關於ros 中msg、srv消息的md5值計算。
官方解釋
http://wiki.ros.org/ROS/Technical%20Overview#Message_serialization_and_msg_MD5_sums
Message serialization and msg MD5 sums
Messages are serialized in a very compact representation that roughly corresponds to a c-struct-like serialization of the message data in little endian format. The compact representation means that two nodes communicating must agree on the layout of the message data.
Message types (msgs) in ROS are versioned using a special MD5 sum calculation of the msg text. In general, client libraries do not implement this MD5 sum calculation directly, instead storing this MD5 sum in auto-generated message source code using the output of roslib/scripts/gendeps. For reference, this MD5 sum is calculated from the MD5 text of the .msg file, where the MD5 text is the .msg text with:
comments removed
whitespace removed
package names of dependencies removed
constants reordered ahead of other declarations
In order to catch changes that occur in embedded message types, the MD5 text is concatenated with the MD5 text of each of the embedded types, in the order that they appear.
翻譯
消息序列化和消息MD5 sums
消息以非常緊湊的表示形式進行序列化，該表示形式相當於以小尾 格式（little endian format）對消息數據進行 c結構體 式的序列化。緊湊表示意味着通信的兩個節點必須就消息數據的佈局達成一致。
ROS中的消息類型（msg）使用msg文本的特殊MD5 sum計算進行版本控制。
一般來說，客戶端庫不會直接實現此MD5 sum計算，
而是使用roslib/scripts/gendeps的輸出，將這個MD5 sum存儲在自動生成的消息源代碼中。作爲參考，此MD5 sum是對.msg文件的 MD5文本 進行計算得出的，其中 MD5文本 就是.msg文本，處理有：
刪除註釋
刪除空白
刪除依賴項的包名稱
常量排列在其他聲明之前
爲了捕捉嵌入消息類型中發生的更改，MD5文本將按照它們出現的順序與每個嵌入類型的MD5文本連接起來。
常量靠前排列舉例
rosmsg show rosserial_msgs/TopicInfo
uint16 ID_PUBLISHER=0
uint16 ID_SUBSCRIBER=1
uint16 ID_SERVICE_SERVER=2
uint16 ID_SERVICE_CLIENT=4
uint16 ID_PARAMETER_REQUEST=6
uint16 ID_LOG=7
uint16 ID_TIME=10
uint16 ID_TX_STOP=11
uint16 topic_id
string topic_name
string message_type
string md5sum
int32 buffer_size
消息爲空時
rosmsg md5 std_msgs/Empty
d41d8cd98f00b204e9800998ecf8427e
即
md5("")=d41d8cd98f00b204e9800998ecf8427e
如包含多個數據的消息
https://github.com/ros-drivers/rosserial/blob/melodic-devel/rosserial_arduino/msg/Adc.msg
md5(
"uint16 adc0
uint16 adc1
uint16 adc2
uint16 adc3
uint16 adc4
uint16 adc5"
)=6d7853a614e2e821319068311f2af25b
消息類型有嵌套時md5的計算
以std_msgs/MultiArrayLayout爲例講解一下
使用rosmsg show std_msgs/MultiArrayLayout命令可以快捷的得到清晰的嵌套關係
rosmsg show std_msgs/MultiArrayLayout
std_msgs/MultiArrayDimension[] dim
string label
uint32 size
uint32 stride
uint32 data_offset
它包含了MultiArrayDimension數組類型的dim 和uint32類型的data_offset
其中MultiArrayDimension的msg文件裏是這樣定義的
string label # label of given dimension 維度標籤
uint32 size # size of given dimension (in type units)包含子維度的個數//E文怎麼理解？
uint32 stride # stride of given dimension 這一維的跨度，這一維包含數據量
MultiArrayDimension的md5
md5(
"string label
uint32 size
uint32 stride"
)=4cd0c83a8683deae40ecdac60e53bfa8 暫時放着待用
我們再看std_msgs/MultiArrayLayout的定義，並解讀註釋
源碼+我的簡譯標註：
# The multiarray declares a generic multi-dimensional array of a
# particular data type. Dimensions are ordered from outer most
# to inner most.
#這個多維數據佈局的定義，是爲了界定多維數組的數據結構而聲明的
#針對特定數據類型的、通用的、多維的數組。
#維度是從最外到內的順序（體->面->線->點）排列。
MultiArrayDimension[] dim # Array of dimension properties維度定義
uint32 data_offset # padding elements at front of data 在多維數組數據之前的 數據量。
# Accessors should ALWAYS be written in terms of dimension stride
# and specified outer-most dimension first.
# 強調順序（體->面->線->點）排列
# multiarray(i,j,k) = data[data_offset + dim_stride[1]*i + dim_stride[2]*j + k]
#算法
# A standard, 3-channel 640x480 image with interleaved color channels
# would be specified as:
#爲了更好理解我們給它的標籤重命名加以解釋
# dim[0].label = "height"#這個名字起 image
# dim[0].size = 480#它包含480個低維度row單位，#跨度921600 個基本單位
# dim[0].stride = 3*640*480 = 921600 (note dim[0] stride is just size of image)跨度是圖片的大小
# dim[1].label = "width" #這個名字起row
# dim[1].size = 640 #它包含640個pix單位
# dim[1].stride = 3*640 = 1920 #跨度1920個基本單位
# dim[2].label = "channel"#這個名字起pix，包含了3個基本單位（是具體什麼單位要看使用包含MultiArrayLayout消息的數據類型？待）
# dim[2].size = 3
# dim[2].stride = 3
#
# multiarray(i,j,k) refers to the ith row, jth column, and kth channel.
這個msg文件內容除去註釋是這樣的。
std_msgs/MultiArrayDimension[] dim
uint32 data_offset
去掉包名、[]，把MultiArrayDimension替換爲那個放着待用的md5字串得到
4cd0c83a8683deae40ecdac60e53bfa8 dim
uint32 data_offset
md5(
"4cd0c83a8683deae40ecdac60e53bfa8 dim
uint32 data_offset"
)=0fed2a11c13e11c5571b4e2a995a91a3
//疑問：替換爲什麼把[]去掉了，消息驗證是否會衝突? where is the "[]"，why dont like this？
4cd0c83a8683deae40ecdac60e53bfa8[] dim
uint32 data_offset
//
命令
rosmsg md5 std_msgs/MultiArrayLayout
0fed2a11c13e11c5571b4e2a995a91a3
srv
實際上它由兩個消息組成
srv分爲請求消息和響應消息兩部分用"---"分割
std_srvs/Empty
rossrv show std_srvs/Empty
---
request：
空字串
md5("")=d41d8cd98f00b204e9800998ecf8427e
response:
空字串
md5("")=d41d8cd98f00b204e9800998ecf8427e
srv：
rossrv md5 std_srvs/Empty
d41d8cd98f00b204e9800998ecf8427e
空字串+空字串還是空字串 也就是說request和response之間沒有"\n"
說明 md5("")=d41d8cd98f00b204e9800998ecf8427e
非空srv RequestParam
https://github.com/ros-drivers/rosserial/blob/melodic-devel/rosserial_msgs/srv/RequestParam.srv
string name
---
int32[] ints
float32[] floats
string[] strings
request：
md5("string name")=c1f3d28f1b044c871e6eff2e9fc3c667
response：
md5(
"int32[] ints
float32[] floats
string[] strings"
)=9f0e98bda65981986ddf53afa7a40e49
srv：
md5(
"string nameint32[] ints
float32[] floats
string[] strings"
)=d7a0c2be00c9fd03cc69f2863de9c4d9
rossrv md5 rosserial_msgs/RequestParam
d7a0c2be00c9fd03cc69f2863de9c4d9
std_srvs/SetBool
rossrv show std_srvs/SetBool
bool data
---
bool success
string message
md5計算
request:
md5("bool data")=8b94c1b53db61fb6aed406028ad6332a
response：
md5(
"bool success
string message"
)=937c9679a518e3a18d831e57125ea522
srv:
md5(
"bool databool success
string message"
)=09fb03525b03e7ea1fd3992bafd87e16
rossrv md5 std_srvs/SetBool
09fb03525b03e7ea1fd3992bafd87e16
總結
md5計算的兩個注意點，過程中 msg 消息遞歸處理時類型的 "[]" 被剔除了，srv中requset和response兩個消息之間沒有"\n",所以空srv和空msg的md5值一樣。
下面是計算md5的部分源碼
def compute_md5_text(msg_context, spec):
"""
Compute the text used for md5 calculation. MD5 spec states that we
removes comments and non-meaningful whitespace. We also strip
packages names from type names. For convenience sake, constants are
reordered ahead of other declarations, in the order that they were
originally defined.
:returns: text for ROS MD5-processing, ``str``
"""
package = spec.package
buff = StringIO()
#常量靠前放
for c in spec.constants:
buff.write("%s %s=%s\n"%(c.type, c.name, c.val_text))
for type_, name in zip(spec.types, spec.names):
msg_type = bare_msg_type(type_) #剔除[],只留類型名
# md5 spec strips package names
if is_builtin(msg_type):
buff.write("%s %s\n"%(type_, name))#內建類型直接寫
else:
# recursively generate md5 for subtype. have to build up
# dependency representation for subtype in order to
# generate md5 遞歸的處理引用類型
sub_pkg, _ = names.package_resource_name(msg_type)
sub_pkg = sub_pkg or package
sub_spec = msg_context.get_registered(msg_type)
sub_md5 = compute_md5(msg_context, sub_spec)
buff.write("%s %s\n"%(sub_md5, name))
return buff.getvalue().strip() # remove trailing new line
def _compute_hash(msg_context, spec, hash):
"""
subroutine of compute_md5()
:param msg_context: :class:`MsgContext` instance to load dependencies into/from.
:param spec: :class:`MsgSpec` to compute hash for.
:param hash: hash instance
"""
# accumulate the hash
# - root file
if isinstance(spec, MsgSpec):#msg類型
hash.update(compute_md5_text(msg_context, spec).encode())
elif isinstance(spec, SrvSpec):#srv類型將request和response的內容連起來做md5(注意它們之間沒有"\n")
hash.update(compute_md5_text(msg_context, spec.request).encode())
hash.update(compute_md5_text(msg_context, spec.response).encode())
else:
raise Exception("[%s] is not a message or service"%spec)
return hash.hexdigest()
def compute_md5(msg_context, spec):
"""
Compute md5 hash for message/service
:param msg_context: :class:`MsgContext` instance to load dependencies into/from.
:param spec: :class:`MsgSpec` to compute md5 for.
:returns: md5 hash, ``str``
"""
return _compute_hash(msg_context, spec, hashlib.md5())
//generator.py一窺調用
def srv_generator(msg_context, spec, search_path):
#...
#... genmsg.msg_loader.load_depends 加載解析到spec,
genmsg.msg_loader.load_depends(msg_context, spec, search_path)
md5 = genmsg.compute_md5(msg_context, spec)
yield "class %s(object):"%name
yield " _type = '%s'"%fulltype
yield " _md5sum = '%s'"%md5
yield " _request_class = %s"%req
yield " _response_class = %s"%resp
opt\ros\melodic\x64\lib\site-packages\genmsg\msg_loader.py節選
def load_msg_from_string(msg_context, text, full_name):
"""
Load message specification from a string.
NOTE: this will register the message in the *msg_context*.
:param msg_context: :class:`MsgContext` for finding loaded dependencies
:param text: .msg text , ``str``
:returns: :class:`MsgSpec` specification
:raises: :exc:`InvalidMsgSpec` If syntax errors or other problems are detected in file
"""
log("load_msg_from_string", full_name)
package_name, short_name = package_resource_name(full_name)
#解析出類型、名稱，常量
types = []
names = []
constants = []
for orig_line in text.split('\n'):
clean_line = _strip_comments(orig_line)
if not clean_line:
continue #ignore empty lines
if CONSTCHAR in clean_line:
constants.append(_load_constant_line(orig_line))#填入常量
else:
field_type, name = _load_field_line(orig_line, package_name)
types.append(field_type)#填入類型、名稱（id對應的）
names.append(name)
spec = MsgSpec(types, names, constants, text, full_name, package_name)#-------------生成MsgSpec類返回
msg_context.register(full_name, spec)
return spec
def load_srv_from_string(msg_context, text, full_name):
"""
Load :class:`SrvSpec` from the .srv file.
"""
text_in = StringIO()
text_out = StringIO()
accum = text_in
for l in text.split('\n'):
l = l.split(COMMENTCHAR)[0].strip() #strip comments
if l.startswith(IODELIM): #lenient, by request
accum = text_out
else:
accum.write(l+'\n')
# create separate MsgSpec objects for each half of file
msg_in = load_msg_from_string(msg_context, text_in.getvalue(), '%sRequest'%(full_name))#-----load_msg_from_string
msg_out = load_msg_from_string(msg_context, text_out.getvalue(), '%sResponse'%(full_name))
return SrvSpec(msg_in, msg_out, text, full_name)#---------------------------------由兩個MsgSpec生成一個SrvSpec返回
def load_srv_from_file(msg_context, file_path, full_name):
"""
Convert the .srv representation in the file to a :class:`SrvSpec` instance.
"""
log("Load spec from %s %s\n"%(file_path, full_name))
#打開並讀出文件 text
with open(file_path, 'r') as f:
text = f.read()
spec = load_srv_from_string(msg_context, text, full_name)#string轉換成SrvSpec類-------load_srv_from_string
msg_context.set_file('%sRequest'%(full_name), file_path)
msg_context.set_file('%sResponse'%(full_name), file_path)
return spec