Conquest error using makefilename.lua

Hello, at the hospital we are using conquest as a backup pacs server, it is really great and I want to congratulate the entire team that works on its development. But I am having a problem, I am trying to generate a specific saving pattern for the files received, and to do this, reading the manual I found that it can be done through a lua script, and as an example it brings makefilename.lua. I have placed this in my dicom.ini file but I get a lua error: return string near <eof>. I clarify that I use the unmodified makefilename.lua and I get the same thing. What I can be doing wrong?

Error imágenes. Please help

Hello Marcel, I have tried with the indicated modification (FilenameSyntax = lua:dofile([[makefilename.lua]]) ), but I still get

*** lua syntax error [string "return dofile([[makefilename.dcm"]:1: unfinished long string near '<eof>' in 'return dofile([[makefilename.dcm'

I also don't understand what the error description refers to.

Juan

Dear Marcel, I tried to do what you asked, but the error is the same and I am not adding any more information.

*** lua syntax error [string "return dofile([[makefilename.dcm"]:1: unfinished long string near '<eof>' in 'return dofile([[makefilename.dcm'

Juan

Hello Marcel,

first of all, thank you very much for your response and dedication. Could you tell me the link to the modified dgate64.

I will use the makefilename.lua as a starting point, but my idea is to generate a hash with the jdb2 algorithm in 8-digit base32 to encode the studyuid and the serieuid in order to store it in order:

studydate/hash_studyuid/hash_serieuid/ imageuid.dcm

The purpose of this is to be able to have an easy order to locate the studies, and at the same time not have very long directory names as would be the case with the direct use of studyuid and serieuid.

If you find this interesting and I manage to make it work I can share it with you.

Dear Marcel,

Excellent, I will try the new .exe.

Regarding dbj2 instead of md5, I thought about this algorithm for the simplicity and speed of execution, since there will surely be a lot of images and speed is essential so as not to slow down the process. On the other hand, I thought about encoding it with base32 since it ensures valid names for directories with a good result space of 8 digits that minimize collisions between 2 hashes.

It would be very nice to integrate the function as you mention. I will do the test and send you the results and code.

Hello Marcel,

the dgate64.exe is from 2 months ago, could it not have been updated? I'm still getting the same error, but the date of it on GitHub caught my attention.

ok Marcel, while I'm giving you my code in Lua with the aforementioned algorithm, I haven't been able to test it yet because of the error we've been having, but if you see something you want to suggest, I'll gladly accept it.

--makefilename.lua
--Create names hash 8 digit from studyUID and SerieUID
--format: studydate/hash_studyuid/hash_serieuid/ imageuid.dcm
--
--Autor: Juan Rivaldez


-- Generate hash jdb2 codificate in 8 digit base32
function hash_name(input)
local hash = 5381
local base32_chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZ234567"
local result = ""


--First generate jdb2 hash
for i = 1, #input do
hash = ((hash << 5) + hash) + string.byte(input, i)
-- fixed to 40 bits for convert to 8 digit base32
hash = hash & 0xFFFFFFFFFF
end


--Codificate en 8 digit base32
for i = 1, 8 do
local index = (hash & 0x1F) + 1
result = base32_chars:sub(index, index) .. result
hash = hash >> 5
end


return result
end


local n = Data.StudyDate
return n .. '/' .. hash_name(Data.SeriesInstanceUID) .. '/' .. hash_name(Data.SeriesInstanceUID) .. '/' .. Data.SOPInstanceUID .. '.dcm'

I have tried to test the code, but I always have an error. I started doing simpler tests and I realized that when using functions within the Lua code I always get the same error. Could it be that functions cannot be used? In fact, the error it gives me has no association with the code, since any code I enter makes no difference.

I have discovered that the problem is not with the functions, apparently it does not like bit operations, not the shift or the and.

Well after haggling a bit with lua (in C it would have been very easy) , I came to the following conclusions:

-in the test runner (image - modify), neither indentations nor comments are supported in the code in lua.

-how lua 5.1 uses bit operations are not available. Therefore I was able to put together the code that is already working, creating the And functions and moving beat using multiplication and division.

function bitand(a, b)
local result = 0
local bitval = 1
while a > 0 and b > 0 do
if a % 2 == 1 and b % 2 == 1 then -- test the rightmost bits
result = result + bitval      -- set the current bit
end
bitval = bitval * 2 -- shift left
a = math.floor(a/2) -- shift right
b = math.floor(b/2)
end
return result
end

function hash_name(input)
local hash = 5381
local base32_chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZ234567"
local result = ""

for i = 1, #input do
hash = (math.floor(hash*32) + hash) + string.byte(input, i)
hash = bitand(hash,0xFFFFFFFFFF)
end

for i = 1, 8 do
local index = bitand(hash,0x1F) + 1
result = base32_chars:sub(index, index) .. result
hash = math.floor(hash/32)
end

return result
end

print(Data.StudyDate .. '/' .. hash_name(Data.StudyInstanceUID) .. '/' .. hash_name(Data.SeriesInstanceUID) .. '/' .. Data.SOPInstanceUID .. '.dcm')function bitand(a, b)
local result = 0
local bitval = 1
while a > 0 and b > 0 do
if a % 2 == 1 and b % 2 == 1 then -- test the rightmost bits
result = result + bitval      -- set the current bit
end
bitval = bitval * 2 -- shift left
a = math.floor(a/2) -- shift right
b = math.floor(b/2)
end
return result
end

function hash_name(input)
local hash = 5381
local base32_chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZ234567"
local result = ""

for i = 1, #input do
hash = (math.floor(hash*32) + hash) + string.byte(input, i)
hash = bitand(hash,0xFFFFFFFFFF)
end

for i = 1, 8 do
local index = bitand(hash,0x1F) + 1
result = base32_chars:sub(index, index) .. result
hash = math.floor(hash/32)
end

return result
end

print(Data.StudyDate .. '/' .. hash_name(Data.StudyInstanceUID) .. '/' .. hash_name(Data.SeriesInstanceUID) .. '/' .. Data.SOPInstanceUID .. '.dcm')

The only thing missing is to be able to launch that script with the arrival of each image at the pacs and everything would be resolved.

Of course, a much better solution would be to integrate this into the C source code, and through some tag to be able to incorporate it directly into the dicom.ini to create the name.