remove MQTT dependency

Reviewed-on: #6

Manifest.toml

@@ -1,8 +1,8 @@
 # This file is machine-generated - editing it directly is not advised
 
-julia_version = "1.11.2"
+julia_version = "1.11.5"
 manifest_format = "2.0"
-project_hash = "75c6a269a13b222c106479d2177b05facfa23f74"
+project_hash = "a942446c2f26ef72d0c4b0ca522e0adcf709ce4e"
 
 [[deps.AliasTables]]
 deps = ["PtrArrays", "Random"]
@@ -22,12 +22,23 @@ version = "1.11.0"
 uuid = "2a0f44e3-6c83-55bd-87e4-b1978d98bd5f"
 version = "1.11.0"
 
+[[deps.BufferedStreams]]
+git-tree-sha1 = "6863c5b7fc997eadcabdbaf6c5f201dc30032643"
+uuid = "e1450e63-4bb3-523b-b2a4-4ffa8c0fd77d"
+version = "1.2.2"
+
 [[deps.CSV]]
 deps = ["CodecZlib", "Dates", "FilePathsBase", "InlineStrings", "Mmap", "Parsers", "PooledArrays", "PrecompileTools", "SentinelArrays", "Tables", "Unicode", "WeakRefStrings", "WorkerUtilities"]
 git-tree-sha1 = "deddd8725e5e1cc49ee205a1964256043720a6c3"
 uuid = "336ed68f-0bac-5ca0-87d4-7b16caf5d00b"
 version = "0.10.15"
 
+[[deps.CodecBase]]
+deps = ["TranscodingStreams"]
+git-tree-sha1 = "40956acdbef3d8c7cc38cba42b56034af8f8581a"
+uuid = "6c391c72-fb7b-5838-ba82-7cfb1bcfecbf"
+version = "0.3.4"
+
 [[deps.CodecZlib]]
 deps = ["TranscodingStreams", "Zlib_jll"]
 git-tree-sha1 = "bce6804e5e6044c6daab27bb533d1295e4a2e759"
@@ -81,11 +92,6 @@ deps = ["Printf"]
 uuid = "ade2ca70-3891-5945-98fb-dc099432e06a"
 version = "1.11.0"
 
-[[deps.Distributed]]
-deps = ["Random", "Serialization", "Sockets"]
-uuid = "8ba89e20-285c-5b6f-9357-94700520ee1b"
-version = "1.11.0"
-
 [[deps.Distributions]]
 deps = ["AliasTables", "FillArrays", "LinearAlgebra", "PDMats", "Printf", "QuadGK", "Random", "SpecialFunctions", "Statistics", "StatsAPI", "StatsBase", "StatsFuns"]
 git-tree-sha1 = "3101c32aab536e7a27b1763c0797dba151b899ad"
@@ -148,6 +154,11 @@ deps = ["Random"]
 uuid = "9fa8497b-333b-5362-9e8d-4d0656e87820"
 version = "1.11.0"
 
+[[deps.HashArrayMappedTries]]
+git-tree-sha1 = "2eaa69a7cab70a52b9687c8bf950a5a93ec895ae"
+uuid = "076d061b-32b6-4027-95e0-9a2c6f6d7e74"
+version = "0.2.0"
+
 [[deps.HypergeometricFunctions]]
 deps = ["LinearAlgebra", "OpenLibm_jll", "SpecialFunctions"]
 git-tree-sha1 = "b1c2585431c382e3fe5805874bda6aea90a95de9"
@@ -264,21 +275,17 @@ version = "0.3.28"
 uuid = "56ddb016-857b-54e1-b83d-db4d58db5568"
 version = "1.11.0"
 
-[[deps.MQTTClient]]
-deps = ["Distributed", "Random", "Sockets"]
-git-tree-sha1 = "f2597b290d4bf17b577346153cd2ddf9accb5c26"
-uuid = "985f35cc-2c3d-4943-b8c1-f0931d5f0959"
-version = "0.3.1"
-weakdeps = ["PrecompileTools"]
-
-    [deps.MQTTClient.extensions]
-    PrecompileMQTT = "PrecompileTools"
-
 [[deps.Markdown]]
 deps = ["Base64"]
 uuid = "d6f4376e-aef5-505a-96c1-9c027394607a"
 version = "1.11.0"
 
+[[deps.MbedTLS]]
+deps = ["Dates", "MbedTLS_jll", "MozillaCACerts_jll", "NetworkOptions", "Random", "Sockets"]
+git-tree-sha1 = "c067a280ddc25f196b5e7df3877c6b226d390aaf"
+uuid = "739be429-bea8-5141-9913-cc70e7f3736d"
+version = "1.1.9"
+
 [[deps.MbedTLS_jll]]
 deps = ["Artifacts", "Libdl"]
 uuid = "c8ffd9c3-330d-5841-b78e-0817d7145fa1"
@@ -298,6 +305,18 @@ version = "1.11.0"
 uuid = "14a3606d-f60d-562e-9121-12d972cd8159"
 version = "2023.12.12"
 
+[[deps.NATS]]
+deps = ["Base64", "BufferedStreams", "CodecBase", "Dates", "DocStringExtensions", "JSON3", "MbedTLS", "NanoDates", "Random", "ScopedValues", "Sockets", "Sodium", "StructTypes", "URIs"]
+git-tree-sha1 = "d9d9a189fb9155a460e6b5e8966bf6a66737abf8"
+uuid = "55e73f9c-eeeb-467f-b4cc-a633fde63d2a"
+version = "0.1.0"
+
+[[deps.NanoDates]]
+deps = ["Dates", "Parsers"]
+git-tree-sha1 = "850a0557ae5934f6e67ac0dc5ca13d0328422d1f"
+uuid = "46f1a544-deae-4307-8689-c12aa3c955c6"
+version = "1.0.3"
+
 [[deps.NetworkOptions]]
 uuid = "ca575930-c2e3-43a9-ace4-1e988b2c1908"
 version = "1.2.0"
@@ -310,7 +329,7 @@ version = "0.3.27+1"
 [[deps.OpenLibm_jll]]
 deps = ["Artifacts", "Libdl"]
 uuid = "05823500-19ac-5b8b-9628-191a04bc5112"
-version = "0.8.1+2"
+version = "0.8.5+0"
 
 [[deps.OpenSpecFun_jll]]
 deps = ["Artifacts", "CompilerSupportLibraries_jll", "JLLWrappers", "Libdl", "Pkg"]
@@ -423,6 +442,12 @@ version = "0.5.1+0"
 uuid = "ea8e919c-243c-51af-8825-aaa63cd721ce"
 version = "0.7.0"
 
+[[deps.ScopedValues]]
+deps = ["HashArrayMappedTries", "Logging"]
+git-tree-sha1 = "1147f140b4c8ddab224c94efa9569fc23d63ab44"
+uuid = "7e506255-f358-4e82-b7e4-beb19740aa63"
+version = "1.3.0"
+
 [[deps.SentinelArrays]]
 deps = ["Dates", "Random"]
 git-tree-sha1 = "d0553ce4031a081cc42387a9b9c8441b7d99f32d"
@@ -437,6 +462,12 @@ version = "1.11.0"
 uuid = "6462fe0b-24de-5631-8697-dd941f90decc"
 version = "1.11.0"
 
+[[deps.Sodium]]
+deps = ["Base64", "libsodium_jll"]
+git-tree-sha1 = "907703e0d50846f300650d7225bdcab145b7bca9"
+uuid = "4f5b5e99-b0ad-42cd-b47a-334e172ec8bd"
+version = "1.1.2"
+
 [[deps.SortingAlgorithms]]
 deps = ["DataStructures"]
 git-tree-sha1 = "66e0a8e672a0bdfca2c3f5937efb8538b9ddc085"
@@ -544,6 +575,11 @@ git-tree-sha1 = "0c45878dcfdcfa8480052b6ab162cdd138781742"
 uuid = "3bb67fe8-82b1-5028-8e26-92a6c54297fa"
 version = "0.11.3"
 
+[[deps.URIs]]
+git-tree-sha1 = "24c1c558881564e2217dcf7840a8b2e10caeb0f9"
+uuid = "5c2747f8-b7ea-4ff2-ba2e-563bfd36b1d4"
+version = "1.6.0"
+
 [[deps.UUIDs]]
 deps = ["Random", "SHA"]
 uuid = "cf7118a7-6976-5b1a-9a39-7adc72f591a4"
@@ -574,6 +610,12 @@ deps = ["Artifacts", "Libdl"]
 uuid = "8e850b90-86db-534c-a0d3-1478176c7d93"
 version = "5.11.0+0"
 
+[[deps.libsodium_jll]]
+deps = ["Artifacts", "JLLWrappers", "Libdl", "Pkg"]
+git-tree-sha1 = "f76d682d87eefadd3f165d8d9fda436464213142"
+uuid = "a9144af2-ca23-56d9-984f-0d03f7b5ccf8"
+version = "1.0.20+3"
+
 [[deps.nghttp2_jll]]
 deps = ["Artifacts", "Libdl"]
 uuid = "8e850ede-7688-5339-a07c-302acd2aaf8d"

Project.toml

@@ -1,7 +1,7 @@
 name = "GeneralUtils"
 uuid = "c6c72f09-b708-4ac8-ac7c-2084d70108fe"
 authors = ["tonaerospace <tonaerospace.etc@gmail.com>"]
-version = "0.2.2"
+version = "0.3.1"
 
 [deps]
 CSV = "336ed68f-0bac-5ca0-87d4-7b16caf5d00b"
@@ -10,8 +10,11 @@ DataStructures = "864edb3b-99cc-5e75-8d2d-829cb0a9cfe8"
 Dates = "ade2ca70-3891-5945-98fb-dc099432e06a"
 Distributions = "31c24e10-a181-5473-b8eb-7969acd0382f"
 JSON3 = "0f8b85d8-7281-11e9-16c2-39a750bddbf1"
-MQTTClient = "985f35cc-2c3d-4943-b8c1-f0931d5f0959"
+NATS = "55e73f9c-eeeb-467f-b4cc-a633fde63d2a"
 PrettyPrinting = "54e16d92-306c-5ea0-a30b-337be88ac337"
 Random = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
 SHA = "ea8e919c-243c-51af-8825-aaa63cd721ce"
 UUIDs = "cf7118a7-6976-5b1a-9a39-7adc72f591a4"
+
+[compat]
+NATS = "0.1.0"

codesnippet/nats.jl

@@ -0,0 +1,79 @@
+
+using NATS, JSON3
+connection  = NATS.connect("nats.yiem.cc:4222")
+sub1 = NATS.reply(connection, "some_subject"; queue_group="group1") do msg
+                  payload = copy(JSON3.read(msg.payload))
+                  println(payload)
+                  println(msg.reply_to)
+                  # publish(connection, msg.reply_to, "ACK")
+                  return JSON3.write(Dict(:a=>"wassup"))
+              end
+
+
+
+
+
+
+
+
+
+
+using NATS, JSON3, GeneralUtils
+connection  = NATS.connect("nats.yiem.cc:4222")
+
+msgMeta = GeneralUtils.generate_msgMeta(
+  "text2textinstruct_medium.inference.api.v1";
+  msgPurpose= "inference",
+  senderName= "yiemagent",
+  senderId= GeneralUtils.uuid4snakecase(),
+  receiverName= "text2textinstruct",
+)
+
+llmHttpTimeout = 60
+outgoingMsg = Dict(
+    :msgMeta=> msgMeta,
+    :payload=> Dict(
+      :text=> "Wassup buddy!",
+      :kwargs=> Dict(
+        :max_tokens=> 2048,
+        :stop=> ["<|im_end|>"],
+        :temperature=> 0.2,
+      ),
+      :llmHttpTimeout=>llmHttpTimeout,
+    )
+  )
+
+r = NATS.request(String, connection, "text2textinstruct_medium.inference.api.v1", 
+                JSON3.write(outgoingMsg); timer=Timer(llmHttpTimeout))
+
+
+
+
+
+
+
+
+
+
+
+using NATS, JSON3, GeneralUtils
+connection  = NATS.connect("nats.yiem.cc:4222")
+
+msgMeta = GeneralUtils.generate_msgMeta(
+  "tonpc.containerServices",
+  msgPurpose="reset container",
+  senderName= "",
+)
+
+outgoingMsg = Dict(
+  :msgMeta=> msgMeta,
+  :payload=> "docker container restart ollama-instance-2",
+)
+
+# may be I can't use NATS request inside NATS reply??
+r = NATS.request(String, connection, msgMeta[:sendTopic], JSON3.write(outgoingMsg); timer=Timer(10))
+
+
+
+
+

src/GeneralUtils.jl

@@ -2,7 +2,7 @@ module GeneralUtils
     
 
 export  # struct
-        mqttClientInstance,
+        # mqttClientInstance,
         # function
         noNegative!, randomWithProb, randomChoiceWithProb, findIndex, limitvalue 
 

src/dbUtil.jl

@@ -2,7 +2,7 @@ module dbUtil
   
 export dictToPostgresKeyValueString, generateInsertSQL, generateUpdateSQL
 
-using JSON3, DataStructures, Distributions, Random, Dates, UUIDs, MQTTClient, DataFrames,
+using JSON3, DataStructures, Distributions, Random, Dates, UUIDs, DataFrames,
       SHA
 using ..util
 

src/interface.jl

@@ -6,9 +6,9 @@ export noNegative!, randomWithProb, randomChoiceWithProb, findIndex, limitvalue,
         matMul_3Dto4D_batchwise, isNotEqual, linearToCartesian, vectorMax, findMax,
         multiply_last, multiplyRandomElements, replaceElements, replaceElements!, isBetween,
         isLess, allTrue, getStringBetweenCharacters, JSON3read_stringKey, mkDictPath!, 
-        getDictPath
+        getDictPath, detectKeywordVariation, textToDict
 
-using JSON3, DataStructures, Distributions, Random, Dates, UUIDs, MQTTClient, DataFrames, CSV
+using JSON3, DataStructures, Distributions, Random, Dates, UUIDs, DataFrames, CSV
 using ..util, ..communication
 
 # ---------------------------------------------- 100 --------------------------------------------- #
@@ -1150,9 +1150,196 @@ end
 
 
 
+"""
+    detectKeywordVariation(keywords::AbstractVector{String}, text::String) -> Dict{String, Union{Array, Nothing}}
+
+Detects and collects all case-variant occurrences of multiple keywords in the text. 
+This function processes each keyword individually and returns an array of matched variations for each keyword.
+
+# Arguments
+- `keywords::AbstractVector{String}` Vector of keywords to search for
+- `text::String` The text to search in
+
+# Returns
+- `Dict{String, Array}` Returns a dictionary mapping each keyword to an array of matched variations found in the text
+
+# Examples
+  ```jldoctest
+  julia> detectKeywordVariation(["test", "example", "cat"], "This is a Test EXAMPLE")
+  Dict{String, Array}("test" => ["Test"], "example" => ["EXAMPLE"], "cat" => nothing)
+"""
+function detectKeywordVariation(keywords::T, text::String)::Dict{String, Union{Array, Nothing}} where {T<:AbstractVector}
+  kw = Dict{String, Union{Array, Nothing}}()
+  
+  # use for loop and detect_keyword function to get the exact variation of each keyword in the text then push to kw list
+  for keyword in keywords
+    ws = detectKeywordVariation.(keyword, text)
+    total = sum(issomething.(ws))
+    if total != 0
+      kw[keyword] = ws
+    else
+      kw[keyword] = nothing
+    end
+  end
+  return kw
+end
 
 
+"""
+    detectKeywordVariation(keyword::String, text::String) -> Union{Nothing, Array{String}}
 
+Detects if a keyword exists in the text in different case variations (lowercase, uppercase first letter, or all uppercase).
+
+# Arguments:
+- `keyword::String` The keyword to search for
+- `text::String` The text to search in
+
+# Returns:
+- `Union{Nothing, Array{String}}` Returns an array of matched keyword variations if found, otherwise returns nothing
+
+# Examples:
+  ```jldoctest
+  julia> detectKeywordVariation("test", "This is a Test case")
+  ["Test"]
+
+  julia> detectKeywordVariation("error", "NO ERRORS FOUND")
+  ["ERRORS"]
+
+  julia> detectKeywordVariation("missing", "complete data")
+  nothing
+"""
+function detectKeywordVariation(keyword::String, text::String)::Union{Nothing, Array{String}}
+  # Define the keyword variations to search for
+  wordVariations = [uppercasefirst(keyword), uppercase(keyword), lowercase(keyword)]
+  # wordVariations may duplicate keyword
+  keyword_variations = [keyword]
+  for i in wordVariations
+    i != keyword ? push!(keyword_variations, i) : nothing
+  end
+
+  _splittext = string.(strip.(split(text, " ")))
+  splittext = String[]
+  # remove . after a word
+  for i in _splittext
+    if length(i) != 0 && i[end] ∈ ['.']
+      word = string(i[1:end-1])
+      push!(splittext, word)
+    else
+      push!(splittext, i)
+    end
+  end
+
+  result = String[]
+  for variation in keyword_variations
+    # if length of both word is equals then it is a whole word otherwise it is part of part of other word
+    r = findIndex(splittext, variation)
+
+    if isempty(r[2])
+      # skip
+    else
+      # if variation > 1 add them all so this function detect duplicate keyword
+      variations = [variation for i in eachindex(r[2])]
+      result = vcat(result, variations)
+    end
+  end
+  return result
+end
+
+
+""" Convert text into a dictionary with a given keywords. This function use keywords to slice
+  a given text into the following format: KW1|kw1_text|KW2|kw2_text|KW3|kw3_text.
+  The left most string which has no keyword will be discarded. WARNING, ordering is important
+
+# Arguments
+  - `text::String`
+    A text to be converted.
+  - `keywords::Vector{String}`
+    A list of keywords to be used to slice the text. 
+    These keywords also be the resulting dict keys.
+# Keyword Arguments
+  - `rightmarker::String`
+    A maker used to make a word to be unique. Ex, A keyword "plan" with rightmarker ":", 
+    the function will search for "plan:" otherwise the function will search for "plan".
+    The marker will not be in the resulting dict keys.
+  - `symbolkey::Bool`
+    If true, resulting dict's key will be Symbols, otherwise string.
+  - `lowercasekey::Bool`
+    set resulting dict's key to be lowercase
+
+# Return
+  - `d::OrderedDict`
+
+# Example
+```jldoctest
+julia> text = "TODAY thought: what to do plan: wake up and going out action: 1. wake up 2. eat 3. sleep"
+julia> sample_keywords = ["thought", "plan", "action"]
+julia> resultdict = GeneralUtils.textToDict(text, sample_keywords; rightmarker=":", symbolkey=true)
+julia> println(resultdict)
+OrderedCollections.OrderedDict{Any, Any}(:thought => "what to do", 
+                                        :plan => "wake up and going out", 
+                                        :action => "1. wake up 2. eat 3. sleep")
+```
+
+# Signature
+"""
+function textToDict(text::String, detectKeywords::Vector{String}; 
+    dictKey::Union{Vector{String}, Nothing}=nothing, 
+    symbolkey::Bool=false, lowercasekey::Bool=false
+    )::OrderedDict
+
+  # make sure this function detect variation of a work e.g. agent, Agent, AGENT
+  kw = []
+  # use for loop and detect_keyword function to get the exact variation of each keyword in the text then push to kw list
+  for keyword in detectKeywords
+    detected = detectKeywordVariation(keyword, text)
+    if detected !== nothing
+      push!(kw, detected)
+    else
+      error("Keyword $keyword not found in text: $text")
+    end
+  end
+  if typeof(kw[1]) <: AbstractArray
+    kw = reduce(vcat, kw)
+  end
+
+  od1, od2 =
+  if symbolkey
+    OrderedDict{Symbol, Any}(), OrderedDict{Symbol, Any}()
+  else
+    OrderedDict{String, Any}(), OrderedDict{String, Any}()
+  end
+
+  remainingtext = text
+  dictKey_ = reverse(dictKey)
+  
+  # process text from back to front
+  rkw = reverse(kw)
+  for (i,keyword) in enumerate(rkw)
+    # Find the position of the keyword in the text
+    keywordidx = findlast(keyword, remainingtext)
+    dKey = dictKey_[i]
+    
+    if keywordidx !== nothing
+      substr = remainingtext[keywordidx[end]+1:end]
+      str = string(strip(substr))  # Removes both leading and trailing whitespace.
+      _key = lowercasekey == true ? lowercase(dKey) : dKey
+      key = symbolkey == true ? Symbol(_key) : _key
+      od1[key] = str
+      remainingtext = remainingtext[1:keywordidx[1]-1]
+    else
+      error("""keyword "$keyword" not found in the provided text: $text </end of error note>""")
+    end
+  end
+
+  # correct the order
+  ks = reverse([i for i in keys(od1)])
+  for k in ks
+    k = symbolkey == true ? Symbol(k) : k
+    od2[k] = od1[k]
+  end
+  
+  return od2
+end
 
 
 

src/llmUtil.jl

@@ -1,6 +1,6 @@
 module llmUtil
 
-export formatLLMtext, formatLLMtext_llama3instruct, jsoncorrection
+export formatLLMtext, formatLLMtext_llama3instruct, jsoncorrection, deFormatLLMtext, extractthink
 
 using UUIDs, JSON3, Dates
 using GeneralUtils
@@ -43,7 +43,7 @@ julia> formattedtext = YiemAgent.formatLLMtext_llama3instruct(d[:name], d[:text]
 Signature
 """
 function formatLLMtext_llama3instruct(name::T, text::T;
-  assistantStarter::Bool=true) where {T<:AbstractString}
+  assistantStarter::Bool=false) where {T<:AbstractString}
   formattedtext = 
   if name == "system"
     """
@@ -68,28 +68,10 @@ function formatLLMtext_llama3instruct(name::T, text::T;
 
   return formattedtext
 end
-# function formatLLMtext_llama3instruct(name::T, text::T) where {T<:AbstractString}
-#   formattedtext = 
-#   if name == "system"
-#     """
-#     <|begin_of_text|>
-#     <|start_header_id|>$name<|end_header_id|>
-#     $text
-#     <|eot_id|>
-#     """
-#   else
-#     """
-#     <|start_header_id|>$name<|end_header_id|>
-#     $text
-#     <|eot_id|>
-#     """
-#   end
 
-#   return formattedtext
-# end
 
 function formatLLMtext_qwen(name::T, text::T;
-  assistantStarter::Bool=true) where {T<:AbstractString}
+  assistantStarter::Bool=false) where {T<:AbstractString}
   formattedtext = 
   if name == "system"
     """
@@ -116,14 +98,94 @@ function formatLLMtext_qwen(name::T, text::T;
 end
 
 
-""" Convert a chat messages in vector of dictionary into LLM model instruct format.
+function formatLLMtext_qwen3(name::T, text::T;
+  assistantStarter::Bool=false) where {T<:AbstractString}
+  formattedtext = 
+  if name == "system"
+    """
+    <|im_start|>$name
+    $text
+    <|im_end|>
+    """
+  else
+    """
+    <|im_start|>$name
+    $text
+    <|im_end|>
+    """
+  end
+
+  if assistantStarter
+    formattedtext *=
+      """
+      <|im_start|>assistant
+      """
+  end
+
+  return formattedtext
+end
+
+
+function formatLLMtext_phi4(name::T, text::T;
+  assistantStarter::Bool=false) where {T<:AbstractString}
+  formattedtext = 
+  if name == "system"
+    """
+    <|system|>
+    $text
+    <|end|>
+    """
+  else
+    """
+    <|assistant|>
+    $text
+    <|end|>
+    """
+  end
+
+  if assistantStarter
+    formattedtext *=
+      """
+      <|assistant|>
+      """
+  end
+
+  return formattedtext
+end
+
+
+function formatLLMtext_granite3(name::T, text::T;
+  assistantStarter::Bool=false) where {T<:AbstractString}
+  formattedtext = 
+  if name == "system"
+    """
+    <|start_of_role|>system<|end_of_role|>{$text}<|end_of_text|>
+    """
+  else
+    """
+    <|start_of_role|>$name<|end_of_role|>{$text}<|end_of_text|>
+    """
+  end
+
+  if assistantStarter
+    formattedtext *=
+      """
+      <|start_of_role|>assistant<|end_of_role|>{
+      """
+  end
+
+  return formattedtext
+end
+
+
+
+""" Convert a vector of chat message dictionaries into LLM model instruct format.
 
 # Arguments
   - `messages::Vector{Dict{Symbol, T}}`
-    message owner name e.f. "system", "user" or "assistant"
+    A vector of dictionaries where each dictionary contains the keys `:name` (the name of the message owner) and `:text` (the text of the message).
   - `formatname::T`
-    format name to be used
+    The name of the format to be used for converting the chat messages.
-
 # Return
   - `formattedtext::String`
     text formatted to model format
@@ -140,12 +202,11 @@ julia> chatmessage = [
 julia> formattedtext = YiemAgent.formatLLMtext(chatmessage, "llama3instruct")
 "<|begin_of_text|>\n  <|start_header_id|>system<|end_header_id|>\n  You are a helpful, respectful and honest assistant.\n  <|eot_id|>\n  <|start_header_id|>user<|end_header_id|>\n  list me all planets in our solar system.\n  <|eot_id|>\n  <|start_header_id|>assistant<|end_header_id|>\n  I'm sorry. I don't know. You tell me.\n  <|eot_id|>\n"
 ```
-
-# Signature
 """
-function formatLLMtext(messages::Vector{Dict{Symbol, T}}; formatname::String="llama3instruct"
+function formatLLMtext(messages::Vector{Dict{Symbol, T}}, formatname::String
-  )::String where {T<:Any}
+  )::String where {T<:AbstractString}
-  f = if formatname == "llama3instruct"
+  f = 
+    if formatname == "llama3instruct"
       formatLLMtext_llama3instruct
     elseif formatname == "mistral"
       # not define yet
@@ -153,18 +214,125 @@ function formatLLMtext(messages::Vector{Dict{Symbol, T}}; formatname::String="ll
       # not define yet
     elseif formatname == "qwen"
       formatLLMtext_qwen
+    elseif formatname == "qwen3"
+      formatLLMtext_qwen3
+    elseif formatname == "phi4"
+      formatLLMtext_phi4
+    elseif formatname == "granite3"
+      formatLLMtext_granite3
     else
       error("$formatname template not define yet")
     end
 
   str = ""
-  for t in messages
+  for (i, t) in enumerate(messages)
+    if i < length(messages)
       str *= f(t[:name], t[:text])
+    else
+      str *= f(t[:name], t[:text]; assistantStarter=true)
+    end
   end
 
   return str
 end
 
+""" Revert LLM-format response back into regular text.
+
+# Arguments
+  - `text::String`
+    The LLM formatted string to be converted.
+
+# Return
+  - `normalText::String`
+    The original plain text extracted from the given LLM-formatted string.
+
+# Example
+```jldoctest
+julia> using Revise
+julia> using YiemAgent
+julia> response = "<|begin_of_text|>This is a sample system instruction.<|eot_id|>"
+julia> normalText = YiemAgent.deFormatLLMtext(response, "granite3")
+"This is a sample system instruction."
+```
+"""
+function deFormatLLMtext(text::String, formatname::String; includethink::Bool=false
+  )::String
+  f = 
+    if formatname == "granite3"
+      deFormatLLMtext_granite3
+    elseif formatname == "qwen3"
+      deFormatLLMtext_qwen3
+    else
+      error("$formatname template not define yet")
+    end
+
+  r = f(text)
+  result = r === nothing ? text : r
+  return result
+end
+
+
+""" Revert LLM-format response back into regular text for Granite 3 format.
+
+# Arguments
+  - `text::String`
+    The LLM formatted string to be converted.
+
+# Return
+  - `normalText::Union{Nothing, String}`
+    The original plain text extracted from the given LLM-formatted string.
+    Returns nothing if the text is not in Granite 3 format.
+
+# Example
+```jldoctest
+julia> using Revise
+julia> using YiemAgent
+julia> response = "{This is a sample LLM response.}"
+julia> normalText = YiemAgent.deFormatLLMtext(response, "granite3")
+"This is a sample LLM response."
+"""
+function deFormatLLMtext_granite3(text::String)::Union{Nothing, String}
+  # check if '{' and '}' are in the text because it's a special format for the LLM response
+  if contains(text, "<|im_start|>assistant")
+    # get the text between '{' and '}'
+    text_between_braces = GeneralUtils.extractTextBetweenCharacter(text, '{', '}')[1]
+    return text_between_braces
+  elseif text[end] == '}'
+    text = "{$text"
+    text_between_braces = GeneralUtils.extractTextBetweenCharacter(text, '{', '}')[1]
+  else
+    return nothing
+  end
+end
+
+
+function deFormatLLMtext_qwen3(text::String)::Union{Nothing, String}
+  return text
+end
+
+# function deFormatLLMtext_qwen3(text::String; includethink::Bool=false)::Union{Nothing, String}
+#   think = nothing
+#   str = nothing
+
+#   if occursin("<think>", text)
+#     r = GeneralUtils.extractTextBetweenString(text, "<think>", "</think>")
+#     if r[:success]
+#       think = r[:text]
+#     end
+#     str = string(split(text, "</think>")[2])
+#   end
+
+#   if includethink == true && occursin("<think>", text)
+#     result = "ModelThought: $think $str"
+#     return result
+#   elseif includethink == false && occursin("<think>", text)
+#     result = str
+#     return result
+#   else
+#     return text
+#   end
+# end
+
 
 """ Attemp to correct LLM response's incorrect JSON response.
 
@@ -255,7 +423,20 @@ function jsoncorrection(config::T1, input::T2, correctJsonExample::T3;
 end
 
 
-
+function extractthink(text::String)
+  think = nothing
+  str = nothing
+  if occursin("<think>", text)
+    r = GeneralUtils.extractTextBetweenString(text, "<think>", "</think>")
+    if r[:success]
+      think = r[:text]
+    end
+    str = string(split(text, "</think>")[2])
+  else
+    str = text
+  end
+  return think, str
+end
 
 
 

src/util.jl

@@ -1,14 +1,15 @@
 module util
   
 export timedifference, showstracktrace, findHighestIndexKey, uuid4snakecase, replaceDictKeys,
-        findMatchingDictKey, textToDict, randstring, randstrings, timeout,
+        findMatchingDictKey, randstring, randstrings, timeout,
         dataframeToCSV, dfToVectorDict, disintegrate_vectorDict, getDataFrameValue, dfRowtoString, 
-        dfToString, dataframe_to_json_list, dictToString, dictToString_noKey, 
+        dfToString, dataframe_to_json_list, dictToString, dictToString_noKey, issomething,
         dictToString_numbering, extract_triple_backtick_text, 
-        countGivenWords, remove_french_accents, detect_keyword, extractTextBetweenCharacter,
+        countGivenWords, remove_french_accents, 
-        convertCamelSnakeKebabCase
+        extractTextBetweenCharacter, extractTextBetweenString, 
+        convertCamelSnakeKebabCase, fitrange, recentElementsIndex, nonRecentElementsIndex
 
-using JSON3, DataStructures, Distributions, Random, Dates, UUIDs, MQTTClient, DataFrames
+using JSON3, DataStructures, Distributions, Random, Dates, UUIDs, DataFrames
 
 # ---------------------------------------------- 100 --------------------------------------------- #
 
@@ -42,6 +43,7 @@ function timedifference(starttime::DateTime, stoptime::DateTime, unit::String)::
   diff = stoptime - starttime
   unit = lowercase(unit)
   
+    # Check the unit and calculate the time difference accordingly
   if unit == "milliseconds"
       return diff.value
   elseif unit == "seconds"
@@ -242,101 +244,6 @@ function replaceDictKeys(d::Dict, replacementMap::Dict)::Dict
 end
 
 
-""" Convert text into a dictionary with a given keywords. This function use keywords to slice
-  a given text into the following format: KW1|kw1_text|KW2|kw2_text|KW3|kw3_text.
-  The left most string which has no keyword will be discarded. WARNING, ordering is important
-
-# Arguments
-  - `text::String`
-    A text to be converted.
-  - `keywords::Vector{String}`
-    A list of keywords to be used to slice the text. 
-    These keywords also be the resulting dict keys.
-# Keyword Arguments
-  - `rightmarker::String`
-    A maker used to make a word to be unique. Ex, A keyword "plan" with rightmarker ":", 
-    the function will search for "plan:" otherwise the function will search for "plan".
-    The marker will not be in the resulting dict keys.
-  - `symbolkey::Bool`
-    If true, resulting dict's key will be Symbols, otherwise string.
-  - `lowercasekey::Bool`
-    set resulting dict's key to be lowercase
-
-# Return
-  - `d::OrderedDict`
-
-# Example
-```jldoctest
-julia> text = "TODAY thought: what to do plan: wake up and going out action: 1. wake up 2. eat 3. sleep"
-julia> sample_keywords = ["thought", "plan", "action"]
-julia> resultdict = GeneralUtils.textToDict(text, sample_keywords; rightmarker=":", symbolkey=true)
-julia> println(resultdict)
-OrderedCollections.OrderedDict{Any, Any}(:thought => "what to do", 
-                                        :plan => "wake up and going out", 
-                                        :action => "1. wake up 2. eat 3. sleep")
-```
-
-# Signature
-"""
-function textToDict(text::String, detectKeywords::Vector{String}; 
-    dictKey::Union{Vector{String}, Nothing}=nothing, 
-    symbolkey::Bool=false, lowercasekey::Bool=false
-    )::OrderedDict
-
-  # make sure this function detect variation of a work e.g. agent, Agent, AGENT
-  kw = []
-  # use for loop and detect_keyword function to get the exact variation of each keyword in the text then push to kw list
-  for keyword in detectKeywords
-    detected = detect_keyword(keyword, text)
-    if detected !== nothing
-      push!(kw, detected)
-    else
-      error("Keyword $keyword not found in text.")
-    end
-  end
-
-  od1, od2 =
-  if symbolkey
-    OrderedDict{Symbol, Any}(), OrderedDict{Symbol, Any}()
-  else
-    OrderedDict{String, Any}(), OrderedDict{String, Any}()
-  end
-
-  remainingtext = text
-  dictKey_ = reverse(dictKey)
-  
-  # process text from back to front
-  for (i,keyword) in enumerate(reverse(kw))
-    # Find the position of the keyword in the text
-    keywordidx = findlast(keyword, remainingtext)
-    dKey = dictKey_[i]
-    
-    if keywordidx !== nothing
-      substr = remainingtext[keywordidx[end]+1:end]
-      str = string(strip(substr))  # Removes both leading and trailing whitespace.
-      _key = lowercasekey == true ? lowercase(dKey) : dKey
-      key = symbolkey == true ? Symbol(_key) : _key
-      od1[key] = str
-      remainingtext = remainingtext[1:keywordidx[1]-1]
-    else
-      error("""keyword "$keyword" not found in the provided text: $text </end of error note>""")
-    end
-  end
-
-  # correct the order
-  ks = reverse([i for i in keys(od1)])
-  for k in ks
-    k = symbolkey == true ? Symbol(k) : k
-    od2[k] = od1[k]
-  end
-  
-  return od2
-end
-
-
-
-
-
 """ Generate a random string
 
 # Arguments
@@ -770,46 +677,17 @@ function extract_triple_backtick_text(input::String)::Vector{String}
 end
 
 
-"""
+wordwindow(word::String, startindex::Integer)::UnitRange = startindex:startindex + length(word) -1
-    detect_keyword(keyword::String, text::String) -> Union{Nothing, String}
 
-Detects if a keyword exists in the text in different case variations (lowercase, uppercase first letter, or all uppercase).
+function cuttext(range, text)
-
+  # check whether range is outside text boundary
-# Arguments:
+  if range.start > length(text) || range.stop > length(text)
-- `keyword::String` The keyword to search for
-- `text::String` The text to search in
-
-# Returns:
-- `Union{Nothing, String}` Returns the matched keyword variation if found, otherwise returns nothing
-
-# Examples:
-  ```jldoctest
-  julia> detect_keyword("test", "This is a Test case")
-  "Test"
-
-  julia> detect_keyword("error", "NO ERRORS FOUND")
-  "ERRORS"
-
-  julia> detect_keyword("missing", "complete data")
-  nothing
-  ```
-
-# Signature
-"""
-function detect_keyword(keyword::String, text::String)::Union{Nothing, String}
-  # Define the keyword variations to search for
-  keyword_variations = [keyword, uppercasefirst(keyword), uppercase(keyword), lowercase(keyword)]
-  
-  # Check if any of the keyword variations are in the text
-  for variation in keyword_variations
-      if occursin(variation, text)
-          return variation
-      end
-  end
-  
-  # Return nothing if no variation is found
     return nothing
+  else
+    return text[range]
   end
+end
+
 
 
 """
@@ -924,9 +802,11 @@ Extracts and returns the text that is enclosed between two specified characters
 
 # Examples
 ```jldoctest
-julia> text = "Hello [World]!"
+julia> text = "Hello [World]! [Yay]"
 julia> extracted_text = extractTextBetweenCharacter(text, '[', ']')
-println(extracted_text)  # Output: "World"
+2-element Vector{Any}:
+ "World"
+ "Yay"
 ```
 """
 function extractTextBetweenCharacter(text::String, startchar::Char, endchar::Char)
@@ -950,6 +830,29 @@ function extractTextBetweenCharacter(text::String, startchar::Char, endchar::Cha
 end
 
 
+function extractTextBetweenString(text::String, startstr::String, endstr::String)
+  # check whether startstr is in the text or not
+  isStartStr = split(text, startstr)
+  if length(isStartStr) > 2
+    return (success=false, error="There are more than one occurrences of the start string '$startstr' in the text. Text must has only one start string", errorcode=2, result=nothing)
+  elseif length(isStartStr) == 1
+    return (success=false, error="There are no start string '$startstr' in the text. Text must has only one start string", errorcode=1, result=nothing)
+  end
+
+  # check whether endstr is in the text or not
+  isEndStr = split(text, endstr)
+  if length(isEndStr) > 2
+    return (success=false, error="There are more than one occurrences of the end string '$endstr' in the text. Text must has only one end string", errorcode=3, result=nothing)
+  elseif length(isStartStr) == 1
+    return (success=false, error="There are no end string '$endstr' in the text. Text must has only one end string", errorcode=4, result=nothing)
+  end
+
+  s = string(split(isStartStr[2], endstr)[1])
+  
+  return (success=true, error=nothing, errorcode=0, text=s)
+end
+
+
 """
 Determines if the given string follows camel case naming convention.
 
@@ -1075,6 +978,140 @@ function convertCamelSnakeKebabCase(text::T, tocase::Symbol)::String where {T<:A
 end
 
 
+""" Check if a value is not `nothing`.
+
+# Arguments
+- `x`: The value to check
+
+# Returns
+- `Bool`: `true` if `x` is not `nothing`, `false` otherwise
+
+# Examples
+```jldoctest
+julia> issomething(1)
+true
+julia> issomething(nothing)
+false
+julia> issomething("test")
+true
+````
+"""
+function issomething(x)
+  return x === nothing ? false : true
+end
+
+
+""" Adjust a given range to fit within the bounds of a vector's length.
+
+# Arguments
+  - `v::T1`
+    the input vector to check against
+  - `range::UnitRange`
+    the original range to be adjusted
+
+# Return
+  - `adjusted_range::UnitRange`
+    a range that is constrained to the vector's length, preventing out-of-bounds indexing
+
+# Example
+
+julia> v = [1, 2, 3, 4, 5]
+julia> fitrange(v, 3:10)
+3:5
+
+"""
+function fitrange(v::T1, range::UnitRange) where {T1<:AbstractVector}
+  totalelements = length(v)
+
+  startind =
+    # check if user put start range greater than total event
+    if range.start > totalelements
+      totalelements
+    else
+      range.start
+    end
+
+  stopind = 
+    if range.stop > totalelements
+      totalelements
+    else
+      range.stop
+    end
+
+  return startind:stopind
+end
+
+
+""" Find a unit range for a vector given a number of the most recent elements of interest.
+
+# Arguments
+  - `vectorLength::Integer`
+    the length of the vector to generate range from
+  - `n::Integer`
+    the number of most recent elements to include in range
+
+
+# Return
+  - `UnitRange`
+    a range representing the n most recent elements of a vector with length vectorLength
+
+# Example
+```jldoctest
+julia> a = [1, 2, 3, 4, 5]
+julia> recentElementsIndex(length(a), 3)
+3:5
+julia> recentElementsIndex(length(a), 0)
+5:5
+```
+"""
+function recentElementsIndex(vectorlength::Integer, n::Integer; includelatest::Bool=false)
+  if n == 0
+    error("n must be greater than 0")
+  end
+  
+  if includelatest
+    start = max(1, vectorlength - n + 1)
+    return start:vectorlength
+  else
+    startind = max(1, vectorlength - n)
+    endind = vectorlength -1
+    return startind:endind
+  end
+end
+
+
+""" Find a unit range for a vector excluding the most recent elements.
+
+# Arguments
+    - `vectorlength::Integer`
+      the length of the vector to generate range from
+    - `n::Integer`
+      the number of most recent elements to exclude from range
+
+# Return
+  - `UnitRange`
+    a range representing the elements of the vector excluding the last `n` elements
+
+# Example
+```jldoctest
+julia> a = [1, 2, 3, 4, 5]
+julia> nonRecentElementsIndex(length(a), 3)
+1:2
+julia> nonRecentElementsIndex(length(a), 1)
+1:4
+julia> nonRecentElementsIndex(length(a), 0)
+1:5
+```
+"""
+function nonRecentElementsIndex(vectorlength::Integer, n::Integer)
+  if n < 0
+      error("n must be non-negative")
+  end
+  if n > vectorlength
+      return 1:0  # empty range
+  end
+  return 1:(vectorlength-n)
+end
 
 
 

test/.vscode/settings.json

@@ -0,0 +1,3 @@
+{
+  "julia.environmentPath": "/appfolder/app/dev/GeneralUtils/test"
+}

test/Manifest.toml

@@ -0,0 +1,41 @@
+# This file is machine-generated - editing it directly is not advised
+
+julia_version = "1.11.4"
+manifest_format = "2.0"
+project_hash = "71d91126b5a1fb1020e1098d9d492de2a4438fd2"
+
+[[deps.Base64]]
+uuid = "2a0f44e3-6c83-55bd-87e4-b1978d98bd5f"
+version = "1.11.0"
+
+[[deps.InteractiveUtils]]
+deps = ["Markdown"]
+uuid = "b77e0a4c-d291-57a0-90e8-8db25a27a240"
+version = "1.11.0"
+
+[[deps.Logging]]
+uuid = "56ddb016-857b-54e1-b83d-db4d58db5568"
+version = "1.11.0"
+
+[[deps.Markdown]]
+deps = ["Base64"]
+uuid = "d6f4376e-aef5-505a-96c1-9c027394607a"
+version = "1.11.0"
+
+[[deps.Random]]
+deps = ["SHA"]
+uuid = "9a3f8284-a2c9-5f02-9a11-845980a1fd5c"
+version = "1.11.0"
+
+[[deps.SHA]]
+uuid = "ea8e919c-243c-51af-8825-aaa63cd721ce"
+version = "0.7.0"
+
+[[deps.Serialization]]
+uuid = "9e88b42a-f829-5b0c-bbe9-9e923198166b"
+version = "1.11.0"
+
+[[deps.Test]]
+deps = ["InteractiveUtils", "Logging", "Random", "Serialization"]
+uuid = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
+version = "1.11.0"

test/Project.toml

@@ -0,0 +1,2 @@
+[deps]
+Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"

test/runtest.jl

@@ -1,44 +0,0 @@
-using Revise
-using GeneralUtils, MQTTClient, JSON3
-
-mqttMsgReceiveTopic = ["/receivetopic_1", "/receivetopic_2"]
-mqttMsgReceiveChannel = (ch1=Channel(8), ch2=Channel(32))
-keepaliveChannel = Channel(8)
-function onMsgCallback(topic, payload)
-            jobj = JSON3.read(String(payload))
-            incomingMqttMsg = copy(jobj)  # convert json object into julia dictionary recursively
-            
-            if occursin("topic_1", topic)
-              put!(mqttMsgReceiveChannel[:ch1], incomingMqttMsg)
-            elseif occursin("topic_2", topic)
-              put!(mqttMsgReceiveChannel[:ch2], incomingMqttMsg)
-            elseif occursin("keepalive", topic)
-              put!(keepaliveChannel, incomingMqttMsg)
-            else
-              println("undefined condition ", @__FILE__, " ", @__LINE__)
-            end
-          end
-mqttInstance = GeneralUtils.mqttClientInstance_v2(
-                          "mqtt.yiem.cc",
-                          mqttMsgReceiveTopic,
-                          mqttMsgReceiveChannel,
-                          keepaliveChannel,
-                          onMsgCallback
-                        )
-
-
-_ = GeneralUtils.checkMqttConnection!(mqttInstance)
-
-
-println("GeneralUtils test done")
-
-
-
-
-
-
-
-
-
-
-

test/runtests.jl

@@ -0,0 +1,39 @@
+using Test
+using GeneralUtils: detect_keyword
+
+@testset "detect_keyword tests" begin
+    @test detect_keyword(["test"], "this is a test") == Dict("test" => 1)
+    
+    @test detect_keyword(["hello", "world"], "hello world hello") == Dict("hello" => 2, "world" => 1)
+    
+    @test detect_keyword(["cat"], "category") == Dict("cat" => 1)
+    
+    @test detect_keyword(["cat"], "category"; mode="individual") == Dict("cat" => 0)
+    
+    @test detect_keyword(["dog"], "dogs and cats"; mode="individual", delimiter=[' ']) == Dict("dog" => 0)
+    
+    @test detect_keyword(["test"], "test.case"; mode="individual", delimiter=['.']) == Dict("test" => 1)
+    
+    @test detect_keyword(["word"], "") == Dict("word" => 0)
+    
+    @test detect_keyword(String[], "some text") == Dict{String, Integer}()
+    
+    @test detect_keyword(["a", "b"], "a.b\nc"; delimiter=['.', '\n']) == Dict("a" => 1, "b" => 1)
+    
+    multiline_text = """
+    first line
+    second line
+    first word
+    """
+    @test detect_keyword(["first"], multiline_text) == Dict("first" => 2)
+    
+    @test detect_keyword(["word"], "word"; mode="individual") == Dict("word" => 1)
+    
+    @test detect_keyword(["test"], "testing.test.tester"; mode="individual", delimiter=['.']) == Dict("test" => 1)
+end
+
+
+
+
+
+