def __convert_eol(self, node, revnr):
"""
Convert EOL of a node.
@type node: SvnDumpNode
@param node: The node to convert.
@type revnr: integer
@param revnr: The current revision number.
@rtype: SvnDumpNode
@return: The converted node.
"""
if not node.has_text():
# no text
print(" selected file, no text changes")
return node
# search for CR
need_conv = False
handle = node.text_open()
data = node.text_read(handle)
while len(data) > 0:
if data.find("\r") != -1:
# found one, need to convert the file
need_conv = True
break
data = node.text_read(handle)
fix = self.__fix
if need_conv:
# special fix option for rev/file ?
key = (revnr, node.get_path())
if self.__fix_rev_path.has_key(key):
fix = self.__fix_rev_path[key]
print(" selected file, convert (fix option %d)" % fix)
if self.__dry_run or fix == 0:
# do not convert with --dry-run or when there's nothing to fix
need_conv = False
else:
print(" selected file, no conversion required")
if need_conv:
# do the conversion
node.text_reopen(handle)
outfilename = self.__temp_file_name()
outfile = open(outfilename, "wb")
outlen = 0
md = sdt_md5()
data = node.text_read(handle)
carry = ""
warning_printed = False
while len(data) > 0:
if len(carry) != 0:
data = carry + data
n = len(data) - 1
carry = data[n]
if carry == "\r":
data = data[:n]
else:
carry = ""
if fix & 1 != 0:
data = data.replace("\r\n", "\n")
if fix & 2 != 0:
data = data.replace("\r", "\n")
if fix & 4 != 0:
data = data.replace("\r", "")
if not warning_printed and data.find("\r") >= 0:
warning_printed = True
print(" WARNING: file still contains CR")
print(" file: '%s'" % node.get_path())
if self.__warning_file is not None:
self.__warning_file.write(
"# WARNING: file still contains CR\n")
file = node.get_path()
while file[0] == "/":
file = file[1:]
tmpfile = file.replace("/", "__")
cmd = '$SVN cat -r %d "$REPOS/%s" > "%s"\n' % \
(revnr, node.get_path(), tmpfile)
self.__warning_file.write(cmd)
self.__warning_count += 1
md.update(data)
outfile.write(data)
outlen = outlen + len(data)
data = node.text_read(handle)
if len(carry) != 0:
if fix & 2 != 0:
carry = "\n"
elif fix & 4 != 0:
carry = ""
outfile.write(carry)
md.update(carry)
outlen += len(carry)
outfile.close()
newnode = SvnDumpNode(node.get_path(), node.get_action(),
node.get_kind())
if node.has_copy_from():
newnode.set_copy_from(node.get_copy_from_path(),
node.get_copy_from_rev())
if node.has_properties():
newnode.set_properties(node.get_properties())
newnode.set_text_file(outfilename, outlen, md.hexdigest())
else:
newnode = node
node.text_close(handle)
return newnode
def __compare_node(self, node1, node2, callback):
"""
Compare two nodes.
@type node1: SvnDumpNode
@param node1: First dump node.
@type node2: SvnDumpNode
@param node2: Second dump node.
@type callback: SvnDumpDiffCallback
@param callback: Callback object for diffs found.
"""
# compare path
if node1.get_path() != node2.get_path():
callback.node_diff("Path", node1.get_path(), node2.get_path())
return
# compare action
if node1.get_action() != node2.get_action():
callback.node_diff("Action", node1.get_action(), node2.get_action())
return
# compare kind
if node1.get_kind() != node2.get_kind():
callback.node_diff("Kind", node1.get_kind(), node2.get_kind())
return
# compare copy-from-path
if node1.get_copy_from_path() != node2.get_copy_from_path():
callback.node_diff("CopyFromPath", node1.get_copy_from_path(), node2.get_copy_from_path())
return
# compare copy-from-rev
if node1.get_copy_from_rev() != node2.get_copy_from_rev():
callback.node_diff("CopyFromRev", node1.get_copy_from_rev(), node2.get_copy_from_rev())
return
# compare props
self.__compare_properties(False, node1.get_properties(),
node2.get_properties(), callback)
# compare text
if node1.has_text() != node2.has_text():
callback.node_diff("HasText", node1.has_text(), node2.has_text())
return
if not node1.has_text():
# no text to compare
return
if node1.get_text_length() != node2.get_text_length():
callback.node_diff("TextLen", node1.get_text_length(), node2.get_text_length())
if node1.get_text_md5() != node2.get_text_md5():
callback.node_diff("TextMD5", node1.get_text_md5(), node2.get_text_md5())
md1 = sdt_md5()
md2 = sdt_md5()
handle1 = node1.text_open()
handle2 = node2.text_open()
str1 = node1.text_read(handle1)
str2 = node2.text_read(handle2)
n1 = len(str1)
n2 = len(str2)
cmpstr1 = ""
cmpstr2 = ""
defreadcount = 16384
readcount1 = defreadcount
readcount2 = defreadcount
# how to compare: 0=normal, 1=eol-check, 2=had a diff
cmpmode = 0
forceloop = False
while n1 > 0 or n2 > 0 or forceloop:
forceloop = False
if n1 > 0:
md1.update(str1)
cmpstr1 += str1
if n2 > 0:
md2.update(str2)
cmpstr2 += str2
if cmpstr1 == cmpstr2:
# save last char for possible eol compare
cmpstr1 = cmpstr1[-1:]
cmpstr2 = cmpstr2[-1:]
elif cmpmode == 0:
# binary compare had a diff
if self.__check_eol:
# start doing eol checks
cmpmode = 1
forceloop = True
else:
# no eol check so files differ
cmpmode = 2
cmpstr1 = ""
cmpstr2 = ""
elif cmpmode == 1:
# eol-diff or real diff?
cn1 = len(cmpstr1) - 1
cn2 = len(cmpstr2) - 1
i1 = 0
i2 = 0
while i1 < cn1 and i2 < cn2:
if ((cmpstr1[i1] == '\n' or cmpstr1[i1] == '\r') and
(cmpstr2[i2] == '\n' or cmpstr2[i2] == '\r')):
# check for LF/CRLF/CR sequence
if cmpstr1[i1] == '\r' and cmpstr1[i1 + 1] == '\n':
i1 += 1
if cmpstr2[i2] == '\r' and cmpstr2[i2 + 1] == '\n':
i2 += 1
elif cmpstr1[i1] != cmpstr2[i2]:
# found a diff
cmpmode = 2
cmpstr1 = ""
cmpstr2 = ""
break
# next...
i1 += 1
i2 += 1
# remove processed data from cmpstr and adjust readcount
cmpstr1 = cmpstr1[i1:]
cmpstr2 = cmpstr2[i2:]
cn1 = len(cmpstr1)
cn2 = len(cmpstr2)
if cmpmode == 1:
# adjust readcount
if cn1 > cn2:
readcount1 = defreadcount + cn1 - cn2
readcount2 = defreadcount
else:
readcount1 = defreadcount
readcount2 = defreadcount + cn2 - cn1
# if n1 > 0 and n2 > 0 and cn1 > 0 and cn2 > 0:
if n1 > 0 or n2 > 0:
forceloop = True
else:
# reset readcount
readcount1 = defreadcount
readcount2 = defreadcount
else:
# cmpmode = 2: just clear compare strings
cmpstr1 = ""
cmpstr2 = ""
# read more text...
if n1 > 0:
str1 = node1.text_read(handle1, readcount1)
n1 = len(str1)
if n2 > 0:
str2 = node2.text_read(handle2, readcount2)
n2 = len(str2)
if cmpmode == 1:
# compare trailing data (can only occur in eol-check mode)
cmpstr1 = cmpstr1.replace("\r\n", "\n").replace("\r", "\n")
cmpstr2 = cmpstr2.replace("\r\n", "\n").replace("\r", "\n")
if cmpstr1 != cmpstr2:
cmpmode = 2
mdstr1 = md1.hexdigest()
mdstr2 = md2.hexdigest()
if node1.get_text_md5() != mdstr1:
callback.wrong_md5(1, node1.get_text_md5(), mdstr1)
if node2.get_text_md5() != mdstr2:
callback.wrong_md5(2, node2.get_text_md5(), mdstr2)
if cmpmode == 1:
callback.text_diff("EOL")
elif cmpmode == 2:
callback.text_diff("Text")
def __compare_node(self, node1, node2, callback):
"""
Compare two nodes.
@type node1: SvnDumpNode
@param node1: First dump node.
@type node2: SvnDumpNode
@param node2: Second dump node.
@type callback: SvnDumpDiffCallback
@param callback: Callback object for diffs found.
"""
# compare path
if node1.get_path() != node2.get_path():
callback.node_diff("Path", node1.get_path(), node2.get_path())
return
# compare action
if node1.get_action() != node2.get_action():
callback.node_diff("Action", node1.get_action(),
node2.get_action())
return
# compare kind
if node1.get_kind() != node2.get_kind():
callback.node_diff("Kind", node1.get_kind(), node2.get_kind())
return
# compare copy-from-path
if node1.get_copy_from_path() != node2.get_copy_from_path():
callback.node_diff("CopyFromPath", node1.get_copy_from_path(),
node2.get_copy_from_path())
return
# compare copy-from-rev
if node1.get_copy_from_rev() != node2.get_copy_from_rev():
callback.node_diff("CopyFromRev", node1.get_copy_from_rev(),
node2.get_copy_from_rev())
return
# compare props
self.__compare_properties(False, node1.get_properties(),
node2.get_properties(), callback)
# compare text
if node1.has_text() != node2.has_text():
callback.node_diff("HasText", node1.has_text(), node2.has_text())
return
if not node1.has_text():
# no text to compare
return
if node1.get_text_length() != node2.get_text_length():
callback.node_diff("TextLen", node1.get_text_length(),
node2.get_text_length())
if node1.get_text_md5() != node2.get_text_md5():
callback.node_diff("TextMD5", node1.get_text_md5(),
node2.get_text_md5())
md1 = sdt_md5()
md2 = sdt_md5()
handle1 = node1.text_open()
handle2 = node2.text_open()
str1 = node1.text_read(handle1)
str2 = node2.text_read(handle2)
n1 = len(str1)
n2 = len(str2)
cmpstr1 = ""
cmpstr2 = ""
defreadcount = 16384
readcount1 = defreadcount
readcount2 = defreadcount
# how to compare: 0=normal, 1=eol-check, 2=had a diff
cmpmode = 0
forceloop = False
while n1 > 0 or n2 > 0 or forceloop:
forceloop = False
if n1 > 0:
md1.update(str1)
cmpstr1 += str1
if n2 > 0:
md2.update(str2)
cmpstr2 += str2
if cmpstr1 == cmpstr2:
# save last char for possible eol compare
cmpstr1 = cmpstr1[-1:]
cmpstr2 = cmpstr2[-1:]
elif cmpmode == 0:
# binary compare had a diff
if self.__check_eol:
# start doing eol checks
cmpmode = 1
forceloop = True
else:
# no eol check so files differ
cmpmode = 2
cmpstr1 = ""
cmpstr2 = ""
elif cmpmode == 1:
# eol-diff or real diff?
cn1 = len(cmpstr1) - 1
cn2 = len(cmpstr2) - 1
i1 = 0
i2 = 0
while i1 < cn1 and i2 < cn2:
if ((cmpstr1[i1] == '\n' or cmpstr1[i1] == '\r')
and (cmpstr2[i2] == '\n' or cmpstr2[i2] == '\r')):
# check for LF/CRLF/CR sequence
if cmpstr1[i1] == '\r' and cmpstr1[i1 + 1] == '\n':
i1 += 1
if cmpstr2[i2] == '\r' and cmpstr2[i2 + 1] == '\n':
i2 += 1
elif cmpstr1[i1] != cmpstr2[i2]:
# found a diff
cmpmode = 2
cmpstr1 = ""
cmpstr2 = ""
break
# next...
i1 += 1
i2 += 1
# remove processed data from cmpstr and adjust readcount
cmpstr1 = cmpstr1[i1:]
cmpstr2 = cmpstr2[i2:]
cn1 = len(cmpstr1)
cn2 = len(cmpstr2)
if cmpmode == 1:
# adjust readcount
if cn1 > cn2:
readcount1 = defreadcount + cn1 - cn2
readcount2 = defreadcount
else:
readcount1 = defreadcount
readcount2 = defreadcount + cn2 - cn1
#if n1 > 0 and n2 > 0 and cn1 > 0 and cn2 > 0:
if n1 > 0 or n2 > 0:
forceloop = True
else:
# reset readcount
readcount1 = defreadcount
readcount2 = defreadcount
else:
# cmpmode = 2: just clear compare strings
cmpstr1 = ""
cmpstr2 = ""
# read more text...
if n1 > 0:
str1 = node1.text_read(handle1, readcount1)
n1 = len(str1)
if n2 > 0:
str2 = node2.text_read(handle2, readcount2)
n2 = len(str2)
if cmpmode == 1:
# compare trailing data (can only occur in eol-check mode)
cmpstr1 = cmpstr1.replace("\r\n", "\n").replace("\r", "\n")
cmpstr2 = cmpstr2.replace("\r\n", "\n").replace("\r", "\n")
if cmpstr1 != cmpstr2:
cmpmode = 2
mdstr1 = md1.hexdigest()
mdstr2 = md2.hexdigest()
if node1.get_text_md5() != mdstr1:
callback.wrong_md5(1, node1.get_text_md5(), mdstr1)
if node2.get_text_md5() != mdstr2:
callback.wrong_md5(2, node2.get_text_md5(), mdstr2)
if cmpmode == 1:
callback.text_diff("EOL")
elif cmpmode == 2:
callback.text_diff("Text")
