try to be smart now and decode part 
-		# of the stream to get the info, problem is that we had trouble finding 
-		# the exact start of the content stream
-		raise BadObjectType(type_id)
-	# END handle exceptions
+	return (type_id, size, i)
+
+def create_pack_object_header(obj_type, obj_size):
+	""":return: string defining the pack header comprised of the object type
+	and its incompressed size in bytes
+	:parmam obj_type: pack type_id of the object
+	:param obj_size: uncompressed size in bytes of the following object stream"""
+	c = 0		# 1 byte
+	hdr = str()	# output string
+
+	c = (obj_type << 4) | (obj_size & 0xf)
+	obj_size >>= 4
+	while obj_size:
+		hdr += chr(c | 0x80)
+		c = obj_size & 0x7f
+		obj_size >>= 7
+	#END until size is consumed
+	hdr += chr(c)
+	return hdr
 
 def msb_size(data, offset=0):
 	"""